Annotation of gforth/doc/gforth.ds, revision 1.89

1.1       anton       1: \input texinfo   @c -*-texinfo-*-
                      2: @comment The source is gforth.ds, from which gforth.texi is generated
1.28      crook       3: 
1.21      crook       4: @comment TODO: nac29jan99 - a list of things to add in the next edit:
1.28      crook       5: @comment 1. x-ref all ambiguous or implementation-defined features?
                      6: @comment 2. Describe the use of Auser Avariable AConstant A, etc.
                      7: @comment 3. words in miscellaneous section need a home.
                      8: @comment 4. search for TODO for other minor and major works required.
                      9: @comment 5. [rats] change all @var to @i in Forth source so that info
                     10: @comment    file looks decent.
1.36      anton      11: @c          Not an improvement IMO - anton
                     12: @c          and anyway, this should be taken up
                     13: @c          with Karl Berry (the texinfo guy) - anton
1.29      crook      14: @comment .. would be useful to have a word that identified all deferred words
                     15: @comment should semantics stuff in intro be moved to another section
                     16: 
1.66      anton      17: @c POSTPONE, COMPILE, [COMPILE], LITERAL should have their own section
1.28      crook      18: 
1.1       anton      19: @comment %**start of header (This is for running Texinfo on a region.)
                     20: @setfilename gforth.info
                     21: @settitle Gforth Manual
                     22: @dircategory GNU programming tools
                     23: @direntry
                     24: * Gforth: (gforth).             A fast interpreter for the Forth language.
                     25: @end direntry
1.49      anton      26: @c The Texinfo manual also recommends doing this, but for Gforth it may
                     27: @c  not make much sense
                     28: @c @dircategory Individual utilities
                     29: @c @direntry
                     30: @c * Gforth: (gforth)Invoking Gforth.      gforth, gforth-fast, gforthmi
                     31: @c @end direntry
                     32: 
1.1       anton      33: @comment @setchapternewpage odd
1.29      crook      34: @comment TODO this gets left in by HTML converter
1.12      anton      35: @macro progstyle {}
                     36: Programming style note:
1.3       anton      37: @end macro
1.48      anton      38: 
                     39: @macro assignment {}
                     40: @table @i
                     41: @item Assignment:
                     42: @end macro
                     43: @macro endassignment {}
                     44: @end table
                     45: @end macro
                     46: 
1.1       anton      47: @comment %**end of header (This is for running Texinfo on a region.)
                     48: 
1.29      crook      49: 
                     50: @comment ----------------------------------------------------------
                     51: @comment macros for beautifying glossary entries
                     52: @comment if these are used, need to strip them out for HTML converter
                     53: @comment else they get repeated verbatim in HTML output.
                     54: @comment .. not working yet.
                     55: 
                     56: @macro GLOSS-START {}
                     57: @iftex
                     58: @ninerm
                     59: @end iftex
                     60: @end macro
                     61: 
                     62: @macro GLOSS-END {}
                     63: @iftex
                     64: @rm
                     65: @end iftex
                     66: @end macro
                     67: 
                     68: @comment ----------------------------------------------------------
                     69: 
                     70: 
1.10      anton      71: @include version.texi
                     72: 
1.49      anton      73: @ifnottex
1.11      anton      74: This file documents Gforth @value{VERSION}
1.1       anton      75: 
1.62      crook      76: Copyright @copyright{} 1995--2000 Free Software Foundation, Inc.
1.1       anton      77: 
                     78:      Permission is granted to make and distribute verbatim copies of
                     79:      this manual provided the copyright notice and this permission notice
                     80:      are preserved on all copies.
                     81:      
                     82: @ignore
                     83:      Permission is granted to process this file through TeX and print the
                     84:      results, provided the printed document carries a copying permission
                     85:      notice identical to this one except for the removal of this paragraph
                     86:      (this paragraph not being relevant to the printed manual).
                     87:      
                     88: @end ignore
                     89:      Permission is granted to copy and distribute modified versions of this
                     90:      manual under the conditions for verbatim copying, provided also that the
                     91:      sections entitled "Distribution" and "General Public License" are
                     92:      included exactly as in the original, and provided that the entire
                     93:      resulting derived work is distributed under the terms of a permission
                     94:      notice identical to this one.
                     95:      
                     96:      Permission is granted to copy and distribute translations of this manual
                     97:      into another language, under the above conditions for modified versions,
                     98:      except that the sections entitled "Distribution" and "General Public
                     99:      License" may be included in a translation approved by the author instead
                    100:      of in the original English.
1.49      anton     101: @end ifnottex
1.1       anton     102: 
                    103: @finalout
                    104: @titlepage
                    105: @sp 10
                    106: @center @titlefont{Gforth Manual}
                    107: @sp 2
1.11      anton     108: @center for version @value{VERSION}
1.1       anton     109: @sp 2
1.34      anton     110: @center Neal Crook
1.1       anton     111: @center Anton Ertl
1.6       pazsan    112: @center Bernd Paysan
1.5       anton     113: @center Jens Wilke
1.1       anton     114: @sp 3
1.47      crook     115: @center This manual is permanently under construction and was last updated on 15-Mar-2000
1.1       anton     116: 
                    117: @comment  The following two commands start the copyright page.
                    118: @page
                    119: @vskip 0pt plus 1filll
1.62      crook     120: Copyright @copyright{} 1995--2000 Free Software Foundation, Inc.
1.1       anton     121: 
                    122: @comment !! Published by ... or You can get a copy of this manual ...
                    123: 
                    124:      Permission is granted to make and distribute verbatim copies of
                    125:      this manual provided the copyright notice and this permission notice
                    126:      are preserved on all copies.
                    127:      
                    128:      Permission is granted to copy and distribute modified versions of this
                    129:      manual under the conditions for verbatim copying, provided also that the
                    130:      sections entitled "Distribution" and "General Public License" are
                    131:      included exactly as in the original, and provided that the entire
                    132:      resulting derived work is distributed under the terms of a permission
                    133:      notice identical to this one.
                    134:      
                    135:      Permission is granted to copy and distribute translations of this manual
                    136:      into another language, under the above conditions for modified versions,
                    137:      except that the sections entitled "Distribution" and "General Public
                    138:      License" may be included in a translation approved by the author instead
                    139:      of in the original English.
                    140: @end titlepage
                    141: 
                    142: @node Top, License, (dir), (dir)
1.49      anton     143: @ifnottex
1.1       anton     144: Gforth is a free implementation of ANS Forth available on many
1.11      anton     145: personal machines. This manual corresponds to version @value{VERSION}.
1.49      anton     146: @end ifnottex
1.1       anton     147: 
                    148: @menu
1.21      crook     149: * License::                     The GPL
1.26      crook     150: * Goals::                       About the Gforth Project
1.29      crook     151: * Gforth Environment::          Starting (and exiting) Gforth
1.48      anton     152: * Tutorial::                    Hands-on Forth Tutorial
1.21      crook     153: * Introduction::                An introduction to ANS Forth
1.1       anton     154: * Words::                       Forth words available in Gforth
1.24      anton     155: * Error messages::              How to interpret them
1.1       anton     156: * Tools::                       Programming tools
                    157: * ANS conformance::             Implementation-defined options etc.
1.65      anton     158: * Standard vs Extensions::      Should I use extensions?
1.1       anton     159: * Model::                       The abstract machine of Gforth
                    160: * Integrating Gforth::          Forth as scripting language for applications
                    161: * Emacs and Gforth::            The Gforth Mode
                    162: * Image Files::                 @code{.fi} files contain compiled code
                    163: * Engine::                      The inner interpreter and the primitives
1.24      anton     164: * Binding to System Library::   
1.13      pazsan    165: * Cross Compiler::              The Cross Compiler
1.1       anton     166: * Bugs::                        How to report them
                    167: * Origin::                      Authors and ancestors of Gforth
1.21      crook     168: * Forth-related information::   Books and places to look on the WWW
1.1       anton     169: * Word Index::                  An item for each Forth word
                    170: * Concept Index::               A menu covering many topics
1.12      anton     171: 
1.78      anton     172: @detailmenu --- The Detailed Node Listing ---
1.12      anton     173: 
1.29      crook     174: Gforth Environment
                    175: 
1.32      anton     176: * Invoking Gforth::             Getting in
                    177: * Leaving Gforth::              Getting out
                    178: * Command-line editing::        
1.48      anton     179: * Environment variables::       that affect how Gforth starts up
1.32      anton     180: * Gforth Files::                What gets installed and where
1.48      anton     181: * Startup speed::               When 35ms is not fast enough ...
                    182: 
                    183: Forth Tutorial
                    184: 
                    185: * Starting Gforth Tutorial::    
                    186: * Syntax Tutorial::             
                    187: * Crash Course Tutorial::       
                    188: * Stack Tutorial::              
                    189: * Arithmetics Tutorial::        
                    190: * Stack Manipulation Tutorial::  
                    191: * Using files for Forth code Tutorial::  
                    192: * Comments Tutorial::           
                    193: * Colon Definitions Tutorial::  
                    194: * Decompilation Tutorial::      
                    195: * Stack-Effect Comments Tutorial::  
                    196: * Types Tutorial::              
                    197: * Factoring Tutorial::          
                    198: * Designing the stack effect Tutorial::  
                    199: * Local Variables Tutorial::    
                    200: * Conditional execution Tutorial::  
                    201: * Flags and Comparisons Tutorial::  
                    202: * General Loops Tutorial::      
                    203: * Counted loops Tutorial::      
                    204: * Recursion Tutorial::          
                    205: * Leaving definitions or loops Tutorial::  
                    206: * Return Stack Tutorial::       
                    207: * Memory Tutorial::             
                    208: * Characters and Strings Tutorial::  
                    209: * Alignment Tutorial::          
1.87      anton     210: * Files Tutorial::              
1.48      anton     211: * Interpretation and Compilation Semantics and Immediacy Tutorial::  
                    212: * Execution Tokens Tutorial::   
                    213: * Exceptions Tutorial::         
                    214: * Defining Words Tutorial::     
                    215: * Arrays and Records Tutorial::  
                    216: * POSTPONE Tutorial::           
                    217: * Literal Tutorial::            
                    218: * Advanced macros Tutorial::    
                    219: * Compilation Tokens Tutorial::  
                    220: * Wordlists and Search Order Tutorial::  
1.29      crook     221: 
1.24      anton     222: An Introduction to ANS Forth
                    223: 
1.67      anton     224: * Introducing the Text Interpreter::  
                    225: * Stacks and Postfix notation::  
                    226: * Your first definition::       
                    227: * How does that work?::         
                    228: * Forth is written in Forth::   
                    229: * Review - elements of a Forth system::  
                    230: * Where to go next::            
                    231: * Exercises::                   
1.24      anton     232: 
1.12      anton     233: Forth Words
                    234: 
                    235: * Notation::                    
1.65      anton     236: * Case insensitivity::          
                    237: * Comments::                    
                    238: * Boolean Flags::               
1.12      anton     239: * Arithmetic::                  
                    240: * Stack Manipulation::          
                    241: * Memory::                      
                    242: * Control Structures::          
                    243: * Defining Words::              
1.65      anton     244: * Interpretation and Compilation Semantics::  
1.47      crook     245: * Tokens for Words::            
1.81      anton     246: * Compiling words::             
1.65      anton     247: * The Text Interpreter::        
                    248: * Word Lists::                  
                    249: * Environmental Queries::       
1.12      anton     250: * Files::                       
                    251: * Blocks::                      
                    252: * Other I/O::                   
1.78      anton     253: * Locals::                      
                    254: * Structures::                  
                    255: * Object-oriented Forth::       
1.12      anton     256: * Programming Tools::           
                    257: * Assembler and Code Words::    
                    258: * Threading Words::             
1.65      anton     259: * Passing Commands to the OS::  
                    260: * Keeping track of Time::       
                    261: * Miscellaneous Words::         
1.12      anton     262: 
                    263: Arithmetic
                    264: 
                    265: * Single precision::            
1.67      anton     266: * Double precision::            Double-cell integer arithmetic
1.12      anton     267: * Bitwise operations::          
1.67      anton     268: * Numeric comparison::          
1.32      anton     269: * Mixed precision::             Operations with single and double-cell integers
1.12      anton     270: * Floating Point::              
                    271: 
                    272: Stack Manipulation
                    273: 
                    274: * Data stack::                  
                    275: * Floating point stack::        
                    276: * Return stack::                
                    277: * Locals stack::                
                    278: * Stack pointer manipulation::  
                    279: 
                    280: Memory
                    281: 
1.32      anton     282: * Memory model::                
                    283: * Dictionary allocation::       
                    284: * Heap Allocation::             
                    285: * Memory Access::               
                    286: * Address arithmetic::          
                    287: * Memory Blocks::               
1.12      anton     288: 
                    289: Control Structures
                    290: 
1.41      anton     291: * Selection::                   IF ... ELSE ... ENDIF
                    292: * Simple Loops::                BEGIN ...
1.32      anton     293: * Counted Loops::               DO
1.67      anton     294: * Arbitrary control structures::  
                    295: * Calls and returns::           
1.12      anton     296: * Exception Handling::          
                    297: 
                    298: Defining Words
                    299: 
1.67      anton     300: * CREATE::                      
1.44      crook     301: * Variables::                   Variables and user variables
1.67      anton     302: * Constants::                   
1.44      crook     303: * Values::                      Initialised variables
1.67      anton     304: * Colon Definitions::           
1.44      crook     305: * Anonymous Definitions::       Definitions without names
1.71      anton     306: * Supplying names::             Passing definition names as strings
1.67      anton     307: * User-defined Defining Words::  
1.44      crook     308: * Deferred words::              Allow forward references
1.67      anton     309: * Aliases::                     
1.47      crook     310: 
1.63      anton     311: User-defined Defining Words
                    312: 
                    313: * CREATE..DOES> applications::  
                    314: * CREATE..DOES> details::       
                    315: * Advanced does> usage example::  
                    316: 
1.47      crook     317: Interpretation and Compilation Semantics
                    318: 
1.67      anton     319: * Combined words::              
1.12      anton     320: 
1.71      anton     321: Tokens for Words
                    322: 
                    323: * Execution token::             represents execution/interpretation semantics
                    324: * Compilation token::           represents compilation semantics
                    325: * Name token::                  represents named words
                    326: 
1.82      anton     327: Compiling words
                    328: 
                    329: * Literals::                    Compiling data values
                    330: * Macros::                      Compiling words
                    331: 
1.21      crook     332: The Text Interpreter
                    333: 
1.67      anton     334: * Input Sources::               
                    335: * Number Conversion::           
                    336: * Interpret/Compile states::    
                    337: * Interpreter Directives::      
1.21      crook     338: 
1.26      crook     339: Word Lists
                    340: 
1.75      anton     341: * Vocabularies::                
1.67      anton     342: * Why use word lists?::         
1.75      anton     343: * Word list example::           
1.26      crook     344: 
                    345: Files
                    346: 
1.48      anton     347: * Forth source files::          
                    348: * General files::               
                    349: * Search Paths::                
                    350: 
                    351: Search Paths
                    352: 
1.75      anton     353: * Source Search Paths::         
1.26      crook     354: * General Search Paths::        
                    355: 
                    356: Other I/O
                    357: 
1.32      anton     358: * Simple numeric output::       Predefined formats
                    359: * Formatted numeric output::    Formatted (pictured) output
                    360: * String Formats::              How Forth stores strings in memory
1.67      anton     361: * Displaying characters and strings::  Other stuff
1.32      anton     362: * Input::                       Input
1.26      crook     363: 
                    364: Locals
                    365: 
                    366: * Gforth locals::               
                    367: * ANS Forth locals::            
                    368: 
                    369: Gforth locals
                    370: 
                    371: * Where are locals visible by name?::  
                    372: * How long do locals live?::    
1.78      anton     373: * Locals programming style::    
                    374: * Locals implementation::       
1.26      crook     375: 
1.12      anton     376: Structures
                    377: 
                    378: * Why explicit structure support?::  
                    379: * Structure Usage::             
                    380: * Structure Naming Convention::  
                    381: * Structure Implementation::    
                    382: * Structure Glossary::          
                    383: 
                    384: Object-oriented Forth
                    385: 
1.48      anton     386: * Why object-oriented programming?::  
                    387: * Object-Oriented Terminology::  
                    388: * Objects::                     
                    389: * OOF::                         
                    390: * Mini-OOF::                    
1.23      crook     391: * Comparison with other object models::  
1.12      anton     392: 
1.24      anton     393: The @file{objects.fs} model
1.12      anton     394: 
                    395: * Properties of the Objects model::  
                    396: * Basic Objects Usage::         
1.41      anton     397: * The Objects base class::      
1.12      anton     398: * Creating objects::            
                    399: * Object-Oriented Programming Style::  
                    400: * Class Binding::               
                    401: * Method conveniences::         
                    402: * Classes and Scoping::         
1.41      anton     403: * Dividing classes::            
1.12      anton     404: * Object Interfaces::           
                    405: * Objects Implementation::      
                    406: * Objects Glossary::            
                    407: 
1.24      anton     408: The @file{oof.fs} model
1.12      anton     409: 
1.67      anton     410: * Properties of the OOF model::  
                    411: * Basic OOF Usage::             
                    412: * The OOF base class::          
                    413: * Class Declaration::           
                    414: * Class Implementation::        
1.12      anton     415: 
1.24      anton     416: The @file{mini-oof.fs} model
1.23      crook     417: 
1.48      anton     418: * Basic Mini-OOF Usage::        
                    419: * Mini-OOF Example::            
                    420: * Mini-OOF Implementation::     
1.23      crook     421: 
1.78      anton     422: Programming Tools
                    423: 
                    424: * Examining::                   
                    425: * Forgetting words::            
                    426: * Debugging::                   Simple and quick.
                    427: * Assertions::                  Making your programs self-checking.
                    428: * Singlestep Debugger::         Executing your program word by word.
                    429: 
                    430: Assembler and Code Words
                    431: 
                    432: * Code and ;code::              
                    433: * Common Assembler::            Assembler Syntax
                    434: * Common Disassembler::         
                    435: * 386 Assembler::               Deviations and special cases
                    436: * Alpha Assembler::             Deviations and special cases
                    437: * MIPS assembler::              Deviations and special cases
                    438: * Other assemblers::            How to write them
                    439: 
1.12      anton     440: Tools
                    441: 
                    442: * ANS Report::                  Report the words used, sorted by wordset.
                    443: 
                    444: ANS conformance
                    445: 
                    446: * The Core Words::              
                    447: * The optional Block word set::  
                    448: * The optional Double Number word set::  
                    449: * The optional Exception word set::  
                    450: * The optional Facility word set::  
                    451: * The optional File-Access word set::  
                    452: * The optional Floating-Point word set::  
                    453: * The optional Locals word set::  
                    454: * The optional Memory-Allocation word set::  
                    455: * The optional Programming-Tools word set::  
                    456: * The optional Search-Order word set::  
                    457: 
                    458: The Core Words
                    459: 
                    460: * core-idef::                   Implementation Defined Options                   
                    461: * core-ambcond::                Ambiguous Conditions                
                    462: * core-other::                  Other System Documentation                  
                    463: 
                    464: The optional Block word set
                    465: 
                    466: * block-idef::                  Implementation Defined Options
                    467: * block-ambcond::               Ambiguous Conditions               
                    468: * block-other::                 Other System Documentation                 
                    469: 
                    470: The optional Double Number word set
                    471: 
                    472: * double-ambcond::              Ambiguous Conditions              
                    473: 
                    474: The optional Exception word set
                    475: 
                    476: * exception-idef::              Implementation Defined Options              
                    477: 
                    478: The optional Facility word set
                    479: 
                    480: * facility-idef::               Implementation Defined Options               
                    481: * facility-ambcond::            Ambiguous Conditions            
                    482: 
                    483: The optional File-Access word set
                    484: 
                    485: * file-idef::                   Implementation Defined Options
                    486: * file-ambcond::                Ambiguous Conditions                
                    487: 
                    488: The optional Floating-Point word set
                    489: 
                    490: * floating-idef::               Implementation Defined Options
                    491: * floating-ambcond::            Ambiguous Conditions            
                    492: 
                    493: The optional Locals word set
                    494: 
                    495: * locals-idef::                 Implementation Defined Options                 
                    496: * locals-ambcond::              Ambiguous Conditions              
                    497: 
                    498: The optional Memory-Allocation word set
                    499: 
                    500: * memory-idef::                 Implementation Defined Options                 
                    501: 
                    502: The optional Programming-Tools word set
                    503: 
                    504: * programming-idef::            Implementation Defined Options            
                    505: * programming-ambcond::         Ambiguous Conditions         
                    506: 
                    507: The optional Search-Order word set
                    508: 
                    509: * search-idef::                 Implementation Defined Options                 
                    510: * search-ambcond::              Ambiguous Conditions              
                    511: 
                    512: Image Files
                    513: 
1.24      anton     514: * Image Licensing Issues::      Distribution terms for images.
                    515: * Image File Background::       Why have image files?
1.67      anton     516: * Non-Relocatable Image Files::  don't always work.
1.24      anton     517: * Data-Relocatable Image Files::  are better.
1.67      anton     518: * Fully Relocatable Image Files::  better yet.
1.24      anton     519: * Stack and Dictionary Sizes::  Setting the default sizes for an image.
1.32      anton     520: * Running Image Files::         @code{gforth -i @i{file}} or @i{file}.
1.24      anton     521: * Modifying the Startup Sequence::  and turnkey applications.
1.12      anton     522: 
                    523: Fully Relocatable Image Files
                    524: 
1.27      crook     525: * gforthmi::                    The normal way
1.12      anton     526: * cross.fs::                    The hard way
                    527: 
                    528: Engine
                    529: 
                    530: * Portability::                 
                    531: * Threading::                   
                    532: * Primitives::                  
                    533: * Performance::                 
                    534: 
                    535: Threading
                    536: 
                    537: * Scheduling::                  
                    538: * Direct or Indirect Threaded?::  
                    539: * DOES>::                       
                    540: 
                    541: Primitives
                    542: 
                    543: * Automatic Generation::        
                    544: * TOS Optimization::            
                    545: * Produced code::               
1.13      pazsan    546: 
                    547: Cross Compiler
                    548: 
1.67      anton     549: * Using the Cross Compiler::    
                    550: * How the Cross Compiler Works::  
1.13      pazsan    551: 
1.24      anton     552: @end detailmenu
1.1       anton     553: @end menu
                    554: 
1.26      crook     555: @node License, Goals, Top, Top
1.1       anton     556: @unnumbered GNU GENERAL PUBLIC LICENSE
                    557: @center Version 2, June 1991
                    558: 
                    559: @display
                    560: Copyright @copyright{} 1989, 1991 Free Software Foundation, Inc.
1.88      anton     561: 59 Temple Place, Suite 330, Boston, MA 02111, USA
1.1       anton     562: 
                    563: Everyone is permitted to copy and distribute verbatim copies
                    564: of this license document, but changing it is not allowed.
                    565: @end display
                    566: 
                    567: @unnumberedsec Preamble
                    568: 
                    569:   The licenses for most software are designed to take away your
                    570: freedom to share and change it.  By contrast, the GNU General Public
                    571: License is intended to guarantee your freedom to share and change free
                    572: software---to make sure the software is free for all its users.  This
                    573: General Public License applies to most of the Free Software
                    574: Foundation's software and to any other program whose authors commit to
                    575: using it.  (Some other Free Software Foundation software is covered by
                    576: the GNU Library General Public License instead.)  You can apply it to
                    577: your programs, too.
                    578: 
                    579:   When we speak of free software, we are referring to freedom, not
                    580: price.  Our General Public Licenses are designed to make sure that you
                    581: have the freedom to distribute copies of free software (and charge for
                    582: this service if you wish), that you receive source code or can get it
                    583: if you want it, that you can change the software or use pieces of it
                    584: in new free programs; and that you know you can do these things.
                    585: 
                    586:   To protect your rights, we need to make restrictions that forbid
                    587: anyone to deny you these rights or to ask you to surrender the rights.
                    588: These restrictions translate to certain responsibilities for you if you
                    589: distribute copies of the software, or if you modify it.
                    590: 
                    591:   For example, if you distribute copies of such a program, whether
                    592: gratis or for a fee, you must give the recipients all the rights that
                    593: you have.  You must make sure that they, too, receive or can get the
                    594: source code.  And you must show them these terms so they know their
                    595: rights.
                    596: 
                    597:   We protect your rights with two steps: (1) copyright the software, and
                    598: (2) offer you this license which gives you legal permission to copy,
                    599: distribute and/or modify the software.
                    600: 
                    601:   Also, for each author's protection and ours, we want to make certain
                    602: that everyone understands that there is no warranty for this free
                    603: software.  If the software is modified by someone else and passed on, we
                    604: want its recipients to know that what they have is not the original, so
                    605: that any problems introduced by others will not reflect on the original
                    606: authors' reputations.
                    607: 
                    608:   Finally, any free program is threatened constantly by software
                    609: patents.  We wish to avoid the danger that redistributors of a free
                    610: program will individually obtain patent licenses, in effect making the
                    611: program proprietary.  To prevent this, we have made it clear that any
                    612: patent must be licensed for everyone's free use or not licensed at all.
                    613: 
                    614:   The precise terms and conditions for copying, distribution and
                    615: modification follow.
                    616: 
                    617: @iftex
                    618: @unnumberedsec TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
                    619: @end iftex
1.49      anton     620: @ifnottex
1.1       anton     621: @center TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
1.49      anton     622: @end ifnottex
1.1       anton     623: 
                    624: @enumerate 0
                    625: @item
                    626: This License applies to any program or other work which contains
                    627: a notice placed by the copyright holder saying it may be distributed
                    628: under the terms of this General Public License.  The ``Program'', below,
                    629: refers to any such program or work, and a ``work based on the Program''
                    630: means either the Program or any derivative work under copyright law:
                    631: that is to say, a work containing the Program or a portion of it,
                    632: either verbatim or with modifications and/or translated into another
                    633: language.  (Hereinafter, translation is included without limitation in
                    634: the term ``modification''.)  Each licensee is addressed as ``you''.
                    635: 
                    636: Activities other than copying, distribution and modification are not
                    637: covered by this License; they are outside its scope.  The act of
                    638: running the Program is not restricted, and the output from the Program
                    639: is covered only if its contents constitute a work based on the
                    640: Program (independent of having been made by running the Program).
                    641: Whether that is true depends on what the Program does.
                    642: 
                    643: @item
                    644: You may copy and distribute verbatim copies of the Program's
                    645: source code as you receive it, in any medium, provided that you
                    646: conspicuously and appropriately publish on each copy an appropriate
                    647: copyright notice and disclaimer of warranty; keep intact all the
                    648: notices that refer to this License and to the absence of any warranty;
                    649: and give any other recipients of the Program a copy of this License
                    650: along with the Program.
                    651: 
                    652: You may charge a fee for the physical act of transferring a copy, and
                    653: you may at your option offer warranty protection in exchange for a fee.
                    654: 
                    655: @item
                    656: You may modify your copy or copies of the Program or any portion
                    657: of it, thus forming a work based on the Program, and copy and
                    658: distribute such modifications or work under the terms of Section 1
                    659: above, provided that you also meet all of these conditions:
                    660: 
                    661: @enumerate a
                    662: @item
                    663: You must cause the modified files to carry prominent notices
                    664: stating that you changed the files and the date of any change.
                    665: 
                    666: @item
                    667: You must cause any work that you distribute or publish, that in
                    668: whole or in part contains or is derived from the Program or any
                    669: part thereof, to be licensed as a whole at no charge to all third
                    670: parties under the terms of this License.
                    671: 
                    672: @item
                    673: If the modified program normally reads commands interactively
                    674: when run, you must cause it, when started running for such
                    675: interactive use in the most ordinary way, to print or display an
                    676: announcement including an appropriate copyright notice and a
                    677: notice that there is no warranty (or else, saying that you provide
                    678: a warranty) and that users may redistribute the program under
                    679: these conditions, and telling the user how to view a copy of this
                    680: License.  (Exception: if the Program itself is interactive but
                    681: does not normally print such an announcement, your work based on
                    682: the Program is not required to print an announcement.)
                    683: @end enumerate
                    684: 
                    685: These requirements apply to the modified work as a whole.  If
                    686: identifiable sections of that work are not derived from the Program,
                    687: and can be reasonably considered independent and separate works in
                    688: themselves, then this License, and its terms, do not apply to those
                    689: sections when you distribute them as separate works.  But when you
                    690: distribute the same sections as part of a whole which is a work based
                    691: on the Program, the distribution of the whole must be on the terms of
                    692: this License, whose permissions for other licensees extend to the
                    693: entire whole, and thus to each and every part regardless of who wrote it.
                    694: 
                    695: Thus, it is not the intent of this section to claim rights or contest
                    696: your rights to work written entirely by you; rather, the intent is to
                    697: exercise the right to control the distribution of derivative or
                    698: collective works based on the Program.
                    699: 
                    700: In addition, mere aggregation of another work not based on the Program
                    701: with the Program (or with a work based on the Program) on a volume of
                    702: a storage or distribution medium does not bring the other work under
                    703: the scope of this License.
                    704: 
                    705: @item
                    706: You may copy and distribute the Program (or a work based on it,
                    707: under Section 2) in object code or executable form under the terms of
                    708: Sections 1 and 2 above provided that you also do one of the following:
                    709: 
                    710: @enumerate a
                    711: @item
                    712: Accompany it with the complete corresponding machine-readable
                    713: source code, which must be distributed under the terms of Sections
                    714: 1 and 2 above on a medium customarily used for software interchange; or,
                    715: 
                    716: @item
                    717: Accompany it with a written offer, valid for at least three
                    718: years, to give any third party, for a charge no more than your
                    719: cost of physically performing source distribution, a complete
                    720: machine-readable copy of the corresponding source code, to be
                    721: distributed under the terms of Sections 1 and 2 above on a medium
                    722: customarily used for software interchange; or,
                    723: 
                    724: @item
                    725: Accompany it with the information you received as to the offer
                    726: to distribute corresponding source code.  (This alternative is
                    727: allowed only for noncommercial distribution and only if you
                    728: received the program in object code or executable form with such
                    729: an offer, in accord with Subsection b above.)
                    730: @end enumerate
                    731: 
                    732: The source code for a work means the preferred form of the work for
                    733: making modifications to it.  For an executable work, complete source
                    734: code means all the source code for all modules it contains, plus any
                    735: associated interface definition files, plus the scripts used to
                    736: control compilation and installation of the executable.  However, as a
                    737: special exception, the source code distributed need not include
                    738: anything that is normally distributed (in either source or binary
                    739: form) with the major components (compiler, kernel, and so on) of the
                    740: operating system on which the executable runs, unless that component
                    741: itself accompanies the executable.
                    742: 
                    743: If distribution of executable or object code is made by offering
                    744: access to copy from a designated place, then offering equivalent
                    745: access to copy the source code from the same place counts as
                    746: distribution of the source code, even though third parties are not
                    747: compelled to copy the source along with the object code.
                    748: 
                    749: @item
                    750: You may not copy, modify, sublicense, or distribute the Program
                    751: except as expressly provided under this License.  Any attempt
                    752: otherwise to copy, modify, sublicense or distribute the Program is
                    753: void, and will automatically terminate your rights under this License.
                    754: However, parties who have received copies, or rights, from you under
                    755: this License will not have their licenses terminated so long as such
                    756: parties remain in full compliance.
                    757: 
                    758: @item
                    759: You are not required to accept this License, since you have not
                    760: signed it.  However, nothing else grants you permission to modify or
                    761: distribute the Program or its derivative works.  These actions are
                    762: prohibited by law if you do not accept this License.  Therefore, by
                    763: modifying or distributing the Program (or any work based on the
                    764: Program), you indicate your acceptance of this License to do so, and
                    765: all its terms and conditions for copying, distributing or modifying
                    766: the Program or works based on it.
                    767: 
                    768: @item
                    769: Each time you redistribute the Program (or any work based on the
                    770: Program), the recipient automatically receives a license from the
                    771: original licensor to copy, distribute or modify the Program subject to
                    772: these terms and conditions.  You may not impose any further
                    773: restrictions on the recipients' exercise of the rights granted herein.
                    774: You are not responsible for enforcing compliance by third parties to
                    775: this License.
                    776: 
                    777: @item
                    778: If, as a consequence of a court judgment or allegation of patent
                    779: infringement or for any other reason (not limited to patent issues),
                    780: conditions are imposed on you (whether by court order, agreement or
                    781: otherwise) that contradict the conditions of this License, they do not
                    782: excuse you from the conditions of this License.  If you cannot
                    783: distribute so as to satisfy simultaneously your obligations under this
                    784: License and any other pertinent obligations, then as a consequence you
                    785: may not distribute the Program at all.  For example, if a patent
                    786: license would not permit royalty-free redistribution of the Program by
                    787: all those who receive copies directly or indirectly through you, then
                    788: the only way you could satisfy both it and this License would be to
                    789: refrain entirely from distribution of the Program.
                    790: 
                    791: If any portion of this section is held invalid or unenforceable under
                    792: any particular circumstance, the balance of the section is intended to
                    793: apply and the section as a whole is intended to apply in other
                    794: circumstances.
                    795: 
                    796: It is not the purpose of this section to induce you to infringe any
                    797: patents or other property right claims or to contest validity of any
                    798: such claims; this section has the sole purpose of protecting the
                    799: integrity of the free software distribution system, which is
                    800: implemented by public license practices.  Many people have made
                    801: generous contributions to the wide range of software distributed
                    802: through that system in reliance on consistent application of that
                    803: system; it is up to the author/donor to decide if he or she is willing
                    804: to distribute software through any other system and a licensee cannot
                    805: impose that choice.
                    806: 
                    807: This section is intended to make thoroughly clear what is believed to
                    808: be a consequence of the rest of this License.
                    809: 
                    810: @item
                    811: If the distribution and/or use of the Program is restricted in
                    812: certain countries either by patents or by copyrighted interfaces, the
                    813: original copyright holder who places the Program under this License
                    814: may add an explicit geographical distribution limitation excluding
                    815: those countries, so that distribution is permitted only in or among
                    816: countries not thus excluded.  In such case, this License incorporates
                    817: the limitation as if written in the body of this License.
                    818: 
                    819: @item
                    820: The Free Software Foundation may publish revised and/or new versions
                    821: of the General Public License from time to time.  Such new versions will
                    822: be similar in spirit to the present version, but may differ in detail to
                    823: address new problems or concerns.
                    824: 
                    825: Each version is given a distinguishing version number.  If the Program
                    826: specifies a version number of this License which applies to it and ``any
                    827: later version'', you have the option of following the terms and conditions
                    828: either of that version or of any later version published by the Free
                    829: Software Foundation.  If the Program does not specify a version number of
                    830: this License, you may choose any version ever published by the Free Software
                    831: Foundation.
                    832: 
                    833: @item
                    834: If you wish to incorporate parts of the Program into other free
                    835: programs whose distribution conditions are different, write to the author
                    836: to ask for permission.  For software which is copyrighted by the Free
                    837: Software Foundation, write to the Free Software Foundation; we sometimes
                    838: make exceptions for this.  Our decision will be guided by the two goals
                    839: of preserving the free status of all derivatives of our free software and
                    840: of promoting the sharing and reuse of software generally.
                    841: 
                    842: @iftex
                    843: @heading NO WARRANTY
                    844: @end iftex
1.49      anton     845: @ifnottex
1.1       anton     846: @center NO WARRANTY
1.49      anton     847: @end ifnottex
1.1       anton     848: 
                    849: @item
                    850: BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
                    851: FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.  EXCEPT WHEN
                    852: OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
                    853: PROVIDE THE PROGRAM ``AS IS'' WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
                    854: OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
                    855: MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.  THE ENTIRE RISK AS
                    856: TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.  SHOULD THE
                    857: PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
                    858: REPAIR OR CORRECTION.
                    859: 
                    860: @item
                    861: IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
                    862: WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
                    863: REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
                    864: INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
                    865: OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED
                    866: TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY
                    867: YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
                    868: PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
                    869: POSSIBILITY OF SUCH DAMAGES.
                    870: @end enumerate
                    871: 
                    872: @iftex
                    873: @heading END OF TERMS AND CONDITIONS
                    874: @end iftex
1.49      anton     875: @ifnottex
1.1       anton     876: @center END OF TERMS AND CONDITIONS
1.49      anton     877: @end ifnottex
1.1       anton     878: 
                    879: @page
                    880: @unnumberedsec How to Apply These Terms to Your New Programs
                    881: 
                    882:   If you develop a new program, and you want it to be of the greatest
                    883: possible use to the public, the best way to achieve this is to make it
                    884: free software which everyone can redistribute and change under these terms.
                    885: 
                    886:   To do so, attach the following notices to the program.  It is safest
                    887: to attach them to the start of each source file to most effectively
                    888: convey the exclusion of warranty; and each file should have at least
                    889: the ``copyright'' line and a pointer to where the full notice is found.
                    890: 
                    891: @smallexample
                    892: @var{one line to give the program's name and a brief idea of what it does.}
                    893: Copyright (C) 19@var{yy}  @var{name of author}
                    894: 
                    895: This program is free software; you can redistribute it and/or modify 
                    896: it under the terms of the GNU General Public License as published by 
                    897: the Free Software Foundation; either version 2 of the License, or 
                    898: (at your option) any later version.
                    899: 
                    900: This program is distributed in the hope that it will be useful,
                    901: but WITHOUT ANY WARRANTY; without even the implied warranty of
                    902: MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
                    903: GNU General Public License for more details.
                    904: 
                    905: You should have received a copy of the GNU General Public License
                    906: along with this program; if not, write to the Free Software
1.88      anton     907: Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111, USA.
1.1       anton     908: @end smallexample
                    909: 
                    910: Also add information on how to contact you by electronic and paper mail.
                    911: 
                    912: If the program is interactive, make it output a short notice like this
                    913: when it starts in an interactive mode:
                    914: 
                    915: @smallexample
                    916: Gnomovision version 69, Copyright (C) 19@var{yy} @var{name of author}
                    917: Gnomovision comes with ABSOLUTELY NO WARRANTY; for details
                    918: type `show w'.  
                    919: This is free software, and you are welcome to redistribute it 
                    920: under certain conditions; type `show c' for details.
                    921: @end smallexample
                    922: 
                    923: The hypothetical commands @samp{show w} and @samp{show c} should show
                    924: the appropriate parts of the General Public License.  Of course, the
                    925: commands you use may be called something other than @samp{show w} and
                    926: @samp{show c}; they could even be mouse-clicks or menu items---whatever
                    927: suits your program.
                    928: 
                    929: You should also get your employer (if you work as a programmer) or your
                    930: school, if any, to sign a ``copyright disclaimer'' for the program, if
                    931: necessary.  Here is a sample; alter the names:
                    932: 
                    933: @smallexample
                    934: Yoyodyne, Inc., hereby disclaims all copyright interest in the program
                    935: `Gnomovision' (which makes passes at compilers) written by James Hacker.
                    936: 
                    937: @var{signature of Ty Coon}, 1 April 1989
                    938: Ty Coon, President of Vice
                    939: @end smallexample
                    940: 
                    941: This General Public License does not permit incorporating your program into
                    942: proprietary programs.  If your program is a subroutine library, you may
                    943: consider it more useful to permit linking proprietary applications with the
                    944: library.  If this is what you want to do, use the GNU Library General
                    945: Public License instead of this License.
                    946: 
                    947: @iftex
                    948: @unnumbered Preface
                    949: @cindex Preface
1.21      crook     950: This manual documents Gforth. Some introductory material is provided for
                    951: readers who are unfamiliar with Forth or who are migrating to Gforth
                    952: from other Forth compilers. However, this manual is primarily a
                    953: reference manual.
1.1       anton     954: @end iftex
                    955: 
1.28      crook     956: @comment TODO much more blurb here.
1.26      crook     957: 
                    958: @c ******************************************************************
1.29      crook     959: @node Goals, Gforth Environment, License, Top
1.26      crook     960: @comment node-name,     next,           previous, up
                    961: @chapter Goals of Gforth
                    962: @cindex goals of the Gforth project
                    963: The goal of the Gforth Project is to develop a standard model for
                    964: ANS Forth. This can be split into several subgoals:
                    965: 
                    966: @itemize @bullet
                    967: @item
                    968: Gforth should conform to the ANS Forth Standard.
                    969: @item
                    970: It should be a model, i.e. it should define all the
                    971: implementation-dependent things.
                    972: @item
                    973: It should become standard, i.e. widely accepted and used. This goal
                    974: is the most difficult one.
                    975: @end itemize
                    976: 
                    977: To achieve these goals Gforth should be
                    978: @itemize @bullet
                    979: @item
                    980: Similar to previous models (fig-Forth, F83)
                    981: @item
                    982: Powerful. It should provide for all the things that are considered
                    983: necessary today and even some that are not yet considered necessary.
                    984: @item
                    985: Efficient. It should not get the reputation of being exceptionally
                    986: slow.
                    987: @item
                    988: Free.
                    989: @item
                    990: Available on many machines/easy to port.
                    991: @end itemize
                    992: 
                    993: Have we achieved these goals? Gforth conforms to the ANS Forth
                    994: standard. It may be considered a model, but we have not yet documented
                    995: which parts of the model are stable and which parts we are likely to
                    996: change. It certainly has not yet become a de facto standard, but it
                    997: appears to be quite popular. It has some similarities to and some
                    998: differences from previous models. It has some powerful features, but not
                    999: yet everything that we envisioned. We certainly have achieved our
1.65      anton    1000: execution speed goals (@pxref{Performance})@footnote{However, in 1998
                   1001: the bar was raised when the major commercial Forth vendors switched to
                   1002: native code compilers.}.  It is free and available on many machines.
1.29      crook    1003: 
1.26      crook    1004: @c ******************************************************************
1.48      anton    1005: @node Gforth Environment, Tutorial, Goals, Top
1.29      crook    1006: @chapter Gforth Environment
                   1007: @cindex Gforth environment
1.21      crook    1008: 
1.45      crook    1009: Note: ultimately, the Gforth man page will be auto-generated from the
1.29      crook    1010: material in this chapter.
1.21      crook    1011: 
                   1012: @menu
1.29      crook    1013: * Invoking Gforth::             Getting in
                   1014: * Leaving Gforth::              Getting out
                   1015: * Command-line editing::        
1.48      anton    1016: * Environment variables::       that affect how Gforth starts up
1.29      crook    1017: * Gforth Files::                What gets installed and where
1.48      anton    1018: * Startup speed::               When 35ms is not fast enough ...
1.21      crook    1019: @end menu
                   1020: 
1.49      anton    1021: For related information about the creation of images see @ref{Image Files}.
1.29      crook    1022: 
1.21      crook    1023: @comment ----------------------------------------------
1.48      anton    1024: @node Invoking Gforth, Leaving Gforth, Gforth Environment, Gforth Environment
1.29      crook    1025: @section Invoking Gforth
                   1026: @cindex invoking Gforth
                   1027: @cindex running Gforth
                   1028: @cindex command-line options
                   1029: @cindex options on the command line
                   1030: @cindex flags on the command line
1.21      crook    1031: 
1.30      anton    1032: Gforth is made up of two parts; an executable ``engine'' (named
                   1033: @file{gforth} or @file{gforth-fast}) and an image file. To start it, you
                   1034: will usually just say @code{gforth} -- this automatically loads the
                   1035: default image file @file{gforth.fi}. In many other cases the default
                   1036: Gforth image will be invoked like this:
1.21      crook    1037: @example
1.30      anton    1038: gforth [file | -e forth-code] ...
1.21      crook    1039: @end example
1.29      crook    1040: @noindent
                   1041: This interprets the contents of the files and the Forth code in the order they
                   1042: are given.
1.21      crook    1043: 
1.30      anton    1044: In addition to the @file{gforth} engine, there is also an engine called
                   1045: @file{gforth-fast}, which is faster, but gives less informative error
1.89    ! anton    1046: messages (@pxref{Error messages}).  You should use it for debugged,
        !          1047: performance-critical programs.
1.30      anton    1048: 
1.29      crook    1049: In general, the command line looks like this:
1.21      crook    1050: 
                   1051: @example
1.30      anton    1052: gforth[-fast] [engine options] [image options]
1.21      crook    1053: @end example
                   1054: 
1.30      anton    1055: The engine options must come before the rest of the command
1.29      crook    1056: line. They are:
1.26      crook    1057: 
1.29      crook    1058: @table @code
                   1059: @cindex -i, command-line option
                   1060: @cindex --image-file, command-line option
                   1061: @item --image-file @i{file}
                   1062: @itemx -i @i{file}
                   1063: Loads the Forth image @i{file} instead of the default
                   1064: @file{gforth.fi} (@pxref{Image Files}).
1.21      crook    1065: 
1.39      anton    1066: @cindex --appl-image, command-line option
                   1067: @item --appl-image @i{file}
                   1068: Loads the image @i{file} and leaves all further command-line arguments
1.65      anton    1069: to the image (instead of processing them as engine options).  This is
                   1070: useful for building executable application images on Unix, built with
1.39      anton    1071: @code{gforthmi --application ...}.
                   1072: 
1.29      crook    1073: @cindex --path, command-line option
                   1074: @cindex -p, command-line option
                   1075: @item --path @i{path}
                   1076: @itemx -p @i{path}
                   1077: Uses @i{path} for searching the image file and Forth source code files
                   1078: instead of the default in the environment variable @code{GFORTHPATH} or
                   1079: the path specified at installation time (e.g.,
                   1080: @file{/usr/local/share/gforth/0.2.0:.}). A path is given as a list of
                   1081: directories, separated by @samp{:} (on Unix) or @samp{;} (on other OSs).
1.21      crook    1082: 
1.29      crook    1083: @cindex --dictionary-size, command-line option
                   1084: @cindex -m, command-line option
                   1085: @cindex @i{size} parameters for command-line options
                   1086: @cindex size of the dictionary and the stacks
                   1087: @item --dictionary-size @i{size}
                   1088: @itemx -m @i{size}
                   1089: Allocate @i{size} space for the Forth dictionary space instead of
                   1090: using the default specified in the image (typically 256K). The
                   1091: @i{size} specification for this and subsequent options consists of
                   1092: an integer and a unit (e.g.,
                   1093: @code{4M}). The unit can be one of @code{b} (bytes), @code{e} (element
                   1094: size, in this case Cells), @code{k} (kilobytes), @code{M} (Megabytes),
                   1095: @code{G} (Gigabytes), and @code{T} (Terabytes). If no unit is specified,
                   1096: @code{e} is used.
1.21      crook    1097: 
1.29      crook    1098: @cindex --data-stack-size, command-line option
                   1099: @cindex -d, command-line option
                   1100: @item --data-stack-size @i{size}
                   1101: @itemx -d @i{size}
                   1102: Allocate @i{size} space for the data stack instead of using the
                   1103: default specified in the image (typically 16K).
1.21      crook    1104: 
1.29      crook    1105: @cindex --return-stack-size, command-line option
                   1106: @cindex -r, command-line option
                   1107: @item --return-stack-size @i{size}
                   1108: @itemx -r @i{size}
                   1109: Allocate @i{size} space for the return stack instead of using the
                   1110: default specified in the image (typically 15K).
1.21      crook    1111: 
1.29      crook    1112: @cindex --fp-stack-size, command-line option
                   1113: @cindex -f, command-line option
                   1114: @item --fp-stack-size @i{size}
                   1115: @itemx -f @i{size}
                   1116: Allocate @i{size} space for the floating point stack instead of
                   1117: using the default specified in the image (typically 15.5K). In this case
                   1118: the unit specifier @code{e} refers to floating point numbers.
1.21      crook    1119: 
1.48      anton    1120: @cindex --locals-stack-size, command-line option
                   1121: @cindex -l, command-line option
                   1122: @item --locals-stack-size @i{size}
                   1123: @itemx -l @i{size}
                   1124: Allocate @i{size} space for the locals stack instead of using the
                   1125: default specified in the image (typically 14.5K).
                   1126: 
                   1127: @cindex -h, command-line option
                   1128: @cindex --help, command-line option
                   1129: @item --help
                   1130: @itemx -h
                   1131: Print a message about the command-line options
                   1132: 
                   1133: @cindex -v, command-line option
                   1134: @cindex --version, command-line option
                   1135: @item --version
                   1136: @itemx -v
                   1137: Print version and exit
                   1138: 
                   1139: @cindex --debug, command-line option
                   1140: @item --debug
                   1141: Print some information useful for debugging on startup.
                   1142: 
                   1143: @cindex --offset-image, command-line option
                   1144: @item --offset-image
                   1145: Start the dictionary at a slightly different position than would be used
                   1146: otherwise (useful for creating data-relocatable images,
                   1147: @pxref{Data-Relocatable Image Files}).
                   1148: 
                   1149: @cindex --no-offset-im, command-line option
                   1150: @item --no-offset-im
                   1151: Start the dictionary at the normal position.
                   1152: 
                   1153: @cindex --clear-dictionary, command-line option
                   1154: @item --clear-dictionary
                   1155: Initialize all bytes in the dictionary to 0 before loading the image
                   1156: (@pxref{Data-Relocatable Image Files}).
                   1157: 
                   1158: @cindex --die-on-signal, command-line-option
                   1159: @item --die-on-signal
                   1160: Normally Gforth handles most signals (e.g., the user interrupt SIGINT,
                   1161: or the segmentation violation SIGSEGV) by translating it into a Forth
                   1162: @code{THROW}. With this option, Gforth exits if it receives such a
                   1163: signal. This option is useful when the engine and/or the image might be
                   1164: severely broken (such that it causes another signal before recovering
                   1165: from the first); this option avoids endless loops in such cases.
                   1166: @end table
                   1167: 
                   1168: @cindex loading files at startup
                   1169: @cindex executing code on startup
                   1170: @cindex batch processing with Gforth
                   1171: As explained above, the image-specific command-line arguments for the
                   1172: default image @file{gforth.fi} consist of a sequence of filenames and
                   1173: @code{-e @var{forth-code}} options that are interpreted in the sequence
                   1174: in which they are given. The @code{-e @var{forth-code}} or
                   1175: @code{--evaluate @var{forth-code}} option evaluates the Forth
                   1176: code. This option takes only one argument; if you want to evaluate more
                   1177: Forth words, you have to quote them or use @code{-e} several times. To exit
                   1178: after processing the command line (instead of entering interactive mode)
                   1179: append @code{-e bye} to the command line.
                   1180: 
                   1181: @cindex versions, invoking other versions of Gforth
                   1182: If you have several versions of Gforth installed, @code{gforth} will
                   1183: invoke the version that was installed last. @code{gforth-@i{version}}
                   1184: invokes a specific version. If your environment contains the variable
                   1185: @code{GFORTHPATH}, you may want to override it by using the
                   1186: @code{--path} option.
                   1187: 
                   1188: Not yet implemented:
                   1189: On startup the system first executes the system initialization file
                   1190: (unless the option @code{--no-init-file} is given; note that the system
                   1191: resulting from using this option may not be ANS Forth conformant). Then
                   1192: the user initialization file @file{.gforth.fs} is executed, unless the
1.62      crook    1193: option @code{--no-rc} is given; this file is searched for in @file{.},
1.48      anton    1194: then in @file{~}, then in the normal path (see above).
                   1195: 
                   1196: 
                   1197: 
                   1198: @comment ----------------------------------------------
                   1199: @node Leaving Gforth, Command-line editing, Invoking Gforth, Gforth Environment
                   1200: @section Leaving Gforth
                   1201: @cindex Gforth - leaving
                   1202: @cindex leaving Gforth
                   1203: 
                   1204: You can leave Gforth by typing @code{bye} or @kbd{Ctrl-d} (at the start
                   1205: of a line) or (if you invoked Gforth with the @code{--die-on-signal}
                   1206: option) @kbd{Ctrl-c}. When you leave Gforth, all of your definitions and
1.49      anton    1207: data are discarded.  For ways of saving the state of the system before
                   1208: leaving Gforth see @ref{Image Files}.
1.48      anton    1209: 
                   1210: doc-bye
                   1211: 
                   1212: 
                   1213: @comment ----------------------------------------------
1.65      anton    1214: @node Command-line editing, Environment variables, Leaving Gforth, Gforth Environment
1.48      anton    1215: @section Command-line editing
                   1216: @cindex command-line editing
                   1217: 
                   1218: Gforth maintains a history file that records every line that you type to
                   1219: the text interpreter. This file is preserved between sessions, and is
                   1220: used to provide a command-line recall facility; if you type @kbd{Ctrl-P}
                   1221: repeatedly you can recall successively older commands from this (or
                   1222: previous) session(s). The full list of command-line editing facilities is:
                   1223: 
                   1224: @itemize @bullet
                   1225: @item
                   1226: @kbd{Ctrl-p} (``previous'') (or up-arrow) to recall successively older
                   1227: commands from the history buffer.
                   1228: @item
                   1229: @kbd{Ctrl-n} (``next'') (or down-arrow) to recall successively newer commands
                   1230: from the history buffer.
                   1231: @item
                   1232: @kbd{Ctrl-f} (or right-arrow) to move the cursor right, non-destructively.
                   1233: @item
                   1234: @kbd{Ctrl-b} (or left-arrow) to move the cursor left, non-destructively.
                   1235: @item
                   1236: @kbd{Ctrl-h} (backspace) to delete the character to the left of the cursor,
                   1237: closing up the line.
                   1238: @item
                   1239: @kbd{Ctrl-k} to delete (``kill'') from the cursor to the end of the line.
                   1240: @item
                   1241: @kbd{Ctrl-a} to move the cursor to the start of the line.
                   1242: @item
                   1243: @kbd{Ctrl-e} to move the cursor to the end of the line.
                   1244: @item
                   1245: @key{RET} (@kbd{Ctrl-m}) or @key{LFD} (@kbd{Ctrl-j}) to submit the current
                   1246: line.
                   1247: @item
                   1248: @key{TAB} to step through all possible full-word completions of the word
                   1249: currently being typed.
                   1250: @item
1.65      anton    1251: @kbd{Ctrl-d} on an empty line line to terminate Gforth (gracefully,
                   1252: using @code{bye}). 
                   1253: @item
                   1254: @kbd{Ctrl-x} (or @code{Ctrl-d} on a non-empty line) to delete the
                   1255: character under the cursor.
1.48      anton    1256: @end itemize
                   1257: 
                   1258: When editing, displayable characters are inserted to the left of the
                   1259: cursor position; the line is always in ``insert'' (as opposed to
                   1260: ``overstrike'') mode.
                   1261: 
                   1262: @cindex history file
                   1263: @cindex @file{.gforth-history}
                   1264: On Unix systems, the history file is @file{~/.gforth-history} by
                   1265: default@footnote{i.e. it is stored in the user's home directory.}. You
                   1266: can find out the name and location of your history file using:
                   1267: 
                   1268: @example 
                   1269: history-file type \ Unix-class systems
                   1270: 
                   1271: history-file type \ Other systems
                   1272: history-dir  type
                   1273: @end example
                   1274: 
                   1275: If you enter long definitions by hand, you can use a text editor to
                   1276: paste them out of the history file into a Forth source file for reuse at
                   1277: a later time.
                   1278: 
                   1279: Gforth never trims the size of the history file, so you should do this
                   1280: periodically, if necessary.
                   1281: 
                   1282: @comment this is all defined in history.fs
                   1283: @comment NAC TODO the ctrl-D behaviour can either do a bye or a beep.. how is that option
                   1284: @comment chosen?
                   1285: 
                   1286: 
                   1287: @comment ----------------------------------------------
1.65      anton    1288: @node Environment variables, Gforth Files, Command-line editing, Gforth Environment
1.48      anton    1289: @section Environment variables
                   1290: @cindex environment variables
                   1291: 
                   1292: Gforth uses these environment variables:
                   1293: 
                   1294: @itemize @bullet
                   1295: @item
                   1296: @cindex @code{GFORTHHIST} -- environment variable
                   1297: @code{GFORTHHIST} -- (Unix systems only) specifies the directory in which to
                   1298: open/create the history file, @file{.gforth-history}. Default:
                   1299: @code{$HOME}.
                   1300: 
                   1301: @item
                   1302: @cindex @code{GFORTHPATH} -- environment variable
                   1303: @code{GFORTHPATH} -- specifies the path used when searching for the gforth image file and
                   1304: for Forth source-code files.
                   1305: 
                   1306: @item
                   1307: @cindex @code{GFORTH} -- environment variable
1.49      anton    1308: @code{GFORTH} -- used by @file{gforthmi}, @xref{gforthmi}.
1.48      anton    1309: 
                   1310: @item
                   1311: @cindex @code{GFORTHD} -- environment variable
1.62      crook    1312: @code{GFORTHD} -- used by @file{gforthmi}, @xref{gforthmi}.
1.48      anton    1313: 
                   1314: @item
                   1315: @cindex @code{TMP}, @code{TEMP} - environment variable
                   1316: @code{TMP}, @code{TEMP} - (non-Unix systems only) used as a potential
                   1317: location for the history file.
                   1318: @end itemize
                   1319: 
                   1320: @comment also POSIXELY_CORRECT LINES COLUMNS HOME but no interest in
                   1321: @comment mentioning these.
                   1322: 
                   1323: All the Gforth environment variables default to sensible values if they
                   1324: are not set.
                   1325: 
                   1326: 
                   1327: @comment ----------------------------------------------
                   1328: @node Gforth Files, Startup speed, Environment variables, Gforth Environment
                   1329: @section Gforth files
                   1330: @cindex Gforth files
                   1331: 
                   1332: When you install Gforth on a Unix system, it installs files in these
                   1333: locations by default:
                   1334: 
                   1335: @itemize @bullet
                   1336: @item
                   1337: @file{/usr/local/bin/gforth}
                   1338: @item
                   1339: @file{/usr/local/bin/gforthmi}
                   1340: @item
                   1341: @file{/usr/local/man/man1/gforth.1} - man page.
                   1342: @item
                   1343: @file{/usr/local/info} - the Info version of this manual.
                   1344: @item
                   1345: @file{/usr/local/lib/gforth/<version>/...} - Gforth @file{.fi} files.
                   1346: @item
                   1347: @file{/usr/local/share/gforth/<version>/TAGS} - Emacs TAGS file.
                   1348: @item
                   1349: @file{/usr/local/share/gforth/<version>/...} - Gforth source files.
                   1350: @item
                   1351: @file{.../emacs/site-lisp/gforth.el} - Emacs gforth mode.
                   1352: @end itemize
                   1353: 
                   1354: You can select different places for installation by using
                   1355: @code{configure} options (listed with @code{configure --help}).
                   1356: 
                   1357: @comment ----------------------------------------------
                   1358: @node Startup speed,  , Gforth Files, Gforth Environment
                   1359: @section Startup speed
                   1360: @cindex Startup speed
                   1361: @cindex speed, startup
                   1362: 
                   1363: If Gforth is used for CGI scripts or in shell scripts, its startup
                   1364: speed may become a problem.  On a 300MHz 21064a under Linux-2.2.13 with
                   1365: glibc-2.0.7, @code{gforth -e bye} takes about 24.6ms user and 11.3ms
                   1366: system time.
                   1367: 
                   1368: If startup speed is a problem, you may consider the following ways to
                   1369: improve it; or you may consider ways to reduce the number of startups
1.62      crook    1370: (for example, by using Fast-CGI).
1.48      anton    1371: 
                   1372: The first step to improve startup speed is to statically link Gforth, by
                   1373: building it with @code{XLDFLAGS=-static}.  This requires more memory for
                   1374: the code and will therefore slow down the first invocation, but
                   1375: subsequent invocations avoid the dynamic linking overhead.  Another
                   1376: disadvantage is that Gforth won't profit from library upgrades.  As a
                   1377: result, @code{gforth-static -e bye} takes about 17.1ms user and
                   1378: 8.2ms system time.
                   1379: 
                   1380: The next step to improve startup speed is to use a non-relocatable image
1.65      anton    1381: (@pxref{Non-Relocatable Image Files}).  You can create this image with
1.48      anton    1382: @code{gforth -e "savesystem gforthnr.fi bye"} and later use it with
                   1383: @code{gforth -i gforthnr.fi ...}.  This avoids the relocation overhead
                   1384: and a part of the copy-on-write overhead.  The disadvantage is that the
1.62      crook    1385: non-relocatable image does not work if the OS gives Gforth a different
1.48      anton    1386: address for the dictionary, for whatever reason; so you better provide a
                   1387: fallback on a relocatable image.  @code{gforth-static -i gforthnr.fi -e
                   1388: bye} takes about 15.3ms user and 7.5ms system time.
                   1389: 
                   1390: The final step is to disable dictionary hashing in Gforth.  Gforth
                   1391: builds the hash table on startup, which takes much of the startup
                   1392: overhead. You can do this by commenting out the @code{include hash.fs}
                   1393: in @file{startup.fs} and everything that requires @file{hash.fs} (at the
                   1394: moment @file{table.fs} and @file{ekey.fs}) and then doing @code{make}.
                   1395: The disadvantages are that functionality like @code{table} and
                   1396: @code{ekey} is missing and that text interpretation (e.g., compiling)
                   1397: now takes much longer. So, you should only use this method if there is
                   1398: no significant text interpretation to perform (the script should be
1.62      crook    1399: compiled into the image, amongst other things).  @code{gforth-static -i
1.48      anton    1400: gforthnrnh.fi -e bye} takes about 2.1ms user and 6.1ms system time.
                   1401: 
                   1402: @c ******************************************************************
                   1403: @node Tutorial, Introduction, Gforth Environment, Top
                   1404: @chapter Forth Tutorial
                   1405: @cindex Tutorial
                   1406: @cindex Forth Tutorial
                   1407: 
1.67      anton    1408: @c Topics from nac's Introduction that could be mentioned:
                   1409: @c press <ret> after each line
                   1410: @c Prompt
                   1411: @c numbers vs. words in dictionary on text interpretation
                   1412: @c what happens on redefinition
                   1413: @c parsing words (in particular, defining words)
                   1414: 
1.83      anton    1415: The difference of this chapter from the Introduction
                   1416: (@pxref{Introduction}) is that this tutorial is more fast-paced, should
                   1417: be used while sitting in front of a computer, and covers much more
                   1418: material, but does not explain how the Forth system works.
                   1419: 
1.62      crook    1420: This tutorial can be used with any ANS-compliant Forth; any
                   1421: Gforth-specific features are marked as such and you can skip them if you
                   1422: work with another Forth.  This tutorial does not explain all features of
                   1423: Forth, just enough to get you started and give you some ideas about the
                   1424: facilities available in Forth.  Read the rest of the manual and the
                   1425: standard when you are through this.
1.48      anton    1426: 
                   1427: The intended way to use this tutorial is that you work through it while
                   1428: sitting in front of the console, take a look at the examples and predict
                   1429: what they will do, then try them out; if the outcome is not as expected,
                   1430: find out why (e.g., by trying out variations of the example), so you
                   1431: understand what's going on.  There are also some assignments that you
                   1432: should solve.
                   1433: 
                   1434: This tutorial assumes that you have programmed before and know what,
                   1435: e.g., a loop is.
                   1436: 
                   1437: @c !! explain compat library
                   1438: 
                   1439: @menu
                   1440: * Starting Gforth Tutorial::    
                   1441: * Syntax Tutorial::             
                   1442: * Crash Course Tutorial::       
                   1443: * Stack Tutorial::              
                   1444: * Arithmetics Tutorial::        
                   1445: * Stack Manipulation Tutorial::  
                   1446: * Using files for Forth code Tutorial::  
                   1447: * Comments Tutorial::           
                   1448: * Colon Definitions Tutorial::  
                   1449: * Decompilation Tutorial::      
                   1450: * Stack-Effect Comments Tutorial::  
                   1451: * Types Tutorial::              
                   1452: * Factoring Tutorial::          
                   1453: * Designing the stack effect Tutorial::  
                   1454: * Local Variables Tutorial::    
                   1455: * Conditional execution Tutorial::  
                   1456: * Flags and Comparisons Tutorial::  
                   1457: * General Loops Tutorial::      
                   1458: * Counted loops Tutorial::      
                   1459: * Recursion Tutorial::          
                   1460: * Leaving definitions or loops Tutorial::  
                   1461: * Return Stack Tutorial::       
                   1462: * Memory Tutorial::             
                   1463: * Characters and Strings Tutorial::  
                   1464: * Alignment Tutorial::          
1.87      anton    1465: * Files Tutorial::              
1.48      anton    1466: * Interpretation and Compilation Semantics and Immediacy Tutorial::  
                   1467: * Execution Tokens Tutorial::   
                   1468: * Exceptions Tutorial::         
                   1469: * Defining Words Tutorial::     
                   1470: * Arrays and Records Tutorial::  
                   1471: * POSTPONE Tutorial::           
                   1472: * Literal Tutorial::            
                   1473: * Advanced macros Tutorial::    
                   1474: * Compilation Tokens Tutorial::  
                   1475: * Wordlists and Search Order Tutorial::  
                   1476: @end menu
                   1477: 
                   1478: @node Starting Gforth Tutorial, Syntax Tutorial, Tutorial, Tutorial
                   1479: @section Starting Gforth
1.66      anton    1480: @cindex starting Gforth tutorial
1.48      anton    1481: You can start Gforth by typing its name:
                   1482: 
                   1483: @example
                   1484: gforth
                   1485: @end example
                   1486: 
                   1487: That puts you into interactive mode; you can leave Gforth by typing
                   1488: @code{bye}.  While in Gforth, you can edit the command line and access
                   1489: the command line history with cursor keys, similar to bash.
                   1490: 
                   1491: 
                   1492: @node Syntax Tutorial, Crash Course Tutorial, Starting Gforth Tutorial, Tutorial
                   1493: @section Syntax
1.66      anton    1494: @cindex syntax tutorial
1.48      anton    1495: 
                   1496: A @dfn{word} is a sequence of arbitrary characters (expcept white
                   1497: space).  Words are separated by white space.  E.g., each of the
                   1498: following lines contains exactly one word:
                   1499: 
                   1500: @example
                   1501: word
                   1502: !@@#$%^&*()
                   1503: 1234567890
                   1504: 5!a
                   1505: @end example
                   1506: 
                   1507: A frequent beginner's error is to leave away necessary white space,
                   1508: resulting in an error like @samp{Undefined word}; so if you see such an
                   1509: error, check if you have put spaces wherever necessary.
                   1510: 
                   1511: @example
                   1512: ." hello, world" \ correct
                   1513: ."hello, world"  \ gives an "Undefined word" error
                   1514: @end example
                   1515: 
1.65      anton    1516: Gforth and most other Forth systems ignore differences in case (they are
1.48      anton    1517: case-insensitive), i.e., @samp{word} is the same as @samp{Word}.  If
                   1518: your system is case-sensitive, you may have to type all the examples
                   1519: given here in upper case.
                   1520: 
                   1521: 
                   1522: @node Crash Course Tutorial, Stack Tutorial, Syntax Tutorial, Tutorial
                   1523: @section Crash Course
                   1524: 
                   1525: Type
                   1526: 
                   1527: @example
                   1528: 0 0 !
                   1529: here execute
                   1530: ' catch >body 20 erase abort
                   1531: ' (quit) >body 20 erase
                   1532: @end example
                   1533: 
                   1534: The last two examples are guaranteed to destroy parts of Gforth (and
                   1535: most other systems), so you better leave Gforth afterwards (if it has
                   1536: not finished by itself).  On some systems you may have to kill gforth
                   1537: from outside (e.g., in Unix with @code{kill}).
                   1538: 
                   1539: Now that you know how to produce crashes (and that there's not much to
                   1540: them), let's learn how to produce meaningful programs.
                   1541: 
                   1542: 
                   1543: @node Stack Tutorial, Arithmetics Tutorial, Crash Course Tutorial, Tutorial
                   1544: @section Stack
1.66      anton    1545: @cindex stack tutorial
1.48      anton    1546: 
                   1547: The most obvious feature of Forth is the stack.  When you type in a
                   1548: number, it is pushed on the stack.  You can display the content of the
                   1549: stack with @code{.s}.
                   1550: 
                   1551: @example
                   1552: 1 2 .s
                   1553: 3 .s
                   1554: @end example
                   1555: 
                   1556: @code{.s} displays the top-of-stack to the right, i.e., the numbers
                   1557: appear in @code{.s} output as they appeared in the input.
                   1558: 
                   1559: You can print the top of stack element with @code{.}.
                   1560: 
                   1561: @example
                   1562: 1 2 3 . . .
                   1563: @end example
                   1564: 
                   1565: In general, words consume their stack arguments (@code{.s} is an
                   1566: exception).
                   1567: 
                   1568: @assignment
                   1569: What does the stack contain after @code{5 6 7 .}?
                   1570: @endassignment
                   1571: 
                   1572: 
                   1573: @node Arithmetics Tutorial, Stack Manipulation Tutorial, Stack Tutorial, Tutorial
                   1574: @section Arithmetics
1.66      anton    1575: @cindex arithmetics tutorial
1.48      anton    1576: 
                   1577: The words @code{+}, @code{-}, @code{*}, @code{/}, and @code{mod} always
                   1578: operate on the top two stack items:
                   1579: 
                   1580: @example
1.67      anton    1581: 2 2 .s
                   1582: + .s
                   1583: .
1.48      anton    1584: 2 1 - .
                   1585: 7 3 mod .
                   1586: @end example
                   1587: 
                   1588: The operands of @code{-}, @code{/}, and @code{mod} are in the same order
                   1589: as in the corresponding infix expression (this is generally the case in
                   1590: Forth).
                   1591: 
                   1592: Parentheses are superfluous (and not available), because the order of
                   1593: the words unambiguously determines the order of evaluation and the
                   1594: operands:
                   1595: 
                   1596: @example
                   1597: 3 4 + 5 * .
                   1598: 3 4 5 * + .
                   1599: @end example
                   1600: 
                   1601: @assignment
                   1602: What are the infix expressions corresponding to the Forth code above?
                   1603: Write @code{6-7*8+9} in Forth notation@footnote{This notation is also
                   1604: known as Postfix or RPN (Reverse Polish Notation).}.
                   1605: @endassignment
                   1606: 
                   1607: To change the sign, use @code{negate}:
                   1608: 
                   1609: @example
                   1610: 2 negate .
                   1611: @end example
                   1612: 
                   1613: @assignment
                   1614: Convert -(-3)*4-5 to Forth.
                   1615: @endassignment
                   1616: 
                   1617: @code{/mod} performs both @code{/} and @code{mod}.
                   1618: 
                   1619: @example
                   1620: 7 3 /mod . .
                   1621: @end example
                   1622: 
1.66      anton    1623: Reference: @ref{Arithmetic}.
                   1624: 
                   1625: 
1.48      anton    1626: @node Stack Manipulation Tutorial, Using files for Forth code Tutorial, Arithmetics Tutorial, Tutorial
                   1627: @section Stack Manipulation
1.66      anton    1628: @cindex stack manipulation tutorial
1.48      anton    1629: 
                   1630: Stack manipulation words rearrange the data on the stack.
                   1631: 
                   1632: @example
                   1633: 1 .s drop .s
                   1634: 1 .s dup .s drop drop .s
                   1635: 1 2 .s over .s drop drop drop
                   1636: 1 2 .s swap .s drop drop
                   1637: 1 2 3 .s rot .s drop drop drop
                   1638: @end example
                   1639: 
                   1640: These are the most important stack manipulation words.  There are also
                   1641: variants that manipulate twice as many stack items:
                   1642: 
                   1643: @example
                   1644: 1 2 3 4 .s 2swap .s 2drop 2drop
                   1645: @end example
                   1646: 
                   1647: Two more stack manipulation words are:
                   1648: 
                   1649: @example
                   1650: 1 2 .s nip .s drop
                   1651: 1 2 .s tuck .s 2drop drop
                   1652: @end example
                   1653: 
                   1654: @assignment
                   1655: Replace @code{nip} and @code{tuck} with combinations of other stack
                   1656: manipulation words.
                   1657: 
                   1658: @example
                   1659: Given:          How do you get:
                   1660: 1 2 3           3 2 1           
                   1661: 1 2 3           1 2 3 2                 
                   1662: 1 2 3           1 2 3 3                 
                   1663: 1 2 3           1 3 3           
                   1664: 1 2 3           2 1 3           
                   1665: 1 2 3 4         4 3 2 1         
                   1666: 1 2 3           1 2 3 1 2 3             
                   1667: 1 2 3 4         1 2 3 4 1 2             
                   1668: 1 2 3
                   1669: 1 2 3           1 2 3 4                 
                   1670: 1 2 3           1 3             
                   1671: @end example
                   1672: @endassignment
                   1673: 
                   1674: @example
                   1675: 5 dup * .
                   1676: @end example
                   1677: 
                   1678: @assignment
                   1679: Write 17^3 and 17^4 in Forth, without writing @code{17} more than once.
                   1680: Write a piece of Forth code that expects two numbers on the stack
                   1681: (@var{a} and @var{b}, with @var{b} on top) and computes
                   1682: @code{(a-b)(a+1)}.
                   1683: @endassignment
                   1684: 
1.66      anton    1685: Reference: @ref{Stack Manipulation}.
                   1686: 
                   1687: 
1.48      anton    1688: @node Using files for Forth code Tutorial, Comments Tutorial, Stack Manipulation Tutorial, Tutorial
                   1689: @section Using files for Forth code
1.66      anton    1690: @cindex loading Forth code, tutorial
                   1691: @cindex files containing Forth code, tutorial
1.48      anton    1692: 
                   1693: While working at the Forth command line is convenient for one-line
                   1694: examples and short one-off code, you probably want to store your source
                   1695: code in files for convenient editing and persistence.  You can use your
                   1696: favourite editor (Gforth includes Emacs support, @pxref{Emacs and
                   1697: Gforth}) to create @var{file} and use
                   1698: 
                   1699: @example
                   1700: s" @var{file}" included
                   1701: @end example
                   1702: 
                   1703: to load it into your Forth system.  The file name extension I use for
                   1704: Forth files is @samp{.fs}.
                   1705: 
                   1706: You can easily start Gforth with some files loaded like this:
                   1707: 
                   1708: @example
                   1709: gforth @var{file1} @var{file2}
                   1710: @end example
                   1711: 
                   1712: If an error occurs during loading these files, Gforth terminates,
                   1713: whereas an error during @code{INCLUDED} within Gforth usually gives you
                   1714: a Gforth command line.  Starting the Forth system every time gives you a
                   1715: clean start every time, without interference from the results of earlier
                   1716: tries.
                   1717: 
                   1718: I often put all the tests in a file, then load the code and run the
                   1719: tests with
                   1720: 
                   1721: @example
                   1722: gforth @var{code} @var{tests} -e bye
                   1723: @end example
                   1724: 
                   1725: (often by performing this command with @kbd{C-x C-e} in Emacs).  The
                   1726: @code{-e bye} ensures that Gforth terminates afterwards so that I can
                   1727: restart this command without ado.
                   1728: 
                   1729: The advantage of this approach is that the tests can be repeated easily
                   1730: every time the program ist changed, making it easy to catch bugs
                   1731: introduced by the change.
                   1732: 
1.66      anton    1733: Reference: @ref{Forth source files}.
                   1734: 
1.48      anton    1735: 
                   1736: @node Comments Tutorial, Colon Definitions Tutorial, Using files for Forth code Tutorial, Tutorial
                   1737: @section Comments
1.66      anton    1738: @cindex comments tutorial
1.48      anton    1739: 
                   1740: @example
                   1741: \ That's a comment; it ends at the end of the line
                   1742: ( Another comment; it ends here: )  .s
                   1743: @end example
                   1744: 
                   1745: @code{\} and @code{(} are ordinary Forth words and therefore have to be
                   1746: separated with white space from the following text.
                   1747: 
                   1748: @example
                   1749: \This gives an "Undefined word" error
                   1750: @end example
                   1751: 
                   1752: The first @code{)} ends a comment started with @code{(}, so you cannot
                   1753: nest @code{(}-comments; and you cannot comment out text containing a
                   1754: @code{)} with @code{( ... )}@footnote{therefore it's a good idea to
                   1755: avoid @code{)} in word names.}.
                   1756: 
                   1757: I use @code{\}-comments for descriptive text and for commenting out code
                   1758: of one or more line; I use @code{(}-comments for describing the stack
                   1759: effect, the stack contents, or for commenting out sub-line pieces of
                   1760: code.
                   1761: 
                   1762: The Emacs mode @file{gforth.el} (@pxref{Emacs and Gforth}) supports
                   1763: these uses by commenting out a region with @kbd{C-x \}, uncommenting a
                   1764: region with @kbd{C-u C-x \}, and filling a @code{\}-commented region
                   1765: with @kbd{M-q}.
                   1766: 
1.66      anton    1767: Reference: @ref{Comments}.
                   1768: 
1.48      anton    1769: 
                   1770: @node Colon Definitions Tutorial, Decompilation Tutorial, Comments Tutorial, Tutorial
                   1771: @section Colon Definitions
1.66      anton    1772: @cindex colon definitions, tutorial
                   1773: @cindex definitions, tutorial
                   1774: @cindex procedures, tutorial
                   1775: @cindex functions, tutorial
1.48      anton    1776: 
                   1777: are similar to procedures and functions in other programming languages.
                   1778: 
                   1779: @example
                   1780: : squared ( n -- n^2 )
                   1781:    dup * ;
                   1782: 5 squared .
                   1783: 7 squared .
                   1784: @end example
                   1785: 
                   1786: @code{:} starts the colon definition; its name is @code{squared}.  The
                   1787: following comment describes its stack effect.  The words @code{dup *}
                   1788: are not executed, but compiled into the definition.  @code{;} ends the
                   1789: colon definition.
                   1790: 
                   1791: The newly-defined word can be used like any other word, including using
                   1792: it in other definitions:
                   1793: 
                   1794: @example
                   1795: : cubed ( n -- n^3 )
                   1796:    dup squared * ;
                   1797: -5 cubed .
                   1798: : fourth-power ( n -- n^4 )
                   1799:    squared squared ;
                   1800: 3 fourth-power .
                   1801: @end example
                   1802: 
                   1803: @assignment
                   1804: Write colon definitions for @code{nip}, @code{tuck}, @code{negate}, and
                   1805: @code{/mod} in terms of other Forth words, and check if they work (hint:
                   1806: test your tests on the originals first).  Don't let the
                   1807: @samp{redefined}-Messages spook you, they are just warnings.
                   1808: @endassignment
                   1809: 
1.66      anton    1810: Reference: @ref{Colon Definitions}.
                   1811: 
1.48      anton    1812: 
                   1813: @node Decompilation Tutorial, Stack-Effect Comments Tutorial, Colon Definitions Tutorial, Tutorial
                   1814: @section Decompilation
1.66      anton    1815: @cindex decompilation tutorial
                   1816: @cindex see tutorial
1.48      anton    1817: 
                   1818: You can decompile colon definitions with @code{see}:
                   1819: 
                   1820: @example
                   1821: see squared
                   1822: see cubed
                   1823: @end example
                   1824: 
                   1825: In Gforth @code{see} shows you a reconstruction of the source code from
                   1826: the executable code.  Informations that were present in the source, but
                   1827: not in the executable code, are lost (e.g., comments).
                   1828: 
1.65      anton    1829: You can also decompile the predefined words:
                   1830: 
                   1831: @example
                   1832: see .
                   1833: see +
                   1834: @end example
                   1835: 
                   1836: 
1.48      anton    1837: @node Stack-Effect Comments Tutorial, Types Tutorial, Decompilation Tutorial, Tutorial
                   1838: @section Stack-Effect Comments
1.66      anton    1839: @cindex stack-effect comments, tutorial
                   1840: @cindex --, tutorial
1.48      anton    1841: By convention the comment after the name of a definition describes the
                   1842: stack effect: The part in from of the @samp{--} describes the state of
                   1843: the stack before the execution of the definition, i.e., the parameters
                   1844: that are passed into the colon definition; the part behind the @samp{--}
                   1845: is the state of the stack after the execution of the definition, i.e.,
                   1846: the results of the definition.  The stack comment only shows the top
                   1847: stack items that the definition accesses and/or changes.
                   1848: 
                   1849: You should put a correct stack effect on every definition, even if it is
                   1850: just @code{( -- )}.  You should also add some descriptive comment to
                   1851: more complicated words (I usually do this in the lines following
                   1852: @code{:}).  If you don't do this, your code becomes unreadable (because
                   1853: you have to work through every definition before you can undertsand
                   1854: any).
                   1855: 
                   1856: @assignment
                   1857: The stack effect of @code{swap} can be written like this: @code{x1 x2 --
                   1858: x2 x1}.  Describe the stack effect of @code{-}, @code{drop}, @code{dup},
                   1859: @code{over}, @code{rot}, @code{nip}, and @code{tuck}.  Hint: When you
1.65      anton    1860: are done, you can compare your stack effects to those in this manual
1.48      anton    1861: (@pxref{Word Index}).
                   1862: @endassignment
                   1863: 
                   1864: Sometimes programmers put comments at various places in colon
                   1865: definitions that describe the contents of the stack at that place (stack
                   1866: comments); i.e., they are like the first part of a stack-effect
                   1867: comment. E.g.,
                   1868: 
                   1869: @example
                   1870: : cubed ( n -- n^3 )
                   1871:    dup squared  ( n n^2 ) * ;
                   1872: @end example
                   1873: 
                   1874: In this case the stack comment is pretty superfluous, because the word
                   1875: is simple enough.  If you think it would be a good idea to add such a
                   1876: comment to increase readability, you should also consider factoring the
                   1877: word into several simpler words (@pxref{Factoring Tutorial,,
1.60      anton    1878: Factoring}), which typically eliminates the need for the stack comment;
1.48      anton    1879: however, if you decide not to refactor it, then having such a comment is
                   1880: better than not having it.
                   1881: 
                   1882: The names of the stack items in stack-effect and stack comments in the
                   1883: standard, in this manual, and in many programs specify the type through
                   1884: a type prefix, similar to Fortran and Hungarian notation.  The most
                   1885: frequent prefixes are:
                   1886: 
                   1887: @table @code
                   1888: @item n
                   1889: signed integer
                   1890: @item u
                   1891: unsigned integer
                   1892: @item c
                   1893: character
                   1894: @item f
                   1895: Boolean flags, i.e. @code{false} or @code{true}.
                   1896: @item a-addr,a-
                   1897: Cell-aligned address
                   1898: @item c-addr,c-
                   1899: Char-aligned address (note that a Char may have two bytes in Windows NT)
                   1900: @item xt
                   1901: Execution token, same size as Cell
                   1902: @item w,x
                   1903: Cell, can contain an integer or an address.  It usually takes 32, 64 or
                   1904: 16 bits (depending on your platform and Forth system). A cell is more
                   1905: commonly known as machine word, but the term @emph{word} already means
                   1906: something different in Forth.
                   1907: @item d
                   1908: signed double-cell integer
                   1909: @item ud
                   1910: unsigned double-cell integer
                   1911: @item r
                   1912: Float (on the FP stack)
                   1913: @end table
                   1914: 
                   1915: You can find a more complete list in @ref{Notation}.
                   1916: 
                   1917: @assignment
                   1918: Write stack-effect comments for all definitions you have written up to
                   1919: now.
                   1920: @endassignment
                   1921: 
                   1922: 
                   1923: @node Types Tutorial, Factoring Tutorial, Stack-Effect Comments Tutorial, Tutorial
                   1924: @section Types
1.66      anton    1925: @cindex types tutorial
1.48      anton    1926: 
                   1927: In Forth the names of the operations are not overloaded; so similar
                   1928: operations on different types need different names; e.g., @code{+} adds
                   1929: integers, and you have to use @code{f+} to add floating-point numbers.
                   1930: The following prefixes are often used for related operations on
                   1931: different types:
                   1932: 
                   1933: @table @code
                   1934: @item (none)
                   1935: signed integer
                   1936: @item u
                   1937: unsigned integer
                   1938: @item c
                   1939: character
                   1940: @item d
                   1941: signed double-cell integer
                   1942: @item ud, du
                   1943: unsigned double-cell integer
                   1944: @item 2
                   1945: two cells (not-necessarily double-cell numbers)
                   1946: @item m, um
                   1947: mixed single-cell and double-cell operations
                   1948: @item f
                   1949: floating-point (note that in stack comments @samp{f} represents flags,
1.66      anton    1950: and @samp{r} represents FP numbers).
1.48      anton    1951: @end table
                   1952: 
                   1953: If there are no differences between the signed and the unsigned variant
                   1954: (e.g., for @code{+}), there is only the prefix-less variant.
                   1955: 
                   1956: Forth does not perform type checking, neither at compile time, nor at
                   1957: run time.  If you use the wrong oeration, the data are interpreted
                   1958: incorrectly:
                   1959: 
                   1960: @example
                   1961: -1 u.
                   1962: @end example
                   1963: 
                   1964: If you have only experience with type-checked languages until now, and
                   1965: have heard how important type-checking is, don't panic!  In my
                   1966: experience (and that of other Forthers), type errors in Forth code are
                   1967: usually easy to find (once you get used to it), the increased vigilance
                   1968: of the programmer tends to catch some harder errors in addition to most
                   1969: type errors, and you never have to work around the type system, so in
                   1970: most situations the lack of type-checking seems to be a win (projects to
                   1971: add type checking to Forth have not caught on).
                   1972: 
                   1973: 
                   1974: @node Factoring Tutorial, Designing the stack effect Tutorial, Types Tutorial, Tutorial
                   1975: @section Factoring
1.66      anton    1976: @cindex factoring tutorial
1.48      anton    1977: 
                   1978: If you try to write longer definitions, you will soon find it hard to
                   1979: keep track of the stack contents.  Therefore, good Forth programmers
                   1980: tend to write only short definitions (e.g., three lines).  The art of
                   1981: finding meaningful short definitions is known as factoring (as in
                   1982: factoring polynomials).
                   1983: 
                   1984: Well-factored programs offer additional advantages: smaller, more
                   1985: general words, are easier to test and debug and can be reused more and
                   1986: better than larger, specialized words.
                   1987: 
                   1988: So, if you run into difficulties with stack management, when writing
                   1989: code, try to define meaningful factors for the word, and define the word
                   1990: in terms of those.  Even if a factor contains only two words, it is
                   1991: often helpful.
                   1992: 
1.65      anton    1993: Good factoring is not easy, and it takes some practice to get the knack
                   1994: for it; but even experienced Forth programmers often don't find the
                   1995: right solution right away, but only when rewriting the program.  So, if
                   1996: you don't come up with a good solution immediately, keep trying, don't
                   1997: despair.
1.48      anton    1998: 
                   1999: @c example !!
                   2000: 
                   2001: 
                   2002: @node Designing the stack effect Tutorial, Local Variables Tutorial, Factoring Tutorial, Tutorial
                   2003: @section Designing the stack effect
1.66      anton    2004: @cindex Stack effect design, tutorial
                   2005: @cindex design of stack effects, tutorial
1.48      anton    2006: 
                   2007: In other languages you can use an arbitrary order of parameters for a
1.65      anton    2008: function; and since there is only one result, you don't have to deal with
1.48      anton    2009: the order of results, either.
                   2010: 
                   2011: In Forth (and other stack-based languages, e.g., Postscript) the
                   2012: parameter and result order of a definition is important and should be
                   2013: designed well.  The general guideline is to design the stack effect such
                   2014: that the word is simple to use in most cases, even if that complicates
                   2015: the implementation of the word.  Some concrete rules are:
                   2016: 
                   2017: @itemize @bullet
                   2018: 
                   2019: @item
                   2020: Words consume all of their parameters (e.g., @code{.}).
                   2021: 
                   2022: @item
                   2023: If there is a convention on the order of parameters (e.g., from
                   2024: mathematics or another programming language), stick with it (e.g.,
                   2025: @code{-}).
                   2026: 
                   2027: @item
                   2028: If one parameter usually requires only a short computation (e.g., it is
                   2029: a constant), pass it on the top of the stack.  Conversely, parameters
                   2030: that usually require a long sequence of code to compute should be passed
                   2031: as the bottom (i.e., first) parameter.  This makes the code easier to
                   2032: read, because reader does not need to keep track of the bottom item
                   2033: through a long sequence of code (or, alternatively, through stack
1.49      anton    2034: manipulations). E.g., @code{!} (store, @pxref{Memory}) expects the
1.48      anton    2035: address on top of the stack because it is usually simpler to compute
                   2036: than the stored value (often the address is just a variable).
                   2037: 
                   2038: @item
                   2039: Similarly, results that are usually consumed quickly should be returned
                   2040: on the top of stack, whereas a result that is often used in long
                   2041: computations should be passed as bottom result.  E.g., the file words
                   2042: like @code{open-file} return the error code on the top of stack, because
                   2043: it is usually consumed quickly by @code{throw}; moreover, the error code
                   2044: has to be checked before doing anything with the other results.
                   2045: 
                   2046: @end itemize
                   2047: 
                   2048: These rules are just general guidelines, don't lose sight of the overall
                   2049: goal to make the words easy to use.  E.g., if the convention rule
                   2050: conflicts with the computation-length rule, you might decide in favour
                   2051: of the convention if the word will be used rarely, and in favour of the
                   2052: computation-length rule if the word will be used frequently (because
                   2053: with frequent use the cost of breaking the computation-length rule would
                   2054: be quite high, and frequent use makes it easier to remember an
                   2055: unconventional order).
                   2056: 
                   2057: @c example !! structure package
                   2058: 
1.65      anton    2059: 
1.48      anton    2060: @node Local Variables Tutorial, Conditional execution Tutorial, Designing the stack effect Tutorial, Tutorial
                   2061: @section Local Variables
1.66      anton    2062: @cindex local variables, tutorial
1.48      anton    2063: 
                   2064: You can define local variables (@emph{locals}) in a colon definition:
                   2065: 
                   2066: @example
                   2067: : swap @{ a b -- b a @}
                   2068:   b a ;
                   2069: 1 2 swap .s 2drop
                   2070: @end example
                   2071: 
                   2072: (If your Forth system does not support this syntax, include
                   2073: @file{compat/anslocals.fs} first).
                   2074: 
                   2075: In this example @code{@{ a b -- b a @}} is the locals definition; it
                   2076: takes two cells from the stack, puts the top of stack in @code{b} and
                   2077: the next stack element in @code{a}.  @code{--} starts a comment ending
                   2078: with @code{@}}.  After the locals definition, using the name of the
                   2079: local will push its value on the stack.  You can leave the comment
                   2080: part (@code{-- b a}) away:
                   2081: 
                   2082: @example
                   2083: : swap ( x1 x2 -- x2 x1 )
                   2084:   @{ a b @} b a ;
                   2085: @end example
                   2086: 
                   2087: In Gforth you can have several locals definitions, anywhere in a colon
                   2088: definition; in contrast, in a standard program you can have only one
                   2089: locals definition per colon definition, and that locals definition must
                   2090: be outside any controll structure.
                   2091: 
                   2092: With locals you can write slightly longer definitions without running
                   2093: into stack trouble.  However, I recommend trying to write colon
                   2094: definitions without locals for exercise purposes to help you gain the
                   2095: essential factoring skills.
                   2096: 
                   2097: @assignment
                   2098: Rewrite your definitions until now with locals
                   2099: @endassignment
                   2100: 
1.66      anton    2101: Reference: @ref{Locals}.
                   2102: 
1.48      anton    2103: 
                   2104: @node Conditional execution Tutorial, Flags and Comparisons Tutorial, Local Variables Tutorial, Tutorial
                   2105: @section Conditional execution
1.66      anton    2106: @cindex conditionals, tutorial
                   2107: @cindex if, tutorial
1.48      anton    2108: 
                   2109: In Forth you can use control structures only inside colon definitions.
                   2110: An @code{if}-structure looks like this:
                   2111: 
                   2112: @example
                   2113: : abs ( n1 -- +n2 )
                   2114:     dup 0 < if
                   2115:         negate
                   2116:     endif ;
                   2117: 5 abs .
                   2118: -5 abs .
                   2119: @end example
                   2120: 
                   2121: @code{if} takes a flag from the stack.  If the flag is non-zero (true),
                   2122: the following code is performed, otherwise execution continues after the
1.51      pazsan   2123: @code{endif} (or @code{else}).  @code{<} compares the top two stack
1.48      anton    2124: elements and prioduces a flag:
                   2125: 
                   2126: @example
                   2127: 1 2 < .
                   2128: 2 1 < .
                   2129: 1 1 < .
                   2130: @end example
                   2131: 
                   2132: Actually the standard name for @code{endif} is @code{then}.  This
                   2133: tutorial presents the examples using @code{endif}, because this is often
                   2134: less confusing for people familiar with other programming languages
                   2135: where @code{then} has a different meaning.  If your system does not have
                   2136: @code{endif}, define it with
                   2137: 
                   2138: @example
                   2139: : endif postpone then ; immediate
                   2140: @end example
                   2141: 
                   2142: You can optionally use an @code{else}-part:
                   2143: 
                   2144: @example
                   2145: : min ( n1 n2 -- n )
                   2146:   2dup < if
                   2147:     drop
                   2148:   else
                   2149:     nip
                   2150:   endif ;
                   2151: 2 3 min .
                   2152: 3 2 min .
                   2153: @end example
                   2154: 
                   2155: @assignment
                   2156: Write @code{min} without @code{else}-part (hint: what's the definition
                   2157: of @code{nip}?).
                   2158: @endassignment
                   2159: 
1.66      anton    2160: Reference: @ref{Selection}.
                   2161: 
1.48      anton    2162: 
                   2163: @node Flags and Comparisons Tutorial, General Loops Tutorial, Conditional execution Tutorial, Tutorial
                   2164: @section Flags and Comparisons
1.66      anton    2165: @cindex flags tutorial
                   2166: @cindex comparison tutorial
1.48      anton    2167: 
                   2168: In a false-flag all bits are clear (0 when interpreted as integer).  In
                   2169: a canonical true-flag all bits are set (-1 as a twos-complement signed
                   2170: integer); in many contexts (e.g., @code{if}) any non-zero value is
                   2171: treated as true flag.
                   2172: 
                   2173: @example
                   2174: false .
                   2175: true .
                   2176: true hex u. decimal
                   2177: @end example
                   2178: 
                   2179: Comparison words produce canonical flags:
                   2180: 
                   2181: @example
                   2182: 1 1 = .
                   2183: 1 0= .
                   2184: 0 1 < .
                   2185: 0 0 < .
                   2186: -1 1 u< . \ type error, u< interprets -1 as large unsigned number
                   2187: -1 1 < .
                   2188: @end example
                   2189: 
1.66      anton    2190: Gforth supports all combinations of the prefixes @code{0 u d d0 du f f0}
                   2191: (or none) and the comparisons @code{= <> < > <= >=}.  Only a part of
                   2192: these combinations are standard (for details see the standard,
                   2193: @ref{Numeric comparison}, @ref{Floating Point} or @ref{Word Index}).
1.48      anton    2194: 
                   2195: You can use @code{and or xor invert} can be used as operations on
                   2196: canonical flags.  Actually they are bitwise operations:
                   2197: 
                   2198: @example
                   2199: 1 2 and .
                   2200: 1 2 or .
                   2201: 1 3 xor .
                   2202: 1 invert .
                   2203: @end example
                   2204: 
                   2205: You can convert a zero/non-zero flag into a canonical flag with
                   2206: @code{0<>} (and complement it on the way with @code{0=}).
                   2207: 
                   2208: @example
                   2209: 1 0= .
                   2210: 1 0<> .
                   2211: @end example
                   2212: 
1.65      anton    2213: You can use the all-bits-set feature of canonical flags and the bitwise
1.48      anton    2214: operation of the Boolean operations to avoid @code{if}s:
                   2215: 
                   2216: @example
                   2217: : foo ( n1 -- n2 )
                   2218:   0= if
                   2219:     14
                   2220:   else
                   2221:     0
                   2222:   endif ;
                   2223: 0 foo .
                   2224: 1 foo .
                   2225: 
                   2226: : foo ( n1 -- n2 )
                   2227:   0= 14 and ;
                   2228: 0 foo .
                   2229: 1 foo .
                   2230: @end example
                   2231: 
                   2232: @assignment
                   2233: Write @code{min} without @code{if}.
                   2234: @endassignment
                   2235: 
1.66      anton    2236: For reference, see @ref{Boolean Flags}, @ref{Numeric comparison}, and
                   2237: @ref{Bitwise operations}.
                   2238: 
1.48      anton    2239: 
                   2240: @node General Loops Tutorial, Counted loops Tutorial, Flags and Comparisons Tutorial, Tutorial
                   2241: @section General Loops
1.66      anton    2242: @cindex loops, indefinite, tutorial
1.48      anton    2243: 
                   2244: The endless loop is the most simple one:
                   2245: 
                   2246: @example
                   2247: : endless ( -- )
                   2248:   0 begin
                   2249:     dup . 1+
                   2250:   again ;
                   2251: endless
                   2252: @end example
                   2253: 
                   2254: Terminate this loop by pressing @kbd{Ctrl-C} (in Gforth).  @code{begin}
                   2255: does nothing at run-time, @code{again} jumps back to @code{begin}.
                   2256: 
                   2257: A loop with one exit at any place looks like this:
                   2258: 
                   2259: @example
                   2260: : log2 ( +n1 -- n2 )
                   2261: \ logarithmus dualis of n1>0, rounded down to the next integer
                   2262:   assert( dup 0> )
                   2263:   2/ 0 begin
                   2264:     over 0> while
                   2265:       1+ swap 2/ swap
                   2266:   repeat
                   2267:   nip ;
                   2268: 7 log2 .
                   2269: 8 log2 .
                   2270: @end example
                   2271: 
                   2272: At run-time @code{while} consumes a flag; if it is 0, execution
1.51      pazsan   2273: continues behind the @code{repeat}; if the flag is non-zero, execution
1.48      anton    2274: continues behind the @code{while}.  @code{Repeat} jumps back to
                   2275: @code{begin}, just like @code{again}.
                   2276: 
                   2277: In Forth there are many combinations/abbreviations, like @code{1+}.
                   2278: However, @code{2/} is not one of them; it shifts it's argument right by
                   2279: one bit (arithmetic shift right):
                   2280: 
                   2281: @example
                   2282: -5 2 / .
                   2283: -5 2/ .
                   2284: @end example
                   2285: 
                   2286: @code{assert(} is no standard word, but you can get it on systems other
                   2287: then Gforth by including @file{compat/assert.fs}.  You can see what it
                   2288: does by trying
                   2289: 
                   2290: @example
                   2291: 0 log2 .
                   2292: @end example
                   2293: 
                   2294: Here's a loop with an exit at the end:
                   2295: 
                   2296: @example
                   2297: : log2 ( +n1 -- n2 )
                   2298: \ logarithmus dualis of n1>0, rounded down to the next integer
                   2299:   assert( dup 0 > )
                   2300:   -1 begin
                   2301:     1+ swap 2/ swap
                   2302:     over 0 <=
                   2303:   until
                   2304:   nip ;
                   2305: @end example
                   2306: 
                   2307: @code{Until} consumes a flag; if it is non-zero, execution continues at
                   2308: the @code{begin}, otherwise after the @code{until}.
                   2309: 
                   2310: @assignment
                   2311: Write a definition for computing the greatest common divisor.
                   2312: @endassignment
                   2313: 
1.66      anton    2314: Reference: @ref{Simple Loops}.
                   2315: 
1.48      anton    2316: 
                   2317: @node Counted loops Tutorial, Recursion Tutorial, General Loops Tutorial, Tutorial
                   2318: @section Counted loops
1.66      anton    2319: @cindex loops, counted, tutorial
1.48      anton    2320: 
                   2321: @example
                   2322: : ^ ( n1 u -- n )
                   2323: \ n = the uth power of u1
                   2324:   1 swap 0 u+do
                   2325:     over *
                   2326:   loop
                   2327:   nip ;
                   2328: 3 2 ^ .
                   2329: 4 3 ^ .
                   2330: @end example
                   2331: 
                   2332: @code{U+do} (from @file{compat/loops.fs}, if your Forth system doesn't
                   2333: have it) takes two numbers of the stack @code{( u3 u4 -- )}, and then
                   2334: performs the code between @code{u+do} and @code{loop} for @code{u3-u4}
                   2335: times (or not at all, if @code{u3-u4<0}).
                   2336: 
                   2337: You can see the stack effect design rules at work in the stack effect of
                   2338: the loop start words: Since the start value of the loop is more
                   2339: frequently constant than the end value, the start value is passed on
                   2340: the top-of-stack.
                   2341: 
                   2342: You can access the counter of a counted loop with @code{i}:
                   2343: 
                   2344: @example
                   2345: : fac ( u -- u! )
                   2346:   1 swap 1+ 1 u+do
                   2347:     i *
                   2348:   loop ;
                   2349: 5 fac .
                   2350: 7 fac .
                   2351: @end example
                   2352: 
                   2353: There is also @code{+do}, which expects signed numbers (important for
                   2354: deciding whether to enter the loop).
                   2355: 
                   2356: @assignment
                   2357: Write a definition for computing the nth Fibonacci number.
                   2358: @endassignment
                   2359: 
1.65      anton    2360: You can also use increments other than 1:
                   2361: 
                   2362: @example
                   2363: : up2 ( n1 n2 -- )
                   2364:   +do
                   2365:     i .
                   2366:   2 +loop ;
                   2367: 10 0 up2
                   2368: 
                   2369: : down2 ( n1 n2 -- )
                   2370:   -do
                   2371:     i .
                   2372:   2 -loop ;
                   2373: 0 10 down2
                   2374: @end example
1.48      anton    2375: 
1.66      anton    2376: Reference: @ref{Counted Loops}.
                   2377: 
1.48      anton    2378: 
                   2379: @node Recursion Tutorial, Leaving definitions or loops Tutorial, Counted loops Tutorial, Tutorial
                   2380: @section Recursion
1.66      anton    2381: @cindex recursion tutorial
1.48      anton    2382: 
                   2383: Usually the name of a definition is not visible in the definition; but
                   2384: earlier definitions are usually visible:
                   2385: 
                   2386: @example
                   2387: 1 0 / . \ "Floating-point unidentified fault" in Gforth on most platforms
                   2388: : / ( n1 n2 -- n )
                   2389:   dup 0= if
                   2390:     -10 throw \ report division by zero
                   2391:   endif
                   2392:   /           \ old version
                   2393: ;
                   2394: 1 0 /
                   2395: @end example
                   2396: 
                   2397: For recursive definitions you can use @code{recursive} (non-standard) or
                   2398: @code{recurse}:
                   2399: 
                   2400: @example
                   2401: : fac1 ( n -- n! ) recursive
                   2402:  dup 0> if
                   2403:    dup 1- fac1 *
                   2404:  else
                   2405:    drop 1
                   2406:  endif ;
                   2407: 7 fac1 .
                   2408: 
                   2409: : fac2 ( n -- n! )
                   2410:  dup 0> if
                   2411:    dup 1- recurse *
                   2412:  else
                   2413:    drop 1
                   2414:  endif ;
                   2415: 8 fac2 .
                   2416: @end example
                   2417: 
                   2418: @assignment
                   2419: Write a recursive definition for computing the nth Fibonacci number.
                   2420: @endassignment
                   2421: 
1.66      anton    2422: Reference (including indirect recursion): @xref{Calls and returns}.
                   2423: 
1.48      anton    2424: 
                   2425: @node Leaving definitions or loops Tutorial, Return Stack Tutorial, Recursion Tutorial, Tutorial
                   2426: @section Leaving definitions or loops
1.66      anton    2427: @cindex leaving definitions, tutorial
                   2428: @cindex leaving loops, tutorial
1.48      anton    2429: 
                   2430: @code{EXIT} exits the current definition right away.  For every counted
                   2431: loop that is left in this way, an @code{UNLOOP} has to be performed
                   2432: before the @code{EXIT}:
                   2433: 
                   2434: @c !! real examples
                   2435: @example
                   2436: : ...
                   2437:  ... u+do
                   2438:    ... if
                   2439:      ... unloop exit
                   2440:    endif
                   2441:    ...
                   2442:  loop
                   2443:  ... ;
                   2444: @end example
                   2445: 
                   2446: @code{LEAVE} leaves the innermost counted loop right away:
                   2447: 
                   2448: @example
                   2449: : ...
                   2450:  ... u+do
                   2451:    ... if
                   2452:      ... leave
                   2453:    endif
                   2454:    ...
                   2455:  loop
                   2456:  ... ;
                   2457: @end example
                   2458: 
1.65      anton    2459: @c !! example
1.48      anton    2460: 
1.66      anton    2461: Reference: @ref{Calls and returns}, @ref{Counted Loops}.
                   2462: 
                   2463: 
1.48      anton    2464: @node Return Stack Tutorial, Memory Tutorial, Leaving definitions or loops Tutorial, Tutorial
                   2465: @section Return Stack
1.66      anton    2466: @cindex return stack tutorial
1.48      anton    2467: 
                   2468: In addition to the data stack Forth also has a second stack, the return
                   2469: stack; most Forth systems store the return addresses of procedure calls
                   2470: there (thus its name).  Programmers can also use this stack:
                   2471: 
                   2472: @example
                   2473: : foo ( n1 n2 -- )
                   2474:  .s
                   2475:  >r .s
1.50      anton    2476:  r@@ .
1.48      anton    2477:  >r .s
1.50      anton    2478:  r@@ .
1.48      anton    2479:  r> .
1.50      anton    2480:  r@@ .
1.48      anton    2481:  r> . ;
                   2482: 1 2 foo
                   2483: @end example
                   2484: 
                   2485: @code{>r} takes an element from the data stack and pushes it onto the
                   2486: return stack; conversely, @code{r>} moves an elementm from the return to
                   2487: the data stack; @code{r@@} pushes a copy of the top of the return stack
                   2488: on the return stack.
                   2489: 
                   2490: Forth programmers usually use the return stack for storing data
                   2491: temporarily, if using the data stack alone would be too complex, and
                   2492: factoring and locals are not an option:
                   2493: 
                   2494: @example
                   2495: : 2swap ( x1 x2 x3 x4 -- x3 x4 x1 x2 )
                   2496:  rot >r rot r> ;
                   2497: @end example
                   2498: 
                   2499: The return address of the definition and the loop control parameters of
                   2500: counted loops usually reside on the return stack, so you have to take
                   2501: all items, that you have pushed on the return stack in a colon
                   2502: definition or counted loop, from the return stack before the definition
                   2503: or loop ends.  You cannot access items that you pushed on the return
                   2504: stack outside some definition or loop within the definition of loop.
                   2505: 
                   2506: If you miscount the return stack items, this usually ends in a crash:
                   2507: 
                   2508: @example
                   2509: : crash ( n -- )
                   2510:   >r ;
                   2511: 5 crash
                   2512: @end example
                   2513: 
                   2514: You cannot mix using locals and using the return stack (according to the
                   2515: standard; Gforth has no problem).  However, they solve the same
                   2516: problems, so this shouldn't be an issue.
                   2517: 
                   2518: @assignment
                   2519: Can you rewrite any of the definitions you wrote until now in a better
                   2520: way using the return stack?
                   2521: @endassignment
                   2522: 
1.66      anton    2523: Reference: @ref{Return stack}.
                   2524: 
1.48      anton    2525: 
                   2526: @node Memory Tutorial, Characters and Strings Tutorial, Return Stack Tutorial, Tutorial
                   2527: @section Memory
1.66      anton    2528: @cindex memory access/allocation tutorial
1.48      anton    2529: 
                   2530: You can create a global variable @code{v} with
                   2531: 
                   2532: @example
                   2533: variable v ( -- addr )
                   2534: @end example
                   2535: 
                   2536: @code{v} pushes the address of a cell in memory on the stack.  This cell
                   2537: was reserved by @code{variable}.  You can use @code{!} (store) to store
                   2538: values into this cell and @code{@@} (fetch) to load the value from the
                   2539: stack into memory:
                   2540: 
                   2541: @example
                   2542: v .
                   2543: 5 v ! .s
1.50      anton    2544: v @@ .
1.48      anton    2545: @end example
                   2546: 
1.65      anton    2547: You can see a raw dump of memory with @code{dump}:
                   2548: 
                   2549: @example
                   2550: v 1 cells .s dump
                   2551: @end example
                   2552: 
                   2553: @code{Cells ( n1 -- n2 )} gives you the number of bytes (or, more
                   2554: generally, address units (aus)) that @code{n1 cells} occupy.  You can
                   2555: also reserve more memory:
1.48      anton    2556: 
                   2557: @example
                   2558: create v2 20 cells allot
1.65      anton    2559: v2 20 cells dump
1.48      anton    2560: @end example
                   2561: 
1.65      anton    2562: creates a word @code{v2} and reserves 20 uninitialized cells; the
                   2563: address pushed by @code{v2} points to the start of these 20 cells.  You
                   2564: can use address arithmetic to access these cells:
1.48      anton    2565: 
                   2566: @example
                   2567: 3 v2 5 cells + !
1.65      anton    2568: v2 20 cells dump
1.48      anton    2569: @end example
                   2570: 
                   2571: You can reserve and initialize memory with @code{,}:
                   2572: 
                   2573: @example
                   2574: create v3
                   2575:   5 , 4 , 3 , 2 , 1 ,
1.50      anton    2576: v3 @@ .
                   2577: v3 cell+ @@ .
                   2578: v3 2 cells + @@ .
1.65      anton    2579: v3 5 cells dump
1.48      anton    2580: @end example
                   2581: 
                   2582: @assignment
                   2583: Write a definition @code{vsum ( addr u -- n )} that computes the sum of
                   2584: @code{u} cells, with the first of these cells at @code{addr}, the next
                   2585: one at @code{addr cell+} etc.
                   2586: @endassignment
                   2587: 
                   2588: You can also reserve memory without creating a new word:
                   2589: 
                   2590: @example
1.60      anton    2591: here 10 cells allot .
                   2592: here .
1.48      anton    2593: @end example
                   2594: 
                   2595: @code{Here} pushes the start address of the memory area.  You should
                   2596: store it somewhere, or you will have a hard time finding the memory area
                   2597: again.
                   2598: 
                   2599: @code{Allot} manages dictionary memory.  The dictionary memory contains
                   2600: the system's data structures for words etc. on Gforth and most other
                   2601: Forth systems.  It is managed like a stack: You can free the memory that
                   2602: you have just @code{allot}ed with
                   2603: 
                   2604: @example
                   2605: -10 cells allot
1.60      anton    2606: here .
1.48      anton    2607: @end example
                   2608: 
                   2609: Note that you cannot do this if you have created a new word in the
                   2610: meantime (because then your @code{allot}ed memory is no longer on the
                   2611: top of the dictionary ``stack'').
                   2612: 
                   2613: Alternatively, you can use @code{allocate} and @code{free} which allow
                   2614: freeing memory in any order:
                   2615: 
                   2616: @example
                   2617: 10 cells allocate throw .s
                   2618: 20 cells allocate throw .s
                   2619: swap
                   2620: free throw
                   2621: free throw
                   2622: @end example
                   2623: 
                   2624: The @code{throw}s deal with errors (e.g., out of memory).
                   2625: 
1.65      anton    2626: And there is also a
                   2627: @uref{http://www.complang.tuwien.ac.at/forth/garbage-collection.zip,
                   2628: garbage collector}, which eliminates the need to @code{free} memory
                   2629: explicitly.
1.48      anton    2630: 
1.66      anton    2631: Reference: @ref{Memory}.
                   2632: 
1.48      anton    2633: 
                   2634: @node Characters and Strings Tutorial, Alignment Tutorial, Memory Tutorial, Tutorial
                   2635: @section Characters and Strings
1.66      anton    2636: @cindex strings tutorial
                   2637: @cindex characters tutorial
1.48      anton    2638: 
                   2639: On the stack characters take up a cell, like numbers.  In memory they
                   2640: have their own size (one 8-bit byte on most systems), and therefore
                   2641: require their own words for memory access:
                   2642: 
                   2643: @example
                   2644: create v4 
                   2645:   104 c, 97 c, 108 c, 108 c, 111 c,
1.50      anton    2646: v4 4 chars + c@@ .
1.65      anton    2647: v4 5 chars dump
1.48      anton    2648: @end example
                   2649: 
                   2650: The preferred representation of strings on the stack is @code{addr
                   2651: u-count}, where @code{addr} is the address of the first character and
                   2652: @code{u-count} is the number of characters in the string.
                   2653: 
                   2654: @example
                   2655: v4 5 type
                   2656: @end example
                   2657: 
                   2658: You get a string constant with
                   2659: 
                   2660: @example
                   2661: s" hello, world" .s
                   2662: type
                   2663: @end example
                   2664: 
                   2665: Make sure you have a space between @code{s"} and the string; @code{s"}
                   2666: is a normal Forth word and must be delimited with white space (try what
                   2667: happens when you remove the space).
                   2668: 
                   2669: However, this interpretive use of @code{s"} is quite restricted: the
                   2670: string exists only until the next call of @code{s"} (some Forth systems
                   2671: keep more than one of these strings, but usually they still have a
1.62      crook    2672: limited lifetime).
1.48      anton    2673: 
                   2674: @example
                   2675: s" hello," s" world" .s
                   2676: type
                   2677: type
                   2678: @end example
                   2679: 
1.62      crook    2680: You can also use @code{s"} in a definition, and the resulting
                   2681: strings then live forever (well, for as long as the definition):
1.48      anton    2682: 
                   2683: @example
                   2684: : foo s" hello," s" world" ;
                   2685: foo .s
                   2686: type
                   2687: type
                   2688: @end example
                   2689: 
                   2690: @assignment
                   2691: @code{Emit ( c -- )} types @code{c} as character (not a number).
                   2692: Implement @code{type ( addr u -- )}.
                   2693: @endassignment
                   2694: 
1.66      anton    2695: Reference: @ref{Memory Blocks}.
                   2696: 
                   2697: 
1.84      pazsan   2698: @node Alignment Tutorial, Files Tutorial, Characters and Strings Tutorial, Tutorial
1.48      anton    2699: @section Alignment
1.66      anton    2700: @cindex alignment tutorial
                   2701: @cindex memory alignment tutorial
1.48      anton    2702: 
                   2703: On many processors cells have to be aligned in memory, if you want to
                   2704: access them with @code{@@} and @code{!} (and even if the processor does
1.62      crook    2705: not require alignment, access to aligned cells is faster).
1.48      anton    2706: 
                   2707: @code{Create} aligns @code{here} (i.e., the place where the next
                   2708: allocation will occur, and that the @code{create}d word points to).
                   2709: Likewise, the memory produced by @code{allocate} starts at an aligned
                   2710: address.  Adding a number of @code{cells} to an aligned address produces
                   2711: another aligned address.
                   2712: 
                   2713: However, address arithmetic involving @code{char+} and @code{chars} can
                   2714: create an address that is not cell-aligned.  @code{Aligned ( addr --
                   2715: a-addr )} produces the next aligned address:
                   2716: 
                   2717: @example
1.50      anton    2718: v3 char+ aligned .s @@ .
                   2719: v3 char+ .s @@ .
1.48      anton    2720: @end example
                   2721: 
                   2722: Similarly, @code{align} advances @code{here} to the next aligned
                   2723: address:
                   2724: 
                   2725: @example
                   2726: create v5 97 c,
                   2727: here .
                   2728: align here .
                   2729: 1000 ,
                   2730: @end example
                   2731: 
                   2732: Note that you should use aligned addresses even if your processor does
                   2733: not require them, if you want your program to be portable.
                   2734: 
1.66      anton    2735: Reference: @ref{Address arithmetic}.
                   2736: 
1.48      anton    2737: 
1.84      pazsan   2738: @node Files Tutorial, Interpretation and Compilation Semantics and Immediacy Tutorial, Alignment Tutorial, Tutorial
                   2739: @section Files
                   2740: @cindex files tutorial
                   2741: 
                   2742: This section gives a short introduction into how to use files inside
                   2743: Forth. It's broken up into five easy steps:
                   2744: 
                   2745: @enumerate 1
                   2746: @item Opened an ASCII text file for input
                   2747: @item Opened a file for output
                   2748: @item Read input file until string matched (or some other condition matched)
                   2749: @item Wrote some lines from input ( modified or not) to output
                   2750: @item Closed the files.
                   2751: @end enumerate
                   2752: 
                   2753: @subsection Open file for input
                   2754: 
                   2755: @example
                   2756: s" foo.in"  r/o open-file throw Value fd-in
                   2757: @end example
                   2758: 
                   2759: @subsection Create file for output
                   2760: 
                   2761: @example
                   2762: s" foo.out" w/o create-file throw Value fd-out
                   2763: @end example
                   2764: 
                   2765: The available file modes are r/o for read-only access, r/w for
                   2766: read-write access, and w/o for write-only access. You could open both
                   2767: files with r/w, too, if you like. All file words return error codes; for
                   2768: most applications, it's best to pass there error codes with @code{throw}
                   2769: to the outer error handler.
                   2770: 
                   2771: If you want words for opening and assigning, define them as follows:
                   2772: 
                   2773: @example
                   2774: 0 Value fd-in
                   2775: 0 Value fd-out
                   2776: : open-input ( addr u -- )  r/o open-file throw to fd-in ;
                   2777: : open-output ( addr u -- )  w/o create-file throw to fd-out ;
                   2778: @end example
                   2779: 
                   2780: Usage example:
                   2781: 
                   2782: @example
                   2783: s" foo.in" open-input
                   2784: s" foo.out" open-output
                   2785: @end example
                   2786: 
                   2787: @subsection Scan file for a particular line
                   2788: 
                   2789: @example
                   2790: 256 Constant max-line
                   2791: Create line-buffer  max-line 2 + allot
                   2792: 
                   2793: : scan-file ( addr u -- )
                   2794:   begin
                   2795:       line-buffer max-line fd-in read-line throw
                   2796:   while
                   2797:          >r 2dup line-buffer r> compare 0=
                   2798:      until
                   2799:   else
                   2800:      drop
                   2801:   then
                   2802:   2drop ;
                   2803: @end example
                   2804: 
                   2805: @code{read-line ( addr u1 fd -- u2 flag ior )} reads up to u1 bytes into
                   2806: the buffer at addr, and returns the number of bytes read, a flag that's
                   2807: true when the end of file is reached, and an error code.
                   2808: 
                   2809: @code{compare ( addr1 u1 addr2 u2 -- n )} compares two strings and
                   2810: returns zero if both strings are equal. It returns a positive number if
                   2811: the first string is lexically greater, a negative if the second string
                   2812: is lexically greater.
                   2813: 
                   2814: We haven't seen this loop here; it has two exits. Since the @code{while}
                   2815: exits with the number of bytes read on the stack, we have to clean up
                   2816: that separately; that's after the @code{else}.
                   2817: 
                   2818: Usage example:
                   2819: 
                   2820: @example
                   2821: s" The text I search is here" scan-file
                   2822: @end example
                   2823: 
                   2824: @subsection Copy input to output
                   2825: 
                   2826: @example
                   2827: : copy-file ( -- )
                   2828:   begin
                   2829:       line-buffer max-line fd-in read-line throw
                   2830:   while
                   2831:       line-buffer swap fd-out write-file throw
                   2832:   repeat ;
                   2833: @end example
                   2834: 
                   2835: @subsection Close files
                   2836: 
                   2837: @example
                   2838: fd-in close-file throw
                   2839: fd-out close-file throw
                   2840: @end example
                   2841: 
                   2842: Likewise, you can put that into definitions, too:
                   2843: 
                   2844: @example
                   2845: : close-input ( -- )  fd-in close-file throw ;
                   2846: : close-output ( -- )  fd-out close-file throw ;
                   2847: @end example
                   2848: 
                   2849: @assignment
                   2850: How could you modify @code{copy-file} so that it copies until a second line is
                   2851: matched? Can you write a program that extracts a section of a text file,
                   2852: given the line that starts and the line that terminates that section?
                   2853: @endassignment
                   2854: 
                   2855: @node Interpretation and Compilation Semantics and Immediacy Tutorial, Execution Tokens Tutorial, Files Tutorial, Tutorial
1.48      anton    2856: @section Interpretation and Compilation Semantics and Immediacy
1.66      anton    2857: @cindex semantics tutorial
                   2858: @cindex interpretation semantics tutorial
                   2859: @cindex compilation semantics tutorial
                   2860: @cindex immediate, tutorial
1.48      anton    2861: 
                   2862: When a word is compiled, it behaves differently from being interpreted.
                   2863: E.g., consider @code{+}:
                   2864: 
                   2865: @example
                   2866: 1 2 + .
                   2867: : foo + ;
                   2868: @end example
                   2869: 
                   2870: These two behaviours are known as compilation and interpretation
                   2871: semantics.  For normal words (e.g., @code{+}), the compilation semantics
                   2872: is to append the interpretation semantics to the currently defined word
                   2873: (@code{foo} in the example above).  I.e., when @code{foo} is executed
                   2874: later, the interpretation semantics of @code{+} (i.e., adding two
                   2875: numbers) will be performed.
                   2876: 
                   2877: However, there are words with non-default compilation semantics, e.g.,
                   2878: the control-flow words like @code{if}.  You can use @code{immediate} to
                   2879: change the compilation semantics of the last defined word to be equal to
                   2880: the interpretation semantics:
                   2881: 
                   2882: @example
                   2883: : [FOO] ( -- )
                   2884:  5 . ; immediate
                   2885: 
                   2886: [FOO]
                   2887: : bar ( -- )
                   2888:   [FOO] ;
                   2889: bar
                   2890: see bar
                   2891: @end example
                   2892: 
                   2893: Two conventions to mark words with non-default compilation semnatics are
                   2894: names with brackets (more frequently used) and to write them all in
                   2895: upper case (less frequently used).
                   2896: 
                   2897: In Gforth (and many other systems) you can also remove the
                   2898: interpretation semantics with @code{compile-only} (the compilation
                   2899: semantics is derived from the original interpretation semantics):
                   2900: 
                   2901: @example
                   2902: : flip ( -- )
                   2903:  6 . ; compile-only \ but not immediate
                   2904: flip
                   2905: 
                   2906: : flop ( -- )
                   2907:  flip ;
                   2908: flop
                   2909: @end example
                   2910: 
                   2911: In this example the interpretation semantics of @code{flop} is equal to
                   2912: the original interpretation semantics of @code{flip}.
                   2913: 
                   2914: The text interpreter has two states: in interpret state, it performs the
                   2915: interpretation semantics of words it encounters; in compile state, it
                   2916: performs the compilation semantics of these words.
                   2917: 
                   2918: Among other things, @code{:} switches into compile state, and @code{;}
                   2919: switches back to interpret state.  They contain the factors @code{]}
                   2920: (switch to compile state) and @code{[} (switch to interpret state), that
                   2921: do nothing but switch the state.
                   2922: 
                   2923: @example
                   2924: : xxx ( -- )
                   2925:   [ 5 . ]
                   2926: ;
                   2927: 
                   2928: xxx
                   2929: see xxx
                   2930: @end example
                   2931: 
                   2932: These brackets are also the source of the naming convention mentioned
                   2933: above.
                   2934: 
1.66      anton    2935: Reference: @ref{Interpretation and Compilation Semantics}.
                   2936: 
1.48      anton    2937: 
                   2938: @node Execution Tokens Tutorial, Exceptions Tutorial, Interpretation and Compilation Semantics and Immediacy Tutorial, Tutorial
                   2939: @section Execution Tokens
1.66      anton    2940: @cindex execution tokens tutorial
                   2941: @cindex XT tutorial
1.48      anton    2942: 
                   2943: @code{' word} gives you the execution token (XT) of a word.  The XT is a
                   2944: cell representing the interpretation semantics of a word.  You can
                   2945: execute this semantics with @code{execute}:
                   2946: 
                   2947: @example
                   2948: ' + .s
                   2949: 1 2 rot execute .
                   2950: @end example
                   2951: 
                   2952: The XT is similar to a function pointer in C.  However, parameter
                   2953: passing through the stack makes it a little more flexible:
                   2954: 
                   2955: @example
                   2956: : map-array ( ... addr u xt -- ... )
1.50      anton    2957: \ executes xt ( ... x -- ... ) for every element of the array starting
                   2958: \ at addr and containing u elements
1.48      anton    2959:   @{ xt @}
                   2960:   cells over + swap ?do
1.50      anton    2961:     i @@ xt execute
1.48      anton    2962:   1 cells +loop ;
                   2963: 
                   2964: create a 3 , 4 , 2 , -1 , 4 ,
                   2965: a 5 ' . map-array .s
                   2966: 0 a 5 ' + map-array .
                   2967: s" max-n" environment? drop .s
                   2968: a 5 ' min map-array .
                   2969: @end example
                   2970: 
                   2971: You can use map-array with the XTs of words that consume one element
                   2972: more than they produce.  In theory you can also use it with other XTs,
                   2973: but the stack effect then depends on the size of the array, which is
                   2974: hard to understand.
                   2975: 
1.51      pazsan   2976: Since XTs are cell-sized, you can store them in memory and manipulate
                   2977: them on the stack like other cells.  You can also compile the XT into a
1.48      anton    2978: word with @code{compile,}:
                   2979: 
                   2980: @example
                   2981: : foo1 ( n1 n2 -- n )
                   2982:    [ ' + compile, ] ;
                   2983: see foo
                   2984: @end example
                   2985: 
                   2986: This is non-standard, because @code{compile,} has no compilation
                   2987: semantics in the standard, but it works in good Forth systems.  For the
                   2988: broken ones, use
                   2989: 
                   2990: @example
                   2991: : [compile,] compile, ; immediate
                   2992: 
                   2993: : foo1 ( n1 n2 -- n )
                   2994:    [ ' + ] [compile,] ;
                   2995: see foo
                   2996: @end example
                   2997: 
                   2998: @code{'} is a word with default compilation semantics; it parses the
                   2999: next word when its interpretation semantics are executed, not during
                   3000: compilation:
                   3001: 
                   3002: @example
                   3003: : foo ( -- xt )
                   3004:   ' ;
                   3005: see foo
                   3006: : bar ( ... "word" -- ... )
                   3007:   ' execute ;
                   3008: see bar
1.60      anton    3009: 1 2 bar + .
1.48      anton    3010: @end example
                   3011: 
                   3012: You often want to parse a word during compilation and compile its XT so
                   3013: it will be pushed on the stack at run-time.  @code{[']} does this:
                   3014: 
                   3015: @example
                   3016: : xt-+ ( -- xt )
                   3017:   ['] + ;
                   3018: see xt-+
                   3019: 1 2 xt-+ execute .
                   3020: @end example
                   3021: 
                   3022: Many programmers tend to see @code{'} and the word it parses as one
                   3023: unit, and expect it to behave like @code{[']} when compiled, and are
                   3024: confused by the actual behaviour.  If you are, just remember that the
                   3025: Forth system just takes @code{'} as one unit and has no idea that it is
                   3026: a parsing word (attempts to convenience programmers in this issue have
                   3027: usually resulted in even worse pitfalls, see
1.66      anton    3028: @uref{http://www.complang.tuwien.ac.at/papers/ertl98.ps.gz,
                   3029: @code{State}-smartness---Why it is evil and How to Exorcise it}).
1.48      anton    3030: 
                   3031: Note that the state of the interpreter does not come into play when
1.51      pazsan   3032: creating and executing XTs.  I.e., even when you execute @code{'} in
1.48      anton    3033: compile state, it still gives you the interpretation semantics.  And
                   3034: whatever that state is, @code{execute} performs the semantics
1.66      anton    3035: represented by the XT (i.e., for XTs produced with @code{'} the
                   3036: interpretation semantics).
                   3037: 
                   3038: Reference: @ref{Tokens for Words}.
1.48      anton    3039: 
                   3040: 
                   3041: @node Exceptions Tutorial, Defining Words Tutorial, Execution Tokens Tutorial, Tutorial
                   3042: @section Exceptions
1.66      anton    3043: @cindex exceptions tutorial
1.48      anton    3044: 
                   3045: @code{throw ( n -- )} causes an exception unless n is zero.
                   3046: 
                   3047: @example
                   3048: 100 throw .s
                   3049: 0 throw .s
                   3050: @end example
                   3051: 
                   3052: @code{catch ( ... xt -- ... n )} behaves similar to @code{execute}, but
                   3053: it catches exceptions and pushes the number of the exception on the
                   3054: stack (or 0, if the xt executed without exception).  If there was an
                   3055: exception, the stacks have the same depth as when entering @code{catch}:
                   3056: 
                   3057: @example
                   3058: .s
                   3059: 3 0 ' / catch .s
                   3060: 3 2 ' / catch .s
                   3061: @end example
                   3062: 
                   3063: @assignment
                   3064: Try the same with @code{execute} instead of @code{catch}.
                   3065: @endassignment
                   3066: 
                   3067: @code{Throw} always jumps to the dynamically next enclosing
                   3068: @code{catch}, even if it has to leave several call levels to achieve
                   3069: this:
                   3070: 
                   3071: @example
                   3072: : foo 100 throw ;
                   3073: : foo1 foo ." after foo" ;
1.51      pazsan   3074: : bar ['] foo1 catch ;
1.60      anton    3075: bar .
1.48      anton    3076: @end example
                   3077: 
                   3078: It is often important to restore a value upon leaving a definition, even
                   3079: if the definition is left through an exception.  You can ensure this
                   3080: like this:
                   3081: 
                   3082: @example
                   3083: : ...
                   3084:    save-x
1.51      pazsan   3085:    ['] word-changing-x catch ( ... n )
1.48      anton    3086:    restore-x
                   3087:    ( ... n ) throw ;
                   3088: @end example
                   3089: 
1.55      anton    3090: Gforth provides an alternative syntax in addition to @code{catch}:
1.48      anton    3091: @code{try ... recover ... endtry}.  If the code between @code{try} and
                   3092: @code{recover} has an exception, the stack depths are restored, the
                   3093: exception number is pushed on the stack, and the code between
                   3094: @code{recover} and @code{endtry} is performed.  E.g., the definition for
                   3095: @code{catch} is
                   3096: 
                   3097: @example
                   3098: : catch ( x1 .. xn xt -- y1 .. ym 0 / z1 .. zn error ) \ exception
                   3099:   try
                   3100:     execute 0
                   3101:   recover
                   3102:     nip
                   3103:   endtry ;
                   3104: @end example
                   3105: 
                   3106: The equivalent to the restoration code above is
                   3107: 
                   3108: @example
                   3109: : ...
                   3110:   save-x
                   3111:   try
                   3112:     word-changing-x
                   3113:   end-try
                   3114:   restore-x
                   3115:   throw ;
                   3116: @end example
                   3117: 
                   3118: As you can see, the @code{recover} part is optional.
                   3119: 
1.66      anton    3120: Reference: @ref{Exception Handling}.
                   3121: 
1.48      anton    3122: 
                   3123: @node Defining Words Tutorial, Arrays and Records Tutorial, Exceptions Tutorial, Tutorial
                   3124: @section Defining Words
1.66      anton    3125: @cindex defining words tutorial
                   3126: @cindex does> tutorial
                   3127: @cindex create...does> tutorial
                   3128: 
                   3129: @c before semantics?
1.48      anton    3130: 
                   3131: @code{:}, @code{create}, and @code{variable} are definition words: They
                   3132: define other words.  @code{Constant} is another definition word:
                   3133: 
                   3134: @example
                   3135: 5 constant foo
                   3136: foo .
                   3137: @end example
                   3138: 
                   3139: You can also use the prefixes @code{2} (double-cell) and @code{f}
                   3140: (floating point) with @code{variable} and @code{constant}.
                   3141: 
                   3142: You can also define your own defining words.  E.g.:
                   3143: 
                   3144: @example
                   3145: : variable ( "name" -- )
                   3146:   create 0 , ;
                   3147: @end example
                   3148: 
                   3149: You can also define defining words that create words that do something
                   3150: other than just producing their address:
                   3151: 
                   3152: @example
                   3153: : constant ( n "name" -- )
                   3154:   create ,
                   3155: does> ( -- n )
1.50      anton    3156:   ( addr ) @@ ;
1.48      anton    3157: 
                   3158: 5 constant foo
                   3159: foo .
                   3160: @end example
                   3161: 
                   3162: The definition of @code{constant} above ends at the @code{does>}; i.e.,
                   3163: @code{does>} replaces @code{;}, but it also does something else: It
                   3164: changes the last defined word such that it pushes the address of the
                   3165: body of the word and then performs the code after the @code{does>}
                   3166: whenever it is called.
                   3167: 
                   3168: In the example above, @code{constant} uses @code{,} to store 5 into the
                   3169: body of @code{foo}.  When @code{foo} executes, it pushes the address of
                   3170: the body onto the stack, then (in the code after the @code{does>})
                   3171: fetches the 5 from there.
                   3172: 
                   3173: The stack comment near the @code{does>} reflects the stack effect of the
                   3174: defined word, not the stack effect of the code after the @code{does>}
                   3175: (the difference is that the code expects the address of the body that
                   3176: the stack comment does not show).
                   3177: 
                   3178: You can use these definition words to do factoring in cases that involve
                   3179: (other) definition words.  E.g., a field offset is always added to an
                   3180: address.  Instead of defining
                   3181: 
                   3182: @example
                   3183: 2 cells constant offset-field1
                   3184: @end example
                   3185: 
                   3186: and using this like
                   3187: 
                   3188: @example
                   3189: ( addr ) offset-field1 +
                   3190: @end example
                   3191: 
                   3192: you can define a definition word
                   3193: 
                   3194: @example
                   3195: : simple-field ( n "name" -- )
                   3196:   create ,
                   3197: does> ( n1 -- n1+n )
1.50      anton    3198:   ( addr ) @@ + ;
1.48      anton    3199: @end example
1.21      crook    3200: 
1.48      anton    3201: Definition and use of field offsets now look like this:
1.21      crook    3202: 
1.48      anton    3203: @example
                   3204: 2 cells simple-field field1
1.60      anton    3205: create mystruct 4 cells allot
                   3206: mystruct .s field1 .s drop
1.48      anton    3207: @end example
1.21      crook    3208: 
1.48      anton    3209: If you want to do something with the word without performing the code
                   3210: after the @code{does>}, you can access the body of a @code{create}d word
                   3211: with @code{>body ( xt -- addr )}:
1.21      crook    3212: 
1.48      anton    3213: @example
                   3214: : value ( n "name" -- )
                   3215:   create ,
                   3216: does> ( -- n1 )
1.50      anton    3217:   @@ ;
1.48      anton    3218: : to ( n "name" -- )
                   3219:   ' >body ! ;
1.21      crook    3220: 
1.48      anton    3221: 5 value foo
                   3222: foo .
                   3223: 7 to foo
                   3224: foo .
                   3225: @end example
1.21      crook    3226: 
1.48      anton    3227: @assignment
                   3228: Define @code{defer ( "name" -- )}, which creates a word that stores an
                   3229: XT (at the start the XT of @code{abort}), and upon execution
                   3230: @code{execute}s the XT.  Define @code{is ( xt "name" -- )} that stores
                   3231: @code{xt} into @code{name}, a word defined with @code{defer}.  Indirect
                   3232: recursion is one application of @code{defer}.
                   3233: @endassignment
1.29      crook    3234: 
1.66      anton    3235: Reference: @ref{User-defined Defining Words}.
                   3236: 
                   3237: 
1.48      anton    3238: @node Arrays and Records Tutorial, POSTPONE Tutorial, Defining Words Tutorial, Tutorial
                   3239: @section Arrays and Records
1.66      anton    3240: @cindex arrays tutorial
                   3241: @cindex records tutorial
                   3242: @cindex structs tutorial
1.29      crook    3243: 
1.48      anton    3244: Forth has no standard words for defining data structures such as arrays
                   3245: and records (structs in C terminology), but you can build them yourself
                   3246: based on address arithmetic.  You can also define words for defining
                   3247: arrays and records (@pxref{Defining Words Tutorial,, Defining Words}).
1.29      crook    3248: 
1.48      anton    3249: One of the first projects a Forth newcomer sets out upon when learning
                   3250: about defining words is an array defining word (possibly for
                   3251: n-dimensional arrays).  Go ahead and do it, I did it, too; you will
                   3252: learn something from it.  However, don't be disappointed when you later
                   3253: learn that you have little use for these words (inappropriate use would
                   3254: be even worse).  I have not yet found a set of useful array words yet;
                   3255: the needs are just too diverse, and named, global arrays (the result of
                   3256: naive use of defining words) are often not flexible enough (e.g.,
1.66      anton    3257: consider how to pass them as parameters).  Another such project is a set
                   3258: of words to help dealing with strings.
1.29      crook    3259: 
1.48      anton    3260: On the other hand, there is a useful set of record words, and it has
                   3261: been defined in @file{compat/struct.fs}; these words are predefined in
                   3262: Gforth.  They are explained in depth elsewhere in this manual (see
                   3263: @pxref{Structures}).  The @code{simple-field} example above is
                   3264: simplified variant of fields in this package.
1.21      crook    3265: 
                   3266: 
1.48      anton    3267: @node POSTPONE Tutorial, Literal Tutorial, Arrays and Records Tutorial, Tutorial
                   3268: @section @code{POSTPONE}
1.66      anton    3269: @cindex postpone tutorial
1.21      crook    3270: 
1.48      anton    3271: You can compile the compilation semantics (instead of compiling the
                   3272: interpretation semantics) of a word with @code{POSTPONE}:
1.21      crook    3273: 
1.48      anton    3274: @example
                   3275: : MY-+ ( Compilation: -- ; Run-time of compiled code: n1 n2 -- n )
1.51      pazsan   3276:  POSTPONE + ; immediate
1.48      anton    3277: : foo ( n1 n2 -- n )
                   3278:  MY-+ ;
                   3279: 1 2 foo .
                   3280: see foo
                   3281: @end example
1.21      crook    3282: 
1.48      anton    3283: During the definition of @code{foo} the text interpreter performs the
                   3284: compilation semantics of @code{MY-+}, which performs the compilation
                   3285: semantics of @code{+}, i.e., it compiles @code{+} into @code{foo}.
                   3286: 
                   3287: This example also displays separate stack comments for the compilation
                   3288: semantics and for the stack effect of the compiled code.  For words with
                   3289: default compilation semantics these stack effects are usually not
                   3290: displayed; the stack effect of the compilation semantics is always
                   3291: @code{( -- )} for these words, the stack effect for the compiled code is
                   3292: the stack effect of the interpretation semantics.
                   3293: 
                   3294: Note that the state of the interpreter does not come into play when
                   3295: performing the compilation semantics in this way.  You can also perform
                   3296: it interpretively, e.g.:
                   3297: 
                   3298: @example
                   3299: : foo2 ( n1 n2 -- n )
                   3300:  [ MY-+ ] ;
                   3301: 1 2 foo .
                   3302: see foo
                   3303: @end example
1.21      crook    3304: 
1.48      anton    3305: However, there are some broken Forth systems where this does not always
1.62      crook    3306: work, and therefore this practice was been declared non-standard in
1.48      anton    3307: 1999.
                   3308: @c !! repair.fs
                   3309: 
                   3310: Here is another example for using @code{POSTPONE}:
1.44      crook    3311: 
1.48      anton    3312: @example
                   3313: : MY-- ( Compilation: -- ; Run-time of compiled code: n1 n2 -- n )
                   3314:  POSTPONE negate POSTPONE + ; immediate compile-only
                   3315: : bar ( n1 n2 -- n )
                   3316:   MY-- ;
                   3317: 2 1 bar .
                   3318: see bar
                   3319: @end example
1.21      crook    3320: 
1.48      anton    3321: You can define @code{ENDIF} in this way:
1.21      crook    3322: 
1.48      anton    3323: @example
                   3324: : ENDIF ( Compilation: orig -- )
                   3325:   POSTPONE then ; immediate
                   3326: @end example
1.21      crook    3327: 
1.48      anton    3328: @assignment
                   3329: Write @code{MY-2DUP} that has compilation semantics equivalent to
                   3330: @code{2dup}, but compiles @code{over over}.
                   3331: @endassignment
1.29      crook    3332: 
1.66      anton    3333: @c !! @xref{Macros} for reference
                   3334: 
                   3335: 
1.48      anton    3336: @node Literal Tutorial, Advanced macros Tutorial, POSTPONE Tutorial, Tutorial
                   3337: @section @code{Literal}
1.66      anton    3338: @cindex literal tutorial
1.29      crook    3339: 
1.48      anton    3340: You cannot @code{POSTPONE} numbers:
1.21      crook    3341: 
1.48      anton    3342: @example
                   3343: : [FOO] POSTPONE 500 ; immediate
1.21      crook    3344: @end example
                   3345: 
1.48      anton    3346: Instead, you can use @code{LITERAL (compilation: n --; run-time: -- n )}:
1.29      crook    3347: 
1.48      anton    3348: @example
                   3349: : [FOO] ( compilation: --; run-time: -- n )
                   3350:   500 POSTPONE literal ; immediate
1.29      crook    3351: 
1.60      anton    3352: : flip [FOO] ;
1.48      anton    3353: flip .
                   3354: see flip
                   3355: @end example
1.29      crook    3356: 
1.48      anton    3357: @code{LITERAL} consumes a number at compile-time (when it's compilation
                   3358: semantics are executed) and pushes it at run-time (when the code it
                   3359: compiled is executed).  A frequent use of @code{LITERAL} is to compile a
                   3360: number computed at compile time into the current word:
1.29      crook    3361: 
1.48      anton    3362: @example
                   3363: : bar ( -- n )
                   3364:   [ 2 2 + ] literal ;
                   3365: see bar
                   3366: @end example
1.29      crook    3367: 
1.48      anton    3368: @assignment
                   3369: Write @code{]L} which allows writing the example above as @code{: bar (
                   3370: -- n ) [ 2 2 + ]L ;}
                   3371: @endassignment
                   3372: 
1.66      anton    3373: @c !! @xref{Macros} for reference
                   3374: 
1.48      anton    3375: 
                   3376: @node Advanced macros Tutorial, Compilation Tokens Tutorial, Literal Tutorial, Tutorial
                   3377: @section Advanced macros
1.66      anton    3378: @cindex macros, advanced tutorial
                   3379: @cindex run-time code generation, tutorial
1.48      anton    3380: 
1.66      anton    3381: Reconsider @code{map-array} from @ref{Execution Tokens Tutorial,,
                   3382: Execution Tokens}.  It frequently performs @code{execute}, a relatively
                   3383: expensive operation in some Forth implementations.  You can use
1.48      anton    3384: @code{compile,} and @code{POSTPONE} to eliminate these @code{execute}s
                   3385: and produce a word that contains the word to be performed directly:
                   3386: 
                   3387: @c use ]] ... [[
                   3388: @example
                   3389: : compile-map-array ( compilation: xt -- ; run-time: ... addr u -- ... )
                   3390: \ at run-time, execute xt ( ... x -- ... ) for each element of the
                   3391: \ array beginning at addr and containing u elements
                   3392:   @{ xt @}
                   3393:   POSTPONE cells POSTPONE over POSTPONE + POSTPONE swap POSTPONE ?do
1.50      anton    3394:     POSTPONE i POSTPONE @@ xt compile,
1.48      anton    3395:   1 cells POSTPONE literal POSTPONE +loop ;
                   3396: 
                   3397: : sum-array ( addr u -- n )
                   3398:  0 rot rot [ ' + compile-map-array ] ;
                   3399: see sum-array
                   3400: a 5 sum-array .
                   3401: @end example
                   3402: 
                   3403: You can use the full power of Forth for generating the code; here's an
                   3404: example where the code is generated in a loop:
                   3405: 
                   3406: @example
                   3407: : compile-vmul-step ( compilation: n --; run-time: n1 addr1 -- n2 addr2 )
                   3408: \ n2=n1+(addr1)*n, addr2=addr1+cell
1.50      anton    3409:   POSTPONE tuck POSTPONE @@
1.48      anton    3410:   POSTPONE literal POSTPONE * POSTPONE +
                   3411:   POSTPONE swap POSTPONE cell+ ;
                   3412: 
                   3413: : compile-vmul ( compilation: addr1 u -- ; run-time: addr2 -- n )
1.51      pazsan   3414: \ n=v1*v2 (inner product), where the v_i are represented as addr_i u
1.48      anton    3415:   0 postpone literal postpone swap
                   3416:   [ ' compile-vmul-step compile-map-array ]
                   3417:   postpone drop ;
                   3418: see compile-vmul
                   3419: 
                   3420: : a-vmul ( addr -- n )
1.51      pazsan   3421: \ n=a*v, where v is a vector that's as long as a and starts at addr
1.48      anton    3422:  [ a 5 compile-vmul ] ;
                   3423: see a-vmul
                   3424: a a-vmul .
                   3425: @end example
                   3426: 
                   3427: This example uses @code{compile-map-array} to show off, but you could
1.66      anton    3428: also use @code{map-array} instead (try it now!).
1.48      anton    3429: 
                   3430: You can use this technique for efficient multiplication of large
                   3431: matrices.  In matrix multiplication, you multiply every line of one
                   3432: matrix with every column of the other matrix.  You can generate the code
                   3433: for one line once, and use it for every column.  The only downside of
                   3434: this technique is that it is cumbersome to recover the memory consumed
                   3435: by the generated code when you are done (and in more complicated cases
                   3436: it is not possible portably).
                   3437: 
1.66      anton    3438: @c !! @xref{Macros} for reference
                   3439: 
                   3440: 
1.48      anton    3441: @node Compilation Tokens Tutorial, Wordlists and Search Order Tutorial, Advanced macros Tutorial, Tutorial
                   3442: @section Compilation Tokens
1.66      anton    3443: @cindex compilation tokens, tutorial
                   3444: @cindex CT, tutorial
1.48      anton    3445: 
                   3446: This section is Gforth-specific.  You can skip it.
                   3447: 
                   3448: @code{' word compile,} compiles the interpretation semantics.  For words
                   3449: with default compilation semantics this is the same as performing the
                   3450: compilation semantics.  To represent the compilation semantics of other
                   3451: words (e.g., words like @code{if} that have no interpretation
                   3452: semantics), Gforth has the concept of a compilation token (CT,
                   3453: consisting of two cells), and words @code{comp'} and @code{[comp']}.
                   3454: You can perform the compilation semantics represented by a CT with
                   3455: @code{execute}:
1.29      crook    3456: 
1.48      anton    3457: @example
                   3458: : foo2 ( n1 n2 -- n )
                   3459:    [ comp' + execute ] ;
                   3460: see foo
                   3461: @end example
1.29      crook    3462: 
1.48      anton    3463: You can compile the compilation semantics represented by a CT with
                   3464: @code{postpone,}:
1.30      anton    3465: 
1.48      anton    3466: @example
                   3467: : foo3 ( -- )
                   3468:   [ comp' + postpone, ] ;
                   3469: see foo3
                   3470: @end example
1.30      anton    3471: 
1.51      pazsan   3472: @code{[ comp' word postpone, ]} is equivalent to @code{POSTPONE word}.
1.48      anton    3473: @code{comp'} is particularly useful for words that have no
                   3474: interpretation semantics:
1.29      crook    3475: 
1.30      anton    3476: @example
1.48      anton    3477: ' if
1.60      anton    3478: comp' if .s 2drop
1.30      anton    3479: @end example
                   3480: 
1.66      anton    3481: Reference: @ref{Tokens for Words}.
                   3482: 
1.29      crook    3483: 
1.48      anton    3484: @node Wordlists and Search Order Tutorial,  , Compilation Tokens Tutorial, Tutorial
                   3485: @section Wordlists and Search Order
1.66      anton    3486: @cindex wordlists tutorial
                   3487: @cindex search order, tutorial
1.48      anton    3488: 
                   3489: The dictionary is not just a memory area that allows you to allocate
                   3490: memory with @code{allot}, it also contains the Forth words, arranged in
                   3491: several wordlists.  When searching for a word in a wordlist,
                   3492: conceptually you start searching at the youngest and proceed towards
                   3493: older words (in reality most systems nowadays use hash-tables); i.e., if
                   3494: you define a word with the same name as an older word, the new word
                   3495: shadows the older word.
                   3496: 
                   3497: Which wordlists are searched in which order is determined by the search
                   3498: order.  You can display the search order with @code{order}.  It displays
                   3499: first the search order, starting with the wordlist searched first, then
                   3500: it displays the wordlist that will contain newly defined words.
1.21      crook    3501: 
1.48      anton    3502: You can create a new, empty wordlist with @code{wordlist ( -- wid )}:
1.21      crook    3503: 
1.48      anton    3504: @example
                   3505: wordlist constant mywords
                   3506: @end example
1.21      crook    3507: 
1.48      anton    3508: @code{Set-current ( wid -- )} sets the wordlist that will contain newly
                   3509: defined words (the @emph{current} wordlist):
1.21      crook    3510: 
1.48      anton    3511: @example
                   3512: mywords set-current
                   3513: order
                   3514: @end example
1.26      crook    3515: 
1.48      anton    3516: Gforth does not display a name for the wordlist in @code{mywords}
                   3517: because this wordlist was created anonymously with @code{wordlist}.
1.21      crook    3518: 
1.48      anton    3519: You can get the current wordlist with @code{get-current ( -- wid)}.  If
                   3520: you want to put something into a specific wordlist without overall
                   3521: effect on the current wordlist, this typically looks like this:
1.21      crook    3522: 
1.48      anton    3523: @example
                   3524: get-current mywords set-current ( wid )
                   3525: create someword
                   3526: ( wid ) set-current
                   3527: @end example
1.21      crook    3528: 
1.48      anton    3529: You can write the search order with @code{set-order ( wid1 .. widn n --
                   3530: )} and read it with @code{get-order ( -- wid1 .. widn n )}.  The first
                   3531: searched wordlist is topmost.
1.21      crook    3532: 
1.48      anton    3533: @example
                   3534: get-order mywords swap 1+ set-order
                   3535: order
                   3536: @end example
1.21      crook    3537: 
1.48      anton    3538: Yes, the order of wordlists in the output of @code{order} is reversed
                   3539: from stack comments and the output of @code{.s} and thus unintuitive.
1.21      crook    3540: 
1.48      anton    3541: @assignment
                   3542: Define @code{>order ( wid -- )} with adds @code{wid} as first searched
                   3543: wordlist to the search order.  Define @code{previous ( -- )}, which
                   3544: removes the first searched wordlist from the search order.  Experiment
                   3545: with boundary conditions (you will see some crashes or situations that
                   3546: are hard or impossible to leave).
                   3547: @endassignment
1.21      crook    3548: 
1.48      anton    3549: The search order is a powerful foundation for providing features similar
                   3550: to Modula-2 modules and C++ namespaces.  However, trying to modularize
                   3551: programs in this way has disadvantages for debugging and reuse/factoring
                   3552: that overcome the advantages in my experience (I don't do huge projects,
1.55      anton    3553: though).  These disadvantages are not so clear in other
1.82      anton    3554: languages/programming environments, because these languages are not so
1.48      anton    3555: strong in debugging and reuse.
1.21      crook    3556: 
1.66      anton    3557: @c !! example
                   3558: 
                   3559: Reference: @ref{Word Lists}.
1.21      crook    3560: 
1.29      crook    3561: @c ******************************************************************
1.48      anton    3562: @node Introduction, Words, Tutorial, Top
1.29      crook    3563: @comment node-name,     next,           previous, up
                   3564: @chapter An Introduction to ANS Forth
                   3565: @cindex Forth - an introduction
1.21      crook    3566: 
1.83      anton    3567: The difference of this chapter from the Tutorial (@pxref{Tutorial}) is
                   3568: that it is slower-paced in its examples, but uses them to dive deep into
                   3569: explaining Forth internals (not covered by the Tutorial).  Apart from
                   3570: that, this chapter covers far less material.  It is suitable for reading
                   3571: without using a computer.
                   3572: 
1.29      crook    3573: The primary purpose of this manual is to document Gforth. However, since
                   3574: Forth is not a widely-known language and there is a lack of up-to-date
                   3575: teaching material, it seems worthwhile to provide some introductory
1.49      anton    3576: material.  For other sources of Forth-related
                   3577: information, see @ref{Forth-related information}.
1.21      crook    3578: 
1.29      crook    3579: The examples in this section should work on any ANS Forth; the
                   3580: output shown was produced using Gforth. Each example attempts to
                   3581: reproduce the exact output that Gforth produces. If you try out the
                   3582: examples (and you should), what you should type is shown @kbd{like this}
                   3583: and Gforth's response is shown @code{like this}. The single exception is
1.30      anton    3584: that, where the example shows @key{RET} it means that you should
1.29      crook    3585: press the ``carriage return'' key. Unfortunately, some output formats for
                   3586: this manual cannot show the difference between @kbd{this} and
                   3587: @code{this} which will make trying out the examples harder (but not
                   3588: impossible).
1.21      crook    3589: 
1.29      crook    3590: Forth is an unusual language. It provides an interactive development
                   3591: environment which includes both an interpreter and compiler. Forth
                   3592: programming style encourages you to break a problem down into many
                   3593: @cindex factoring
                   3594: small fragments (@dfn{factoring}), and then to develop and test each
                   3595: fragment interactively. Forth advocates assert that breaking the
                   3596: edit-compile-test cycle used by conventional programming languages can
                   3597: lead to great productivity improvements.
1.21      crook    3598: 
1.29      crook    3599: @menu
1.67      anton    3600: * Introducing the Text Interpreter::  
                   3601: * Stacks and Postfix notation::  
                   3602: * Your first definition::       
                   3603: * How does that work?::         
                   3604: * Forth is written in Forth::   
                   3605: * Review - elements of a Forth system::  
                   3606: * Where to go next::            
                   3607: * Exercises::                   
1.29      crook    3608: @end menu
1.21      crook    3609: 
1.29      crook    3610: @comment ----------------------------------------------
                   3611: @node Introducing the Text Interpreter, Stacks and Postfix notation, Introduction, Introduction
                   3612: @section Introducing the Text Interpreter
                   3613: @cindex text interpreter
                   3614: @cindex outer interpreter
1.21      crook    3615: 
1.30      anton    3616: @c IMO this is too detailed and the pace is too slow for
                   3617: @c an introduction.  If you know German, take a look at
                   3618: @c http://www.complang.tuwien.ac.at/anton/lvas/skriptum-stack.html 
                   3619: @c to see how I do it - anton 
                   3620: 
1.44      crook    3621: @c nac-> Where I have accepted your comments 100% and modified the text
                   3622: @c accordingly, I have deleted your comments. Elsewhere I have added a
                   3623: @c response like this to attempt to rationalise what I have done. Of
                   3624: @c course, this is a very clumsy mechanism for something that would be
                   3625: @c done far more efficiently over a beer. Please delete any dialogue
                   3626: @c you consider closed.
                   3627: 
1.29      crook    3628: When you invoke the Forth image, you will see a startup banner printed
                   3629: and nothing else (if you have Gforth installed on your system, try
1.30      anton    3630: invoking it now, by typing @kbd{gforth@key{RET}}). Forth is now running
1.29      crook    3631: its command line interpreter, which is called the @dfn{Text Interpreter}
                   3632: (also known as the @dfn{Outer Interpreter}).  (You will learn a lot
1.49      anton    3633: about the text interpreter as you read through this chapter, for more
                   3634: detail @pxref{The Text Interpreter}).
1.21      crook    3635: 
1.29      crook    3636: Although it's not obvious, Forth is actually waiting for your
1.30      anton    3637: input. Type a number and press the @key{RET} key:
1.21      crook    3638: 
1.26      crook    3639: @example
1.30      anton    3640: @kbd{45@key{RET}}  ok
1.26      crook    3641: @end example
1.21      crook    3642: 
1.29      crook    3643: Rather than give you a prompt to invite you to input something, the text
                   3644: interpreter prints a status message @i{after} it has processed a line
                   3645: of input. The status message in this case (``@code{ ok}'' followed by
                   3646: carriage-return) indicates that the text interpreter was able to process
                   3647: all of your input successfully. Now type something illegal:
                   3648: 
                   3649: @example
1.30      anton    3650: @kbd{qwer341@key{RET}}
1.29      crook    3651: :1: Undefined word
                   3652: qwer341
                   3653: ^^^^^^^
                   3654: $400D2BA8 Bounce
                   3655: $400DBDA8 no.extensions
                   3656: @end example
1.23      crook    3657: 
1.29      crook    3658: The exact text, other than the ``Undefined word'' may differ slightly on
                   3659: your system, but the effect is the same; when the text interpreter
                   3660: detects an error, it discards any remaining text on a line, resets
1.49      anton    3661: certain internal state and prints an error message. For a detailed description of error messages see @ref{Error
                   3662: messages}.
1.23      crook    3663: 
1.29      crook    3664: The text interpreter waits for you to press carriage-return, and then
                   3665: processes your input line. Starting at the beginning of the line, it
                   3666: breaks the line into groups of characters separated by spaces. For each
                   3667: group of characters in turn, it makes two attempts to do something:
1.23      crook    3668: 
1.29      crook    3669: @itemize @bullet
                   3670: @item
1.44      crook    3671: @cindex name dictionary
1.29      crook    3672: It tries to treat it as a command. It does this by searching a @dfn{name
                   3673: dictionary}. If the group of characters matches an entry in the name
                   3674: dictionary, the name dictionary provides the text interpreter with
                   3675: information that allows the text interpreter perform some actions. In
                   3676: Forth jargon, we say that the group
                   3677: @cindex word
                   3678: @cindex definition
                   3679: @cindex execution token
                   3680: @cindex xt
                   3681: of characters names a @dfn{word}, that the dictionary search returns an
                   3682: @dfn{execution token (xt)} corresponding to the @dfn{definition} of the
                   3683: word, and that the text interpreter executes the xt. Often, the terms
                   3684: @dfn{word} and @dfn{definition} are used interchangeably.
                   3685: @item
                   3686: If the text interpreter fails to find a match in the name dictionary, it
                   3687: tries to treat the group of characters as a number in the current number
                   3688: base (when you start up Forth, the current number base is base 10). If
                   3689: the group of characters legitimately represents a number, the text
                   3690: interpreter pushes the number onto a stack (we'll learn more about that
                   3691: in the next section).
                   3692: @end itemize
1.23      crook    3693: 
1.29      crook    3694: If the text interpreter is unable to do either of these things with any
                   3695: group of characters, it discards the group of characters and the rest of
                   3696: the line, then prints an error message. If the text interpreter reaches
                   3697: the end of the line without error, it prints the status message ``@code{ ok}''
                   3698: followed by carriage-return.
1.21      crook    3699: 
1.29      crook    3700: This is the simplest command we can give to the text interpreter:
1.23      crook    3701: 
                   3702: @example
1.30      anton    3703: @key{RET}  ok
1.23      crook    3704: @end example
1.21      crook    3705: 
1.29      crook    3706: The text interpreter did everything we asked it to do (nothing) without
                   3707: an error, so it said that everything is ``@code{ ok}''. Try a slightly longer
                   3708: command:
1.21      crook    3709: 
1.23      crook    3710: @example
1.30      anton    3711: @kbd{12 dup fred dup@key{RET}}
1.29      crook    3712: :1: Undefined word
                   3713: 12 dup fred dup
                   3714:        ^^^^
                   3715: $400D2BA8 Bounce
                   3716: $400DBDA8 no.extensions
1.23      crook    3717: @end example
1.21      crook    3718: 
1.29      crook    3719: When you press the carriage-return key, the text interpreter starts to
                   3720: work its way along the line:
1.21      crook    3721: 
1.29      crook    3722: @itemize @bullet
                   3723: @item
                   3724: When it gets to the space after the @code{2}, it takes the group of
                   3725: characters @code{12} and looks them up in the name
                   3726: dictionary@footnote{We can't tell if it found them or not, but assume
                   3727: for now that it did not}. There is no match for this group of characters
                   3728: in the name dictionary, so it tries to treat them as a number. It is
                   3729: able to do this successfully, so it puts the number, 12, ``on the stack''
                   3730: (whatever that means).
                   3731: @item
                   3732: The text interpreter resumes scanning the line and gets the next group
                   3733: of characters, @code{dup}. It looks it up in the name dictionary and
                   3734: (you'll have to take my word for this) finds it, and executes the word
                   3735: @code{dup} (whatever that means).
                   3736: @item
                   3737: Once again, the text interpreter resumes scanning the line and gets the
                   3738: group of characters @code{fred}. It looks them up in the name
                   3739: dictionary, but can't find them. It tries to treat them as a number, but
                   3740: they don't represent any legal number.
                   3741: @end itemize
1.21      crook    3742: 
1.29      crook    3743: At this point, the text interpreter gives up and prints an error
                   3744: message. The error message shows exactly how far the text interpreter
                   3745: got in processing the line. In particular, it shows that the text
                   3746: interpreter made no attempt to do anything with the final character
                   3747: group, @code{dup}, even though we have good reason to believe that the
                   3748: text interpreter would have no problem looking that word up and
                   3749: executing it a second time.
1.21      crook    3750: 
                   3751: 
1.29      crook    3752: @comment ----------------------------------------------
                   3753: @node Stacks and Postfix notation, Your first definition, Introducing the Text Interpreter, Introduction
                   3754: @section Stacks, postfix notation and parameter passing
                   3755: @cindex text interpreter
                   3756: @cindex outer interpreter
1.21      crook    3757: 
1.29      crook    3758: In procedural programming languages (like C and Pascal), the
                   3759: building-block of programs is the @dfn{function} or @dfn{procedure}. These
                   3760: functions or procedures are called with @dfn{explicit parameters}. For
                   3761: example, in C we might write:
1.21      crook    3762: 
1.23      crook    3763: @example
1.29      crook    3764: total = total + new_volume(length,height,depth);
1.23      crook    3765: @end example
1.21      crook    3766: 
1.23      crook    3767: @noindent
1.29      crook    3768: where new_volume is a function-call to another piece of code, and total,
                   3769: length, height and depth are all variables. length, height and depth are
                   3770: parameters to the function-call.
1.21      crook    3771: 
1.29      crook    3772: In Forth, the equivalent of the function or procedure is the
                   3773: @dfn{definition} and parameters are implicitly passed between
                   3774: definitions using a shared stack that is visible to the
                   3775: programmer. Although Forth does support variables, the existence of the
                   3776: stack means that they are used far less often than in most other
                   3777: programming languages. When the text interpreter encounters a number, it
                   3778: will place (@dfn{push}) it on the stack. There are several stacks (the
1.30      anton    3779: actual number is implementation-dependent ...) and the particular stack
1.29      crook    3780: used for any operation is implied unambiguously by the operation being
                   3781: performed. The stack used for all integer operations is called the @dfn{data
                   3782: stack} and, since this is the stack used most commonly, references to
                   3783: ``the data stack'' are often abbreviated to ``the stack''.
1.21      crook    3784: 
1.29      crook    3785: The stacks have a last-in, first-out (LIFO) organisation. If you type:
1.21      crook    3786: 
1.23      crook    3787: @example
1.30      anton    3788: @kbd{1 2 3@key{RET}}  ok
1.23      crook    3789: @end example
1.21      crook    3790: 
1.29      crook    3791: Then this instructs the text interpreter to placed three numbers on the
                   3792: (data) stack. An analogy for the behaviour of the stack is to take a
                   3793: pack of playing cards and deal out the ace (1), 2 and 3 into a pile on
                   3794: the table. The 3 was the last card onto the pile (``last-in'') and if
                   3795: you take a card off the pile then, unless you're prepared to fiddle a
                   3796: bit, the card that you take off will be the 3 (``first-out''). The
                   3797: number that will be first-out of the stack is called the @dfn{top of
                   3798: stack}, which
                   3799: @cindex TOS definition
                   3800: is often abbreviated to @dfn{TOS}.
1.21      crook    3801: 
1.29      crook    3802: To understand how parameters are passed in Forth, consider the
                   3803: behaviour of the definition @code{+} (pronounced ``plus''). You will not
                   3804: be surprised to learn that this definition performs addition. More
                   3805: precisely, it adds two number together and produces a result. Where does
                   3806: it get the two numbers from? It takes the top two numbers off the
                   3807: stack. Where does it place the result? On the stack. You can act-out the
                   3808: behaviour of @code{+} with your playing cards like this:
1.21      crook    3809: 
                   3810: @itemize @bullet
                   3811: @item
1.29      crook    3812: Pick up two cards from the stack on the table
1.21      crook    3813: @item
1.29      crook    3814: Stare at them intently and ask yourself ``what @i{is} the sum of these two
                   3815: numbers''
1.21      crook    3816: @item
1.29      crook    3817: Decide that the answer is 5
1.21      crook    3818: @item
1.29      crook    3819: Shuffle the two cards back into the pack and find a 5
1.21      crook    3820: @item
1.29      crook    3821: Put a 5 on the remaining ace that's on the table.
1.21      crook    3822: @end itemize
                   3823: 
1.29      crook    3824: If you don't have a pack of cards handy but you do have Forth running,
                   3825: you can use the definition @code{.s} to show the current state of the stack,
                   3826: without affecting the stack. Type:
1.21      crook    3827: 
                   3828: @example
1.30      anton    3829: @kbd{clearstack 1 2 3@key{RET}} ok
                   3830: @kbd{.s@key{RET}} <3> 1 2 3  ok
1.23      crook    3831: @end example
                   3832: 
1.29      crook    3833: The text interpreter looks up the word @code{clearstack} and executes
                   3834: it; it tidies up the stack and removes any entries that may have been
                   3835: left on it by earlier examples. The text interpreter pushes each of the
                   3836: three numbers in turn onto the stack. Finally, the text interpreter
                   3837: looks up the word @code{.s} and executes it. The effect of executing
                   3838: @code{.s} is to print the ``<3>'' (the total number of items on the stack)
                   3839: followed by a list of all the items on the stack; the item on the far
                   3840: right-hand side is the TOS.
1.21      crook    3841: 
1.29      crook    3842: You can now type:
1.21      crook    3843: 
1.29      crook    3844: @example
1.30      anton    3845: @kbd{+ .s@key{RET}} <2> 1 5  ok
1.29      crook    3846: @end example
1.21      crook    3847: 
1.29      crook    3848: @noindent
                   3849: which is correct; there are now 2 items on the stack and the result of
                   3850: the addition is 5.
1.23      crook    3851: 
1.29      crook    3852: If you're playing with cards, try doing a second addition: pick up the
                   3853: two cards, work out that their sum is 6, shuffle them into the pack,
                   3854: look for a 6 and place that on the table. You now have just one item on
                   3855: the stack. What happens if you try to do a third addition? Pick up the
                   3856: first card, pick up the second card -- ah! There is no second card. This
                   3857: is called a @dfn{stack underflow} and consitutes an error. If you try to
                   3858: do the same thing with Forth it will report an error (probably a Stack
                   3859: Underflow or an Invalid Memory Address error).
1.23      crook    3860: 
1.29      crook    3861: The opposite situation to a stack underflow is a @dfn{stack overflow},
                   3862: which simply accepts that there is a finite amount of storage space
                   3863: reserved for the stack. To stretch the playing card analogy, if you had
                   3864: enough packs of cards and you piled the cards up on the table, you would
                   3865: eventually be unable to add another card; you'd hit the ceiling. Gforth
                   3866: allows you to set the maximum size of the stacks. In general, the only
                   3867: time that you will get a stack overflow is because a definition has a
                   3868: bug in it and is generating data on the stack uncontrollably.
1.23      crook    3869: 
1.29      crook    3870: There's one final use for the playing card analogy. If you model your
                   3871: stack using a pack of playing cards, the maximum number of items on
                   3872: your stack will be 52 (I assume you didn't use the Joker). The maximum
                   3873: @i{value} of any item on the stack is 13 (the King). In fact, the only
                   3874: possible numbers are positive integer numbers 1 through 13; you can't
                   3875: have (for example) 0 or 27 or 3.52 or -2. If you change the way you
                   3876: think about some of the cards, you can accommodate different
                   3877: numbers. For example, you could think of the Jack as representing 0,
                   3878: the Queen as representing -1 and the King as representing -2. Your
1.45      crook    3879: @i{range} remains unchanged (you can still only represent a total of 13
1.29      crook    3880: numbers) but the numbers that you can represent are -2 through 10.
1.28      crook    3881: 
1.29      crook    3882: In that analogy, the limit was the amount of information that a single
                   3883: stack entry could hold, and Forth has a similar limit. In Forth, the
                   3884: size of a stack entry is called a @dfn{cell}. The actual size of a cell is
                   3885: implementation dependent and affects the maximum value that a stack
                   3886: entry can hold. A Standard Forth provides a cell size of at least
                   3887: 16-bits, and most desktop systems use a cell size of 32-bits.
1.21      crook    3888: 
1.29      crook    3889: Forth does not do any type checking for you, so you are free to
                   3890: manipulate and combine stack items in any way you wish. A convenient way
                   3891: of treating stack items is as 2's complement signed integers, and that
                   3892: is what Standard words like @code{+} do. Therefore you can type:
1.21      crook    3893: 
1.29      crook    3894: @example
1.30      anton    3895: @kbd{-5 12 + .s@key{RET}} <1> 7  ok
1.29      crook    3896: @end example
1.21      crook    3897: 
1.29      crook    3898: If you use numbers and definitions like @code{+} in order to turn Forth
                   3899: into a great big pocket calculator, you will realise that it's rather
                   3900: different from a normal calculator. Rather than typing 2 + 3 = you had
                   3901: to type 2 3 + (ignore the fact that you had to use @code{.s} to see the
                   3902: result). The terminology used to describe this difference is to say that
                   3903: your calculator uses @dfn{Infix Notation} (parameters and operators are
                   3904: mixed) whilst Forth uses @dfn{Postfix Notation} (parameters and
                   3905: operators are separate), also called @dfn{Reverse Polish Notation}.
1.21      crook    3906: 
1.29      crook    3907: Whilst postfix notation might look confusing to begin with, it has
                   3908: several important advantages:
1.21      crook    3909: 
1.23      crook    3910: @itemize @bullet
                   3911: @item
1.29      crook    3912: it is unambiguous
1.23      crook    3913: @item
1.29      crook    3914: it is more concise
1.23      crook    3915: @item
1.29      crook    3916: it fits naturally with a stack-based system
1.23      crook    3917: @end itemize
1.21      crook    3918: 
1.29      crook    3919: To examine these claims in more detail, consider these sums:
1.21      crook    3920: 
1.29      crook    3921: @example
                   3922: 6 + 5 * 4 =
                   3923: 4 * 5 + 6 =
                   3924: @end example
1.21      crook    3925: 
1.29      crook    3926: If you're just learning maths or your maths is very rusty, you will
                   3927: probably come up with the answer 44 for the first and 26 for the
                   3928: second. If you are a bit of a whizz at maths you will remember the
                   3929: @i{convention} that multiplication takes precendence over addition, and
                   3930: you'd come up with the answer 26 both times. To explain the answer 26
                   3931: to someone who got the answer 44, you'd probably rewrite the first sum
                   3932: like this:
1.21      crook    3933: 
1.29      crook    3934: @example
                   3935: 6 + (5 * 4) =
                   3936: @end example
1.21      crook    3937: 
1.29      crook    3938: If what you really wanted was to perform the addition before the
                   3939: multiplication, you would have to use parentheses to force it.
1.21      crook    3940: 
1.29      crook    3941: If you did the first two sums on a pocket calculator you would probably
                   3942: get the right answers, unless you were very cautious and entered them using
                   3943: these keystroke sequences:
1.21      crook    3944: 
1.29      crook    3945: 6 + 5 = * 4 =
                   3946: 4 * 5 = + 6 =
1.21      crook    3947: 
1.29      crook    3948: Postfix notation is unambiguous because the order that the operators
                   3949: are applied is always explicit; that also means that parentheses are
                   3950: never required. The operators are @i{active} (the act of quoting the
                   3951: operator makes the operation occur) which removes the need for ``=''.
1.28      crook    3952: 
1.29      crook    3953: The sum 6 + 5 * 4 can be written (in postfix notation) in two
                   3954: equivalent ways:
1.26      crook    3955: 
                   3956: @example
1.29      crook    3957: 6 5 4 * +      or:
                   3958: 5 4 * 6 +
1.26      crook    3959: @end example
1.23      crook    3960: 
1.29      crook    3961: An important thing that you should notice about this notation is that
                   3962: the @i{order} of the numbers does not change; if you want to subtract
                   3963: 2 from 10 you type @code{10 2 -}.
1.1       anton    3964: 
1.29      crook    3965: The reason that Forth uses postfix notation is very simple to explain: it
                   3966: makes the implementation extremely simple, and it follows naturally from
                   3967: using the stack as a mechanism for passing parameters. Another way of
                   3968: thinking about this is to realise that all Forth definitions are
                   3969: @i{active}; they execute as they are encountered by the text
                   3970: interpreter. The result of this is that the syntax of Forth is trivially
                   3971: simple.
1.1       anton    3972: 
                   3973: 
                   3974: 
1.29      crook    3975: @comment ----------------------------------------------
                   3976: @node Your first definition, How does that work?, Stacks and Postfix notation, Introduction
                   3977: @section Your first Forth definition
                   3978: @cindex first definition
1.1       anton    3979: 
1.29      crook    3980: Until now, the examples we've seen have been trivial; we've just been
                   3981: using Forth as a bigger-than-pocket calculator. Also, each calculation
                   3982: we've shown has been a ``one-off'' -- to repeat it we'd need to type it in
                   3983: again@footnote{That's not quite true. If you press the up-arrow key on
                   3984: your keyboard you should be able to scroll back to any earlier command,
                   3985: edit it and re-enter it.} In this section we'll see how to add new
                   3986: words to Forth's vocabulary.
1.1       anton    3987: 
1.29      crook    3988: The easiest way to create a new word is to use a @dfn{colon
                   3989: definition}. We'll define a few and try them out before worrying too
                   3990: much about how they work. Try typing in these examples; be careful to
                   3991: copy the spaces accurately:
1.1       anton    3992: 
1.29      crook    3993: @example
                   3994: : add-two 2 + . ;
                   3995: : greet ." Hello and welcome" ;
                   3996: : demo 5 add-two ;
                   3997: @end example
1.1       anton    3998: 
1.29      crook    3999: @noindent
                   4000: Now try them out:
1.1       anton    4001: 
1.29      crook    4002: @example
1.30      anton    4003: @kbd{greet@key{RET}} Hello and welcome  ok
                   4004: @kbd{greet greet@key{RET}} Hello and welcomeHello and welcome  ok
                   4005: @kbd{4 add-two@key{RET}} 6  ok
                   4006: @kbd{demo@key{RET}} 7  ok
                   4007: @kbd{9 greet demo add-two@key{RET}} Hello and welcome7 11  ok
1.29      crook    4008: @end example
1.1       anton    4009: 
1.29      crook    4010: The first new thing that we've introduced here is the pair of words
                   4011: @code{:} and @code{;}. These are used to start and terminate a new
                   4012: definition, respectively. The first word after the @code{:} is the name
                   4013: for the new definition.
1.1       anton    4014: 
1.29      crook    4015: As you can see from the examples, a definition is built up of words that
                   4016: have already been defined; Forth makes no distinction between
                   4017: definitions that existed when you started the system up, and those that
                   4018: you define yourself.
1.1       anton    4019: 
1.29      crook    4020: The examples also introduce the words @code{.} (dot), @code{."}
                   4021: (dot-quote) and @code{dup} (dewp). Dot takes the value from the top of
                   4022: the stack and displays it. It's like @code{.s} except that it only
                   4023: displays the top item of the stack and it is destructive; after it has
                   4024: executed, the number is no longer on the stack. There is always one
                   4025: space printed after the number, and no spaces before it. Dot-quote
                   4026: defines a string (a sequence of characters) that will be printed when
                   4027: the word is executed. The string can contain any printable characters
                   4028: except @code{"}. A @code{"} has a special function; it is not a Forth
                   4029: word but it acts as a delimiter (the way that delimiters work is
                   4030: described in the next section). Finally, @code{dup} duplicates the value
                   4031: at the top of the stack. Try typing @code{5 dup .s} to see what it does.
1.1       anton    4032: 
1.29      crook    4033: We already know that the text interpreter searches through the
                   4034: dictionary to locate names. If you've followed the examples earlier, you
                   4035: will already have a definition called @code{add-two}. Lets try modifying
                   4036: it by typing in a new definition:
1.1       anton    4037: 
1.29      crook    4038: @example
1.30      anton    4039: @kbd{: add-two dup . ." + 2 =" 2 + . ;@key{RET}} redefined add-two  ok
1.29      crook    4040: @end example
1.5       anton    4041: 
1.29      crook    4042: Forth recognised that we were defining a word that already exists, and
                   4043: printed a message to warn us of that fact. Let's try out the new
                   4044: definition:
1.5       anton    4045: 
1.29      crook    4046: @example
1.30      anton    4047: @kbd{9 add-two@key{RET}} 9 + 2 =11  ok
1.29      crook    4048: @end example
1.1       anton    4049: 
1.29      crook    4050: @noindent
                   4051: All that we've actually done here, though, is to create a new
                   4052: definition, with a particular name. The fact that there was already a
                   4053: definition with the same name did not make any difference to the way
                   4054: that the new definition was created (except that Forth printed a warning
                   4055: message). The old definition of add-two still exists (try @code{demo}
                   4056: again to see that this is true). Any new definition will use the new
                   4057: definition of @code{add-two}, but old definitions continue to use the
                   4058: version that already existed at the time that they were @code{compiled}.
1.1       anton    4059: 
1.29      crook    4060: Before you go on to the next section, try defining and redefining some
                   4061: words of your own.
1.1       anton    4062: 
1.29      crook    4063: @comment ----------------------------------------------
                   4064: @node How does that work?, Forth is written in Forth, Your first definition, Introduction
                   4065: @section How does that work?
                   4066: @cindex parsing words
1.1       anton    4067: 
1.30      anton    4068: @c That's pretty deep (IMO way too deep) for an introduction. - anton
                   4069: 
                   4070: @c Is it a good idea to talk about the interpretation semantics of a
                   4071: @c number? We don't have an xt to go along with it. - anton
                   4072: 
                   4073: @c Now that I have eliminated execution semantics, I wonder if it would not
                   4074: @c be better to keep them (or add run-time semantics), to make it easier to
                   4075: @c explain what compilation semantics usually does. - anton
                   4076: 
1.44      crook    4077: @c nac-> I removed the term ``default compilation sematics'' from the
                   4078: @c introductory chapter. Removing ``execution semantics'' was making
                   4079: @c everything simpler to explain, then I think the use of this term made
                   4080: @c everything more complex again. I replaced it with ``default
                   4081: @c semantics'' (which is used elsewhere in the manual) by which I mean
                   4082: @c ``a definition that has neither the immediate nor the compile-only
1.83      anton    4083: @c flag set''.
                   4084: 
                   4085: @c anton: I have eliminated default semantics (except in one place where it
                   4086: @c means "default interpretation and compilation semantics"), because it
                   4087: @c makes no sense in the presence of combined words.  I reverted to
                   4088: @c "execution semantics" where necessary.
                   4089: 
                   4090: @c nac-> I reworded big chunks of the ``how does that work''
1.44      crook    4091: @c section (and, unusually for me, I think I even made it shorter!).  See
                   4092: @c what you think -- I know I have not addressed your primary concern
                   4093: @c that it is too heavy-going for an introduction. From what I understood
                   4094: @c of your course notes it looks as though they might be a good framework. 
                   4095: @c Things that I've tried to capture here are some things that came as a
                   4096: @c great revelation here when I first understood them. Also, I like the
                   4097: @c fact that a very simple code example shows up almost all of the issues
                   4098: @c that you need to understand to see how Forth works. That's unique and
                   4099: @c worthwhile to emphasise.
                   4100: 
1.83      anton    4101: @c anton: I think it's a good idea to present the details, especially those
                   4102: @c that you found to be a revelation, and probably the tutorial tries to be
                   4103: @c too superficial and does not get some of the things across that make
                   4104: @c Forth special.  I do believe that most of the time these things should
                   4105: @c be discussed at the end of a section or in separate sections instead of
                   4106: @c in the middle of a section (e.g., the stuff you added in "User-defined
                   4107: @c defining words" leads in a completely different direction from the rest
                   4108: @c of the section).
                   4109: 
1.29      crook    4110: Now we're going to take another look at the definition of @code{add-two}
                   4111: from the previous section. From our knowledge of the way that the text
                   4112: interpreter works, we would have expected this result when we tried to
                   4113: define @code{add-two}:
1.21      crook    4114: 
1.29      crook    4115: @example
1.44      crook    4116: @kbd{: add-two 2 + . ;@key{RET}}
1.29      crook    4117:   ^^^^^^^
                   4118: Error: Undefined word
                   4119: @end example
1.28      crook    4120: 
1.29      crook    4121: The reason that this didn't happen is bound up in the way that @code{:}
                   4122: works. The word @code{:} does two special things. The first special
                   4123: thing that it does prevents the text interpreter from ever seeing the
                   4124: characters @code{add-two}. The text interpreter uses a variable called
                   4125: @cindex modifying >IN
1.44      crook    4126: @code{>IN} (pronounced ``to-in'') to keep track of where it is in the
1.29      crook    4127: input line. When it encounters the word @code{:} it behaves in exactly
                   4128: the same way as it does for any other word; it looks it up in the name
                   4129: dictionary, finds its xt and executes it. When @code{:} executes, it
                   4130: looks at the input buffer, finds the word @code{add-two} and advances the
                   4131: value of @code{>IN} to point past it. It then does some other stuff
                   4132: associated with creating the new definition (including creating an entry
                   4133: for @code{add-two} in the name dictionary). When the execution of @code{:}
                   4134: completes, control returns to the text interpreter, which is oblivious
                   4135: to the fact that it has been tricked into ignoring part of the input
                   4136: line.
1.21      crook    4137: 
1.29      crook    4138: @cindex parsing words
                   4139: Words like @code{:} -- words that advance the value of @code{>IN} and so
                   4140: prevent the text interpreter from acting on the whole of the input line
                   4141: -- are called @dfn{parsing words}.
1.21      crook    4142: 
1.29      crook    4143: @cindex @code{state} - effect on the text interpreter
                   4144: @cindex text interpreter - effect of state
                   4145: The second special thing that @code{:} does is change the value of a
                   4146: variable called @code{state}, which affects the way that the text
                   4147: interpreter behaves. When Gforth starts up, @code{state} has the value
                   4148: 0, and the text interpreter is said to be @dfn{interpreting}. During a
                   4149: colon definition (started with @code{:}), @code{state} is set to -1 and
1.44      crook    4150: the text interpreter is said to be @dfn{compiling}.
                   4151: 
                   4152: In this example, the text interpreter is compiling when it processes the
                   4153: string ``@code{2 + . ;}''. It still breaks the string down into
                   4154: character sequences in the same way. However, instead of pushing the
                   4155: number @code{2} onto the stack, it lays down (@dfn{compiles}) some magic
                   4156: into the definition of @code{add-two} that will make the number @code{2} get
                   4157: pushed onto the stack when @code{add-two} is @dfn{executed}. Similarly,
                   4158: the behaviours of @code{+} and @code{.} are also compiled into the
                   4159: definition.
                   4160: 
                   4161: One category of words don't get compiled. These so-called @dfn{immediate
                   4162: words} get executed (performed @i{now}) regardless of whether the text
                   4163: interpreter is interpreting or compiling. The word @code{;} is an
                   4164: immediate word. Rather than being compiled into the definition, it
                   4165: executes. Its effect is to terminate the current definition, which
                   4166: includes changing the value of @code{state} back to 0.
                   4167: 
                   4168: When you execute @code{add-two}, it has a @dfn{run-time effect} that is
                   4169: exactly the same as if you had typed @code{2 + . @key{RET}} outside of a
                   4170: definition.
1.28      crook    4171: 
1.30      anton    4172: In Forth, every word or number can be described in terms of two
1.29      crook    4173: properties:
1.28      crook    4174: 
                   4175: @itemize @bullet
                   4176: @item
1.29      crook    4177: @cindex interpretation semantics
1.44      crook    4178: Its @dfn{interpretation semantics} describe how it will behave when the
                   4179: text interpreter encounters it in @dfn{interpret} state. The
                   4180: interpretation semantics of a word are represented by an @dfn{execution
                   4181: token}.
1.28      crook    4182: @item
1.29      crook    4183: @cindex compilation semantics
1.44      crook    4184: Its @dfn{compilation semantics} describe how it will behave when the
                   4185: text interpreter encounters it in @dfn{compile} state. The compilation
                   4186: semantics of a word are represented in an implementation-dependent way;
                   4187: Gforth uses a @dfn{compilation token}.
1.29      crook    4188: @end itemize
                   4189: 
                   4190: @noindent
                   4191: Numbers are always treated in a fixed way:
                   4192: 
                   4193: @itemize @bullet
1.28      crook    4194: @item
1.44      crook    4195: When the number is @dfn{interpreted}, its behaviour is to push the
                   4196: number onto the stack.
1.28      crook    4197: @item
1.30      anton    4198: When the number is @dfn{compiled}, a piece of code is appended to the
                   4199: current definition that pushes the number when it runs. (In other words,
                   4200: the compilation semantics of a number are to postpone its interpretation
                   4201: semantics until the run-time of the definition that it is being compiled
                   4202: into.)
1.29      crook    4203: @end itemize
                   4204: 
1.44      crook    4205: Words don't behave in such a regular way, but most have @i{default
                   4206: semantics} which means that they behave like this:
1.29      crook    4207: 
                   4208: @itemize @bullet
1.28      crook    4209: @item
1.30      anton    4210: The @dfn{interpretation semantics} of the word are to do something useful.
                   4211: @item
1.29      crook    4212: The @dfn{compilation semantics} of the word are to append its
1.30      anton    4213: @dfn{interpretation semantics} to the current definition (so that its
                   4214: run-time behaviour is to do something useful).
1.28      crook    4215: @end itemize
                   4216: 
1.30      anton    4217: @cindex immediate words
1.44      crook    4218: The actual behaviour of any particular word can be controlled by using
                   4219: the words @code{immediate} and @code{compile-only} when the word is
                   4220: defined. These words set flags in the name dictionary entry of the most
                   4221: recently defined word, and these flags are retrieved by the text
                   4222: interpreter when it finds the word in the name dictionary.
                   4223: 
                   4224: A word that is marked as @dfn{immediate} has compilation semantics that
                   4225: are identical to its interpretation semantics. In other words, it
                   4226: behaves like this:
1.29      crook    4227: 
                   4228: @itemize @bullet
                   4229: @item
1.30      anton    4230: The @dfn{interpretation semantics} of the word are to do something useful.
1.29      crook    4231: @item
1.30      anton    4232: The @dfn{compilation semantics} of the word are to do something useful
                   4233: (and actually the same thing); i.e., it is executed during compilation.
1.29      crook    4234: @end itemize
1.28      crook    4235: 
1.44      crook    4236: Marking a word as @dfn{compile-only} prohibits the text interpreter from
                   4237: performing the interpretation semantics of the word directly; an attempt
                   4238: to do so will generate an error. It is never necessary to use
                   4239: @code{compile-only} (and it is not even part of ANS Forth, though it is
                   4240: provided by many implementations) but it is good etiquette to apply it
                   4241: to a word that will not behave correctly (and might have unexpected
                   4242: side-effects) in interpret state. For example, it is only legal to use
                   4243: the conditional word @code{IF} within a definition. If you forget this
                   4244: and try to use it elsewhere, the fact that (in Gforth) it is marked as
                   4245: @code{compile-only} allows the text interpreter to generate a helpful
                   4246: error message rather than subjecting you to the consequences of your
                   4247: folly.
                   4248: 
1.29      crook    4249: This example shows the difference between an immediate and a
                   4250: non-immediate word:
1.28      crook    4251: 
1.29      crook    4252: @example
                   4253: : show-state state @@ . ;
                   4254: : show-state-now show-state ; immediate
                   4255: : word1 show-state ;
                   4256: : word2 show-state-now ;
1.28      crook    4257: @end example
1.23      crook    4258: 
1.29      crook    4259: The word @code{immediate} after the definition of @code{show-state-now}
                   4260: makes that word an immediate word. These definitions introduce a new
                   4261: word: @code{@@} (pronounced ``fetch''). This word fetches the value of a
                   4262: variable, and leaves it on the stack. Therefore, the behaviour of
                   4263: @code{show-state} is to print a number that represents the current value
                   4264: of @code{state}.
1.28      crook    4265: 
1.29      crook    4266: When you execute @code{word1}, it prints the number 0, indicating that
                   4267: the system is interpreting. When the text interpreter compiled the
                   4268: definition of @code{word1}, it encountered @code{show-state} whose
1.30      anton    4269: compilation semantics are to append its interpretation semantics to the
1.29      crook    4270: current definition. When you execute @code{word1}, it performs the
1.30      anton    4271: interpretation semantics of @code{show-state}.  At the time that @code{word1}
1.29      crook    4272: (and therefore @code{show-state}) are executed, the system is
                   4273: interpreting.
1.28      crook    4274: 
1.30      anton    4275: When you pressed @key{RET} after entering the definition of @code{word2},
1.29      crook    4276: you should have seen the number -1 printed, followed by ``@code{
                   4277: ok}''. When the text interpreter compiled the definition of
                   4278: @code{word2}, it encountered @code{show-state-now}, an immediate word,
1.30      anton    4279: whose compilation semantics are therefore to perform its interpretation
1.29      crook    4280: semantics. It is executed straight away (even before the text
                   4281: interpreter has moved on to process another group of characters; the
                   4282: @code{;} in this example). The effect of executing it are to display the
                   4283: value of @code{state} @i{at the time that the definition of}
                   4284: @code{word2} @i{is being defined}. Printing -1 demonstrates that the
                   4285: system is compiling at this time. If you execute @code{word2} it does
                   4286: nothing at all.
1.28      crook    4287: 
1.29      crook    4288: @cindex @code{."}, how it works
                   4289: Before leaving the subject of immediate words, consider the behaviour of
                   4290: @code{."} in the definition of @code{greet}, in the previous
                   4291: section. This word is both a parsing word and an immediate word. Notice
                   4292: that there is a space between @code{."} and the start of the text
                   4293: @code{Hello and welcome}, but that there is no space between the last
                   4294: letter of @code{welcome} and the @code{"} character. The reason for this
                   4295: is that @code{."} is a Forth word; it must have a space after it so that
                   4296: the text interpreter can identify it. The @code{"} is not a Forth word;
                   4297: it is a @dfn{delimiter}. The examples earlier show that, when the string
                   4298: is displayed, there is neither a space before the @code{H} nor after the
                   4299: @code{e}. Since @code{."} is an immediate word, it executes at the time
                   4300: that @code{greet} is defined. When it executes, its behaviour is to
                   4301: search forward in the input line looking for the delimiter. When it
                   4302: finds the delimiter, it updates @code{>IN} to point past the
                   4303: delimiter. It also compiles some magic code into the definition of
                   4304: @code{greet}; the xt of a run-time routine that prints a text string. It
                   4305: compiles the string @code{Hello and welcome} into memory so that it is
                   4306: available to be printed later. When the text interpreter gains control,
                   4307: the next word it finds in the input stream is @code{;} and so it
                   4308: terminates the definition of @code{greet}.
1.28      crook    4309: 
                   4310: 
                   4311: @comment ----------------------------------------------
1.29      crook    4312: @node Forth is written in Forth, Review - elements of a Forth system, How does that work?, Introduction
                   4313: @section Forth is written in Forth
                   4314: @cindex structure of Forth programs
                   4315: 
                   4316: When you start up a Forth compiler, a large number of definitions
                   4317: already exist. In Forth, you develop a new application using bottom-up
                   4318: programming techniques to create new definitions that are defined in
                   4319: terms of existing definitions. As you create each definition you can
                   4320: test and debug it interactively.
                   4321: 
                   4322: If you have tried out the examples in this section, you will probably
                   4323: have typed them in by hand; when you leave Gforth, your definitions will
                   4324: be lost. You can avoid this by using a text editor to enter Forth source
                   4325: code into a file, and then loading code from the file using
1.49      anton    4326: @code{include} (@pxref{Forth source files}). A Forth source file is
1.29      crook    4327: processed by the text interpreter, just as though you had typed it in by
                   4328: hand@footnote{Actually, there are some subtle differences -- see
                   4329: @ref{The Text Interpreter}.}.
                   4330: 
                   4331: Gforth also supports the traditional Forth alternative to using text
1.49      anton    4332: files for program entry (@pxref{Blocks}).
1.28      crook    4333: 
1.29      crook    4334: In common with many, if not most, Forth compilers, most of Gforth is
                   4335: actually written in Forth. All of the @file{.fs} files in the
                   4336: installation directory@footnote{For example,
1.30      anton    4337: @file{/usr/local/share/gforth...}} are Forth source files, which you can
1.29      crook    4338: study to see examples of Forth programming.
1.28      crook    4339: 
1.29      crook    4340: Gforth maintains a history file that records every line that you type to
                   4341: the text interpreter. This file is preserved between sessions, and is
                   4342: used to provide a command-line recall facility. If you enter long
                   4343: definitions by hand, you can use a text editor to paste them out of the
                   4344: history file into a Forth source file for reuse at a later time
1.49      anton    4345: (for more information @pxref{Command-line editing}).
1.28      crook    4346: 
                   4347: 
                   4348: @comment ----------------------------------------------
1.29      crook    4349: @node Review - elements of a Forth system, Where to go next, Forth is written in Forth, Introduction
                   4350: @section Review - elements of a Forth system
                   4351: @cindex elements of a Forth system
1.28      crook    4352: 
1.29      crook    4353: To summarise this chapter:
1.28      crook    4354: 
                   4355: @itemize @bullet
                   4356: @item
1.29      crook    4357: Forth programs use @dfn{factoring} to break a problem down into small
                   4358: fragments called @dfn{words} or @dfn{definitions}.
                   4359: @item
                   4360: Forth program development is an interactive process.
                   4361: @item
                   4362: The main command loop that accepts input, and controls both
                   4363: interpretation and compilation, is called the @dfn{text interpreter}
                   4364: (also known as the @dfn{outer interpreter}).
                   4365: @item
                   4366: Forth has a very simple syntax, consisting of words and numbers
                   4367: separated by spaces or carriage-return characters. Any additional syntax
                   4368: is imposed by @dfn{parsing words}.
                   4369: @item
                   4370: Forth uses a stack to pass parameters between words. As a result, it
                   4371: uses postfix notation.
                   4372: @item
                   4373: To use a word that has previously been defined, the text interpreter
                   4374: searches for the word in the @dfn{name dictionary}.
                   4375: @item
1.30      anton    4376: Words have @dfn{interpretation semantics} and @dfn{compilation semantics}.
1.28      crook    4377: @item
1.29      crook    4378: The text interpreter uses the value of @code{state} to select between
                   4379: the use of the @dfn{interpretation semantics} and the  @dfn{compilation
                   4380: semantics} of a word that it encounters.
1.28      crook    4381: @item
1.30      anton    4382: The relationship between the @dfn{interpretation semantics} and
                   4383: @dfn{compilation semantics} for a word
1.29      crook    4384: depend upon the way in which the word was defined (for example, whether
                   4385: it is an @dfn{immediate} word).
1.28      crook    4386: @item
1.29      crook    4387: Forth definitions can be implemented in Forth (called @dfn{high-level
                   4388: definitions}) or in some other way (usually a lower-level language and
                   4389: as a result often called @dfn{low-level definitions}, @dfn{code
                   4390: definitions} or @dfn{primitives}).
1.28      crook    4391: @item
1.29      crook    4392: Many Forth systems are implemented mainly in Forth.
1.28      crook    4393: @end itemize
                   4394: 
                   4395: 
1.29      crook    4396: @comment ----------------------------------------------
1.48      anton    4397: @node Where to go next, Exercises, Review - elements of a Forth system, Introduction
1.29      crook    4398: @section Where To Go Next
                   4399: @cindex where to go next
1.28      crook    4400: 
1.29      crook    4401: Amazing as it may seem, if you have read (and understood) this far, you
                   4402: know almost all the fundamentals about the inner workings of a Forth
                   4403: system. You certainly know enough to be able to read and understand the
                   4404: rest of this manual and the ANS Forth document, to learn more about the
                   4405: facilities that Forth in general and Gforth in particular provide. Even
                   4406: scarier, you know almost enough to implement your own Forth system.
1.30      anton    4407: However, that's not a good idea just yet... better to try writing some
1.29      crook    4408: programs in Gforth.
1.28      crook    4409: 
1.29      crook    4410: Forth has such a rich vocabulary that it can be hard to know where to
                   4411: start in learning it. This section suggests a few sets of words that are
                   4412: enough to write small but useful programs. Use the word index in this
                   4413: document to learn more about each word, then try it out and try to write
                   4414: small definitions using it. Start by experimenting with these words:
1.28      crook    4415: 
                   4416: @itemize @bullet
                   4417: @item
1.29      crook    4418: Arithmetic: @code{+ - * / /MOD */ ABS INVERT}
                   4419: @item
                   4420: Comparison: @code{MIN MAX =}
                   4421: @item
                   4422: Logic: @code{AND OR XOR NOT}
                   4423: @item
                   4424: Stack manipulation: @code{DUP DROP SWAP OVER}
1.28      crook    4425: @item
1.29      crook    4426: Loops and decisions: @code{IF ELSE ENDIF ?DO I LOOP}
1.28      crook    4427: @item
1.29      crook    4428: Input/Output: @code{. ." EMIT CR KEY}
1.28      crook    4429: @item
1.29      crook    4430: Defining words: @code{: ; CREATE}
1.28      crook    4431: @item
1.29      crook    4432: Memory allocation words: @code{ALLOT ,}
1.28      crook    4433: @item
1.29      crook    4434: Tools: @code{SEE WORDS .S MARKER}
                   4435: @end itemize
                   4436: 
                   4437: When you have mastered those, go on to:
                   4438: 
                   4439: @itemize @bullet
1.28      crook    4440: @item
1.29      crook    4441: More defining words: @code{VARIABLE CONSTANT VALUE TO CREATE DOES>}
1.28      crook    4442: @item
1.29      crook    4443: Memory access: @code{@@ !}
1.28      crook    4444: @end itemize
1.23      crook    4445: 
1.29      crook    4446: When you have mastered these, there's nothing for it but to read through
                   4447: the whole of this manual and find out what you've missed.
                   4448: 
                   4449: @comment ----------------------------------------------
1.48      anton    4450: @node Exercises,  , Where to go next, Introduction
1.29      crook    4451: @section Exercises
                   4452: @cindex exercises
                   4453: 
                   4454: TODO: provide a set of programming excercises linked into the stuff done
                   4455: already and into other sections of the manual. Provide solutions to all
                   4456: the exercises in a .fs file in the distribution.
                   4457: 
                   4458: @c Get some inspiration from Starting Forth and Kelly&Spies.
                   4459: 
                   4460: @c excercises:
                   4461: @c 1. take inches and convert to feet and inches.
                   4462: @c 2. take temperature and convert from fahrenheight to celcius;
                   4463: @c    may need to care about symmetric vs floored??
                   4464: @c 3. take input line and do character substitution
                   4465: @c    to encipher or decipher
                   4466: @c 4. as above but work on a file for in and out
                   4467: @c 5. take input line and convert to pig-latin 
                   4468: @c
                   4469: @c thing of sets of things to exercise then come up with
                   4470: @c problems that need those things.
                   4471: 
                   4472: 
1.26      crook    4473: @c ******************************************************************
1.29      crook    4474: @node Words, Error messages, Introduction, Top
1.1       anton    4475: @chapter Forth Words
1.26      crook    4476: @cindex words
1.1       anton    4477: 
                   4478: @menu
                   4479: * Notation::                    
1.65      anton    4480: * Case insensitivity::          
                   4481: * Comments::                    
                   4482: * Boolean Flags::               
1.1       anton    4483: * Arithmetic::                  
                   4484: * Stack Manipulation::          
1.5       anton    4485: * Memory::                      
1.1       anton    4486: * Control Structures::          
                   4487: * Defining Words::              
1.65      anton    4488: * Interpretation and Compilation Semantics::  
1.47      crook    4489: * Tokens for Words::            
1.81      anton    4490: * Compiling words::             
1.65      anton    4491: * The Text Interpreter::        
                   4492: * Word Lists::                  
                   4493: * Environmental Queries::       
1.12      anton    4494: * Files::                       
                   4495: * Blocks::                      
                   4496: * Other I/O::                   
1.78      anton    4497: * Locals::                      
                   4498: * Structures::                  
                   4499: * Object-oriented Forth::       
1.12      anton    4500: * Programming Tools::           
                   4501: * Assembler and Code Words::    
                   4502: * Threading Words::             
1.65      anton    4503: * Passing Commands to the OS::  
                   4504: * Keeping track of Time::       
                   4505: * Miscellaneous Words::         
1.1       anton    4506: @end menu
                   4507: 
1.65      anton    4508: @node Notation, Case insensitivity, Words, Words
1.1       anton    4509: @section Notation
                   4510: @cindex notation of glossary entries
                   4511: @cindex format of glossary entries
                   4512: @cindex glossary notation format
                   4513: @cindex word glossary entry format
                   4514: 
                   4515: The Forth words are described in this section in the glossary notation
1.67      anton    4516: that has become a de-facto standard for Forth texts:
1.1       anton    4517: 
                   4518: @format
1.29      crook    4519: @i{word}     @i{Stack effect}   @i{wordset}   @i{pronunciation}
1.1       anton    4520: @end format
1.29      crook    4521: @i{Description}
1.1       anton    4522: 
                   4523: @table @var
                   4524: @item word
1.28      crook    4525: The name of the word.
1.1       anton    4526: 
                   4527: @item Stack effect
                   4528: @cindex stack effect
1.29      crook    4529: The stack effect is written in the notation @code{@i{before} --
                   4530: @i{after}}, where @i{before} and @i{after} describe the top of
1.1       anton    4531: stack entries before and after the execution of the word. The rest of
                   4532: the stack is not touched by the word. The top of stack is rightmost,
                   4533: i.e., a stack sequence is written as it is typed in. Note that Gforth
                   4534: uses a separate floating point stack, but a unified stack
1.29      crook    4535: notation. Also, return stack effects are not shown in @i{stack
                   4536: effect}, but in @i{Description}. The name of a stack item describes
1.1       anton    4537: the type and/or the function of the item. See below for a discussion of
                   4538: the types.
                   4539: 
                   4540: All words have two stack effects: A compile-time stack effect and a
                   4541: run-time stack effect. The compile-time stack-effect of most words is
1.29      crook    4542: @i{ -- }. If the compile-time stack-effect of a word deviates from
1.1       anton    4543: this standard behaviour, or the word does other unusual things at
                   4544: compile time, both stack effects are shown; otherwise only the run-time
                   4545: stack effect is shown.
                   4546: 
                   4547: @cindex pronounciation of words
                   4548: @item pronunciation
                   4549: How the word is pronounced.
                   4550: 
                   4551: @cindex wordset
1.67      anton    4552: @cindex environment wordset
1.1       anton    4553: @item wordset
1.21      crook    4554: The ANS Forth standard is divided into several word sets. A standard
                   4555: system need not support all of them. Therefore, in theory, the fewer
                   4556: word sets your program uses the more portable it will be. However, we
                   4557: suspect that most ANS Forth systems on personal machines will feature
1.26      crook    4558: all word sets. Words that are not defined in ANS Forth have
1.21      crook    4559: @code{gforth} or @code{gforth-internal} as word set. @code{gforth}
1.1       anton    4560: describes words that will work in future releases of Gforth;
                   4561: @code{gforth-internal} words are more volatile. Environmental query
                   4562: strings are also displayed like words; you can recognize them by the
1.21      crook    4563: @code{environment} in the word set field.
1.1       anton    4564: 
                   4565: @item Description
                   4566: A description of the behaviour of the word.
                   4567: @end table
                   4568: 
                   4569: @cindex types of stack items
                   4570: @cindex stack item types
                   4571: The type of a stack item is specified by the character(s) the name
                   4572: starts with:
                   4573: 
                   4574: @table @code
                   4575: @item f
                   4576: @cindex @code{f}, stack item type
                   4577: Boolean flags, i.e. @code{false} or @code{true}.
                   4578: @item c
                   4579: @cindex @code{c}, stack item type
                   4580: Char
                   4581: @item w
                   4582: @cindex @code{w}, stack item type
                   4583: Cell, can contain an integer or an address
                   4584: @item n
                   4585: @cindex @code{n}, stack item type
                   4586: signed integer
                   4587: @item u
                   4588: @cindex @code{u}, stack item type
                   4589: unsigned integer
                   4590: @item d
                   4591: @cindex @code{d}, stack item type
                   4592: double sized signed integer
                   4593: @item ud
                   4594: @cindex @code{ud}, stack item type
                   4595: double sized unsigned integer
                   4596: @item r
                   4597: @cindex @code{r}, stack item type
                   4598: Float (on the FP stack)
1.21      crook    4599: @item a-
1.1       anton    4600: @cindex @code{a_}, stack item type
                   4601: Cell-aligned address
1.21      crook    4602: @item c-
1.1       anton    4603: @cindex @code{c_}, stack item type
                   4604: Char-aligned address (note that a Char may have two bytes in Windows NT)
1.21      crook    4605: @item f-
1.1       anton    4606: @cindex @code{f_}, stack item type
                   4607: Float-aligned address
1.21      crook    4608: @item df-
1.1       anton    4609: @cindex @code{df_}, stack item type
                   4610: Address aligned for IEEE double precision float
1.21      crook    4611: @item sf-
1.1       anton    4612: @cindex @code{sf_}, stack item type
                   4613: Address aligned for IEEE single precision float
                   4614: @item xt
                   4615: @cindex @code{xt}, stack item type
                   4616: Execution token, same size as Cell
                   4617: @item wid
                   4618: @cindex @code{wid}, stack item type
1.21      crook    4619: Word list ID, same size as Cell
1.68      anton    4620: @item ior, wior
                   4621: @cindex ior type description
                   4622: @cindex wior type description
                   4623: I/O result code, cell-sized.  In Gforth, you can @code{throw} iors.
1.1       anton    4624: @item f83name
                   4625: @cindex @code{f83name}, stack item type
                   4626: Pointer to a name structure
                   4627: @item "
                   4628: @cindex @code{"}, stack item type
1.12      anton    4629: string in the input stream (not on the stack). The terminating character
                   4630: is a blank by default. If it is not a blank, it is shown in @code{<>}
1.1       anton    4631: quotes.
                   4632: @end table
                   4633: 
1.65      anton    4634: @comment ----------------------------------------------
                   4635: @node Case insensitivity, Comments, Notation, Words
                   4636: @section Case insensitivity
                   4637: @cindex case sensitivity
                   4638: @cindex upper and lower case
                   4639: 
                   4640: Gforth is case-insensitive; you can enter definitions and invoke
                   4641: Standard words using upper, lower or mixed case (however,
                   4642: @pxref{core-idef, Implementation-defined options, Implementation-defined
                   4643: options}).
                   4644: 
                   4645: ANS Forth only @i{requires} implementations to recognise Standard words
                   4646: when they are typed entirely in upper case. Therefore, a Standard
                   4647: program must use upper case for all Standard words. You can use whatever
                   4648: case you like for words that you define, but in a Standard program you
                   4649: have to use the words in the same case that you defined them.
                   4650: 
                   4651: Gforth supports case sensitivity through @code{table}s (case-sensitive
                   4652: wordlists, @pxref{Word Lists}).
                   4653: 
                   4654: Two people have asked how to convert Gforth to be case-sensitive; while
                   4655: we think this is a bad idea, you can change all wordlists into tables
                   4656: like this:
                   4657: 
                   4658: @example
                   4659: ' table-find forth-wordlist wordlist-map @ !
                   4660: @end example
                   4661: 
                   4662: Note that you now have to type the predefined words in the same case
                   4663: that we defined them, which are varying.  You may want to convert them
                   4664: to your favourite case before doing this operation (I won't explain how,
                   4665: because if you are even contemplating doing this, you'd better have
                   4666: enough knowledge of Forth systems to know this already).
                   4667: 
                   4668: @node Comments, Boolean Flags, Case insensitivity, Words
1.21      crook    4669: @section Comments
1.26      crook    4670: @cindex comments
1.21      crook    4671: 
1.29      crook    4672: Forth supports two styles of comment; the traditional @i{in-line} comment,
                   4673: @code{(} and its modern cousin, the @i{comment to end of line}; @code{\}.
1.21      crook    4674: 
1.44      crook    4675: 
1.23      crook    4676: doc-(
1.21      crook    4677: doc-\
1.23      crook    4678: doc-\G
1.21      crook    4679: 
1.44      crook    4680: 
1.21      crook    4681: @node Boolean Flags, Arithmetic, Comments, Words
                   4682: @section Boolean Flags
1.26      crook    4683: @cindex Boolean flags
1.21      crook    4684: 
                   4685: A Boolean flag is cell-sized. A cell with all bits clear represents the
                   4686: flag @code{false} and a flag with all bits set represents the flag
1.26      crook    4687: @code{true}. Words that check a flag (for example, @code{IF}) will treat
1.29      crook    4688: a cell that has @i{any} bit set as @code{true}.
1.67      anton    4689: @c on and off to Memory? 
                   4690: @c true and false to "Bitwise operations" or "Numeric comparison"?
1.44      crook    4691: 
1.21      crook    4692: doc-true
                   4693: doc-false
1.29      crook    4694: doc-on
                   4695: doc-off
1.21      crook    4696: 
1.44      crook    4697: 
1.21      crook    4698: @node Arithmetic, Stack Manipulation, Boolean Flags, Words
1.1       anton    4699: @section Arithmetic
                   4700: @cindex arithmetic words
                   4701: 
                   4702: @cindex division with potentially negative operands
                   4703: Forth arithmetic is not checked, i.e., you will not hear about integer
                   4704: overflow on addition or multiplication, you may hear about division by
                   4705: zero if you are lucky. The operator is written after the operands, but
                   4706: the operands are still in the original order. I.e., the infix @code{2-1}
                   4707: corresponds to @code{2 1 -}. Forth offers a variety of division
                   4708: operators. If you perform division with potentially negative operands,
                   4709: you do not want to use @code{/} or @code{/mod} with its undefined
                   4710: behaviour, but rather @code{fm/mod} or @code{sm/mod} (probably the
                   4711: former, @pxref{Mixed precision}).
1.26      crook    4712: @comment TODO discuss the different division forms and the std approach
1.1       anton    4713: 
                   4714: @menu
                   4715: * Single precision::            
1.67      anton    4716: * Double precision::            Double-cell integer arithmetic
1.1       anton    4717: * Bitwise operations::          
1.67      anton    4718: * Numeric comparison::          
1.29      crook    4719: * Mixed precision::             Operations with single and double-cell integers
1.1       anton    4720: * Floating Point::              
                   4721: @end menu
                   4722: 
1.67      anton    4723: @node Single precision, Double precision, Arithmetic, Arithmetic
1.1       anton    4724: @subsection Single precision
                   4725: @cindex single precision arithmetic words
                   4726: 
1.67      anton    4727: @c !! cell undefined
                   4728: 
                   4729: By default, numbers in Forth are single-precision integers that are one
1.26      crook    4730: cell in size. They can be signed or unsigned, depending upon how you
1.49      anton    4731: treat them. For the rules used by the text interpreter for recognising
                   4732: single-precision integers see @ref{Number Conversion}.
1.21      crook    4733: 
1.67      anton    4734: These words are all defined for signed operands, but some of them also
                   4735: work for unsigned numbers: @code{+}, @code{1+}, @code{-}, @code{1-},
                   4736: @code{*}.
1.44      crook    4737: 
1.1       anton    4738: doc-+
1.21      crook    4739: doc-1+
1.1       anton    4740: doc--
1.21      crook    4741: doc-1-
1.1       anton    4742: doc-*
                   4743: doc-/
                   4744: doc-mod
                   4745: doc-/mod
                   4746: doc-negate
                   4747: doc-abs
                   4748: doc-min
                   4749: doc-max
1.27      crook    4750: doc-floored
1.1       anton    4751: 
1.44      crook    4752: 
1.67      anton    4753: @node Double precision, Bitwise operations, Single precision, Arithmetic
1.21      crook    4754: @subsection Double precision
                   4755: @cindex double precision arithmetic words
                   4756: 
1.49      anton    4757: For the rules used by the text interpreter for
                   4758: recognising double-precision integers, see @ref{Number Conversion}.
1.21      crook    4759: 
                   4760: A double precision number is represented by a cell pair, with the most
1.67      anton    4761: significant cell at the TOS. It is trivial to convert an unsigned single
                   4762: to a double: simply push a @code{0} onto the TOS. Since numbers are
                   4763: represented by Gforth using 2's complement arithmetic, converting a
                   4764: signed single to a (signed) double requires sign-extension across the
                   4765: most significant cell. This can be achieved using @code{s>d}. The moral
                   4766: of the story is that you cannot convert a number without knowing whether
                   4767: it represents an unsigned or a signed number.
1.21      crook    4768: 
1.67      anton    4769: These words are all defined for signed operands, but some of them also
                   4770: work for unsigned numbers: @code{d+}, @code{d-}.
1.44      crook    4771: 
1.21      crook    4772: doc-s>d
1.67      anton    4773: doc-d>s
1.21      crook    4774: doc-d+
                   4775: doc-d-
                   4776: doc-dnegate
                   4777: doc-dabs
                   4778: doc-dmin
                   4779: doc-dmax
                   4780: 
1.44      crook    4781: 
1.67      anton    4782: @node Bitwise operations, Numeric comparison, Double precision, Arithmetic
                   4783: @subsection Bitwise operations
                   4784: @cindex bitwise operation words
                   4785: 
                   4786: 
                   4787: doc-and
                   4788: doc-or
                   4789: doc-xor
                   4790: doc-invert
                   4791: doc-lshift
                   4792: doc-rshift
                   4793: doc-2*
                   4794: doc-d2*
                   4795: doc-2/
                   4796: doc-d2/
                   4797: 
                   4798: 
                   4799: @node Numeric comparison, Mixed precision, Bitwise operations, Arithmetic
1.21      crook    4800: @subsection Numeric comparison
                   4801: @cindex numeric comparison words
                   4802: 
1.67      anton    4803: Note that the words that compare for equality (@code{= <> 0= 0<> d= d<>
                   4804: d0= d0<>}) work for for both signed and unsigned numbers.
1.44      crook    4805: 
1.28      crook    4806: doc-<
                   4807: doc-<=
                   4808: doc-<>
                   4809: doc-=
                   4810: doc->
                   4811: doc->=
                   4812: 
1.21      crook    4813: doc-0<
1.23      crook    4814: doc-0<=
1.21      crook    4815: doc-0<>
                   4816: doc-0=
1.23      crook    4817: doc-0>
                   4818: doc-0>=
1.28      crook    4819: 
                   4820: doc-u<
                   4821: doc-u<=
1.44      crook    4822: @c u<> and u= exist but are the same as <> and =
1.31      anton    4823: @c doc-u<>
                   4824: @c doc-u=
1.28      crook    4825: doc-u>
                   4826: doc-u>=
                   4827: 
                   4828: doc-within
                   4829: 
                   4830: doc-d<
                   4831: doc-d<=
                   4832: doc-d<>
                   4833: doc-d=
                   4834: doc-d>
                   4835: doc-d>=
1.23      crook    4836: 
1.21      crook    4837: doc-d0<
1.23      crook    4838: doc-d0<=
                   4839: doc-d0<>
1.21      crook    4840: doc-d0=
1.23      crook    4841: doc-d0>
                   4842: doc-d0>=
                   4843: 
1.21      crook    4844: doc-du<
1.28      crook    4845: doc-du<=
1.44      crook    4846: @c du<> and du= exist but are the same as d<> and d=
1.31      anton    4847: @c doc-du<>
                   4848: @c doc-du=
1.28      crook    4849: doc-du>
                   4850: doc-du>=
1.1       anton    4851: 
1.44      crook    4852: 
1.21      crook    4853: @node Mixed precision, Floating Point, Numeric comparison, Arithmetic
1.1       anton    4854: @subsection Mixed precision
                   4855: @cindex mixed precision arithmetic words
                   4856: 
1.44      crook    4857: 
1.1       anton    4858: doc-m+
                   4859: doc-*/
                   4860: doc-*/mod
                   4861: doc-m*
                   4862: doc-um*
                   4863: doc-m*/
                   4864: doc-um/mod
                   4865: doc-fm/mod
                   4866: doc-sm/rem
                   4867: 
1.44      crook    4868: 
1.21      crook    4869: @node Floating Point,  , Mixed precision, Arithmetic
1.1       anton    4870: @subsection Floating Point
                   4871: @cindex floating point arithmetic words
                   4872: 
1.49      anton    4873: For the rules used by the text interpreter for
                   4874: recognising floating-point numbers see @ref{Number Conversion}.
1.1       anton    4875: 
1.67      anton    4876: Gforth has a separate floating point stack, but the documentation uses
                   4877: the unified notation.@footnote{It's easy to generate the separate
                   4878: notation from that by just separating the floating-point numbers out:
                   4879: e.g. @code{( n r1 u r2 -- r3 )} becomes @code{( n u -- ) ( F: r1 r2 --
                   4880: r3 )}.}
1.1       anton    4881: 
                   4882: @cindex floating-point arithmetic, pitfalls
                   4883: Floating point numbers have a number of unpleasant surprises for the
                   4884: unwary (e.g., floating point addition is not associative) and even a few
                   4885: for the wary. You should not use them unless you know what you are doing
                   4886: or you don't care that the results you get are totally bogus. If you
                   4887: want to learn about the problems of floating point numbers (and how to
1.66      anton    4888: avoid them), you might start with @cite{David Goldberg,
                   4889: @uref{http://www.validgh.com/goldberg/paper.ps,What Every Computer
                   4890: Scientist Should Know About Floating-Point Arithmetic}, ACM Computing
                   4891: Surveys 23(1):5@minus{}48, March 1991}.
1.1       anton    4892: 
1.44      crook    4893: 
1.21      crook    4894: doc-d>f
                   4895: doc-f>d
1.1       anton    4896: doc-f+
                   4897: doc-f-
                   4898: doc-f*
                   4899: doc-f/
                   4900: doc-fnegate
                   4901: doc-fabs
                   4902: doc-fmax
                   4903: doc-fmin
                   4904: doc-floor
                   4905: doc-fround
                   4906: doc-f**
                   4907: doc-fsqrt
                   4908: doc-fexp
                   4909: doc-fexpm1
                   4910: doc-fln
                   4911: doc-flnp1
                   4912: doc-flog
                   4913: doc-falog
1.32      anton    4914: doc-f2*
                   4915: doc-f2/
                   4916: doc-1/f
                   4917: doc-precision
                   4918: doc-set-precision
                   4919: 
                   4920: @cindex angles in trigonometric operations
                   4921: @cindex trigonometric operations
                   4922: Angles in floating point operations are given in radians (a full circle
                   4923: has 2 pi radians).
                   4924: 
1.1       anton    4925: doc-fsin
                   4926: doc-fcos
                   4927: doc-fsincos
                   4928: doc-ftan
                   4929: doc-fasin
                   4930: doc-facos
                   4931: doc-fatan
                   4932: doc-fatan2
                   4933: doc-fsinh
                   4934: doc-fcosh
                   4935: doc-ftanh
                   4936: doc-fasinh
                   4937: doc-facosh
                   4938: doc-fatanh
1.21      crook    4939: doc-pi
1.28      crook    4940: 
1.32      anton    4941: @cindex equality of floats
                   4942: @cindex floating-point comparisons
1.31      anton    4943: One particular problem with floating-point arithmetic is that comparison
                   4944: for equality often fails when you would expect it to succeed.  For this
                   4945: reason approximate equality is often preferred (but you still have to
1.67      anton    4946: know what you are doing).  Also note that IEEE NaNs may compare
1.68      anton    4947: differently from what you might expect.  The comparison words are:
1.31      anton    4948: 
                   4949: doc-f~rel
                   4950: doc-f~abs
1.68      anton    4951: doc-f~
1.31      anton    4952: doc-f=
                   4953: doc-f<>
                   4954: 
                   4955: doc-f<
                   4956: doc-f<=
                   4957: doc-f>
                   4958: doc-f>=
                   4959: 
1.21      crook    4960: doc-f0<
1.28      crook    4961: doc-f0<=
                   4962: doc-f0<>
1.21      crook    4963: doc-f0=
1.28      crook    4964: doc-f0>
                   4965: doc-f0>=
                   4966: 
1.1       anton    4967: 
                   4968: @node Stack Manipulation, Memory, Arithmetic, Words
                   4969: @section Stack Manipulation
                   4970: @cindex stack manipulation words
                   4971: 
                   4972: @cindex floating-point stack in the standard
1.21      crook    4973: Gforth maintains a number of separate stacks:
                   4974: 
1.29      crook    4975: @cindex data stack
                   4976: @cindex parameter stack
1.21      crook    4977: @itemize @bullet
                   4978: @item
1.29      crook    4979: A data stack (also known as the @dfn{parameter stack}) -- for
                   4980: characters, cells, addresses, and double cells.
1.21      crook    4981: 
1.29      crook    4982: @cindex floating-point stack
1.21      crook    4983: @item
1.44      crook    4984: A floating point stack -- for holding floating point (FP) numbers.
1.21      crook    4985: 
1.29      crook    4986: @cindex return stack
1.21      crook    4987: @item
1.44      crook    4988: A return stack -- for holding the return addresses of colon
1.32      anton    4989: definitions and other (non-FP) data.
1.21      crook    4990: 
1.29      crook    4991: @cindex locals stack
1.21      crook    4992: @item
1.44      crook    4993: A locals stack -- for holding local variables.
1.21      crook    4994: @end itemize
                   4995: 
1.1       anton    4996: @menu
                   4997: * Data stack::                  
                   4998: * Floating point stack::        
                   4999: * Return stack::                
                   5000: * Locals stack::                
                   5001: * Stack pointer manipulation::  
                   5002: @end menu
                   5003: 
                   5004: @node Data stack, Floating point stack, Stack Manipulation, Stack Manipulation
                   5005: @subsection Data stack
                   5006: @cindex data stack manipulation words
                   5007: @cindex stack manipulations words, data stack
                   5008: 
1.44      crook    5009: 
1.1       anton    5010: doc-drop
                   5011: doc-nip
                   5012: doc-dup
                   5013: doc-over
                   5014: doc-tuck
                   5015: doc-swap
1.21      crook    5016: doc-pick
1.1       anton    5017: doc-rot
                   5018: doc--rot
                   5019: doc-?dup
                   5020: doc-roll
                   5021: doc-2drop
                   5022: doc-2nip
                   5023: doc-2dup
                   5024: doc-2over
                   5025: doc-2tuck
                   5026: doc-2swap
                   5027: doc-2rot
                   5028: 
1.44      crook    5029: 
1.1       anton    5030: @node Floating point stack, Return stack, Data stack, Stack Manipulation
                   5031: @subsection Floating point stack
                   5032: @cindex floating-point stack manipulation words
                   5033: @cindex stack manipulation words, floating-point stack
                   5034: 
1.32      anton    5035: Whilst every sane Forth has a separate floating-point stack, it is not
                   5036: strictly required; an ANS Forth system could theoretically keep
                   5037: floating-point numbers on the data stack. As an additional difficulty,
                   5038: you don't know how many cells a floating-point number takes. It is
                   5039: reportedly possible to write words in a way that they work also for a
                   5040: unified stack model, but we do not recommend trying it. Instead, just
                   5041: say that your program has an environmental dependency on a separate
                   5042: floating-point stack.
                   5043: 
                   5044: doc-floating-stack
                   5045: 
1.1       anton    5046: doc-fdrop
                   5047: doc-fnip
                   5048: doc-fdup
                   5049: doc-fover
                   5050: doc-ftuck
                   5051: doc-fswap
1.21      crook    5052: doc-fpick
1.1       anton    5053: doc-frot
                   5054: 
1.44      crook    5055: 
1.1       anton    5056: @node Return stack, Locals stack, Floating point stack, Stack Manipulation
                   5057: @subsection Return stack
                   5058: @cindex return stack manipulation words
                   5059: @cindex stack manipulation words, return stack
                   5060: 
1.32      anton    5061: @cindex return stack and locals
                   5062: @cindex locals and return stack
                   5063: A Forth system is allowed to keep local variables on the
                   5064: return stack. This is reasonable, as local variables usually eliminate
                   5065: the need to use the return stack explicitly. So, if you want to produce
                   5066: a standard compliant program and you are using local variables in a
                   5067: word, forget about return stack manipulations in that word (refer to the
                   5068: standard document for the exact rules).
                   5069: 
1.1       anton    5070: doc->r
                   5071: doc-r>
                   5072: doc-r@
                   5073: doc-rdrop
                   5074: doc-2>r
                   5075: doc-2r>
                   5076: doc-2r@
                   5077: doc-2rdrop
                   5078: 
1.44      crook    5079: 
1.1       anton    5080: @node Locals stack, Stack pointer manipulation, Return stack, Stack Manipulation
                   5081: @subsection Locals stack
                   5082: 
1.78      anton    5083: Gforth uses an extra locals stack.  It is described, along with the
                   5084: reasons for its existence, in @ref{Locals implementation}.
1.21      crook    5085: 
1.1       anton    5086: @node Stack pointer manipulation,  , Locals stack, Stack Manipulation
                   5087: @subsection Stack pointer manipulation
                   5088: @cindex stack pointer manipulation words
                   5089: 
1.44      crook    5090: @c removed s0 r0 l0 -- they are obsolete aliases for sp0 rp0 lp0
1.21      crook    5091: doc-sp0
1.1       anton    5092: doc-sp@
                   5093: doc-sp!
1.21      crook    5094: doc-fp0
1.1       anton    5095: doc-fp@
                   5096: doc-fp!
1.21      crook    5097: doc-rp0
1.1       anton    5098: doc-rp@
                   5099: doc-rp!
1.21      crook    5100: doc-lp0
1.1       anton    5101: doc-lp@
                   5102: doc-lp!
                   5103: 
1.44      crook    5104: 
1.1       anton    5105: @node Memory, Control Structures, Stack Manipulation, Words
                   5106: @section Memory
1.26      crook    5107: @cindex memory words
1.1       anton    5108: 
1.32      anton    5109: @menu
                   5110: * Memory model::                
                   5111: * Dictionary allocation::       
                   5112: * Heap Allocation::             
                   5113: * Memory Access::               
                   5114: * Address arithmetic::          
                   5115: * Memory Blocks::               
                   5116: @end menu
                   5117: 
1.67      anton    5118: In addition to the standard Forth memory allocation words, there is also
                   5119: a @uref{http://www.complang.tuwien.ac.at/forth/garbage-collection.zip,
                   5120: garbage collector}.
                   5121: 
1.32      anton    5122: @node Memory model, Dictionary allocation, Memory, Memory
                   5123: @subsection ANS Forth and Gforth memory models
                   5124: 
                   5125: @c The ANS Forth description is a mess (e.g., is the heap part of
                   5126: @c the dictionary?), so let's not stick to closely with it.
                   5127: 
1.67      anton    5128: ANS Forth considers a Forth system as consisting of several address
                   5129: spaces, of which only @dfn{data space} is managed and accessible with
                   5130: the memory words.  Memory not necessarily in data space includes the
                   5131: stacks, the code (called code space) and the headers (called name
                   5132: space). In Gforth everything is in data space, but the code for the
                   5133: primitives is usually read-only.
1.32      anton    5134: 
                   5135: Data space is divided into a number of areas: The (data space portion of
                   5136: the) dictionary@footnote{Sometimes, the term @dfn{dictionary} is used to
                   5137: refer to the search data structure embodied in word lists and headers,
                   5138: because it is used for looking up names, just as you would in a
                   5139: conventional dictionary.}, the heap, and a number of system-allocated
                   5140: buffers.
                   5141: 
1.68      anton    5142: @cindex address arithmetic restrictions, ANS vs. Gforth
                   5143: @cindex contiguous regions, ANS vs. Gforth
1.32      anton    5144: In ANS Forth data space is also divided into contiguous regions.  You
                   5145: can only use address arithmetic within a contiguous region, not between
                   5146: them.  Usually each allocation gives you one contiguous region, but the
1.33      anton    5147: dictionary allocation words have additional rules (@pxref{Dictionary
1.32      anton    5148: allocation}).
                   5149: 
                   5150: Gforth provides one big address space, and address arithmetic can be
                   5151: performed between any addresses. However, in the dictionary headers or
                   5152: code are interleaved with data, so almost the only contiguous data space
                   5153: regions there are those described by ANS Forth as contiguous; but you
                   5154: can be sure that the dictionary is allocated towards increasing
                   5155: addresses even between contiguous regions.  The memory order of
                   5156: allocations in the heap is platform-dependent (and possibly different
                   5157: from one run to the next).
                   5158: 
1.27      crook    5159: 
1.32      anton    5160: @node Dictionary allocation, Heap Allocation, Memory model, Memory
                   5161: @subsection Dictionary allocation
1.27      crook    5162: @cindex reserving data space
                   5163: @cindex data space - reserving some
                   5164: 
1.32      anton    5165: Dictionary allocation is a stack-oriented allocation scheme, i.e., if
                   5166: you want to deallocate X, you also deallocate everything
                   5167: allocated after X.
                   5168: 
1.68      anton    5169: @cindex contiguous regions in dictionary allocation
1.32      anton    5170: The allocations using the words below are contiguous and grow the region
                   5171: towards increasing addresses.  Other words that allocate dictionary
                   5172: memory of any kind (i.e., defining words including @code{:noname}) end
                   5173: the contiguous region and start a new one.
                   5174: 
                   5175: In ANS Forth only @code{create}d words are guaranteed to produce an
                   5176: address that is the start of the following contiguous region.  In
                   5177: particular, the cell allocated by @code{variable} is not guaranteed to
                   5178: be contiguous with following @code{allot}ed memory.
                   5179: 
                   5180: You can deallocate memory by using @code{allot} with a negative argument
                   5181: (with some restrictions, see @code{allot}). For larger deallocations use
                   5182: @code{marker}.
1.27      crook    5183: 
1.29      crook    5184: 
1.27      crook    5185: doc-here
                   5186: doc-unused
                   5187: doc-allot
                   5188: doc-c,
1.29      crook    5189: doc-f,
1.27      crook    5190: doc-,
                   5191: doc-2,
                   5192: 
1.32      anton    5193: Memory accesses have to be aligned (@pxref{Address arithmetic}). So of
                   5194: course you should allocate memory in an aligned way, too. I.e., before
                   5195: allocating allocating a cell, @code{here} must be cell-aligned, etc.
                   5196: The words below align @code{here} if it is not already.  Basically it is
                   5197: only already aligned for a type, if the last allocation was a multiple
                   5198: of the size of this type and if @code{here} was aligned for this type
                   5199: before.
                   5200: 
                   5201: After freshly @code{create}ing a word, @code{here} is @code{align}ed in
                   5202: ANS Forth (@code{maxalign}ed in Gforth).
                   5203: 
                   5204: doc-align
                   5205: doc-falign
                   5206: doc-sfalign
                   5207: doc-dfalign
                   5208: doc-maxalign
                   5209: doc-cfalign
                   5210: 
                   5211: 
                   5212: @node Heap Allocation, Memory Access, Dictionary allocation, Memory
                   5213: @subsection Heap allocation
                   5214: @cindex heap allocation
                   5215: @cindex dynamic allocation of memory
                   5216: @cindex memory-allocation word set
                   5217: 
1.68      anton    5218: @cindex contiguous regions and heap allocation
1.32      anton    5219: Heap allocation supports deallocation of allocated memory in any
                   5220: order. Dictionary allocation is not affected by it (i.e., it does not
                   5221: end a contiguous region). In Gforth, these words are implemented using
                   5222: the standard C library calls malloc(), free() and resize().
                   5223: 
1.68      anton    5224: The memory region produced by one invocation of @code{allocate} or
                   5225: @code{resize} is internally contiguous.  There is no contiguity between
                   5226: such a region and any other region (including others allocated from the
                   5227: heap).
                   5228: 
1.32      anton    5229: doc-allocate
                   5230: doc-free
                   5231: doc-resize
                   5232: 
1.27      crook    5233: 
1.32      anton    5234: @node Memory Access, Address arithmetic, Heap Allocation, Memory
1.1       anton    5235: @subsection Memory Access
                   5236: @cindex memory access words
                   5237: 
                   5238: doc-@
                   5239: doc-!
                   5240: doc-+!
                   5241: doc-c@
                   5242: doc-c!
                   5243: doc-2@
                   5244: doc-2!
                   5245: doc-f@
                   5246: doc-f!
                   5247: doc-sf@
                   5248: doc-sf!
                   5249: doc-df@
                   5250: doc-df!
                   5251: 
1.68      anton    5252: 
1.32      anton    5253: @node Address arithmetic, Memory Blocks, Memory Access, Memory
                   5254: @subsection Address arithmetic
1.1       anton    5255: @cindex address arithmetic words
                   5256: 
1.67      anton    5257: Address arithmetic is the foundation on which you can build data
                   5258: structures like arrays, records (@pxref{Structures}) and objects
                   5259: (@pxref{Object-oriented Forth}).
1.32      anton    5260: 
1.68      anton    5261: @cindex address unit
                   5262: @cindex au (address unit)
1.1       anton    5263: ANS Forth does not specify the sizes of the data types. Instead, it
                   5264: offers a number of words for computing sizes and doing address
1.29      crook    5265: arithmetic. Address arithmetic is performed in terms of address units
                   5266: (aus); on most systems the address unit is one byte. Note that a
                   5267: character may have more than one au, so @code{chars} is no noop (on
1.68      anton    5268: platforms where it is a noop, it compiles to nothing).
1.1       anton    5269: 
1.67      anton    5270: The basic address arithmetic words are @code{+} and @code{-}.  E.g., if
                   5271: you have the address of a cell, perform @code{1 cells +}, and you will
                   5272: have the address of the next cell.
                   5273: 
1.68      anton    5274: @cindex contiguous regions and address arithmetic
1.67      anton    5275: In ANS Forth you can perform address arithmetic only within a contiguous
                   5276: region, i.e., if you have an address into one region, you can only add
                   5277: and subtract such that the result is still within the region; you can
                   5278: only subtract or compare addresses from within the same contiguous
                   5279: region.  Reasons: several contiguous regions can be arranged in memory
                   5280: in any way; on segmented systems addresses may have unusual
                   5281: representations, such that address arithmetic only works within a
                   5282: region.  Gforth provides a few more guarantees (linear address space,
                   5283: dictionary grows upwards), but in general I have found it easy to stay
                   5284: within contiguous regions (exception: computing and comparing to the
                   5285: address just beyond the end of an array).
                   5286: 
1.1       anton    5287: @cindex alignment of addresses for types
                   5288: ANS Forth also defines words for aligning addresses for specific
                   5289: types. Many computers require that accesses to specific data types
                   5290: must only occur at specific addresses; e.g., that cells may only be
                   5291: accessed at addresses divisible by 4. Even if a machine allows unaligned
                   5292: accesses, it can usually perform aligned accesses faster. 
                   5293: 
                   5294: For the performance-conscious: alignment operations are usually only
                   5295: necessary during the definition of a data structure, not during the
                   5296: (more frequent) accesses to it.
                   5297: 
                   5298: ANS Forth defines no words for character-aligning addresses. This is not
                   5299: an oversight, but reflects the fact that addresses that are not
                   5300: char-aligned have no use in the standard and therefore will not be
                   5301: created.
                   5302: 
                   5303: @cindex @code{CREATE} and alignment
1.29      crook    5304: ANS Forth guarantees that addresses returned by @code{CREATE}d words
1.1       anton    5305: are cell-aligned; in addition, Gforth guarantees that these addresses
                   5306: are aligned for all purposes.
                   5307: 
1.26      crook    5308: Note that the ANS Forth word @code{char} has nothing to do with address
                   5309: arithmetic.
1.1       anton    5310: 
1.44      crook    5311: 
1.1       anton    5312: doc-chars
                   5313: doc-char+
                   5314: doc-cells
                   5315: doc-cell+
                   5316: doc-cell
                   5317: doc-aligned
                   5318: doc-floats
                   5319: doc-float+
                   5320: doc-float
                   5321: doc-faligned
                   5322: doc-sfloats
                   5323: doc-sfloat+
                   5324: doc-sfaligned
                   5325: doc-dfloats
                   5326: doc-dfloat+
                   5327: doc-dfaligned
                   5328: doc-maxaligned
                   5329: doc-cfaligned
                   5330: doc-address-unit-bits
                   5331: 
1.44      crook    5332: 
1.32      anton    5333: @node Memory Blocks,  , Address arithmetic, Memory
1.1       anton    5334: @subsection Memory Blocks
                   5335: @cindex memory block words
1.27      crook    5336: @cindex character strings - moving and copying
                   5337: 
1.49      anton    5338: Memory blocks often represent character strings; For ways of storing
                   5339: character strings in memory see @ref{String Formats}.  For other
                   5340: string-processing words see @ref{Displaying characters and strings}.
1.1       anton    5341: 
1.67      anton    5342: A few of these words work on address unit blocks.  In that case, you
                   5343: usually have to insert @code{CHARS} before the word when working on
                   5344: character strings.  Most words work on character blocks, and expect a
                   5345: char-aligned address.
                   5346: 
                   5347: When copying characters between overlapping memory regions, use
                   5348: @code{chars move} or choose carefully between @code{cmove} and
                   5349: @code{cmove>}.
1.44      crook    5350: 
1.1       anton    5351: doc-move
                   5352: doc-erase
                   5353: doc-cmove
                   5354: doc-cmove>
                   5355: doc-fill
                   5356: doc-blank
1.21      crook    5357: doc-compare
                   5358: doc-search
1.27      crook    5359: doc--trailing
                   5360: doc-/string
1.82      anton    5361: doc-bounds
1.44      crook    5362: 
1.27      crook    5363: @comment TODO examples
                   5364: 
1.1       anton    5365: 
1.26      crook    5366: @node Control Structures, Defining Words, Memory, Words
1.1       anton    5367: @section Control Structures
                   5368: @cindex control structures
                   5369: 
1.33      anton    5370: Control structures in Forth cannot be used interpretively, only in a
                   5371: colon definition@footnote{To be precise, they have no interpretation
                   5372: semantics (@pxref{Interpretation and Compilation Semantics}).}. We do
                   5373: not like this limitation, but have not seen a satisfying way around it
                   5374: yet, although many schemes have been proposed.
1.1       anton    5375: 
                   5376: @menu
1.33      anton    5377: * Selection::                   IF ... ELSE ... ENDIF
                   5378: * Simple Loops::                BEGIN ...
1.29      crook    5379: * Counted Loops::               DO
1.67      anton    5380: * Arbitrary control structures::  
                   5381: * Calls and returns::           
1.1       anton    5382: * Exception Handling::          
                   5383: @end menu
                   5384: 
                   5385: @node Selection, Simple Loops, Control Structures, Control Structures
                   5386: @subsection Selection
                   5387: @cindex selection control structures
                   5388: @cindex control structures for selection
                   5389: 
                   5390: @cindex @code{IF} control structure
                   5391: @example
1.29      crook    5392: @i{flag}
1.1       anton    5393: IF
1.29      crook    5394:   @i{code}
1.1       anton    5395: ENDIF
                   5396: @end example
1.21      crook    5397: @noindent
1.33      anton    5398: 
1.44      crook    5399: If @i{flag} is non-zero (as far as @code{IF} etc. are concerned, a cell
                   5400: with any bit set represents truth) @i{code} is executed.
1.33      anton    5401: 
1.1       anton    5402: @example
1.29      crook    5403: @i{flag}
1.1       anton    5404: IF
1.29      crook    5405:   @i{code1}
1.1       anton    5406: ELSE
1.29      crook    5407:   @i{code2}
1.1       anton    5408: ENDIF
                   5409: @end example
                   5410: 
1.44      crook    5411: If @var{flag} is true, @i{code1} is executed, otherwise @i{code2} is
                   5412: executed.
1.33      anton    5413: 
1.1       anton    5414: You can use @code{THEN} instead of @code{ENDIF}. Indeed, @code{THEN} is
                   5415: standard, and @code{ENDIF} is not, although it is quite popular. We
                   5416: recommend using @code{ENDIF}, because it is less confusing for people
                   5417: who also know other languages (and is not prone to reinforcing negative
                   5418: prejudices against Forth in these people). Adding @code{ENDIF} to a
                   5419: system that only supplies @code{THEN} is simple:
                   5420: @example
1.82      anton    5421: : ENDIF   POSTPONE then ; immediate
1.1       anton    5422: @end example
                   5423: 
                   5424: [According to @cite{Webster's New Encyclopedic Dictionary}, @dfn{then
                   5425: (adv.)}  has the following meanings:
                   5426: @quotation
                   5427: ... 2b: following next after in order ... 3d: as a necessary consequence
                   5428: (if you were there, then you saw them).
                   5429: @end quotation
                   5430: Forth's @code{THEN} has the meaning 2b, whereas @code{THEN} in Pascal
                   5431: and many other programming languages has the meaning 3d.]
                   5432: 
1.21      crook    5433: Gforth also provides the words @code{?DUP-IF} and @code{?DUP-0=-IF}, so
1.1       anton    5434: you can avoid using @code{?dup}. Using these alternatives is also more
1.26      crook    5435: efficient than using @code{?dup}. Definitions in ANS Forth
1.1       anton    5436: for @code{ENDIF}, @code{?DUP-IF} and @code{?DUP-0=-IF} are provided in
                   5437: @file{compat/control.fs}.
                   5438: 
                   5439: @cindex @code{CASE} control structure
                   5440: @example
1.29      crook    5441: @i{n}
1.1       anton    5442: CASE
1.29      crook    5443:   @i{n1} OF @i{code1} ENDOF
                   5444:   @i{n2} OF @i{code2} ENDOF
1.1       anton    5445:   @dots{}
1.68      anton    5446:   ( n ) @i{default-code} ( n )
1.1       anton    5447: ENDCASE
                   5448: @end example
                   5449: 
1.68      anton    5450: Executes the first @i{codei}, where the @i{ni} is equal to @i{n}.  If no
                   5451: @i{ni} matches, the optional @i{default-code} is executed. The optional
                   5452: default case can be added by simply writing the code after the last
                   5453: @code{ENDOF}. It may use @i{n}, which is on top of the stack, but must
                   5454: not consume it.
1.1       anton    5455: 
1.69      anton    5456: @progstyle
                   5457: To keep the code understandable, you should ensure that on all paths
                   5458: through a selection construct the stack is changed in the same way
                   5459: (wrt. number and types of stack items consumed and pushed).
                   5460: 
1.1       anton    5461: @node Simple Loops, Counted Loops, Selection, Control Structures
                   5462: @subsection Simple Loops
                   5463: @cindex simple loops
                   5464: @cindex loops without count 
                   5465: 
                   5466: @cindex @code{WHILE} loop
                   5467: @example
                   5468: BEGIN
1.29      crook    5469:   @i{code1}
                   5470:   @i{flag}
1.1       anton    5471: WHILE
1.29      crook    5472:   @i{code2}
1.1       anton    5473: REPEAT
                   5474: @end example
                   5475: 
1.29      crook    5476: @i{code1} is executed and @i{flag} is computed. If it is true,
                   5477: @i{code2} is executed and the loop is restarted; If @i{flag} is
1.1       anton    5478: false, execution continues after the @code{REPEAT}.
                   5479: 
                   5480: @cindex @code{UNTIL} loop
                   5481: @example
                   5482: BEGIN
1.29      crook    5483:   @i{code}
                   5484:   @i{flag}
1.1       anton    5485: UNTIL
                   5486: @end example
                   5487: 
1.29      crook    5488: @i{code} is executed. The loop is restarted if @code{flag} is false.
1.1       anton    5489: 
1.69      anton    5490: @progstyle
                   5491: To keep the code understandable, a complete iteration of the loop should
                   5492: not change the number and types of the items on the stacks.
                   5493: 
1.1       anton    5494: @cindex endless loop
                   5495: @cindex loops, endless
                   5496: @example
                   5497: BEGIN
1.29      crook    5498:   @i{code}
1.1       anton    5499: AGAIN
                   5500: @end example
                   5501: 
                   5502: This is an endless loop.
                   5503: 
                   5504: @node Counted Loops, Arbitrary control structures, Simple Loops, Control Structures
                   5505: @subsection Counted Loops
                   5506: @cindex counted loops
                   5507: @cindex loops, counted
                   5508: @cindex @code{DO} loops
                   5509: 
                   5510: The basic counted loop is:
                   5511: @example
1.29      crook    5512: @i{limit} @i{start}
1.1       anton    5513: ?DO
1.29      crook    5514:   @i{body}
1.1       anton    5515: LOOP
                   5516: @end example
                   5517: 
1.29      crook    5518: This performs one iteration for every integer, starting from @i{start}
                   5519: and up to, but excluding @i{limit}. The counter, or @i{index}, can be
1.21      crook    5520: accessed with @code{i}. For example, the loop:
1.1       anton    5521: @example
                   5522: 10 0 ?DO
                   5523:   i .
                   5524: LOOP
                   5525: @end example
1.21      crook    5526: @noindent
                   5527: prints @code{0 1 2 3 4 5 6 7 8 9}
                   5528: 
1.1       anton    5529: The index of the innermost loop can be accessed with @code{i}, the index
                   5530: of the next loop with @code{j}, and the index of the third loop with
                   5531: @code{k}.
                   5532: 
1.44      crook    5533: 
1.1       anton    5534: doc-i
                   5535: doc-j
                   5536: doc-k
                   5537: 
1.44      crook    5538: 
1.1       anton    5539: The loop control data are kept on the return stack, so there are some
1.21      crook    5540: restrictions on mixing return stack accesses and counted loop words. In
                   5541: particuler, if you put values on the return stack outside the loop, you
                   5542: cannot read them inside the loop@footnote{well, not in a way that is
                   5543: portable.}. If you put values on the return stack within a loop, you
                   5544: have to remove them before the end of the loop and before accessing the
                   5545: index of the loop.
1.1       anton    5546: 
                   5547: There are several variations on the counted loop:
                   5548: 
1.21      crook    5549: @itemize @bullet
                   5550: @item
                   5551: @code{LEAVE} leaves the innermost counted loop immediately; execution
                   5552: continues after the associated @code{LOOP} or @code{NEXT}. For example:
                   5553: 
                   5554: @example
                   5555: 10 0 ?DO  i DUP . 3 = IF LEAVE THEN LOOP
                   5556: @end example
                   5557: prints @code{0 1 2 3}
                   5558: 
1.1       anton    5559: 
1.21      crook    5560: @item
                   5561: @code{UNLOOP} prepares for an abnormal loop exit, e.g., via
                   5562: @code{EXIT}. @code{UNLOOP} removes the loop control parameters from the
                   5563: return stack so @code{EXIT} can get to its return address. For example:
                   5564: 
                   5565: @example
                   5566: : demo 10 0 ?DO i DUP . 3 = IF UNLOOP EXIT THEN LOOP ." Done" ;
                   5567: @end example
                   5568: prints @code{0 1 2 3}
                   5569: 
                   5570: 
                   5571: @item
1.29      crook    5572: If @i{start} is greater than @i{limit}, a @code{?DO} loop is entered
1.1       anton    5573: (and @code{LOOP} iterates until they become equal by wrap-around
                   5574: arithmetic). This behaviour is usually not what you want. Therefore,
                   5575: Gforth offers @code{+DO} and @code{U+DO} (as replacements for
1.29      crook    5576: @code{?DO}), which do not enter the loop if @i{start} is greater than
                   5577: @i{limit}; @code{+DO} is for signed loop parameters, @code{U+DO} for
1.1       anton    5578: unsigned loop parameters.
                   5579: 
1.21      crook    5580: @item
                   5581: @code{?DO} can be replaced by @code{DO}. @code{DO} always enters
                   5582: the loop, independent of the loop parameters. Do not use @code{DO}, even
                   5583: if you know that the loop is entered in any case. Such knowledge tends
                   5584: to become invalid during maintenance of a program, and then the
                   5585: @code{DO} will make trouble.
                   5586: 
                   5587: @item
1.29      crook    5588: @code{LOOP} can be replaced with @code{@i{n} +LOOP}; this updates the
                   5589: index by @i{n} instead of by 1. The loop is terminated when the border
                   5590: between @i{limit-1} and @i{limit} is crossed. E.g.:
1.1       anton    5591: 
1.21      crook    5592: @example
                   5593: 4 0 +DO  i .  2 +LOOP
                   5594: @end example
                   5595: @noindent
                   5596: prints @code{0 2}
                   5597: 
                   5598: @example
                   5599: 4 1 +DO  i .  2 +LOOP
                   5600: @end example
                   5601: @noindent
                   5602: prints @code{1 3}
1.1       anton    5603: 
1.68      anton    5604: @item
1.1       anton    5605: @cindex negative increment for counted loops
                   5606: @cindex counted loops with negative increment
1.29      crook    5607: The behaviour of @code{@i{n} +LOOP} is peculiar when @i{n} is negative:
1.1       anton    5608: 
1.21      crook    5609: @example
                   5610: -1 0 ?DO  i .  -1 +LOOP
                   5611: @end example
                   5612: @noindent
                   5613: prints @code{0 -1}
1.1       anton    5614: 
1.21      crook    5615: @example
                   5616: 0 0 ?DO  i .  -1 +LOOP
                   5617: @end example
                   5618: prints nothing.
1.1       anton    5619: 
1.29      crook    5620: Therefore we recommend avoiding @code{@i{n} +LOOP} with negative
                   5621: @i{n}. One alternative is @code{@i{u} -LOOP}, which reduces the
                   5622: index by @i{u} each iteration. The loop is terminated when the border
                   5623: between @i{limit+1} and @i{limit} is crossed. Gforth also provides
1.1       anton    5624: @code{-DO} and @code{U-DO} for down-counting loops. E.g.:
                   5625: 
1.21      crook    5626: @example
                   5627: -2 0 -DO  i .  1 -LOOP
                   5628: @end example
                   5629: @noindent
                   5630: prints @code{0 -1}
1.1       anton    5631: 
1.21      crook    5632: @example
                   5633: -1 0 -DO  i .  1 -LOOP
                   5634: @end example
                   5635: @noindent
                   5636: prints @code{0}
                   5637: 
                   5638: @example
                   5639: 0 0 -DO  i .  1 -LOOP
                   5640: @end example
                   5641: @noindent
                   5642: prints nothing.
1.1       anton    5643: 
1.21      crook    5644: @end itemize
1.1       anton    5645: 
                   5646: Unfortunately, @code{+DO}, @code{U+DO}, @code{-DO}, @code{U-DO} and
1.26      crook    5647: @code{-LOOP} are not defined in ANS Forth. However, an implementation
                   5648: for these words that uses only standard words is provided in
                   5649: @file{compat/loops.fs}.
1.1       anton    5650: 
                   5651: 
                   5652: @cindex @code{FOR} loops
1.26      crook    5653: Another counted loop is:
1.1       anton    5654: @example
1.29      crook    5655: @i{n}
1.1       anton    5656: FOR
1.29      crook    5657:   @i{body}
1.1       anton    5658: NEXT
                   5659: @end example
                   5660: This is the preferred loop of native code compiler writers who are too
1.26      crook    5661: lazy to optimize @code{?DO} loops properly. This loop structure is not
1.29      crook    5662: defined in ANS Forth. In Gforth, this loop iterates @i{n+1} times;
                   5663: @code{i} produces values starting with @i{n} and ending with 0. Other
1.26      crook    5664: Forth systems may behave differently, even if they support @code{FOR}
                   5665: loops. To avoid problems, don't use @code{FOR} loops.
1.1       anton    5666: 
                   5667: @node Arbitrary control structures, Calls and returns, Counted Loops, Control Structures
                   5668: @subsection Arbitrary control structures
                   5669: @cindex control structures, user-defined
                   5670: 
                   5671: @cindex control-flow stack
                   5672: ANS Forth permits and supports using control structures in a non-nested
                   5673: way. Information about incomplete control structures is stored on the
                   5674: control-flow stack. This stack may be implemented on the Forth data
                   5675: stack, and this is what we have done in Gforth.
                   5676: 
                   5677: @cindex @code{orig}, control-flow stack item
                   5678: @cindex @code{dest}, control-flow stack item
                   5679: An @i{orig} entry represents an unresolved forward branch, a @i{dest}
                   5680: entry represents a backward branch target. A few words are the basis for
                   5681: building any control structure possible (except control structures that
                   5682: need storage, like calls, coroutines, and backtracking).
                   5683: 
1.44      crook    5684: 
1.1       anton    5685: doc-if
                   5686: doc-ahead
                   5687: doc-then
                   5688: doc-begin
                   5689: doc-until
                   5690: doc-again
                   5691: doc-cs-pick
                   5692: doc-cs-roll
                   5693: 
1.44      crook    5694: 
1.21      crook    5695: The Standard words @code{CS-PICK} and @code{CS-ROLL} allow you to
                   5696: manipulate the control-flow stack in a portable way. Without them, you
                   5697: would need to know how many stack items are occupied by a control-flow
                   5698: entry (many systems use one cell. In Gforth they currently take three,
                   5699: but this may change in the future).
                   5700: 
1.1       anton    5701: Some standard control structure words are built from these words:
                   5702: 
1.44      crook    5703: 
1.1       anton    5704: doc-else
                   5705: doc-while
                   5706: doc-repeat
                   5707: 
1.44      crook    5708: 
                   5709: @noindent
1.1       anton    5710: Gforth adds some more control-structure words:
                   5711: 
1.44      crook    5712: 
1.1       anton    5713: doc-endif
                   5714: doc-?dup-if
                   5715: doc-?dup-0=-if
                   5716: 
1.44      crook    5717: 
                   5718: @noindent
1.1       anton    5719: Counted loop words constitute a separate group of words:
                   5720: 
1.44      crook    5721: 
1.1       anton    5722: doc-?do
                   5723: doc-+do
                   5724: doc-u+do
                   5725: doc--do
                   5726: doc-u-do
                   5727: doc-do
                   5728: doc-for
                   5729: doc-loop
                   5730: doc-+loop
                   5731: doc--loop
                   5732: doc-next
                   5733: doc-leave
                   5734: doc-?leave
                   5735: doc-unloop
                   5736: doc-done
                   5737: 
1.44      crook    5738: 
1.21      crook    5739: The standard does not allow using @code{CS-PICK} and @code{CS-ROLL} on
                   5740: @i{do-sys}. Gforth allows it, but it's your job to ensure that for
1.1       anton    5741: every @code{?DO} etc. there is exactly one @code{UNLOOP} on any path
                   5742: through the definition (@code{LOOP} etc. compile an @code{UNLOOP} on the
                   5743: fall-through path). Also, you have to ensure that all @code{LEAVE}s are
                   5744: resolved (by using one of the loop-ending words or @code{DONE}).
                   5745: 
1.44      crook    5746: @noindent
1.26      crook    5747: Another group of control structure words are:
1.1       anton    5748: 
1.44      crook    5749: 
1.1       anton    5750: doc-case
                   5751: doc-endcase
                   5752: doc-of
                   5753: doc-endof
                   5754: 
1.44      crook    5755: 
1.21      crook    5756: @i{case-sys} and @i{of-sys} cannot be processed using @code{CS-PICK} and
                   5757: @code{CS-ROLL}.
1.1       anton    5758: 
                   5759: @subsubsection Programming Style
1.47      crook    5760: @cindex control structures programming style
                   5761: @cindex programming style, arbitrary control structures
1.1       anton    5762: 
                   5763: In order to ensure readability we recommend that you do not create
                   5764: arbitrary control structures directly, but define new control structure
                   5765: words for the control structure you want and use these words in your
1.26      crook    5766: program. For example, instead of writing:
1.1       anton    5767: 
                   5768: @example
1.26      crook    5769: BEGIN
1.1       anton    5770:   ...
1.26      crook    5771: IF [ 1 CS-ROLL ]
1.1       anton    5772:   ...
1.26      crook    5773: AGAIN THEN
1.1       anton    5774: @end example
                   5775: 
1.21      crook    5776: @noindent
1.1       anton    5777: we recommend defining control structure words, e.g.,
                   5778: 
                   5779: @example
1.26      crook    5780: : WHILE ( DEST -- ORIG DEST )
                   5781:  POSTPONE IF
                   5782:  1 CS-ROLL ; immediate
                   5783: 
                   5784: : REPEAT ( orig dest -- )
                   5785:  POSTPONE AGAIN
                   5786:  POSTPONE THEN ; immediate
1.1       anton    5787: @end example
                   5788: 
1.21      crook    5789: @noindent
1.1       anton    5790: and then using these to create the control structure:
                   5791: 
                   5792: @example
1.26      crook    5793: BEGIN
1.1       anton    5794:   ...
1.26      crook    5795: WHILE
1.1       anton    5796:   ...
1.26      crook    5797: REPEAT
1.1       anton    5798: @end example
                   5799: 
                   5800: That's much easier to read, isn't it? Of course, @code{REPEAT} and
                   5801: @code{WHILE} are predefined, so in this example it would not be
                   5802: necessary to define them.
                   5803: 
                   5804: @node Calls and returns, Exception Handling, Arbitrary control structures, Control Structures
                   5805: @subsection Calls and returns
                   5806: @cindex calling a definition
                   5807: @cindex returning from a definition
                   5808: 
1.3       anton    5809: @cindex recursive definitions
                   5810: A definition can be called simply be writing the name of the definition
1.26      crook    5811: to be called. Normally a definition is invisible during its own
1.3       anton    5812: definition. If you want to write a directly recursive definition, you
1.26      crook    5813: can use @code{recursive} to make the current definition visible, or
                   5814: @code{recurse} to call the current definition directly.
1.3       anton    5815: 
1.44      crook    5816: 
1.3       anton    5817: doc-recursive
                   5818: doc-recurse
                   5819: 
1.44      crook    5820: 
1.21      crook    5821: @comment TODO add example of the two recursion methods
1.12      anton    5822: @quotation
                   5823: @progstyle
                   5824: I prefer using @code{recursive} to @code{recurse}, because calling the
                   5825: definition by name is more descriptive (if the name is well-chosen) than
                   5826: the somewhat cryptic @code{recurse}.  E.g., in a quicksort
                   5827: implementation, it is much better to read (and think) ``now sort the
                   5828: partitions'' than to read ``now do a recursive call''.
                   5829: @end quotation
1.3       anton    5830: 
1.29      crook    5831: For mutual recursion, use @code{Defer}red words, like this:
1.3       anton    5832: 
                   5833: @example
1.28      crook    5834: Defer foo
1.3       anton    5835: 
                   5836: : bar ( ... -- ... )
                   5837:  ... foo ... ;
                   5838: 
                   5839: :noname ( ... -- ... )
                   5840:  ... bar ... ;
                   5841: IS foo
                   5842: @end example
                   5843: 
1.44      crook    5844: Deferred words are discussed in more detail in @ref{Deferred words}.
1.33      anton    5845: 
1.26      crook    5846: The current definition returns control to the calling definition when
1.33      anton    5847: the end of the definition is reached or @code{EXIT} is encountered.
1.1       anton    5848: 
                   5849: doc-exit
                   5850: doc-;s
                   5851: 
1.44      crook    5852: 
1.1       anton    5853: @node Exception Handling,  , Calls and returns, Control Structures
                   5854: @subsection Exception Handling
1.26      crook    5855: @cindex exceptions
1.1       anton    5856: 
1.68      anton    5857: @c quit is a very bad idea for error handling, 
                   5858: @c because it does not translate into a THROW
                   5859: @c it also does not belong into this chapter
                   5860: 
                   5861: If a word detects an error condition that it cannot handle, it can
                   5862: @code{throw} an exception.  In the simplest case, this will terminate
                   5863: your program, and report an appropriate error.
1.21      crook    5864: 
1.68      anton    5865: doc-throw
1.1       anton    5866: 
1.69      anton    5867: @code{Throw} consumes a cell-sized error number on the stack. There are
                   5868: some predefined error numbers in ANS Forth (see @file{errors.fs}).  In
                   5869: Gforth (and most other systems) you can use the iors produced by various
                   5870: words as error numbers (e.g., a typical use of @code{allocate} is
                   5871: @code{allocate throw}).  Gforth also provides the word @code{exception}
                   5872: to define your own error numbers (with decent error reporting); an ANS
                   5873: Forth version of this word (but without the error messages) is available
                   5874: in @code{compat/except.fs}.  And finally, you can use your own error
1.68      anton    5875: numbers (anything outside the range -4095..0), but won't get nice error
                   5876: messages, only numbers.  For example, try:
                   5877: 
                   5878: @example
1.69      anton    5879: -10 throw                    \ ANS defined
                   5880: -267 throw                   \ system defined
                   5881: s" my error" exception throw \ user defined
                   5882: 7 throw                      \ arbitrary number
1.68      anton    5883: @end example
                   5884: 
                   5885: doc---exception-exception
1.1       anton    5886: 
1.69      anton    5887: A common idiom to @code{THROW} a specific error if a flag is true is
                   5888: this:
                   5889: 
                   5890: @example
                   5891: @code{( flag ) 0<> @i{errno} and throw}
                   5892: @end example
                   5893: 
                   5894: Your program can provide exception handlers to catch exceptions.  An
                   5895: exception handler can be used to correct the problem, or to clean up
                   5896: some data structures and just throw the exception to the next exception
                   5897: handler.  Note that @code{throw} jumps to the dynamically innermost
                   5898: exception handler.  The system's exception handler is outermost, and just
                   5899: prints an error and restarts command-line interpretation (or, in batch
                   5900: mode (i.e., while processing the shell command line), leaves Gforth).
1.1       anton    5901: 
1.68      anton    5902: The ANS Forth way to catch exceptions is @code{catch}:
1.1       anton    5903: 
1.68      anton    5904: doc-catch
                   5905: 
                   5906: The most common use of exception handlers is to clean up the state when
                   5907: an error happens.  E.g.,
1.1       anton    5908: 
1.26      crook    5909: @example
1.68      anton    5910: base @ >r hex \ actually the hex should be inside foo, or we h
                   5911: ['] foo catch ( nerror|0 )
                   5912: r> base !
1.69      anton    5913: ( nerror|0 ) throw \ pass it on
1.26      crook    5914: @end example
1.1       anton    5915: 
1.69      anton    5916: A use of @code{catch} for handling the error @code{myerror} might look
                   5917: like this:
1.44      crook    5918: 
1.68      anton    5919: @example
1.69      anton    5920: ['] foo catch
                   5921: CASE
                   5922:   myerror OF ... ( do something about it ) ENDOF
                   5923:   dup throw \ default: pass other errors on, do nothing on non-errors
                   5924: ENDCASE
1.68      anton    5925: @end example
1.44      crook    5926: 
1.68      anton    5927: Having to wrap the code into a separate word is often cumbersome,
                   5928: therefore Gforth provides an alternative syntax:
1.1       anton    5929: 
                   5930: @example
1.69      anton    5931: TRY
1.68      anton    5932:   @i{code1}
1.69      anton    5933: RECOVER     \ optional
1.68      anton    5934:   @i{code2} \ optional
1.69      anton    5935: ENDTRY
1.1       anton    5936: @end example
                   5937: 
1.68      anton    5938: This performs @i{Code1}.  If @i{code1} completes normally, execution
                   5939: continues after the @code{endtry}.  If @i{Code1} throws, the stacks are
                   5940: reset to the state during @code{try}, the throw value is pushed on the
                   5941: data stack, and execution constinues at @i{code2}, and finally falls
                   5942: through the @code{endtry} into the following code. If there is no
                   5943: @code{recover} clause, this works like an empty recover clause.
1.26      crook    5944: 
1.68      anton    5945: doc-try
                   5946: doc-recover
                   5947: doc-endtry
1.26      crook    5948: 
1.69      anton    5949: The cleanup example from above in this syntax:
1.26      crook    5950: 
1.68      anton    5951: @example
1.69      anton    5952: base @ >r TRY
1.68      anton    5953:   hex foo \ now the hex is placed correctly
1.69      anton    5954:   0       \ value for throw
                   5955: ENDTRY
1.68      anton    5956: r> base ! throw
1.1       anton    5957: @end example
                   5958: 
1.69      anton    5959: And here's the error handling example:
1.1       anton    5960: 
1.68      anton    5961: @example
1.69      anton    5962: TRY
1.68      anton    5963:   foo
1.69      anton    5964: RECOVER
                   5965:   CASE
                   5966:     myerror OF ... ( do something about it ) ENDOF
                   5967:     throw \ pass other errors on
                   5968:   ENDCASE
                   5969: ENDTRY
1.68      anton    5970: @end example
1.1       anton    5971: 
1.69      anton    5972: @progstyle
                   5973: As usual, you should ensure that the stack depth is statically known at
                   5974: the end: either after the @code{throw} for passing on errors, or after
                   5975: the @code{ENDTRY} (or, if you use @code{catch}, after the end of the
                   5976: selection construct for handling the error).
                   5977: 
1.68      anton    5978: There are two alternatives to @code{throw}: @code{Abort"} is conditional
                   5979: and you can provide an error message.  @code{Abort} just produces an
                   5980: ``Aborted'' error.
1.1       anton    5981: 
1.68      anton    5982: The problem with these words is that exception handlers cannot
                   5983: differentiate between different @code{abort"}s; they just look like
                   5984: @code{-2 throw} to them (the error message cannot be accessed by
                   5985: standard programs).  Similar @code{abort} looks like @code{-1 throw} to
                   5986: exception handlers.
1.44      crook    5987: 
1.68      anton    5988: doc-abort"
1.26      crook    5989: doc-abort
1.29      crook    5990: 
                   5991: 
1.44      crook    5992: 
1.29      crook    5993: @c -------------------------------------------------------------
1.47      crook    5994: @node Defining Words, Interpretation and Compilation Semantics, Control Structures, Words
1.29      crook    5995: @section Defining Words
                   5996: @cindex defining words
                   5997: 
1.47      crook    5998: Defining words are used to extend Forth by creating new entries in the dictionary.
                   5999: 
1.29      crook    6000: @menu
1.67      anton    6001: * CREATE::                      
1.44      crook    6002: * Variables::                   Variables and user variables
1.67      anton    6003: * Constants::                   
1.44      crook    6004: * Values::                      Initialised variables
1.67      anton    6005: * Colon Definitions::           
1.44      crook    6006: * Anonymous Definitions::       Definitions without names
1.69      anton    6007: * Supplying names::             Passing definition names as strings
1.67      anton    6008: * User-defined Defining Words::  
1.44      crook    6009: * Deferred words::              Allow forward references
1.67      anton    6010: * Aliases::                     
1.29      crook    6011: @end menu
                   6012: 
1.44      crook    6013: @node CREATE, Variables, Defining Words, Defining Words
                   6014: @subsection @code{CREATE}
1.29      crook    6015: @cindex simple defining words
                   6016: @cindex defining words, simple
                   6017: 
                   6018: Defining words are used to create new entries in the dictionary. The
                   6019: simplest defining word is @code{CREATE}. @code{CREATE} is used like
                   6020: this:
                   6021: 
                   6022: @example
                   6023: CREATE new-word1
                   6024: @end example
                   6025: 
1.69      anton    6026: @code{CREATE} is a parsing word, i.e., it takes an argument from the
                   6027: input stream (@code{new-word1} in our example).  It generates a
                   6028: dictionary entry for @code{new-word1}. When @code{new-word1} is
                   6029: executed, all that it does is leave an address on the stack. The address
                   6030: represents the value of the data space pointer (@code{HERE}) at the time
                   6031: that @code{new-word1} was defined. Therefore, @code{CREATE} is a way of
                   6032: associating a name with the address of a region of memory.
1.29      crook    6033: 
1.34      anton    6034: doc-create
                   6035: 
1.69      anton    6036: Note that in ANS Forth guarantees only for @code{create} that its body
                   6037: is in dictionary data space (i.e., where @code{here}, @code{allot}
                   6038: etc. work, @pxref{Dictionary allocation}).  Also, in ANS Forth only
                   6039: @code{create}d words can be modified with @code{does>}
                   6040: (@pxref{User-defined Defining Words}).  And in ANS Forth @code{>body}
                   6041: can only be applied to @code{create}d words.
                   6042: 
1.29      crook    6043: By extending this example to reserve some memory in data space, we end
1.69      anton    6044: up with something like a @i{variable}. Here are two different ways to do
                   6045: it:
1.29      crook    6046: 
                   6047: @example
                   6048: CREATE new-word2 1 cells allot  \ reserve 1 cell - initial value undefined
                   6049: CREATE new-word3 4 ,            \ reserve 1 cell and initialise it (to 4)
                   6050: @end example
                   6051: 
                   6052: The variable can be examined and modified using @code{@@} (``fetch'') and
                   6053: @code{!} (``store'') like this:
                   6054: 
                   6055: @example
                   6056: new-word2 @@ .      \ get address, fetch from it and display
                   6057: 1234 new-word2 !   \ new value, get address, store to it
                   6058: @end example
                   6059: 
1.44      crook    6060: @cindex arrays
                   6061: A similar mechanism can be used to create arrays. For example, an
                   6062: 80-character text input buffer:
1.29      crook    6063: 
                   6064: @example
1.44      crook    6065: CREATE text-buf 80 chars allot
                   6066: 
                   6067: text-buf 0 chars c@@ \ the 1st character (offset 0)
                   6068: text-buf 3 chars c@@ \ the 4th character (offset 3)
                   6069: @end example
1.29      crook    6070: 
1.44      crook    6071: You can build arbitrarily complex data structures by allocating
1.49      anton    6072: appropriate areas of memory. For further discussions of this, and to
1.66      anton    6073: learn about some Gforth tools that make it easier,
1.49      anton    6074: @xref{Structures}.
1.44      crook    6075: 
                   6076: 
                   6077: @node Variables, Constants, CREATE, Defining Words
                   6078: @subsection Variables
                   6079: @cindex variables
                   6080: 
                   6081: The previous section showed how a sequence of commands could be used to
                   6082: generate a variable.  As a final refinement, the whole code sequence can
                   6083: be wrapped up in a defining word (pre-empting the subject of the next
                   6084: section), making it easier to create new variables:
                   6085: 
                   6086: @example
                   6087: : myvariableX ( "name" -- a-addr ) CREATE 1 cells allot ;
                   6088: : myvariable0 ( "name" -- a-addr ) CREATE 0 , ;
                   6089: 
                   6090: myvariableX foo \ variable foo starts off with an unknown value
                   6091: myvariable0 joe \ whilst joe is initialised to 0
1.29      crook    6092: 
                   6093: 45 3 * foo !   \ set foo to 135
                   6094: 1234 joe !     \ set joe to 1234
                   6095: 3 joe +!       \ increment joe by 3.. to 1237
                   6096: @end example
                   6097: 
                   6098: Not surprisingly, there is no need to define @code{myvariable}, since
1.44      crook    6099: Forth already has a definition @code{Variable}. ANS Forth does not
1.69      anton    6100: guarantee that a @code{Variable} is initialised when it is created
                   6101: (i.e., it may behave like @code{myvariableX}). In contrast, Gforth's
                   6102: @code{Variable} initialises the variable to 0 (i.e., it behaves exactly
                   6103: like @code{myvariable0}). Forth also provides @code{2Variable} and
1.47      crook    6104: @code{fvariable} for double and floating-point variables, respectively
1.69      anton    6105: -- they are initialised to 0. and 0e in Gforth. If you use a @code{Variable} to
1.47      crook    6106: store a boolean, you can use @code{on} and @code{off} to toggle its
                   6107: state.
1.29      crook    6108: 
1.34      anton    6109: doc-variable
                   6110: doc-2variable
                   6111: doc-fvariable
                   6112: 
1.29      crook    6113: @cindex user variables
                   6114: @cindex user space
                   6115: The defining word @code{User} behaves in the same way as @code{Variable}.
                   6116: The difference is that it reserves space in @i{user (data) space} rather
                   6117: than normal data space. In a Forth system that has a multi-tasker, each
                   6118: task has its own set of user variables.
                   6119: 
1.34      anton    6120: doc-user
1.67      anton    6121: @c doc-udp
                   6122: @c doc-uallot
1.34      anton    6123: 
1.29      crook    6124: @comment TODO is that stuff about user variables strictly correct? Is it
                   6125: @comment just terminal tasks that have user variables?
                   6126: @comment should document tasker.fs (with some examples) elsewhere
                   6127: @comment in this manual, then expand on user space and user variables.
                   6128: 
1.44      crook    6129: @node Constants, Values, Variables, Defining Words
                   6130: @subsection Constants
                   6131: @cindex constants
                   6132: 
                   6133: @code{Constant} allows you to declare a fixed value and refer to it by
                   6134: name. For example:
1.29      crook    6135: 
                   6136: @example
                   6137: 12 Constant INCHES-PER-FOOT
                   6138: 3E+08 fconstant SPEED-O-LIGHT
                   6139: @end example
                   6140: 
                   6141: A @code{Variable} can be both read and written, so its run-time
                   6142: behaviour is to supply an address through which its current value can be
                   6143: manipulated. In contrast, the value of a @code{Constant} cannot be
                   6144: changed once it has been declared@footnote{Well, often it can be -- but
                   6145: not in a Standard, portable way. It's safer to use a @code{Value} (read
                   6146: on).} so it's not necessary to supply the address -- it is more
                   6147: efficient to return the value of the constant directly. That's exactly
                   6148: what happens; the run-time effect of a constant is to put its value on
1.49      anton    6149: the top of the stack (You can find one
                   6150: way of implementing @code{Constant} in @ref{User-defined Defining Words}).
1.29      crook    6151: 
1.69      anton    6152: Forth also provides @code{2Constant} and @code{fconstant} for defining
1.29      crook    6153: double and floating-point constants, respectively.
                   6154: 
1.34      anton    6155: doc-constant
                   6156: doc-2constant
                   6157: doc-fconstant
                   6158: 
                   6159: @c that's too deep, and it's not necessarily true for all ANS Forths. - anton
1.44      crook    6160: @c nac-> How could that not be true in an ANS Forth? You can't define a
                   6161: @c constant, use it and then delete the definition of the constant..
1.69      anton    6162: 
                   6163: @c anton->An ANS Forth system can compile a constant to a literal; On
                   6164: @c decompilation you would see only the number, just as if it had been used
                   6165: @c in the first place.  The word will stay, of course, but it will only be
                   6166: @c used by the text interpreter (no run-time duties, except when it is 
                   6167: @c POSTPONEd or somesuch).
                   6168: 
                   6169: @c nac:
1.44      crook    6170: @c I agree that it's rather deep, but IMO it is an important difference
                   6171: @c relative to other programming languages.. often it's annoying: it
                   6172: @c certainly changes my programming style relative to C.
                   6173: 
1.69      anton    6174: @c anton: In what way?
                   6175: 
1.29      crook    6176: Constants in Forth behave differently from their equivalents in other
                   6177: programming languages. In other languages, a constant (such as an EQU in
                   6178: assembler or a #define in C) only exists at compile-time; in the
                   6179: executable program the constant has been translated into an absolute
                   6180: number and, unless you are using a symbolic debugger, it's impossible to
                   6181: know what abstract thing that number represents. In Forth a constant has
1.44      crook    6182: an entry in the header space and remains there after the code that uses
                   6183: it has been defined. In fact, it must remain in the dictionary since it
                   6184: has run-time duties to perform. For example:
1.29      crook    6185: 
                   6186: @example
                   6187: 12 Constant INCHES-PER-FOOT
                   6188: : FEET-TO-INCHES ( n1 -- n2 ) INCHES-PER-FOOT * ;
                   6189: @end example
                   6190: 
                   6191: @cindex in-lining of constants
                   6192: When @code{FEET-TO-INCHES} is executed, it will in turn execute the xt
                   6193: associated with the constant @code{INCHES-PER-FOOT}. If you use
                   6194: @code{see} to decompile the definition of @code{FEET-TO-INCHES}, you can
                   6195: see that it makes a call to @code{INCHES-PER-FOOT}. Some Forth compilers
                   6196: attempt to optimise constants by in-lining them where they are used. You
                   6197: can force Gforth to in-line a constant like this:
                   6198: 
                   6199: @example
                   6200: : FEET-TO-INCHES ( n1 -- n2 ) [ INCHES-PER-FOOT ] LITERAL * ;
                   6201: @end example
                   6202: 
                   6203: If you use @code{see} to decompile @i{this} version of
                   6204: @code{FEET-TO-INCHES}, you can see that @code{INCHES-PER-FOOT} is no
1.49      anton    6205: longer present. To understand how this works, read
                   6206: @ref{Interpret/Compile states}, and @ref{Literals}.
1.29      crook    6207: 
                   6208: In-lining constants in this way might improve execution time
                   6209: fractionally, and can ensure that a constant is now only referenced at
                   6210: compile-time. However, the definition of the constant still remains in
                   6211: the dictionary. Some Forth compilers provide a mechanism for controlling
                   6212: a second dictionary for holding transient words such that this second
                   6213: dictionary can be deleted later in order to recover memory
                   6214: space. However, there is no standard way of doing this.
                   6215: 
                   6216: 
1.44      crook    6217: @node Values, Colon Definitions, Constants, Defining Words
                   6218: @subsection Values
                   6219: @cindex values
1.34      anton    6220: 
1.69      anton    6221: A @code{Value} behaves like a @code{Constant}, but it can be changed.
                   6222: @code{TO} is a parsing word that changes a @code{Values}.  In Gforth
                   6223: (not in ANS Forth) you can access (and change) a @code{value} also with
                   6224: @code{>body}.
                   6225: 
                   6226: Here are some
                   6227: examples:
1.29      crook    6228: 
                   6229: @example
1.69      anton    6230: 12 Value APPLES     \ Define APPLES with an initial value of 12
                   6231: 34 TO APPLES        \ Change the value of APPLES. TO is a parsing word
                   6232: 1 ' APPLES >body +! \ Increment APPLES.  Non-standard usage.
                   6233: APPLES              \ puts 35 on the top of the stack.
1.29      crook    6234: @end example
                   6235: 
1.44      crook    6236: doc-value
                   6237: doc-to
1.29      crook    6238: 
1.35      anton    6239: 
1.69      anton    6240: 
1.44      crook    6241: @node Colon Definitions, Anonymous Definitions, Values, Defining Words
                   6242: @subsection Colon Definitions
                   6243: @cindex colon definitions
1.35      anton    6244: 
                   6245: @example
1.44      crook    6246: : name ( ... -- ... )
                   6247:     word1 word2 word3 ;
1.29      crook    6248: @end example
                   6249: 
1.44      crook    6250: @noindent
                   6251: Creates a word called @code{name} that, upon execution, executes
                   6252: @code{word1 word2 word3}. @code{name} is a @dfn{(colon) definition}.
1.29      crook    6253: 
1.49      anton    6254: The explanation above is somewhat superficial. For simple examples of
                   6255: colon definitions see @ref{Your first definition}.  For an in-depth
1.66      anton    6256: discussion of some of the issues involved, @xref{Interpretation and
1.49      anton    6257: Compilation Semantics}.
1.29      crook    6258: 
1.44      crook    6259: doc-:
                   6260: doc-;
1.1       anton    6261: 
1.34      anton    6262: 
1.69      anton    6263: @node Anonymous Definitions, Supplying names, Colon Definitions, Defining Words
1.44      crook    6264: @subsection Anonymous Definitions
                   6265: @cindex colon definitions
                   6266: @cindex defining words without name
1.34      anton    6267: 
1.44      crook    6268: Sometimes you want to define an @dfn{anonymous word}; a word without a
                   6269: name. You can do this with:
1.1       anton    6270: 
1.44      crook    6271: doc-:noname
1.1       anton    6272: 
1.44      crook    6273: This leaves the execution token for the word on the stack after the
                   6274: closing @code{;}. Here's an example in which a deferred word is
                   6275: initialised with an @code{xt} from an anonymous colon definition:
1.1       anton    6276: 
1.29      crook    6277: @example
1.44      crook    6278: Defer deferred
                   6279: :noname ( ... -- ... )
                   6280:   ... ;
                   6281: IS deferred
1.29      crook    6282: @end example
1.26      crook    6283: 
1.44      crook    6284: @noindent
                   6285: Gforth provides an alternative way of doing this, using two separate
                   6286: words:
1.27      crook    6287: 
1.44      crook    6288: doc-noname
                   6289: @cindex execution token of last defined word
                   6290: doc-lastxt
1.1       anton    6291: 
1.44      crook    6292: @noindent
                   6293: The previous example can be rewritten using @code{noname} and
                   6294: @code{lastxt}:
1.1       anton    6295: 
1.26      crook    6296: @example
1.44      crook    6297: Defer deferred
                   6298: noname : ( ... -- ... )
                   6299:   ... ;
                   6300: lastxt IS deferred
1.26      crook    6301: @end example
1.1       anton    6302: 
1.29      crook    6303: @noindent
1.44      crook    6304: @code{noname} works with any defining word, not just @code{:}.
                   6305: 
                   6306: @code{lastxt} also works when the last word was not defined as
1.71      anton    6307: @code{noname}.  It does not work for combined words, though.  It also has
                   6308: the useful property that is is valid as soon as the header for a
                   6309: definition has been built. Thus:
1.44      crook    6310: 
                   6311: @example
                   6312: lastxt . : foo [ lastxt . ] ; ' foo .
                   6313: @end example
1.1       anton    6314: 
1.44      crook    6315: @noindent
                   6316: prints 3 numbers; the last two are the same.
1.26      crook    6317: 
1.69      anton    6318: @node Supplying names, User-defined Defining Words, Anonymous Definitions, Defining Words
                   6319: @subsection Supplying the name of a defined word
                   6320: @cindex names for defined words
                   6321: @cindex defining words, name given in a string
                   6322: 
                   6323: By default, a defining word takes the name for the defined word from the
                   6324: input stream. Sometimes you want to supply the name from a string. You
                   6325: can do this with:
                   6326: 
                   6327: doc-nextname
                   6328: 
                   6329: For example:
                   6330: 
                   6331: @example
                   6332: s" foo" nextname create
                   6333: @end example
                   6334: 
                   6335: @noindent
                   6336: is equivalent to:
                   6337: 
                   6338: @example
                   6339: create foo
                   6340: @end example
                   6341: 
                   6342: @noindent
                   6343: @code{nextname} works with any defining word.
                   6344: 
1.1       anton    6345: 
1.69      anton    6346: @node User-defined Defining Words, Deferred words, Supplying names, Defining Words
1.26      crook    6347: @subsection User-defined Defining Words
                   6348: @cindex user-defined defining words
                   6349: @cindex defining words, user-defined
1.1       anton    6350: 
1.29      crook    6351: You can create a new defining word by wrapping defining-time code around
                   6352: an existing defining word and putting the sequence in a colon
1.69      anton    6353: definition. 
                   6354: 
                   6355: @c anton: This example is very complex and leads in a quite different
                   6356: @c direction from the CREATE-DOES> stuff that follows.  It should probably
                   6357: @c be done elsewhere, or as a subsubsection of this subsection (or as a
                   6358: @c subsection of Defining Words)
                   6359: 
                   6360: For example, suppose that you have a word @code{stats} that
1.29      crook    6361: gathers statistics about colon definitions given the @i{xt} of the
                   6362: definition, and you want every colon definition in your application to
                   6363: make a call to @code{stats}. You can define and use a new version of
                   6364: @code{:} like this:
                   6365: 
                   6366: @example
                   6367: : stats ( xt -- ) DUP ." (Gathering statistics for " . ." )"
                   6368:   ... ;  \ other code
                   6369: 
                   6370: : my: : lastxt postpone literal ['] stats compile, ;
                   6371: 
                   6372: my: foo + - ;
                   6373: @end example
                   6374: 
                   6375: When @code{foo} is defined using @code{my:} these steps occur:
                   6376: 
                   6377: @itemize @bullet
                   6378: @item
                   6379: @code{my:} is executed.
                   6380: @item
                   6381: The @code{:} within the definition (the one between @code{my:} and
                   6382: @code{lastxt}) is executed, and does just what it always does; it parses
                   6383: the input stream for a name, builds a dictionary header for the name
                   6384: @code{foo} and switches @code{state} from interpret to compile.
                   6385: @item
                   6386: The word @code{lastxt} is executed. It puts the @i{xt} for the word that is
                   6387: being defined -- @code{foo} -- onto the stack.
                   6388: @item
                   6389: The code that was produced by @code{postpone literal} is executed; this
                   6390: causes the value on the stack to be compiled as a literal in the code
                   6391: area of @code{foo}.
                   6392: @item
                   6393: The code @code{['] stats} compiles a literal into the definition of
                   6394: @code{my:}. When @code{compile,} is executed, that literal -- the
                   6395: execution token for @code{stats} -- is layed down in the code area of
                   6396: @code{foo} , following the literal@footnote{Strictly speaking, the
                   6397: mechanism that @code{compile,} uses to convert an @i{xt} into something
                   6398: in the code area is implementation-dependent. A threaded implementation
                   6399: might spit out the execution token directly whilst another
                   6400: implementation might spit out a native code sequence.}.
                   6401: @item
                   6402: At this point, the execution of @code{my:} is complete, and control
                   6403: returns to the text interpreter. The text interpreter is in compile
                   6404: state, so subsequent text @code{+ -} is compiled into the definition of
                   6405: @code{foo} and the @code{;} terminates the definition as always.
                   6406: @end itemize
                   6407: 
                   6408: You can use @code{see} to decompile a word that was defined using
                   6409: @code{my:} and see how it is different from a normal @code{:}
                   6410: definition. For example:
                   6411: 
                   6412: @example
                   6413: : bar + - ;  \ like foo but using : rather than my:
                   6414: see bar
                   6415: : bar
                   6416:   + - ;
                   6417: see foo
                   6418: : foo
                   6419:   107645672 stats + - ;
                   6420: 
                   6421: \ use ' stats . to show that 107645672 is the xt for stats
                   6422: @end example
                   6423: 
                   6424: You can use techniques like this to make new defining words in terms of
                   6425: @i{any} existing defining word.
1.1       anton    6426: 
                   6427: 
1.29      crook    6428: @cindex defining defining words
1.26      crook    6429: @cindex @code{CREATE} ... @code{DOES>}
                   6430: If you want the words defined with your defining words to behave
                   6431: differently from words defined with standard defining words, you can
                   6432: write your defining word like this:
1.1       anton    6433: 
                   6434: @example
1.26      crook    6435: : def-word ( "name" -- )
1.29      crook    6436:     CREATE @i{code1}
1.26      crook    6437: DOES> ( ... -- ... )
1.29      crook    6438:     @i{code2} ;
1.26      crook    6439: 
                   6440: def-word name
1.1       anton    6441: @end example
                   6442: 
1.29      crook    6443: @cindex child words
                   6444: This fragment defines a @dfn{defining word} @code{def-word} and then
                   6445: executes it.  When @code{def-word} executes, it @code{CREATE}s a new
                   6446: word, @code{name}, and executes the code @i{code1}. The code @i{code2}
                   6447: is not executed at this time. The word @code{name} is sometimes called a
                   6448: @dfn{child} of @code{def-word}.
                   6449: 
                   6450: When you execute @code{name}, the address of the body of @code{name} is
                   6451: put on the data stack and @i{code2} is executed (the address of the body
                   6452: of @code{name} is the address @code{HERE} returns immediately after the
1.69      anton    6453: @code{CREATE}, i.e., the address a @code{create}d word returns by
                   6454: default).
                   6455: 
                   6456: @c anton:
                   6457: @c www.dictionary.com says:
                   6458: @c at·a·vism: 1.The reappearance of a characteristic in an organism after
                   6459: @c several generations of absence, usually caused by the chance
                   6460: @c recombination of genes.  2.An individual or a part that exhibits
                   6461: @c atavism. Also called throwback.  3.The return of a trait or recurrence
                   6462: @c of previous behavior after a period of absence.
                   6463: @c
                   6464: @c Doesn't seem to fit.
1.29      crook    6465: 
1.69      anton    6466: @c @cindex atavism in child words
1.33      anton    6467: You can use @code{def-word} to define a set of child words that behave
1.69      anton    6468: similarly; they all have a common run-time behaviour determined by
                   6469: @i{code2}. Typically, the @i{code1} sequence builds a data area in the
                   6470: body of the child word. The structure of the data is common to all
                   6471: children of @code{def-word}, but the data values are specific -- and
                   6472: private -- to each child word. When a child word is executed, the
                   6473: address of its private data area is passed as a parameter on TOS to be
                   6474: used and manipulated@footnote{It is legitimate both to read and write to
                   6475: this data area.} by @i{code2}.
1.29      crook    6476: 
                   6477: The two fragments of code that make up the defining words act (are
                   6478: executed) at two completely separate times:
1.1       anton    6479: 
1.29      crook    6480: @itemize @bullet
                   6481: @item
                   6482: At @i{define time}, the defining word executes @i{code1} to generate a
                   6483: child word
                   6484: @item
                   6485: At @i{child execution time}, when a child word is invoked, @i{code2}
                   6486: is executed, using parameters (data) that are private and specific to
                   6487: the child word.
                   6488: @end itemize
                   6489: 
1.44      crook    6490: Another way of understanding the behaviour of @code{def-word} and
                   6491: @code{name} is to say that, if you make the following definitions:
1.33      anton    6492: @example
                   6493: : def-word1 ( "name" -- )
                   6494:     CREATE @i{code1} ;
                   6495: 
                   6496: : action1 ( ... -- ... )
                   6497:     @i{code2} ;
                   6498: 
                   6499: def-word1 name1
                   6500: @end example
                   6501: 
1.44      crook    6502: @noindent
                   6503: Then using @code{name1 action1} is equivalent to using @code{name}.
1.1       anton    6504: 
1.29      crook    6505: The classic example is that you can define @code{CONSTANT} in this way:
1.26      crook    6506: 
1.1       anton    6507: @example
1.29      crook    6508: : CONSTANT ( w "name" -- )
                   6509:     CREATE ,
1.26      crook    6510: DOES> ( -- w )
                   6511:     @@ ;
1.1       anton    6512: @end example
                   6513: 
1.29      crook    6514: @comment There is a beautiful description of how this works and what
                   6515: @comment it does in the Forthwrite 100th edition.. as well as an elegant
                   6516: @comment commentary on the Counting Fruits problem.
                   6517: 
                   6518: When you create a constant with @code{5 CONSTANT five}, a set of
                   6519: define-time actions take place; first a new word @code{five} is created,
                   6520: then the value 5 is laid down in the body of @code{five} with
1.44      crook    6521: @code{,}. When @code{five} is executed, the address of the body is put on
1.29      crook    6522: the stack, and @code{@@} retrieves the value 5. The word @code{five} has
                   6523: no code of its own; it simply contains a data field and a pointer to the
                   6524: code that follows @code{DOES>} in its defining word. That makes words
                   6525: created in this way very compact.
                   6526: 
                   6527: The final example in this section is intended to remind you that space
                   6528: reserved in @code{CREATE}d words is @i{data} space and therefore can be
                   6529: both read and written by a Standard program@footnote{Exercise: use this
                   6530: example as a starting point for your own implementation of @code{Value}
                   6531: and @code{TO} -- if you get stuck, investigate the behaviour of @code{'} and
                   6532: @code{[']}.}:
                   6533: 
                   6534: @example
                   6535: : foo ( "name" -- )
                   6536:     CREATE -1 ,
                   6537: DOES> ( -- )
1.33      anton    6538:     @@ . ;
1.29      crook    6539: 
                   6540: foo first-word
                   6541: foo second-word
                   6542: 
                   6543: 123 ' first-word >BODY !
                   6544: @end example
                   6545: 
                   6546: If @code{first-word} had been a @code{CREATE}d word, we could simply
                   6547: have executed it to get the address of its data field. However, since it
                   6548: was defined to have @code{DOES>} actions, its execution semantics are to
                   6549: perform those @code{DOES>} actions. To get the address of its data field
                   6550: it's necessary to use @code{'} to get its xt, then @code{>BODY} to
                   6551: translate the xt into the address of the data field.  When you execute
                   6552: @code{first-word}, it will display @code{123}. When you execute
                   6553: @code{second-word} it will display @code{-1}.
1.26      crook    6554: 
                   6555: @cindex stack effect of @code{DOES>}-parts
                   6556: @cindex @code{DOES>}-parts, stack effect
1.29      crook    6557: In the examples above the stack comment after the @code{DOES>} specifies
1.26      crook    6558: the stack effect of the defined words, not the stack effect of the
                   6559: following code (the following code expects the address of the body on
                   6560: the top of stack, which is not reflected in the stack comment). This is
                   6561: the convention that I use and recommend (it clashes a bit with using
                   6562: locals declarations for stack effect specification, though).
1.1       anton    6563: 
1.53      anton    6564: @menu
                   6565: * CREATE..DOES> applications::  
                   6566: * CREATE..DOES> details::       
1.63      anton    6567: * Advanced does> usage example::  
1.53      anton    6568: @end menu
                   6569: 
                   6570: @node CREATE..DOES> applications, CREATE..DOES> details, User-defined Defining Words, User-defined Defining Words
1.26      crook    6571: @subsubsection Applications of @code{CREATE..DOES>}
                   6572: @cindex @code{CREATE} ... @code{DOES>}, applications
1.1       anton    6573: 
1.26      crook    6574: You may wonder how to use this feature. Here are some usage patterns:
1.1       anton    6575: 
1.26      crook    6576: @cindex factoring similar colon definitions
                   6577: When you see a sequence of code occurring several times, and you can
                   6578: identify a meaning, you will factor it out as a colon definition. When
                   6579: you see similar colon definitions, you can factor them using
                   6580: @code{CREATE..DOES>}. E.g., an assembler usually defines several words
                   6581: that look very similar:
1.1       anton    6582: @example
1.26      crook    6583: : ori, ( reg-target reg-source n -- )
                   6584:     0 asm-reg-reg-imm ;
                   6585: : andi, ( reg-target reg-source n -- )
                   6586:     1 asm-reg-reg-imm ;
1.1       anton    6587: @end example
                   6588: 
1.26      crook    6589: @noindent
                   6590: This could be factored with:
                   6591: @example
                   6592: : reg-reg-imm ( op-code -- )
                   6593:     CREATE ,
                   6594: DOES> ( reg-target reg-source n -- )
                   6595:     @@ asm-reg-reg-imm ;
                   6596: 
                   6597: 0 reg-reg-imm ori,
                   6598: 1 reg-reg-imm andi,
                   6599: @end example
1.1       anton    6600: 
1.26      crook    6601: @cindex currying
                   6602: Another view of @code{CREATE..DOES>} is to consider it as a crude way to
                   6603: supply a part of the parameters for a word (known as @dfn{currying} in
                   6604: the functional language community). E.g., @code{+} needs two
                   6605: parameters. Creating versions of @code{+} with one parameter fixed can
                   6606: be done like this:
1.82      anton    6607: 
1.1       anton    6608: @example
1.82      anton    6609: : curry+ ( n1 "name" -- )
1.26      crook    6610:     CREATE ,
                   6611: DOES> ( n2 -- n1+n2 )
                   6612:     @@ + ;
                   6613: 
                   6614:  3 curry+ 3+
                   6615: -2 curry+ 2-
1.1       anton    6616: @end example
                   6617: 
1.63      anton    6618: @node CREATE..DOES> details, Advanced does> usage example, CREATE..DOES> applications, User-defined Defining Words
1.26      crook    6619: @subsubsection The gory details of @code{CREATE..DOES>}
                   6620: @cindex @code{CREATE} ... @code{DOES>}, details
1.1       anton    6621: 
1.26      crook    6622: doc-does>
1.1       anton    6623: 
1.26      crook    6624: @cindex @code{DOES>} in a separate definition
                   6625: This means that you need not use @code{CREATE} and @code{DOES>} in the
                   6626: same definition; you can put the @code{DOES>}-part in a separate
1.29      crook    6627: definition. This allows us to, e.g., select among different @code{DOES>}-parts:
1.26      crook    6628: @example
                   6629: : does1 
                   6630: DOES> ( ... -- ... )
1.44      crook    6631:     ... ;
                   6632: 
                   6633: : does2
                   6634: DOES> ( ... -- ... )
                   6635:     ... ;
                   6636: 
                   6637: : def-word ( ... -- ... )
                   6638:     create ...
                   6639:     IF
                   6640:        does1
                   6641:     ELSE
                   6642:        does2
                   6643:     ENDIF ;
                   6644: @end example
                   6645: 
                   6646: In this example, the selection of whether to use @code{does1} or
1.69      anton    6647: @code{does2} is made at definition-time; at the time that the child word is
1.44      crook    6648: @code{CREATE}d.
                   6649: 
                   6650: @cindex @code{DOES>} in interpretation state
                   6651: In a standard program you can apply a @code{DOES>}-part only if the last
                   6652: word was defined with @code{CREATE}. In Gforth, the @code{DOES>}-part
                   6653: will override the behaviour of the last word defined in any case. In a
                   6654: standard program, you can use @code{DOES>} only in a colon
                   6655: definition. In Gforth, you can also use it in interpretation state, in a
                   6656: kind of one-shot mode; for example:
                   6657: @example
                   6658: CREATE name ( ... -- ... )
                   6659:   @i{initialization}
                   6660: DOES>
                   6661:   @i{code} ;
                   6662: @end example
                   6663: 
                   6664: @noindent
                   6665: is equivalent to the standard:
                   6666: @example
                   6667: :noname
                   6668: DOES>
                   6669:     @i{code} ;
                   6670: CREATE name EXECUTE ( ... -- ... )
                   6671:     @i{initialization}
                   6672: @end example
                   6673: 
1.53      anton    6674: doc->body
                   6675: 
1.63      anton    6676: @node Advanced does> usage example,  , CREATE..DOES> details, User-defined Defining Words
                   6677: @subsubsection Advanced does> usage example
                   6678: 
                   6679: The MIPS disassembler (@file{arch/mips/disasm.fs}) contains many words
                   6680: for disassembling instructions, that follow a very repetetive scheme:
                   6681: 
                   6682: @example
                   6683: :noname @var{disasm-operands} s" @var{inst-name}" type ;
                   6684: @var{entry-num} cells @var{table} + !
                   6685: @end example
                   6686: 
                   6687: Of course, this inspires the idea to factor out the commonalities to
                   6688: allow a definition like
                   6689: 
                   6690: @example
                   6691: @var{disasm-operands} @var{entry-num} @var{table} define-inst @var{inst-name}
                   6692: @end example
                   6693: 
                   6694: The parameters @var{disasm-operands} and @var{table} are usually
1.69      anton    6695: correlated.  Moreover, before I wrote the disassembler, there already
                   6696: existed code that defines instructions like this:
1.63      anton    6697: 
                   6698: @example
                   6699: @var{entry-num} @var{inst-format} @var{inst-name}
                   6700: @end example
                   6701: 
                   6702: This code comes from the assembler and resides in
                   6703: @file{arch/mips/insts.fs}.
                   6704: 
                   6705: So I had to define the @var{inst-format} words that performed the scheme
                   6706: above when executed.  At first I chose to use run-time code-generation:
                   6707: 
                   6708: @example
                   6709: : @var{inst-format} ( entry-num "name" -- ; compiled code: addr w -- )
                   6710:   :noname Postpone @var{disasm-operands}
                   6711:   name Postpone sliteral Postpone type Postpone ;
                   6712:   swap cells @var{table} + ! ;
                   6713: @end example
                   6714: 
                   6715: Note that this supplies the other two parameters of the scheme above.
1.44      crook    6716: 
1.63      anton    6717: An alternative would have been to write this using
                   6718: @code{create}/@code{does>}:
                   6719: 
                   6720: @example
                   6721: : @var{inst-format} ( entry-num "name" -- )
                   6722:   here name string, ( entry-num c-addr ) \ parse and save "name"
                   6723:   noname create , ( entry-num )
                   6724:   lastxt swap cells @var{table} + !
                   6725: does> ( addr w -- )
                   6726:   \ disassemble instruction w at addr
                   6727:   @@ >r 
                   6728:   @var{disasm-operands}
                   6729:   r> count type ;
                   6730: @end example
                   6731: 
                   6732: Somehow the first solution is simpler, mainly because it's simpler to
                   6733: shift a string from definition-time to use-time with @code{sliteral}
                   6734: than with @code{string,} and friends.
                   6735: 
                   6736: I wrote a lot of words following this scheme and soon thought about
                   6737: factoring out the commonalities among them.  Note that this uses a
                   6738: two-level defining word, i.e., a word that defines ordinary defining
                   6739: words.
                   6740: 
                   6741: This time a solution involving @code{postpone} and friends seemed more
                   6742: difficult (try it as an exercise), so I decided to use a
                   6743: @code{create}/@code{does>} word; since I was already at it, I also used
                   6744: @code{create}/@code{does>} for the lower level (try using
                   6745: @code{postpone} etc. as an exercise), resulting in the following
                   6746: definition:
                   6747: 
                   6748: @example
                   6749: : define-format ( disasm-xt table-xt -- )
                   6750:     \ define an instruction format that uses disasm-xt for
                   6751:     \ disassembling and enters the defined instructions into table
                   6752:     \ table-xt
                   6753:     create 2,
                   6754: does> ( u "inst" -- )
                   6755:     \ defines an anonymous word for disassembling instruction inst,
                   6756:     \ and enters it as u-th entry into table-xt
                   6757:     2@@ swap here name string, ( u table-xt disasm-xt c-addr ) \ remember string
                   6758:     noname create 2,      \ define anonymous word
                   6759:     execute lastxt swap ! \ enter xt of defined word into table-xt
                   6760: does> ( addr w -- )
                   6761:     \ disassemble instruction w at addr
                   6762:     2@@ >r ( addr w disasm-xt R: c-addr )
                   6763:     execute ( R: c-addr ) \ disassemble operands
                   6764:     r> count type ; \ print name 
                   6765: @end example
                   6766: 
                   6767: Note that the tables here (in contrast to above) do the @code{cells +}
                   6768: by themselves (that's why you have to pass an xt).  This word is used in
                   6769: the following way:
                   6770: 
                   6771: @example
                   6772: ' @var{disasm-operands} ' @var{table} define-format @var{inst-format}
                   6773: @end example
                   6774: 
1.71      anton    6775: As shown above, the defined instruction format is then used like this:
                   6776: 
                   6777: @example
                   6778: @var{entry-num} @var{inst-format} @var{inst-name}
                   6779: @end example
                   6780: 
1.63      anton    6781: In terms of currying, this kind of two-level defining word provides the
                   6782: parameters in three stages: first @var{disasm-operands} and @var{table},
                   6783: then @var{entry-num} and @var{inst-name}, finally @code{addr w}, i.e.,
                   6784: the instruction to be disassembled.  
                   6785: 
                   6786: Of course this did not quite fit all the instruction format names used
                   6787: in @file{insts.fs}, so I had to define a few wrappers that conditioned
                   6788: the parameters into the right form.
                   6789: 
                   6790: If you have trouble following this section, don't worry.  First, this is
                   6791: involved and takes time (and probably some playing around) to
                   6792: understand; second, this is the first two-level
                   6793: @code{create}/@code{does>} word I have written in seventeen years of
                   6794: Forth; and if I did not have @file{insts.fs} to start with, I may well
                   6795: have elected to use just a one-level defining word (with some repeating
                   6796: of parameters when using the defining word). So it is not necessary to
                   6797: understand this, but it may improve your understanding of Forth.
1.44      crook    6798: 
                   6799: 
                   6800: @node Deferred words, Aliases, User-defined Defining Words, Defining Words
                   6801: @subsection Deferred words
                   6802: @cindex deferred words
                   6803: 
                   6804: The defining word @code{Defer} allows you to define a word by name
                   6805: without defining its behaviour; the definition of its behaviour is
                   6806: deferred. Here are two situation where this can be useful:
                   6807: 
                   6808: @itemize @bullet
                   6809: @item
                   6810: Where you want to allow the behaviour of a word to be altered later, and
                   6811: for all precompiled references to the word to change when its behaviour
                   6812: is changed.
                   6813: @item
                   6814: For mutual recursion; @xref{Calls and returns}.
                   6815: @end itemize
                   6816: 
                   6817: In the following example, @code{foo} always invokes the version of
                   6818: @code{greet} that prints ``@code{Good morning}'' whilst @code{bar}
                   6819: always invokes the version that prints ``@code{Hello}''. There is no way
                   6820: of getting @code{foo} to use the later version without re-ordering the
                   6821: source code and recompiling it.
                   6822: 
                   6823: @example
                   6824: : greet ." Good morning" ;
                   6825: : foo ... greet ... ;
                   6826: : greet ." Hello" ;
                   6827: : bar ... greet ... ;
                   6828: @end example
                   6829: 
                   6830: This problem can be solved by defining @code{greet} as a @code{Defer}red
                   6831: word. The behaviour of a @code{Defer}red word can be defined and
                   6832: redefined at any time by using @code{IS} to associate the xt of a
                   6833: previously-defined word with it. The previous example becomes:
                   6834: 
                   6835: @example
1.69      anton    6836: Defer greet ( -- )
1.44      crook    6837: : foo ... greet ... ;
                   6838: : bar ... greet ... ;
1.69      anton    6839: : greet1 ( -- ) ." Good morning" ;
                   6840: : greet2 ( -- ) ." Hello" ;
1.44      crook    6841: ' greet2 <IS> greet  \ make greet behave like greet2
                   6842: @end example
                   6843: 
1.69      anton    6844: @progstyle
                   6845: You should write a stack comment for every deferred word, and put only
                   6846: XTs into deferred words that conform to this stack effect.  Otherwise
                   6847: it's too difficult to use the deferred word.
                   6848: 
1.44      crook    6849: A deferred word can be used to improve the statistics-gathering example
                   6850: from @ref{User-defined Defining Words}; rather than edit the
                   6851: application's source code to change every @code{:} to a @code{my:}, do
                   6852: this:
                   6853: 
                   6854: @example
                   6855: : real: : ;     \ retain access to the original
                   6856: defer :         \ redefine as a deferred word
1.69      anton    6857: ' my: <IS> :      \ use special version of :
1.44      crook    6858: \
                   6859: \ load application here
                   6860: \
1.69      anton    6861: ' real: <IS> :    \ go back to the original
1.44      crook    6862: @end example
                   6863: 
                   6864: 
                   6865: One thing to note is that @code{<IS>} consumes its name when it is
                   6866: executed.  If you want to specify the name at compile time, use
                   6867: @code{[IS]}:
                   6868: 
                   6869: @example
                   6870: : set-greet ( xt -- )
                   6871:   [IS] greet ;
                   6872: 
                   6873: ' greet1 set-greet
                   6874: @end example
                   6875: 
1.69      anton    6876: A deferred word can only inherit execution semantics from the xt
                   6877: (because that is all that an xt can represent -- for more discussion of
                   6878: this @pxref{Tokens for Words}); by default it will have default
                   6879: interpretation and compilation semantics deriving from this execution
                   6880: semantics.  However, you can change the interpretation and compilation
                   6881: semantics of the deferred word in the usual ways:
1.44      crook    6882: 
                   6883: @example
                   6884: : bar .... ; compile-only
                   6885: Defer fred immediate
                   6886: Defer jim
                   6887: 
                   6888: ' bar <IS> jim  \ jim has default semantics
                   6889: ' bar <IS> fred \ fred is immediate
                   6890: @end example
                   6891: 
                   6892: doc-defer
                   6893: doc-<is>
                   6894: doc-[is]
                   6895: doc-is
                   6896: @comment TODO document these: what's defers [is]
                   6897: doc-what's
                   6898: doc-defers
                   6899: 
                   6900: @c Use @code{words-deferred} to see a list of deferred words.
                   6901: 
                   6902: Definitions in ANS Forth for @code{defer}, @code{<is>} and @code{[is]}
                   6903: are provided in @file{compat/defer.fs}.
                   6904: 
                   6905: 
1.69      anton    6906: @node Aliases,  , Deferred words, Defining Words
1.44      crook    6907: @subsection Aliases
                   6908: @cindex aliases
1.1       anton    6909: 
1.44      crook    6910: The defining word @code{Alias} allows you to define a word by name that
                   6911: has the same behaviour as some other word. Here are two situation where
                   6912: this can be useful:
1.1       anton    6913: 
1.44      crook    6914: @itemize @bullet
                   6915: @item
                   6916: When you want access to a word's definition from a different word list
                   6917: (for an example of this, see the definition of the @code{Root} word list
                   6918: in the Gforth source).
                   6919: @item
                   6920: When you want to create a synonym; a definition that can be known by
                   6921: either of two names (for example, @code{THEN} and @code{ENDIF} are
                   6922: aliases).
                   6923: @end itemize
1.1       anton    6924: 
1.69      anton    6925: Like deferred words, an alias has default compilation and interpretation
                   6926: semantics at the beginning (not the modifications of the other word),
                   6927: but you can change them in the usual ways (@code{immediate},
                   6928: @code{compile-only}). For example:
1.1       anton    6929: 
                   6930: @example
1.44      crook    6931: : foo ... ; immediate
                   6932: 
                   6933: ' foo Alias bar \ bar is not an immediate word
                   6934: ' foo Alias fooby immediate \ fooby is an immediate word
1.1       anton    6935: @end example
                   6936: 
1.44      crook    6937: Words that are aliases have the same xt, different headers in the
                   6938: dictionary, and consequently different name tokens (@pxref{Tokens for
                   6939: Words}) and possibly different immediate flags.  An alias can only have
                   6940: default or immediate compilation semantics; you can define aliases for
                   6941: combined words with @code{interpret/compile:} -- see @ref{Combined words}.
1.1       anton    6942: 
1.44      crook    6943: doc-alias
1.1       anton    6944: 
                   6945: 
1.47      crook    6946: @node Interpretation and Compilation Semantics, Tokens for Words, Defining Words, Words
                   6947: @section Interpretation and Compilation Semantics
1.26      crook    6948: @cindex semantics, interpretation and compilation
1.1       anton    6949: 
1.71      anton    6950: @c !! state and ' are used without explanation
                   6951: @c example for immediate/compile-only? or is the tutorial enough
                   6952: 
1.26      crook    6953: @cindex interpretation semantics
1.71      anton    6954: The @dfn{interpretation semantics} of a (named) word are what the text
1.26      crook    6955: interpreter does when it encounters the word in interpret state. It also
                   6956: appears in some other contexts, e.g., the execution token returned by
1.71      anton    6957: @code{' @i{word}} identifies the interpretation semantics of @i{word}
                   6958: (in other words, @code{' @i{word} execute} is equivalent to
1.29      crook    6959: interpret-state text interpretation of @code{@i{word}}).
1.1       anton    6960: 
1.26      crook    6961: @cindex compilation semantics
1.71      anton    6962: The @dfn{compilation semantics} of a (named) word are what the text
                   6963: interpreter does when it encounters the word in compile state. It also
                   6964: appears in other contexts, e.g, @code{POSTPONE @i{word}}
                   6965: compiles@footnote{In standard terminology, ``appends to the current
                   6966: definition''.} the compilation semantics of @i{word}.
1.1       anton    6967: 
1.26      crook    6968: @cindex execution semantics
                   6969: The standard also talks about @dfn{execution semantics}. They are used
                   6970: only for defining the interpretation and compilation semantics of many
                   6971: words. By default, the interpretation semantics of a word are to
                   6972: @code{execute} its execution semantics, and the compilation semantics of
                   6973: a word are to @code{compile,} its execution semantics.@footnote{In
                   6974: standard terminology: The default interpretation semantics are its
                   6975: execution semantics; the default compilation semantics are to append its
                   6976: execution semantics to the execution semantics of the current
                   6977: definition.}
                   6978: 
1.71      anton    6979: Unnamed words (@pxref{Anonymous Definitions}) cannot be encountered by
                   6980: the text interpreter, ticked, or @code{postpone}d, so they have no
                   6981: interpretation or compilation semantics.  Their behaviour is represented
                   6982: by their XT (@pxref{Tokens for Words}), and we call it execution
                   6983: semantics, too.
                   6984: 
1.26      crook    6985: @comment TODO expand, make it co-operate with new sections on text interpreter.
                   6986: 
                   6987: @cindex immediate words
                   6988: @cindex compile-only words
                   6989: You can change the semantics of the most-recently defined word:
                   6990: 
1.44      crook    6991: 
1.26      crook    6992: doc-immediate
                   6993: doc-compile-only
                   6994: doc-restrict
                   6995: 
1.82      anton    6996: By convention, words with non-default compilation semantics (e.g.,
                   6997: immediate words) often have names surrounded with brackets (e.g.,
                   6998: @code{[']}, @pxref{Execution token}).
1.44      crook    6999: 
1.26      crook    7000: Note that ticking (@code{'}) a compile-only word gives an error
                   7001: (``Interpreting a compile-only word'').
1.1       anton    7002: 
1.47      crook    7003: @menu
1.67      anton    7004: * Combined words::              
1.47      crook    7005: @end menu
1.44      crook    7006: 
1.71      anton    7007: 
1.48      anton    7008: @node Combined words,  , Interpretation and Compilation Semantics, Interpretation and Compilation Semantics
1.44      crook    7009: @subsection Combined Words
                   7010: @cindex combined words
                   7011: 
                   7012: Gforth allows you to define @dfn{combined words} -- words that have an
                   7013: arbitrary combination of interpretation and compilation semantics.
                   7014: 
1.26      crook    7015: doc-interpret/compile:
1.1       anton    7016: 
1.26      crook    7017: This feature was introduced for implementing @code{TO} and @code{S"}. I
                   7018: recommend that you do not define such words, as cute as they may be:
                   7019: they make it hard to get at both parts of the word in some contexts.
                   7020: E.g., assume you want to get an execution token for the compilation
                   7021: part. Instead, define two words, one that embodies the interpretation
                   7022: part, and one that embodies the compilation part.  Once you have done
                   7023: that, you can define a combined word with @code{interpret/compile:} for
                   7024: the convenience of your users.
1.1       anton    7025: 
1.26      crook    7026: You might try to use this feature to provide an optimizing
                   7027: implementation of the default compilation semantics of a word. For
                   7028: example, by defining:
1.1       anton    7029: @example
1.26      crook    7030: :noname
                   7031:    foo bar ;
                   7032: :noname
                   7033:    POSTPONE foo POSTPONE bar ;
1.29      crook    7034: interpret/compile: opti-foobar
1.1       anton    7035: @end example
1.26      crook    7036: 
1.23      crook    7037: @noindent
1.26      crook    7038: as an optimizing version of:
                   7039: 
1.1       anton    7040: @example
1.26      crook    7041: : foobar
                   7042:     foo bar ;
1.1       anton    7043: @end example
                   7044: 
1.26      crook    7045: Unfortunately, this does not work correctly with @code{[compile]},
                   7046: because @code{[compile]} assumes that the compilation semantics of all
                   7047: @code{interpret/compile:} words are non-default. I.e., @code{[compile]
1.29      crook    7048: opti-foobar} would compile compilation semantics, whereas
                   7049: @code{[compile] foobar} would compile interpretation semantics.
1.1       anton    7050: 
1.26      crook    7051: @cindex state-smart words (are a bad idea)
1.82      anton    7052: @anchor{state-smartness}
1.29      crook    7053: Some people try to use @dfn{state-smart} words to emulate the feature provided
1.26      crook    7054: by @code{interpret/compile:} (words are state-smart if they check
                   7055: @code{STATE} during execution). E.g., they would try to code
                   7056: @code{foobar} like this:
1.1       anton    7057: 
1.26      crook    7058: @example
                   7059: : foobar
                   7060:   STATE @@
                   7061:   IF ( compilation state )
                   7062:     POSTPONE foo POSTPONE bar
                   7063:   ELSE
                   7064:     foo bar
                   7065:   ENDIF ; immediate
                   7066: @end example
1.1       anton    7067: 
1.26      crook    7068: Although this works if @code{foobar} is only processed by the text
                   7069: interpreter, it does not work in other contexts (like @code{'} or
                   7070: @code{POSTPONE}). E.g., @code{' foobar} will produce an execution token
                   7071: for a state-smart word, not for the interpretation semantics of the
                   7072: original @code{foobar}; when you execute this execution token (directly
                   7073: with @code{EXECUTE} or indirectly through @code{COMPILE,}) in compile
                   7074: state, the result will not be what you expected (i.e., it will not
                   7075: perform @code{foo bar}). State-smart words are a bad idea. Simply don't
                   7076: write them@footnote{For a more detailed discussion of this topic, see
1.66      anton    7077: M. Anton Ertl,
                   7078: @cite{@uref{http://www.complang.tuwien.ac.at/papers/ertl98.ps.gz,@code{State}-smartness---Why
                   7079: it is Evil and How to Exorcise it}}, EuroForth '98.}!
1.1       anton    7080: 
1.26      crook    7081: @cindex defining words with arbitrary semantics combinations
                   7082: It is also possible to write defining words that define words with
                   7083: arbitrary combinations of interpretation and compilation semantics. In
                   7084: general, they look like this:
1.1       anton    7085: 
1.26      crook    7086: @example
                   7087: : def-word
                   7088:     create-interpret/compile
1.29      crook    7089:     @i{code1}
1.26      crook    7090: interpretation>
1.29      crook    7091:     @i{code2}
1.26      crook    7092: <interpretation
                   7093: compilation>
1.29      crook    7094:     @i{code3}
1.26      crook    7095: <compilation ;
                   7096: @end example
1.1       anton    7097: 
1.29      crook    7098: For a @i{word} defined with @code{def-word}, the interpretation
                   7099: semantics are to push the address of the body of @i{word} and perform
                   7100: @i{code2}, and the compilation semantics are to push the address of
                   7101: the body of @i{word} and perform @i{code3}. E.g., @code{constant}
1.26      crook    7102: can also be defined like this (except that the defined constants don't
                   7103: behave correctly when @code{[compile]}d):
1.1       anton    7104: 
1.26      crook    7105: @example
                   7106: : constant ( n "name" -- )
                   7107:     create-interpret/compile
                   7108:     ,
                   7109: interpretation> ( -- n )
                   7110:     @@
                   7111: <interpretation
                   7112: compilation> ( compilation. -- ; run-time. -- n )
                   7113:     @@ postpone literal
                   7114: <compilation ;
                   7115: @end example
1.1       anton    7116: 
1.44      crook    7117: 
1.26      crook    7118: doc-create-interpret/compile
                   7119: doc-interpretation>
                   7120: doc-<interpretation
                   7121: doc-compilation>
                   7122: doc-<compilation
1.1       anton    7123: 
1.44      crook    7124: 
1.29      crook    7125: Words defined with @code{interpret/compile:} and
1.26      crook    7126: @code{create-interpret/compile} have an extended header structure that
                   7127: differs from other words; however, unless you try to access them with
                   7128: plain address arithmetic, you should not notice this. Words for
                   7129: accessing the header structure usually know how to deal with this; e.g.,
1.29      crook    7130: @code{'} @i{word} @code{>body} also gives you the body of a word created
                   7131: with @code{create-interpret/compile}.
1.1       anton    7132: 
1.44      crook    7133: 
1.47      crook    7134: @c -------------------------------------------------------------
1.81      anton    7135: @node Tokens for Words, Compiling words, Interpretation and Compilation Semantics, Words
1.47      crook    7136: @section Tokens for Words
                   7137: @cindex tokens for words
                   7138: 
                   7139: This section describes the creation and use of tokens that represent
                   7140: words.
                   7141: 
1.71      anton    7142: @menu
                   7143: * Execution token::             represents execution/interpretation semantics
                   7144: * Compilation token::           represents compilation semantics
                   7145: * Name token::                  represents named words
                   7146: @end menu
1.47      crook    7147: 
1.71      anton    7148: @node Execution token, Compilation token, Tokens for Words, Tokens for Words
                   7149: @subsection Execution token
1.47      crook    7150: 
                   7151: @cindex xt
                   7152: @cindex execution token
1.71      anton    7153: An @dfn{execution token} (@i{XT}) represents some behaviour of a word.
                   7154: You can use @code{execute} to invoke this behaviour.
1.47      crook    7155: 
1.71      anton    7156: @cindex tick (')
                   7157: You can use @code{'} to get an execution token that represents the
                   7158: interpretation semantics of a named word:
1.47      crook    7159: 
                   7160: @example
1.71      anton    7161: 5 ' .
                   7162: execute
                   7163: @end example
1.47      crook    7164: 
1.71      anton    7165: doc-'
                   7166: 
                   7167: @code{'} parses at run-time; there is also a word @code{[']} that parses
                   7168: when it is compiled, and compiles the resulting XT:
                   7169: 
                   7170: @example
                   7171: : foo ['] . execute ;
                   7172: 5 foo
                   7173: : bar ' execute ; \ by contrast,
                   7174: 5 bar .           \ ' parses "." when bar executes
                   7175: @end example
                   7176: 
                   7177: doc-[']
                   7178: 
                   7179: If you want the execution token of @i{word}, write @code{['] @i{word}}
                   7180: in compiled code and @code{' @i{word}} in interpreted code.  Gforth's
                   7181: @code{'} and @code{[']} behave somewhat unusually by complaining about
                   7182: compile-only words (because these words have no interpretation
                   7183: semantics).  You might get what you want by using @code{COMP' @i{word}
                   7184: DROP} or @code{[COMP'] @i{word} DROP} (for details @pxref{Compilation
                   7185: token}).
                   7186: 
                   7187: Another way to get an XT is @code{:noname} or @code{lastxt}
                   7188: (@pxref{Anonymous Definitions}).  For anonymous words this gives an xt
                   7189: for the only behaviour the word has (the execution semantics).  For
                   7190: named words, @code{lastxt} produces an XT for the same behaviour it
                   7191: would produce if the word was defined anonymously.
                   7192: 
                   7193: @example
                   7194: :noname ." hello" ;
                   7195: execute
1.47      crook    7196: @end example
                   7197: 
1.71      anton    7198: An XT occupies one cell and can be manipulated like any other cell.
                   7199: 
1.47      crook    7200: @cindex code field address
                   7201: @cindex CFA
1.71      anton    7202: In ANS Forth the XT is just an abstract data type (i.e., defined by the
                   7203: operations that produce or consume it).  For old hands: In Gforth, the
                   7204: XT is implemented as a code field address (CFA).
                   7205: 
                   7206: doc-execute
                   7207: doc-perform
                   7208: 
                   7209: @node Compilation token, Name token, Execution token, Tokens for Words
                   7210: @subsection Compilation token
1.47      crook    7211: 
                   7212: @cindex compilation token
1.71      anton    7213: @cindex CT (compilation token)
                   7214: Gforth represents the compilation semantics of a named word by a
1.47      crook    7215: @dfn{compilation token} consisting of two cells: @i{w xt}. The top cell
                   7216: @i{xt} is an execution token. The compilation semantics represented by
                   7217: the compilation token can be performed with @code{execute}, which
                   7218: consumes the whole compilation token, with an additional stack effect
                   7219: determined by the represented compilation semantics.
                   7220: 
                   7221: At present, the @i{w} part of a compilation token is an execution token,
                   7222: and the @i{xt} part represents either @code{execute} or
                   7223: @code{compile,}@footnote{Depending upon the compilation semantics of the
                   7224: word. If the word has default compilation semantics, the @i{xt} will
                   7225: represent @code{compile,}. Otherwise (e.g., for immediate words), the
                   7226: @i{xt} will represent @code{execute}.}. However, don't rely on that
                   7227: knowledge, unless necessary; future versions of Gforth may introduce
                   7228: unusual compilation tokens (e.g., a compilation token that represents
                   7229: the compilation semantics of a literal).
                   7230: 
1.71      anton    7231: You can perform the compilation semantics represented by the compilation
                   7232: token with @code{execute}.  You can compile the compilation semantics
                   7233: with @code{postpone,}. I.e., @code{COMP' @i{word} postpone,} is
                   7234: equivalent to @code{postpone @i{word}}.
                   7235: 
                   7236: doc-[comp']
                   7237: doc-comp'
                   7238: doc-postpone,
                   7239: 
                   7240: @node Name token,  , Compilation token, Tokens for Words
                   7241: @subsection Name token
1.47      crook    7242: 
                   7243: @cindex name token
                   7244: @cindex name field address
                   7245: @cindex NFA
1.71      anton    7246: Gforth represents named words by the @dfn{name token}, (@i{nt}). In
1.47      crook    7247: Gforth, the abstract data type @emph{name token} is implemented as a
                   7248: name field address (NFA).
                   7249: 
                   7250: doc-find-name
                   7251: doc-name>int
                   7252: doc-name?int
                   7253: doc-name>comp
                   7254: doc-name>string
                   7255: 
1.81      anton    7256: @c ----------------------------------------------------------
                   7257: @node Compiling words, The Text Interpreter, Tokens for Words, Words
                   7258: @section Compiling words
                   7259: @cindex compiling words
                   7260: @cindex macros
                   7261: 
                   7262: In contrast to most other languages, Forth has no strict boundary
1.82      anton    7263: between compilation and run-time.  E.g., you can run arbitrary code
                   7264: between defining words (or for computing data used by defining words
                   7265: like @code{constant}). Moreover, @code{Immediate} (@pxref{Interpretation
                   7266: and Compilation Semantics} and @code{[}...@code{]} (see below) allow
                   7267: running arbitrary code while compiling a colon definition (exception:
                   7268: you must not allot dictionary space).
                   7269: 
                   7270: @menu
                   7271: * Literals::                    Compiling data values
                   7272: * Macros::                      Compiling words
                   7273: @end menu
                   7274: 
                   7275: @node Literals, Macros, Compiling words, Compiling words
                   7276: @subsection Literals
                   7277: @cindex Literals
                   7278: 
                   7279: The simplest and most frequent example is to compute a literal during
                   7280: compilation.  E.g., the following definition prints an array of strings,
                   7281: one string per line:
                   7282: 
                   7283: @example
                   7284: : .strings ( addr u -- ) \ gforth
                   7285:     2* cells bounds U+DO
                   7286:        cr i 2@@ type
                   7287:     2 cells +LOOP ;  
                   7288: @end example
1.81      anton    7289: 
1.82      anton    7290: With a simple-minded compiler like Gforth's, this computes @code{2
                   7291: cells} on every loop iteration.  You can compute this value once and for
                   7292: all at compile time and compile it into the definition like this:
                   7293: 
                   7294: @example
                   7295: : .strings ( addr u -- ) \ gforth
                   7296:     2* cells bounds U+DO
                   7297:        cr i 2@@ type
                   7298:     [ 2 cells ] literal +LOOP ;  
                   7299: @end example
                   7300: 
                   7301: @code{[} switches the text interpreter to interpret state (you will get
                   7302: an @code{ok} prompt if you type this example interactively and insert a
                   7303: newline between @code{[} and @code{]}), so it performs the
                   7304: interpretation semantics of @code{2 cells}; this computes a number.
                   7305: @code{]} switches the text interpreter back into compile state.  It then
                   7306: performs @code{Literal}'s compilation semantics, which are to compile
                   7307: this number into the current word.  You can decompile the word with
                   7308: @code{see .strings} to see the effect on the compiled code.
1.81      anton    7309: 
1.82      anton    7310: You can also optimize the @code{2* cells} into @code{[ 2 cells ] literal
                   7311: *} in this way.
1.81      anton    7312: 
1.82      anton    7313: doc-[
                   7314: doc-]
1.81      anton    7315: doc-literal
                   7316: doc-]L
1.82      anton    7317: 
                   7318: There are also words for compiling other data types than single cells as
                   7319: literals:
                   7320: 
1.81      anton    7321: doc-2literal
                   7322: doc-fliteral
1.82      anton    7323: doc-sliteral
                   7324: 
                   7325: @cindex colon-sys, passing data across @code{:}
                   7326: @cindex @code{:}, passing data across
                   7327: You might be tempted to pass data from outside a colon definition to the
                   7328: inside on the data stack.  This does not work, because @code{:} puhes a
                   7329: colon-sys, making stuff below unaccessible.  E.g., this does not work:
                   7330: 
                   7331: @example
                   7332: 5 : foo literal ; \ error: "unstructured"
                   7333: @end example
                   7334: 
                   7335: Instead, you have to pass the value in some other way, e.g., through a
                   7336: variable:
                   7337: 
                   7338: @example
                   7339: variable temp
                   7340: 5 temp !
                   7341: : foo [ temp @@ ] literal ;
                   7342: @end example
                   7343: 
                   7344: 
                   7345: @node Macros,  , Literals, Compiling words
                   7346: @subsection Macros
                   7347: @cindex Macros
                   7348: @cindex compiling compilation semantics
                   7349: 
                   7350: @code{Literal} and friends compile data values into the current
                   7351: definition.  You can also write words that compile other words into the
                   7352: current definition.  E.g.,
                   7353: 
                   7354: @example
                   7355: : compile-+ ( -- ) \ compiled code: ( n1 n2 -- n )
                   7356:   POSTPONE + ;
                   7357: 
                   7358: : foo ( n1 n2 -- n )
                   7359:   [ compile-+ ] ;
                   7360: 1 2 foo .
                   7361: @end example
                   7362: 
                   7363: This is equivalent to @code{: foo + ;} (@code{see foo} to check this).
                   7364: What happens in this example?  @code{Postpone} compiles the compilation
                   7365: semantics of @code{+} into @code{compile-+}; later the text interpreter
                   7366: executes @code{compile-+} and thus the compilation semantics of +, which
                   7367: compile (the execution semantics of) @code{+} into
                   7368: @code{foo}.@footnote{A recent RFI answer requires that compiling words
                   7369: should only be executed in compile state, so this example is not
                   7370: guaranteed to work on all standard systems, but on any decent system it
                   7371: will work.}
                   7372: 
                   7373: doc-postpone
                   7374: doc-[compile]
                   7375: 
                   7376: Compiling words like @code{compile-+} are usually immediate (or similar)
                   7377: so you do not have to switch to interpret state to execute them;
                   7378: mopifying the last example accordingly produces:
                   7379: 
                   7380: @example
                   7381: : [compile-+] ( compilation: --; interpretation: -- )
                   7382:   \ compiled code: ( n1 n2 -- n )
                   7383:   POSTPONE + ; immediate
                   7384: 
                   7385: : foo ( n1 n2 -- n )
                   7386:   [compile-+] ;
                   7387: 1 2 foo .
                   7388: @end example
                   7389: 
                   7390: Immediate compiling words are similar to macros in other languages (in
                   7391: particular, Lisp).  The important differences to macros in, e.g., C are:
                   7392: 
                   7393: @itemize @bullet
                   7394: 
                   7395: @item
                   7396: You use the same language for defining and processing macros, not a
                   7397: separate preprocessing language and processor.
                   7398: 
                   7399: @item
                   7400: Consequently, the full power of Forth is available in macro definitions.
                   7401: E.g., you can perform arbitrarily complex computations, or generate
                   7402: different code conditionally or in a loop (e.g., @pxref{Advanced macros
                   7403: Tutorial}).  This power is very useful when writing a parser generators
                   7404: or other code-generating software.
                   7405: 
                   7406: @item
                   7407: Macros defined using @code{postpone} etc. deal with the language at a
                   7408: higher level than strings; name binding happens at macro definition
                   7409: time, so you can avoid the pitfalls of name collisions that can happen
                   7410: in C macros.  Of course, Forth is a liberal language and also allows to
                   7411: shoot yourself in the foot with text-interpreted macros like
                   7412: 
                   7413: @example
                   7414: : [compile-+] s" +" evaluate ; immediate
                   7415: @end example
                   7416: 
                   7417: Apart from binding the name at macro use time, using @code{evaluate}
                   7418: also makes your definition @code{state}-smart (@pxref{state-smartness}).
                   7419: @end itemize
                   7420: 
                   7421: You may want the macro to compile a number into a word.  The word to do
                   7422: it is @code{literal}, but you have to @code{postpone} it, so its
                   7423: compilation semantics take effect when the macro is executed, not when
                   7424: it is compiled:
                   7425: 
                   7426: @example
                   7427: : [compile-5] ( -- ) \ compiled code: ( -- n )
                   7428:   5 POSTPONE literal ; immediate
                   7429: 
                   7430: : foo [compile-5] ;
                   7431: foo .
                   7432: @end example
                   7433: 
                   7434: You may want to pass parameters to a macro, that the macro should
                   7435: compile into the current definition.  If the parameter is a number, then
                   7436: you can use @code{postpone literal} (similar for other values).
                   7437: 
                   7438: If you want to pass a word that is to be compiled, the usual way is to
                   7439: pass an execution token and @code{compile,} it:
                   7440: 
                   7441: @example
                   7442: : twice1 ( xt -- ) \ compiled code: ... -- ...
                   7443:   dup compile, compile, ;
                   7444: 
                   7445: : 2+ ( n1 -- n2 )
                   7446:   [ ' 1+ twice1 ] ;
                   7447: @end example
                   7448: 
                   7449: doc-compile,
                   7450: 
                   7451: An alternative available in Gforth, that allows you to pass compile-only
                   7452: words as parameters is to use the compilation token (@pxref{Compilation
                   7453: token}).  The same example in this technique:
                   7454: 
                   7455: @example
                   7456: : twice ( ... ct -- ... ) \ compiled code: ... -- ...
                   7457:   2dup 2>r execute 2r> execute ;
                   7458: 
                   7459: : 2+ ( n1 -- n2 )
                   7460:   [ comp' 1+ twice ] ;
                   7461: @end example
                   7462: 
                   7463: In the example above @code{2>r} and @code{2r>} ensure that @code{twice}
                   7464: works even if the executed compilation semantics has an effect on the
                   7465: data stack.
                   7466: 
                   7467: You can also define complete definitions with these words; this provides
                   7468: an alternative to using @code{does>} (@pxref{User-defined Defining
                   7469: Words}).  E.g., instead of
                   7470: 
                   7471: @example
                   7472: : curry+ ( n1 "name" -- )
                   7473:     CREATE ,
                   7474: DOES> ( n2 -- n1+n2 )
                   7475:     @@ + ;
                   7476: @end example
                   7477: 
                   7478: you could define
                   7479: 
                   7480: @example
                   7481: : curry+ ( n1 "name" -- )
                   7482:   \ name execution: ( n2 -- n1+n2 )
                   7483:   >r : r> POSTPONE literal POSTPONE + POSTPONE ; ;
1.81      anton    7484: 
1.82      anton    7485: -3 curry+ 3-
                   7486: see 3-
                   7487: @end example
1.81      anton    7488: 
1.82      anton    7489: The sequence @code{>r : r>} is necessary, because @code{:} puts a
                   7490: colon-sys on the data stack that makes everything below it unaccessible.
1.81      anton    7491: 
1.82      anton    7492: This way of writing defining words is sometimes more, sometimes less
                   7493: convenient than using @code{does>} (@pxref{Advanced does> usage
                   7494: example}).  One advantage of this method is that it can be optimized
                   7495: better, because the compiler knows that the value compiled with
                   7496: @code{literal} is fixed, whereas the data associated with a
                   7497: @code{create}d word can be changed.
1.47      crook    7498: 
1.26      crook    7499: @c ----------------------------------------------------------
1.81      anton    7500: @node The Text Interpreter, Word Lists, Compiling words, Words
1.26      crook    7501: @section  The Text Interpreter
                   7502: @cindex interpreter - outer
                   7503: @cindex text interpreter
                   7504: @cindex outer interpreter
1.1       anton    7505: 
1.34      anton    7506: @c Should we really describe all these ugly details?  IMO the text
                   7507: @c interpreter should be much cleaner, but that may not be possible within
                   7508: @c ANS Forth. - anton
1.44      crook    7509: @c nac-> I wanted to explain how it works to show how you can exploit
                   7510: @c it in your own programs. When I was writing a cross-compiler, figuring out
                   7511: @c some of these gory details was very helpful to me. None of the textbooks
                   7512: @c I've seen cover it, and the most modern Forth textbook -- Forth Inc's,
                   7513: @c seems to positively avoid going into too much detail for some of
                   7514: @c the internals.
1.34      anton    7515: 
1.71      anton    7516: @c anton: ok.  I wonder, though, if this is the right place; for some stuff
                   7517: @c it is; for the ugly details, I would prefer another place.  I wonder
                   7518: @c whether we should have a chapter before "Words" that describes some
                   7519: @c basic concepts referred to in words, and a chapter after "Words" that
                   7520: @c describes implementation details.
                   7521: 
1.29      crook    7522: The text interpreter@footnote{This is an expanded version of the
                   7523: material in @ref{Introducing the Text Interpreter}.} is an endless loop
1.34      anton    7524: that processes input from the current input device. It is also called
                   7525: the outer interpreter, in contrast to the inner interpreter
                   7526: (@pxref{Engine}) which executes the compiled Forth code on interpretive
                   7527: implementations.
1.27      crook    7528: 
1.29      crook    7529: @cindex interpret state
                   7530: @cindex compile state
                   7531: The text interpreter operates in one of two states: @dfn{interpret
                   7532: state} and @dfn{compile state}. The current state is defined by the
1.71      anton    7533: aptly-named variable @code{state}.
1.29      crook    7534: 
                   7535: This section starts by describing how the text interpreter behaves when
                   7536: it is in interpret state, processing input from the user input device --
                   7537: the keyboard. This is the mode that a Forth system is in after it starts
                   7538: up.
                   7539: 
                   7540: @cindex input buffer
                   7541: @cindex terminal input buffer
                   7542: The text interpreter works from an area of memory called the @dfn{input
                   7543: buffer}@footnote{When the text interpreter is processing input from the
                   7544: keyboard, this area of memory is called the @dfn{terminal input buffer}
                   7545: (TIB) and is addressed by the (obsolescent) words @code{TIB} and
                   7546: @code{#TIB}.}, which stores your keyboard input when you press the
1.30      anton    7547: @key{RET} key. Starting at the beginning of the input buffer, it skips
1.29      crook    7548: leading spaces (called @dfn{delimiters}) then parses a string (a
                   7549: sequence of non-space characters) until it reaches either a space
                   7550: character or the end of the buffer. Having parsed a string, it makes two
                   7551: attempts to process it:
1.27      crook    7552: 
1.29      crook    7553: @cindex dictionary
1.27      crook    7554: @itemize @bullet
                   7555: @item
1.29      crook    7556: It looks for the string in a @dfn{dictionary} of definitions. If the
                   7557: string is found, the string names a @dfn{definition} (also known as a
                   7558: @dfn{word}) and the dictionary search returns information that allows
                   7559: the text interpreter to perform the word's @dfn{interpretation
                   7560: semantics}. In most cases, this simply means that the word will be
                   7561: executed.
1.27      crook    7562: @item
                   7563: If the string is not found in the dictionary, the text interpreter
1.29      crook    7564: attempts to treat it as a number, using the rules described in
                   7565: @ref{Number Conversion}. If the string represents a legal number in the
                   7566: current radix, the number is pushed onto a parameter stack (the data
                   7567: stack for integers, the floating-point stack for floating-point
                   7568: numbers).
                   7569: @end itemize
                   7570: 
                   7571: If both attempts fail, or if the word is found in the dictionary but has
                   7572: no interpretation semantics@footnote{This happens if the word was
                   7573: defined as @code{COMPILE-ONLY}.} the text interpreter discards the
                   7574: remainder of the input buffer, issues an error message and waits for
                   7575: more input. If one of the attempts succeeds, the text interpreter
                   7576: repeats the parsing process until the whole of the input buffer has been
                   7577: processed, at which point it prints the status message ``@code{ ok}''
                   7578: and waits for more input.
                   7579: 
1.71      anton    7580: @c anton: this should be in the input stream subsection (or below it)
                   7581: 
1.29      crook    7582: @cindex parse area
                   7583: The text interpreter keeps track of its position in the input buffer by
                   7584: updating a variable called @code{>IN} (pronounced ``to-in''). The value
                   7585: of @code{>IN} starts out as 0, indicating an offset of 0 from the start
                   7586: of the input buffer. The region from offset @code{>IN @@} to the end of
                   7587: the input buffer is called the @dfn{parse area}@footnote{In other words,
                   7588: the text interpreter processes the contents of the input buffer by
                   7589: parsing strings from the parse area until the parse area is empty.}.
                   7590: This example shows how @code{>IN} changes as the text interpreter parses
                   7591: the input buffer:
                   7592: 
                   7593: @example
                   7594: : remaining >IN @@ SOURCE 2 PICK - -ROT + SWAP
                   7595:   CR ." ->" TYPE ." <-" ; IMMEDIATE 
                   7596: 
                   7597: 1 2 3 remaining + remaining . 
                   7598: 
                   7599: : foo 1 2 3 remaining SWAP remaining ;
                   7600: @end example
                   7601: 
                   7602: @noindent
                   7603: The result is:
                   7604: 
                   7605: @example
                   7606: ->+ remaining .<-
                   7607: ->.<-5  ok
                   7608: 
                   7609: ->SWAP remaining ;-<
                   7610: ->;<-  ok
                   7611: @end example
                   7612: 
                   7613: @cindex parsing words
                   7614: The value of @code{>IN} can also be modified by a word in the input
                   7615: buffer that is executed by the text interpreter.  This means that a word
                   7616: can ``trick'' the text interpreter into either skipping a section of the
                   7617: input buffer@footnote{This is how parsing words work.} or into parsing a
                   7618: section twice. For example:
1.27      crook    7619: 
1.29      crook    7620: @example
1.71      anton    7621: : lat ." <<foo>>" ;
                   7622: : flat ." <<bar>>" >IN DUP @@ 3 - SWAP ! ;
1.29      crook    7623: @end example
                   7624: 
                   7625: @noindent
                   7626: When @code{flat} is executed, this output is produced@footnote{Exercise
                   7627: for the reader: what would happen if the @code{3} were replaced with
                   7628: @code{4}?}:
                   7629: 
                   7630: @example
1.71      anton    7631: <<bar>><<foo>>
1.29      crook    7632: @end example
                   7633: 
1.71      anton    7634: This technique can be used to work around some of the interoperability
                   7635: problems of parsing words.  Of course, it's better to avoid parsing
                   7636: words where possible.
                   7637: 
1.29      crook    7638: @noindent
                   7639: Two important notes about the behaviour of the text interpreter:
1.27      crook    7640: 
                   7641: @itemize @bullet
                   7642: @item
                   7643: It processes each input string to completion before parsing additional
1.29      crook    7644: characters from the input buffer.
                   7645: @item
                   7646: It treats the input buffer as a read-only region (and so must your code).
                   7647: @end itemize
                   7648: 
                   7649: @noindent
                   7650: When the text interpreter is in compile state, its behaviour changes in
                   7651: these ways:
                   7652: 
                   7653: @itemize @bullet
                   7654: @item
                   7655: If a parsed string is found in the dictionary, the text interpreter will
                   7656: perform the word's @dfn{compilation semantics}. In most cases, this
                   7657: simply means that the execution semantics of the word will be appended
                   7658: to the current definition.
1.27      crook    7659: @item
1.29      crook    7660: When a number is encountered, it is compiled into the current definition
                   7661: (as a literal) rather than being pushed onto a parameter stack.
                   7662: @item
                   7663: If an error occurs, @code{state} is modified to put the text interpreter
                   7664: back into interpret state.
                   7665: @item
                   7666: Each time a line is entered from the keyboard, Gforth prints
                   7667: ``@code{ compiled}'' rather than `` @code{ok}''.
                   7668: @end itemize
                   7669: 
                   7670: @cindex text interpreter - input sources
                   7671: When the text interpreter is using an input device other than the
                   7672: keyboard, its behaviour changes in these ways:
                   7673: 
                   7674: @itemize @bullet
                   7675: @item
                   7676: When the parse area is empty, the text interpreter attempts to refill
                   7677: the input buffer from the input source. When the input source is
1.71      anton    7678: exhausted, the input source is set back to the previous input source.
1.29      crook    7679: @item
                   7680: It doesn't print out ``@code{ ok}'' or ``@code{ compiled}'' messages each
                   7681: time the parse area is emptied.
                   7682: @item
                   7683: If an error occurs, the input source is set back to the user input
                   7684: device.
1.27      crook    7685: @end itemize
1.21      crook    7686: 
1.49      anton    7687: You can read about this in more detail in @ref{Input Sources}.
1.44      crook    7688: 
1.26      crook    7689: doc->in
1.27      crook    7690: doc-source
                   7691: 
1.26      crook    7692: doc-tib
                   7693: doc-#tib
1.1       anton    7694: 
1.44      crook    7695: 
1.26      crook    7696: @menu
1.67      anton    7697: * Input Sources::               
                   7698: * Number Conversion::           
                   7699: * Interpret/Compile states::    
                   7700: * Interpreter Directives::      
1.26      crook    7701: @end menu
1.1       anton    7702: 
1.29      crook    7703: @node Input Sources, Number Conversion, The Text Interpreter, The Text Interpreter
                   7704: @subsection Input Sources
                   7705: @cindex input sources
                   7706: @cindex text interpreter - input sources
                   7707: 
1.44      crook    7708: By default, the text interpreter processes input from the user input
1.29      crook    7709: device (the keyboard) when Forth starts up. The text interpreter can
                   7710: process input from any of these sources:
                   7711: 
                   7712: @itemize @bullet
                   7713: @item
                   7714: The user input device -- the keyboard.
                   7715: @item
                   7716: A file, using the words described in @ref{Forth source files}.
                   7717: @item
                   7718: A block, using the words described in @ref{Blocks}.
                   7719: @item
                   7720: A text string, using @code{evaluate}.
                   7721: @end itemize
                   7722: 
                   7723: A program can identify the current input device from the values of
                   7724: @code{source-id} and @code{blk}.
                   7725: 
1.44      crook    7726: 
1.29      crook    7727: doc-source-id
                   7728: doc-blk
                   7729: 
                   7730: doc-save-input
                   7731: doc-restore-input
                   7732: 
                   7733: doc-evaluate
1.1       anton    7734: 
1.29      crook    7735: 
1.44      crook    7736: 
1.29      crook    7737: @node Number Conversion, Interpret/Compile states, Input Sources, The Text Interpreter
1.26      crook    7738: @subsection Number Conversion
                   7739: @cindex number conversion
                   7740: @cindex double-cell numbers, input format
                   7741: @cindex input format for double-cell numbers
                   7742: @cindex single-cell numbers, input format
                   7743: @cindex input format for single-cell numbers
                   7744: @cindex floating-point numbers, input format
                   7745: @cindex input format for floating-point numbers
1.1       anton    7746: 
1.29      crook    7747: This section describes the rules that the text interpreter uses when it
                   7748: tries to convert a string into a number.
1.1       anton    7749: 
1.26      crook    7750: Let <digit> represent any character that is a legal digit in the current
1.29      crook    7751: number base@footnote{For example, 0-9 when the number base is decimal or
                   7752: 0-9, A-F when the number base is hexadecimal.}.
1.1       anton    7753: 
1.26      crook    7754: Let <decimal digit> represent any character in the range 0-9.
1.1       anton    7755: 
1.29      crook    7756: Let @{@i{a b}@} represent the @i{optional} presence of any of the characters
                   7757: in the braces (@i{a} or @i{b} or neither).
1.1       anton    7758: 
1.26      crook    7759: Let * represent any number of instances of the previous character
                   7760: (including none).
1.1       anton    7761: 
1.26      crook    7762: Let any other character represent itself.
1.1       anton    7763: 
1.29      crook    7764: @noindent
1.26      crook    7765: Now, the conversion rules are:
1.21      crook    7766: 
1.26      crook    7767: @itemize @bullet
                   7768: @item
                   7769: A string of the form <digit><digit>* is treated as a single-precision
1.29      crook    7770: (cell-sized) positive integer. Examples are 0 123 6784532 32343212343456 42
1.26      crook    7771: @item
                   7772: A string of the form -<digit><digit>* is treated as a single-precision
1.29      crook    7773: (cell-sized) negative integer, and is represented using 2's-complement
1.26      crook    7774: arithmetic. Examples are -45 -5681 -0
                   7775: @item
                   7776: A string of the form <digit><digit>*.<digit>* is treated as a double-precision
1.29      crook    7777: (double-cell-sized) positive integer. Examples are 3465. 3.465 34.65
                   7778: (all three of these represent the same number).
1.26      crook    7779: @item
                   7780: A string of the form -<digit><digit>*.<digit>* is treated as a
1.29      crook    7781: double-precision (double-cell-sized) negative integer, and is
1.26      crook    7782: represented using 2's-complement arithmetic. Examples are -3465. -3.465
1.29      crook    7783: -34.65 (all three of these represent the same number).
1.26      crook    7784: @item
1.29      crook    7785: A string of the form @{+ -@}<decimal digit>@{.@}<decimal digit>*@{e
                   7786: E@}@{+ -@}<decimal digit><decimal digit>* is treated as a floating-point
1.35      anton    7787: number. Examples are 1e 1e0 1.e 1.e0 +1e+0 (which all represent the same
1.29      crook    7788: number) +12.E-4
1.26      crook    7789: @end itemize
1.1       anton    7790: 
1.26      crook    7791: By default, the number base used for integer number conversion is given
1.35      anton    7792: by the contents of the variable @code{base}.  Note that a lot of
                   7793: confusion can result from unexpected values of @code{base}.  If you
                   7794: change @code{base} anywhere, make sure to save the old value and restore
                   7795: it afterwards.  In general I recommend keeping @code{base} decimal, and
                   7796: using the prefixes described below for the popular non-decimal bases.
1.1       anton    7797: 
1.29      crook    7798: doc-dpl
1.26      crook    7799: doc-base
                   7800: doc-hex
                   7801: doc-decimal
1.1       anton    7802: 
1.44      crook    7803: 
1.26      crook    7804: @cindex '-prefix for character strings
                   7805: @cindex &-prefix for decimal numbers
                   7806: @cindex %-prefix for binary numbers
                   7807: @cindex $-prefix for hexadecimal numbers
1.35      anton    7808: Gforth allows you to override the value of @code{base} by using a
1.29      crook    7809: prefix@footnote{Some Forth implementations provide a similar scheme by
                   7810: implementing @code{$} etc. as parsing words that process the subsequent
                   7811: number in the input stream and push it onto the stack. For example, see
                   7812: @cite{Number Conversion and Literals}, by Wil Baden; Forth Dimensions
                   7813: 20(3) pages 26--27. In such implementations, unlike in Gforth, a space
                   7814: is required between the prefix and the number.} before the first digit
                   7815: of an (integer) number. Four prefixes are supported:
1.1       anton    7816: 
1.26      crook    7817: @itemize @bullet
                   7818: @item
1.35      anton    7819: @code{&} -- decimal
1.26      crook    7820: @item
1.35      anton    7821: @code{%} -- binary
1.26      crook    7822: @item
1.35      anton    7823: @code{$} -- hexadecimal
1.26      crook    7824: @item
1.35      anton    7825: @code{'} -- base @code{max-char+1}
1.26      crook    7826: @end itemize
1.1       anton    7827: 
1.26      crook    7828: Here are some examples, with the equivalent decimal number shown after
                   7829: in braces:
1.1       anton    7830: 
1.26      crook    7831: -$41 (-65), %1001101 (205), %1001.0001 (145 - a double-precision number),
                   7832: 'AB (16706; ascii A is 65, ascii B is 66, number is 65*256 + 66),
                   7833: 'ab (24930; ascii a is 97, ascii B is 98, number is 97*256 + 98),
                   7834: &905 (905), $abc (2478), $ABC (2478).
1.1       anton    7835: 
1.26      crook    7836: @cindex number conversion - traps for the unwary
1.29      crook    7837: @noindent
1.26      crook    7838: Number conversion has a number of traps for the unwary:
1.1       anton    7839: 
1.26      crook    7840: @itemize @bullet
                   7841: @item
                   7842: You cannot determine the current number base using the code sequence
1.35      anton    7843: @code{base @@ .} -- the number base is always 10 in the current number
                   7844: base. Instead, use something like @code{base @@ dec.}
1.26      crook    7845: @item
                   7846: If the number base is set to a value greater than 14 (for example,
                   7847: hexadecimal), the number 123E4 is ambiguous; the conversion rules allow
                   7848: it to be intepreted as either a single-precision integer or a
                   7849: floating-point number (Gforth treats it as an integer). The ambiguity
                   7850: can be resolved by explicitly stating the sign of the mantissa and/or
                   7851: exponent: 123E+4 or +123E4 -- if the number base is decimal, no
                   7852: ambiguity arises; either representation will be treated as a
                   7853: floating-point number.
                   7854: @item
1.29      crook    7855: There is a word @code{bin} but it does @i{not} set the number base!
1.26      crook    7856: It is used to specify file types.
                   7857: @item
1.72      anton    7858: ANS Forth requires the @code{.} of a double-precision number to be the
                   7859: final character in the string.  Gforth allows the @code{.} to be
                   7860: anywhere after the first digit.
1.26      crook    7861: @item
                   7862: The number conversion process does not check for overflow.
                   7863: @item
1.72      anton    7864: In an ANS Forth program @code{base} is required to be decimal when
                   7865: converting floating-point numbers.  In Gforth, number conversion to
                   7866: floating-point numbers always uses base &10, irrespective of the value
                   7867: of @code{base}.
1.26      crook    7868: @end itemize
1.1       anton    7869: 
1.49      anton    7870: You can read numbers into your programs with the words described in
                   7871: @ref{Input}.
1.1       anton    7872: 
1.82      anton    7873: @node Interpret/Compile states, Interpreter Directives, Number Conversion, The Text Interpreter
1.26      crook    7874: @subsection Interpret/Compile states
                   7875: @cindex Interpret/Compile states
1.1       anton    7876: 
1.29      crook    7877: A standard program is not permitted to change @code{state}
                   7878: explicitly. However, it can change @code{state} implicitly, using the
                   7879: words @code{[} and @code{]}. When @code{[} is executed it switches
                   7880: @code{state} to interpret state, and therefore the text interpreter
                   7881: starts interpreting. When @code{]} is executed it switches @code{state}
                   7882: to compile state and therefore the text interpreter starts
1.44      crook    7883: compiling. The most common usage for these words is for switching into
                   7884: interpret state and back from within a colon definition; this technique
1.49      anton    7885: can be used to compile a literal (for an example, @pxref{Literals}) or
                   7886: for conditional compilation (for an example, @pxref{Interpreter
                   7887: Directives}).
1.44      crook    7888: 
1.35      anton    7889: 
                   7890: @c This is a bad example: It's non-standard, and it's not necessary.
                   7891: @c However, I can't think of a good example for switching into compile
                   7892: @c state when there is no current word (@code{state}-smart words are not a
                   7893: @c good reason).  So maybe we should use an example for switching into
                   7894: @c interpret @code{state} in a colon def. - anton
1.44      crook    7895: @c nac-> I agree. I started out by putting in the example, then realised
                   7896: @c that it was non-ANS, so wrote more words around it. I hope this
                   7897: @c re-written version is acceptable to you. I do want to keep the example
                   7898: @c as it is helpful for showing what is and what is not portable, particularly
                   7899: @c where it outlaws a style in common use.
                   7900: 
1.72      anton    7901: @c anton: it's more important to show what's portable.  After we have done
1.83      anton    7902: @c that, we can also show what's not.  In any case, I have written a
                   7903: @c section Compiling Words which also deals with [ ].
1.35      anton    7904: 
1.44      crook    7905: @code{[} and @code{]} also give you the ability to switch into compile
                   7906: state and back, but we cannot think of any useful Standard application
                   7907: for this ability. Pre-ANS Forth textbooks have examples like this:
1.29      crook    7908: 
                   7909: @example
                   7910: : AA ." this is A" ;
                   7911: : BB ." this is B" ;
                   7912: : CC ." this is C" ;
                   7913: 
1.44      crook    7914: create table ] aa bb cc [
                   7915: 
1.29      crook    7916: : go ( n -- ) \ n is offset into table.. 0 for 1st entry
                   7917:   cells table + @ execute ;
                   7918: @end example
                   7919: 
1.44      crook    7920: This example builds a jump table; @code{0 go} will display ``@code{this
                   7921: is A}''. Using @code{[} and @code{]} in this example is equivalent to
                   7922: defining @code{table} like this:
1.29      crook    7923: 
                   7924: @example
1.44      crook    7925: create table ' aa COMPILE, ' bb COMPILE, ' cc COMPILE,
1.29      crook    7926: @end example
                   7927: 
1.44      crook    7928: The problem with this code is that the definition of @code{table} is not
                   7929: portable -- it @i{compile}s execution tokens into code space. Whilst it
                   7930: @i{may} work on systems where code space and data space co-incide, the
1.29      crook    7931: Standard only allows data space to be assigned for a @code{CREATE}d
                   7932: word. In addition, the Standard only allows @code{@@} to access data
                   7933: space, whilst this example is using it to access code space. The only
                   7934: portable, Standard way to build this table is to build it in data space,
                   7935: like this:
                   7936: 
                   7937: @example
                   7938: create table ' aa , ' bb , ' cc ,
                   7939: @end example
                   7940: 
1.26      crook    7941: doc-state
1.44      crook    7942: 
1.29      crook    7943: 
1.82      anton    7944: @node Interpreter Directives,  , Interpret/Compile states, The Text Interpreter
1.26      crook    7945: @subsection Interpreter Directives
                   7946: @cindex interpreter directives
1.72      anton    7947: @cindex conditional compilation
1.1       anton    7948: 
1.29      crook    7949: These words are usually used in interpret state; typically to control
                   7950: which parts of a source file are processed by the text
1.26      crook    7951: interpreter. There are only a few ANS Forth Standard words, but Gforth
                   7952: supplements these with a rich set of immediate control structure words
                   7953: to compensate for the fact that the non-immediate versions can only be
1.29      crook    7954: used in compile state (@pxref{Control Structures}). Typical usages:
                   7955: 
                   7956: @example
1.72      anton    7957: FALSE Constant HAVE-ASSEMBLER
1.29      crook    7958: .
                   7959: .
1.72      anton    7960: HAVE-ASSEMBLER [IF]
1.29      crook    7961: : ASSEMBLER-FEATURE
                   7962:   ...
                   7963: ;
                   7964: [ENDIF]
                   7965: .
                   7966: .
                   7967: : SEE
                   7968:   ... \ general-purpose SEE code
1.72      anton    7969:   [ HAVE-ASSEMBLER [IF] ]
1.29      crook    7970:   ... \ assembler-specific SEE code
                   7971:   [ [ENDIF] ]
                   7972: ;
                   7973: @end example
1.1       anton    7974: 
1.44      crook    7975: 
1.26      crook    7976: doc-[IF]
                   7977: doc-[ELSE]
                   7978: doc-[THEN]
                   7979: doc-[ENDIF]
1.1       anton    7980: 
1.26      crook    7981: doc-[IFDEF]
                   7982: doc-[IFUNDEF]
1.1       anton    7983: 
1.26      crook    7984: doc-[?DO]
                   7985: doc-[DO]
                   7986: doc-[FOR]
                   7987: doc-[LOOP]
                   7988: doc-[+LOOP]
                   7989: doc-[NEXT]
1.1       anton    7990: 
1.26      crook    7991: doc-[BEGIN]
                   7992: doc-[UNTIL]
                   7993: doc-[AGAIN]
                   7994: doc-[WHILE]
                   7995: doc-[REPEAT]
1.1       anton    7996: 
1.27      crook    7997: 
1.26      crook    7998: @c -------------------------------------------------------------
1.47      crook    7999: @node Word Lists, Environmental Queries, The Text Interpreter, Words
1.26      crook    8000: @section Word Lists
                   8001: @cindex word lists
1.32      anton    8002: @cindex header space
1.1       anton    8003: 
1.36      anton    8004: A wordlist is a list of named words; you can add new words and look up
                   8005: words by name (and you can remove words in a restricted way with
                   8006: markers).  Every named (and @code{reveal}ed) word is in one wordlist.
                   8007: 
                   8008: @cindex search order stack
                   8009: The text interpreter searches the wordlists present in the search order
                   8010: (a stack of wordlists), from the top to the bottom.  Within each
                   8011: wordlist, the search starts conceptually at the newest word; i.e., if
                   8012: two words in a wordlist have the same name, the newer word is found.
1.1       anton    8013: 
1.26      crook    8014: @cindex compilation word list
1.36      anton    8015: New words are added to the @dfn{compilation wordlist} (aka current
                   8016: wordlist).
1.1       anton    8017: 
1.36      anton    8018: @cindex wid
                   8019: A word list is identified by a cell-sized word list identifier (@i{wid})
                   8020: in much the same way as a file is identified by a file handle. The
                   8021: numerical value of the wid has no (portable) meaning, and might change
                   8022: from session to session.
1.1       anton    8023: 
1.29      crook    8024: The ANS Forth ``Search order'' word set is intended to provide a set of
                   8025: low-level tools that allow various different schemes to be
1.74      anton    8026: implemented. Gforth also provides @code{vocabulary}, a traditional Forth
1.26      crook    8027: word.  @file{compat/vocabulary.fs} provides an implementation in ANS
1.45      crook    8028: Forth.
1.1       anton    8029: 
1.27      crook    8030: @comment TODO: locals section refers to here, saying that every word list (aka
                   8031: @comment vocabulary) has its own methods for searching etc. Need to document that.
1.78      anton    8032: @c anton: but better in a separate subsection on wordlist internals
1.1       anton    8033: 
1.45      crook    8034: @comment TODO: document markers, reveal, tables, mappedwordlist
                   8035: 
                   8036: @comment the gforthman- prefix is used to pick out the true definition of a
1.27      crook    8037: @comment word from the source files, rather than some alias.
1.44      crook    8038: 
1.26      crook    8039: doc-forth-wordlist
                   8040: doc-definitions
                   8041: doc-get-current
                   8042: doc-set-current
                   8043: doc-get-order
1.45      crook    8044: doc---gforthman-set-order
1.26      crook    8045: doc-wordlist
1.30      anton    8046: doc-table
1.79      anton    8047: doc->order
1.36      anton    8048: doc-previous
1.26      crook    8049: doc-also
1.45      crook    8050: doc---gforthman-forth
1.26      crook    8051: doc-only
1.45      crook    8052: doc---gforthman-order
1.15      anton    8053: 
1.26      crook    8054: doc-find
                   8055: doc-search-wordlist
1.15      anton    8056: 
1.26      crook    8057: doc-words
                   8058: doc-vlist
1.44      crook    8059: @c doc-words-deferred
1.1       anton    8060: 
1.74      anton    8061: @c doc-mappedwordlist @c map-structure undefined, implemantation-specific
1.26      crook    8062: doc-root
                   8063: doc-vocabulary
                   8064: doc-seal
                   8065: doc-vocs
                   8066: doc-current
                   8067: doc-context
1.1       anton    8068: 
1.44      crook    8069: 
1.26      crook    8070: @menu
1.75      anton    8071: * Vocabularies::                
1.67      anton    8072: * Why use word lists?::         
1.75      anton    8073: * Word list example::           
1.26      crook    8074: @end menu
                   8075: 
1.75      anton    8076: @node Vocabularies, Why use word lists?, Word Lists, Word Lists
                   8077: @subsection Vocabularies
                   8078: @cindex Vocabularies, detailed explanation
                   8079: 
                   8080: Here is an example of creating and using a new wordlist using ANS
                   8081: Forth words:
                   8082: 
                   8083: @example
                   8084: wordlist constant my-new-words-wordlist
                   8085: : my-new-words get-order nip my-new-words-wordlist swap set-order ;
                   8086: 
                   8087: \ add it to the search order
                   8088: also my-new-words
                   8089: 
                   8090: \ alternatively, add it to the search order and make it
                   8091: \ the compilation word list
                   8092: also my-new-words definitions
                   8093: \ type "order" to see the problem
                   8094: @end example
                   8095: 
                   8096: The problem with this example is that @code{order} has no way to
                   8097: associate the name @code{my-new-words} with the wid of the word list (in
                   8098: Gforth, @code{order} and @code{vocs} will display @code{???}  for a wid
                   8099: that has no associated name). There is no Standard way of associating a
                   8100: name with a wid.
                   8101: 
                   8102: In Gforth, this example can be re-coded using @code{vocabulary}, which
                   8103: associates a name with a wid:
                   8104: 
                   8105: @example
                   8106: vocabulary my-new-words
                   8107: 
                   8108: \ add it to the search order
                   8109: also my-new-words
                   8110: 
                   8111: \ alternatively, add it to the search order and make it
                   8112: \ the compilation word list
                   8113: my-new-words definitions
                   8114: \ type "order" to see that the problem is solved
                   8115: @end example
                   8116: 
                   8117: 
                   8118: @node Why use word lists?, Word list example, Vocabularies, Word Lists
1.26      crook    8119: @subsection Why use word lists?
                   8120: @cindex word lists - why use them?
                   8121: 
1.74      anton    8122: Here are some reasons why people use wordlists:
1.26      crook    8123: 
                   8124: @itemize @bullet
1.74      anton    8125: 
                   8126: @c anton: Gforth's hashing implementation makes the search speed
                   8127: @c independent from the number of words.  But it is linear with the number
                   8128: @c of wordlists that have to be searched, so in effect using more wordlists
                   8129: @c actually slows down compilation.
                   8130: 
                   8131: @c @item
                   8132: @c To improve compilation speed by reducing the number of header space
                   8133: @c entries that must be searched. This is achieved by creating a new
                   8134: @c word list that contains all of the definitions that are used in the
                   8135: @c definition of a Forth system but which would not usually be used by
                   8136: @c programs running on that system. That word list would be on the search
                   8137: @c list when the Forth system was compiled but would be removed from the
                   8138: @c search list for normal operation. This can be a useful technique for
                   8139: @c low-performance systems (for example, 8-bit processors in embedded
                   8140: @c systems) but is unlikely to be necessary in high-performance desktop
                   8141: @c systems.
                   8142: 
1.26      crook    8143: @item
                   8144: To prevent a set of words from being used outside the context in which
                   8145: they are valid. Two classic examples of this are an integrated editor
                   8146: (all of the edit commands are defined in a separate word list; the
                   8147: search order is set to the editor word list when the editor is invoked;
                   8148: the old search order is restored when the editor is terminated) and an
                   8149: integrated assembler (the op-codes for the machine are defined in a
                   8150: separate word list which is used when a @code{CODE} word is defined).
1.74      anton    8151: 
                   8152: @item
                   8153: To organize the words of an application or library into a user-visible
                   8154: set (in @code{forth-wordlist} or some other common wordlist) and a set
                   8155: of helper words used just for the implementation (hidden in a separate
1.75      anton    8156: wordlist).  This keeps @code{words}' output smaller, separates
                   8157: implementation and interface, and reduces the chance of name conflicts
                   8158: within the common wordlist.
1.74      anton    8159: 
1.26      crook    8160: @item
                   8161: To prevent a name-space clash between multiple definitions with the same
                   8162: name. For example, when building a cross-compiler you might have a word
                   8163: @code{IF} that generates conditional code for your target system. By
                   8164: placing this definition in a different word list you can control whether
                   8165: the host system's @code{IF} or the target system's @code{IF} get used in
                   8166: any particular context by controlling the order of the word lists on the
                   8167: search order stack.
1.74      anton    8168: 
1.26      crook    8169: @end itemize
1.1       anton    8170: 
1.74      anton    8171: The downsides of using wordlists are:
                   8172: 
                   8173: @itemize
                   8174: 
                   8175: @item
                   8176: Debugging becomes more cumbersome.
                   8177: 
                   8178: @item
                   8179: Name conflicts worked around with wordlists are still there, and you
                   8180: have to arrange the search order carefully to get the desired results;
                   8181: if you forget to do that, you get hard-to-find errors (as in any case
                   8182: where you read the code differently from the compiler; @code{see} can
1.75      anton    8183: help seeing which of several possible words the name resolves to in such
                   8184: cases).  @code{See} displays just the name of the words, not what
                   8185: wordlist they belong to, so it might be misleading.  Using unique names
                   8186: is a better approach to avoid name conflicts.
1.74      anton    8187: 
                   8188: @item
                   8189: You have to explicitly undo any changes to the search order.  In many
                   8190: cases it would be more convenient if this happened implicitly.  Gforth
                   8191: currently does not provide such a feature, but it may do so in the
                   8192: future.
                   8193: @end itemize
                   8194: 
                   8195: 
1.75      anton    8196: @node Word list example,  , Why use word lists?, Word Lists
                   8197: @subsection Word list example
                   8198: @cindex word lists - example
1.1       anton    8199: 
1.74      anton    8200: The following example is from the
                   8201: @uref{http://www.complang.tuwien.ac.at/forth/garbage-collection.zip,
                   8202: garbage collector} and uses wordlists to separate public words from
                   8203: helper words:
                   8204: 
                   8205: @example
                   8206: get-current ( wid )
                   8207: vocabulary garbage-collector also garbage-collector definitions
                   8208: ... \ define helper words
                   8209: ( wid ) set-current \ restore original (i.e., public) compilation wordlist
                   8210: ... \ define the public (i.e., API) words
                   8211:     \ they can refer to the helper words
                   8212: previous \ restore original search order (helper words become invisible)
                   8213: @end example
                   8214: 
1.26      crook    8215: @c -------------------------------------------------------------
                   8216: @node Environmental Queries, Files, Word Lists, Words
                   8217: @section Environmental Queries
                   8218: @cindex environmental queries
1.21      crook    8219: 
1.26      crook    8220: ANS Forth introduced the idea of ``environmental queries'' as a way
                   8221: for a program running on a system to determine certain characteristics of the system.
                   8222: The Standard specifies a number of strings that might be recognised by a system.
1.21      crook    8223: 
1.32      anton    8224: The Standard requires that the header space used for environmental queries
                   8225: be distinct from the header space used for definitions.
1.21      crook    8226: 
1.26      crook    8227: Typically, environmental queries are supported by creating a set of
1.29      crook    8228: definitions in a word list that is @i{only} used during environmental
1.26      crook    8229: queries; that is what Gforth does. There is no Standard way of adding
                   8230: definitions to the set of recognised environmental queries, but any
                   8231: implementation that supports the loading of optional word sets must have
                   8232: some mechanism for doing this (after loading the word set, the
                   8233: associated environmental query string must return @code{true}). In
                   8234: Gforth, the word list used to honour environmental queries can be
                   8235: manipulated just like any other word list.
1.21      crook    8236: 
1.44      crook    8237: 
1.26      crook    8238: doc-environment?
                   8239: doc-environment-wordlist
1.21      crook    8240: 
1.26      crook    8241: doc-gforth
                   8242: doc-os-class
1.21      crook    8243: 
1.44      crook    8244: 
1.26      crook    8245: Note that, whilst the documentation for (e.g.) @code{gforth} shows it
                   8246: returning two items on the stack, querying it using @code{environment?}
                   8247: will return an additional item; the @code{true} flag that shows that the
                   8248: string was recognised.
1.21      crook    8249: 
1.26      crook    8250: @comment TODO Document the standard strings or note where they are documented herein
1.21      crook    8251: 
1.26      crook    8252: Here are some examples of using environmental queries:
1.21      crook    8253: 
1.26      crook    8254: @example
                   8255: s" address-unit-bits" environment? 0=
                   8256: [IF]
                   8257:      cr .( environmental attribute address-units-bits unknown... ) cr
1.75      anton    8258: [ELSE]
                   8259:      drop \ ensure balanced stack effect
1.26      crook    8260: [THEN]
1.21      crook    8261: 
1.75      anton    8262: \ this might occur in the prelude of a standard program that uses THROW
                   8263: s" exception" environment? [IF]
                   8264:    0= [IF]
                   8265:       : throw abort" exception thrown" ;
                   8266:    [THEN]
                   8267: [ELSE] \ we don't know, so make sure
                   8268:    : throw abort" exception thrown" ;
                   8269: [THEN]
1.21      crook    8270: 
1.26      crook    8271: s" gforth" environment? [IF] .( Gforth version ) TYPE
                   8272:                         [ELSE] .( Not Gforth..) [THEN]
1.75      anton    8273: 
                   8274: \ a program using v*
                   8275: s" gforth" environment? [IF]
                   8276:   s" 0.5.0" compare 0< [IF] \ v* is a primitive since 0.5.0
                   8277:    : v* ( f_addr1 nstride1 f_addr2 nstride2 ucount -- r )
                   8278:      >r swap 2swap swap 0e r> 0 ?DO
                   8279:        dup f@ over + 2swap dup f@ f* f+ over + 2swap
                   8280:      LOOP
                   8281:      2drop 2drop ; 
                   8282:   [THEN]
                   8283: [ELSE] \ 
                   8284:   : v* ( f_addr1 nstride1 f_addr2 nstride2 ucount -- r )
                   8285:   ...
                   8286: [THEN]
1.26      crook    8287: @end example
1.21      crook    8288: 
1.26      crook    8289: Here is an example of adding a definition to the environment word list:
1.21      crook    8290: 
1.26      crook    8291: @example
                   8292: get-current environment-wordlist set-current
                   8293: true constant block
                   8294: true constant block-ext
                   8295: set-current
                   8296: @end example
1.21      crook    8297: 
1.26      crook    8298: You can see what definitions are in the environment word list like this:
1.21      crook    8299: 
1.26      crook    8300: @example
1.79      anton    8301: environment-wordlist >order words previous
1.26      crook    8302: @end example
1.21      crook    8303: 
                   8304: 
1.26      crook    8305: @c -------------------------------------------------------------
                   8306: @node Files, Blocks, Environmental Queries, Words
                   8307: @section Files
1.28      crook    8308: @cindex files
                   8309: @cindex I/O - file-handling
1.21      crook    8310: 
1.26      crook    8311: Gforth provides facilities for accessing files that are stored in the
                   8312: host operating system's file-system. Files that are processed by Gforth
                   8313: can be divided into two categories:
1.21      crook    8314: 
1.23      crook    8315: @itemize @bullet
                   8316: @item
1.29      crook    8317: Files that are processed by the Text Interpreter (@dfn{Forth source files}).
1.23      crook    8318: @item
1.29      crook    8319: Files that are processed by some other program (@dfn{general files}).
1.26      crook    8320: @end itemize
                   8321: 
                   8322: @menu
1.48      anton    8323: * Forth source files::          
                   8324: * General files::               
                   8325: * Search Paths::                
1.26      crook    8326: @end menu
                   8327: 
                   8328: @c -------------------------------------------------------------
                   8329: @node Forth source files, General files, Files, Files
                   8330: @subsection Forth source files
                   8331: @cindex including files
                   8332: @cindex Forth source files
1.21      crook    8333: 
1.26      crook    8334: The simplest way to interpret the contents of a file is to use one of
                   8335: these two formats:
1.21      crook    8336: 
1.26      crook    8337: @example
                   8338: include mysource.fs
                   8339: s" mysource.fs" included
                   8340: @end example
1.21      crook    8341: 
1.75      anton    8342: You usually want to include a file only if it is not included already
1.26      crook    8343: (by, say, another source file). In that case, you can use one of these
1.45      crook    8344: three formats:
1.21      crook    8345: 
1.26      crook    8346: @example
                   8347: require mysource.fs
                   8348: needs mysource.fs
                   8349: s" mysource.fs" required
                   8350: @end example
1.21      crook    8351: 
1.26      crook    8352: @cindex stack effect of included files
                   8353: @cindex including files, stack effect
1.45      crook    8354: It is good practice to write your source files such that interpreting them
                   8355: does not change the stack. Source files designed in this way can be used with
1.26      crook    8356: @code{required} and friends without complications. For example:
1.21      crook    8357: 
1.26      crook    8358: @example
1.75      anton    8359: 1024 require foo.fs drop
1.26      crook    8360: @end example
1.21      crook    8361: 
1.75      anton    8362: Here you want to pass the argument 1024 (e.g., a buffer size) to
                   8363: @file{foo.fs}.  Interpreting @file{foo.fs} has the stack effect ( n -- n
                   8364: ), which allows its use with @code{require}.  Of course with such
                   8365: parameters to required files, you have to ensure that the first
                   8366: @code{require} fits for all uses (i.e., @code{require} it early in the
                   8367: master load file).
1.44      crook    8368: 
1.26      crook    8369: doc-include-file
                   8370: doc-included
1.28      crook    8371: doc-included?
1.26      crook    8372: doc-include
                   8373: doc-required
                   8374: doc-require
                   8375: doc-needs
1.75      anton    8376: @c doc-init-included-files @c internal
                   8377: doc-sourcefilename
                   8378: doc-sourceline#
1.44      crook    8379: 
1.26      crook    8380: A definition in ANS Forth for @code{required} is provided in
                   8381: @file{compat/required.fs}.
1.21      crook    8382: 
1.26      crook    8383: @c -------------------------------------------------------------
                   8384: @node General files, Search Paths, Forth source files, Files
                   8385: @subsection General files
                   8386: @cindex general files
                   8387: @cindex file-handling
1.21      crook    8388: 
1.75      anton    8389: Files are opened/created by name and type. The following file access
                   8390: methods (FAMs) are recognised:
1.44      crook    8391: 
1.75      anton    8392: @cindex fam (file access method)
1.26      crook    8393: doc-r/o
                   8394: doc-r/w
                   8395: doc-w/o
                   8396: doc-bin
1.1       anton    8397: 
1.44      crook    8398: 
1.26      crook    8399: When a file is opened/created, it returns a file identifier,
1.29      crook    8400: @i{wfileid} that is used for all other file commands. All file
                   8401: commands also return a status value, @i{wior}, that is 0 for a
1.26      crook    8402: successful operation and an implementation-defined non-zero value in the
                   8403: case of an error.
1.21      crook    8404: 
1.44      crook    8405: 
1.26      crook    8406: doc-open-file
                   8407: doc-create-file
1.21      crook    8408: 
1.26      crook    8409: doc-close-file
                   8410: doc-delete-file
                   8411: doc-rename-file
                   8412: doc-read-file
                   8413: doc-read-line
                   8414: doc-write-file
                   8415: doc-write-line
                   8416: doc-emit-file
                   8417: doc-flush-file
1.21      crook    8418: 
1.26      crook    8419: doc-file-status
                   8420: doc-file-position
                   8421: doc-reposition-file
                   8422: doc-file-size
                   8423: doc-resize-file
1.21      crook    8424: 
1.44      crook    8425: 
1.26      crook    8426: @c ---------------------------------------------------------
1.48      anton    8427: @node Search Paths,  , General files, Files
1.26      crook    8428: @subsection Search Paths
                   8429: @cindex path for @code{included}
                   8430: @cindex file search path
                   8431: @cindex @code{include} search path
                   8432: @cindex search path for files
1.21      crook    8433: 
1.26      crook    8434: If you specify an absolute filename (i.e., a filename starting with
                   8435: @file{/} or @file{~}, or with @file{:} in the second position (as in
                   8436: @samp{C:...})) for @code{included} and friends, that file is included
                   8437: just as you would expect.
1.21      crook    8438: 
1.75      anton    8439: If the filename starts with @file{./}, this refers to the directory that
                   8440: the present file was @code{included} from.  This allows files to include
                   8441: other files relative to their own position (irrespective of the current
                   8442: working directory or the absolute position).  This feature is essential
                   8443: for libraries consisting of several files, where a file may include
                   8444: other files from the library.  It corresponds to @code{#include "..."}
                   8445: in C. If the current input source is not a file, @file{.} refers to the
                   8446: directory of the innermost file being included, or, if there is no file
                   8447: being included, to the current working directory.
                   8448: 
                   8449: For relative filenames (not starting with @file{./}), Gforth uses a
                   8450: search path similar to Forth's search order (@pxref{Word Lists}). It
                   8451: tries to find the given filename in the directories present in the path,
                   8452: and includes the first one it finds. There are separate search paths for
                   8453: Forth source files and general files.  If the search path contains the
                   8454: directory @file{.}, this refers to the directory of the current file, or
                   8455: the working directory, as if the file had been specified with @file{./}.
1.21      crook    8456: 
1.26      crook    8457: Use @file{~+} to refer to the current working directory (as in the
                   8458: @code{bash}).
1.1       anton    8459: 
1.75      anton    8460: @c anton: fold the following subsubsections into this subsection?
1.1       anton    8461: 
1.48      anton    8462: @menu
1.75      anton    8463: * Source Search Paths::         
1.48      anton    8464: * General Search Paths::        
                   8465: @end menu
                   8466: 
1.26      crook    8467: @c ---------------------------------------------------------
1.75      anton    8468: @node Source Search Paths, General Search Paths, Search Paths, Search Paths
                   8469: @subsubsection Source Search Paths
                   8470: @cindex search path control, source files
1.5       anton    8471: 
1.26      crook    8472: The search path is initialized when you start Gforth (@pxref{Invoking
1.75      anton    8473: Gforth}). You can display it and change it using @code{fpath} in
                   8474: combination with the general path handling words.
1.5       anton    8475: 
1.75      anton    8476: doc-fpath
                   8477: @c the functionality of the following words is easily available through
                   8478: @c   fpath and the general path words.  The may go away.
                   8479: @c doc-.fpath
                   8480: @c doc-fpath+
                   8481: @c doc-fpath=
                   8482: @c doc-open-fpath-file
1.44      crook    8483: 
                   8484: @noindent
1.26      crook    8485: Here is an example of using @code{fpath} and @code{require}:
1.5       anton    8486: 
1.26      crook    8487: @example
1.75      anton    8488: fpath path= /usr/lib/forth/|./
1.26      crook    8489: require timer.fs
                   8490: @end example
1.5       anton    8491: 
1.75      anton    8492: 
1.26      crook    8493: @c ---------------------------------------------------------
1.75      anton    8494: @node General Search Paths,  , Source Search Paths, Search Paths
1.26      crook    8495: @subsubsection General Search Paths
1.75      anton    8496: @cindex search path control, source files
1.5       anton    8497: 
1.26      crook    8498: Your application may need to search files in several directories, like
                   8499: @code{included} does. To facilitate this, Gforth allows you to define
                   8500: and use your own search paths, by providing generic equivalents of the
                   8501: Forth search path words:
1.5       anton    8502: 
1.75      anton    8503: doc-open-path-file
                   8504: doc-path-allot
                   8505: doc-clear-path
                   8506: doc-also-path
1.26      crook    8507: doc-.path
                   8508: doc-path+
                   8509: doc-path=
1.5       anton    8510: 
1.75      anton    8511: @c anton: better define a word for it, say "path-allot ( ucount -- path-addr )
1.44      crook    8512: 
1.75      anton    8513: Here's an example of creating an empty search path:
                   8514: @c
1.26      crook    8515: @example
1.75      anton    8516: create mypath 500 path-allot \ maximum length 500 chars (is checked)
1.26      crook    8517: @end example
1.5       anton    8518: 
1.26      crook    8519: @c -------------------------------------------------------------
                   8520: @node Blocks, Other I/O, Files, Words
                   8521: @section Blocks
1.28      crook    8522: @cindex I/O - blocks
                   8523: @cindex blocks
                   8524: 
                   8525: When you run Gforth on a modern desk-top computer, it runs under the
                   8526: control of an operating system which provides certain services.  One of
                   8527: these services is @var{file services}, which allows Forth source code
                   8528: and data to be stored in files and read into Gforth (@pxref{Files}).
                   8529: 
                   8530: Traditionally, Forth has been an important programming language on
                   8531: systems where it has interfaced directly to the underlying hardware with
                   8532: no intervening operating system. Forth provides a mechanism, called
1.29      crook    8533: @dfn{blocks}, for accessing mass storage on such systems.
1.28      crook    8534: 
                   8535: A block is a 1024-byte data area, which can be used to hold data or
                   8536: Forth source code. No structure is imposed on the contents of the
                   8537: block. A block is identified by its number; blocks are numbered
                   8538: contiguously from 1 to an implementation-defined maximum.
                   8539: 
                   8540: A typical system that used blocks but no operating system might use a
                   8541: single floppy-disk drive for mass storage, with the disks formatted to
                   8542: provide 256-byte sectors. Blocks would be implemented by assigning the
                   8543: first four sectors of the disk to block 1, the second four sectors to
                   8544: block 2 and so on, up to the limit of the capacity of the disk. The disk
                   8545: would not contain any file system information, just the set of blocks.
                   8546: 
1.29      crook    8547: @cindex blocks file
1.28      crook    8548: On systems that do provide file services, blocks are typically
1.29      crook    8549: implemented by storing a sequence of blocks within a single @dfn{blocks
1.28      crook    8550: file}.  The size of the blocks file will be an exact multiple of 1024
                   8551: bytes, corresponding to the number of blocks it contains. This is the
                   8552: mechanism that Gforth uses.
                   8553: 
1.29      crook    8554: @cindex @file{blocks.fb}
1.75      anton    8555: Only one blocks file can be open at a time. If you use block words without
1.28      crook    8556: having specified a blocks file, Gforth defaults to the blocks file
                   8557: @file{blocks.fb}. Gforth uses the Forth search path when attempting to
1.75      anton    8558: locate a blocks file (@pxref{Source Search Paths}).
1.28      crook    8559: 
1.29      crook    8560: @cindex block buffers
1.28      crook    8561: When you read and write blocks under program control, Gforth uses a
1.29      crook    8562: number of @dfn{block buffers} as intermediate storage. These buffers are
1.28      crook    8563: not used when you use @code{load} to interpret the contents of a block.
                   8564: 
1.75      anton    8565: The behaviour of the block buffers is analagous to that of a cache.
                   8566: Each block buffer has three states:
1.28      crook    8567: 
                   8568: @itemize @bullet
                   8569: @item
                   8570: Unassigned
                   8571: @item
                   8572: Assigned-clean
                   8573: @item
                   8574: Assigned-dirty
                   8575: @end itemize
                   8576: 
1.29      crook    8577: Initially, all block buffers are @i{unassigned}. In order to access a
1.28      crook    8578: block, the block (specified by its block number) must be assigned to a
                   8579: block buffer.
                   8580: 
                   8581: The assignment of a block to a block buffer is performed by @code{block}
                   8582: or @code{buffer}. Use @code{block} when you wish to modify the existing
                   8583: contents of a block. Use @code{buffer} when you don't care about the
                   8584: existing contents of the block@footnote{The ANS Forth definition of
1.35      anton    8585: @code{buffer} is intended not to cause disk I/O; if the data associated
1.28      crook    8586: with the particular block is already stored in a block buffer due to an
                   8587: earlier @code{block} command, @code{buffer} will return that block
                   8588: buffer and the existing contents of the block will be
                   8589: available. Otherwise, @code{buffer} will simply assign a new, empty
1.29      crook    8590: block buffer for the block.}.
1.28      crook    8591: 
1.47      crook    8592: Once a block has been assigned to a block buffer using @code{block} or
1.75      anton    8593: @code{buffer}, that block buffer becomes the @i{current block
                   8594: buffer}. Data may only be manipulated (read or written) within the
                   8595: current block buffer.
1.47      crook    8596: 
                   8597: When the contents of the current block buffer has been modified it is
1.48      anton    8598: necessary, @emph{before calling @code{block} or @code{buffer} again}, to
1.75      anton    8599: either abandon the changes (by doing nothing) or mark the block as
                   8600: changed (assigned-dirty), using @code{update}. Using @code{update} does
                   8601: not change the blocks file; it simply changes a block buffer's state to
                   8602: @i{assigned-dirty}.  The block will be written implicitly when it's
                   8603: buffer is needed for another block, or explicitly by @code{flush} or
                   8604: @code{save-buffers}.
                   8605: 
                   8606: word @code{Flush} writes all @i{assigned-dirty} blocks back to the
                   8607: blocks file on disk. Leaving Gforth with @code{bye} also performs a
                   8608: @code{flush}.
1.28      crook    8609: 
1.29      crook    8610: In Gforth, @code{block} and @code{buffer} use a @i{direct-mapped}
1.28      crook    8611: algorithm to assign a block buffer to a block. That means that any
                   8612: particular block can only be assigned to one specific block buffer,
1.29      crook    8613: called (for the particular operation) the @i{victim buffer}. If the
1.47      crook    8614: victim buffer is @i{unassigned} or @i{assigned-clean} it is allocated to
                   8615: the new block immediately. If it is @i{assigned-dirty} its current
                   8616: contents are written back to the blocks file on disk before it is
1.28      crook    8617: allocated to the new block.
                   8618: 
                   8619: Although no structure is imposed on the contents of a block, it is
                   8620: traditional to display the contents as 16 lines each of 64 characters.  A
                   8621: block provides a single, continuous stream of input (for example, it
                   8622: acts as a single parse area) -- there are no end-of-line characters
                   8623: within a block, and no end-of-file character at the end of a
                   8624: block. There are two consequences of this:
1.26      crook    8625: 
1.28      crook    8626: @itemize @bullet
                   8627: @item
                   8628: The last character of one line wraps straight into the first character
                   8629: of the following line
                   8630: @item
                   8631: The word @code{\} -- comment to end of line -- requires special
                   8632: treatment; in the context of a block it causes all characters until the
                   8633: end of the current 64-character ``line'' to be ignored.
                   8634: @end itemize
                   8635: 
                   8636: In Gforth, when you use @code{block} with a non-existent block number,
1.45      crook    8637: the current blocks file will be extended to the appropriate size and the
1.28      crook    8638: block buffer will be initialised with spaces.
                   8639: 
1.47      crook    8640: Gforth includes a simple block editor (type @code{use blocked.fb 0 list}
                   8641: for details) but doesn't encourage the use of blocks; the mechanism is
                   8642: only provided for backward compatibility -- ANS Forth requires blocks to
                   8643: be available when files are.
1.28      crook    8644: 
                   8645: Common techniques that are used when working with blocks include:
                   8646: 
                   8647: @itemize @bullet
                   8648: @item
                   8649: A screen editor that allows you to edit blocks without leaving the Forth
                   8650: environment.
                   8651: @item
                   8652: Shadow screens; where every code block has an associated block
                   8653: containing comments (for example: code in odd block numbers, comments in
                   8654: even block numbers). Typically, the block editor provides a convenient
                   8655: mechanism to toggle between code and comments.
                   8656: @item
                   8657: Load blocks; a single block (typically block 1) contains a number of
                   8658: @code{thru} commands which @code{load} the whole of the application.
                   8659: @end itemize
1.26      crook    8660: 
1.29      crook    8661: See Frank Sergeant's Pygmy Forth to see just how well blocks can be
                   8662: integrated into a Forth programming environment.
1.26      crook    8663: 
                   8664: @comment TODO what about errors on open-blocks?
1.44      crook    8665: 
1.26      crook    8666: doc-open-blocks
                   8667: doc-use
1.75      anton    8668: doc-block-offset
1.26      crook    8669: doc-get-block-fid
                   8670: doc-block-position
1.28      crook    8671: 
1.75      anton    8672: doc-list
1.28      crook    8673: doc-scr
                   8674: 
1.45      crook    8675: doc---gforthman-block
1.28      crook    8676: doc-buffer
                   8677: 
1.75      anton    8678: doc-empty-buffers
                   8679: doc-empty-buffer
1.26      crook    8680: doc-update
1.28      crook    8681: doc-updated?
1.26      crook    8682: doc-save-buffers
1.75      anton    8683: doc-save-buffer
1.26      crook    8684: doc-flush
1.28      crook    8685: 
1.26      crook    8686: doc-load
                   8687: doc-thru
                   8688: doc-+load
                   8689: doc-+thru
1.45      crook    8690: doc---gforthman--->
1.26      crook    8691: doc-block-included
                   8692: 
1.44      crook    8693: 
1.26      crook    8694: @c -------------------------------------------------------------
1.78      anton    8695: @node Other I/O, Locals, Blocks, Words
1.26      crook    8696: @section Other I/O
1.28      crook    8697: @cindex I/O - keyboard and display
1.26      crook    8698: 
                   8699: @menu
                   8700: * Simple numeric output::       Predefined formats
                   8701: * Formatted numeric output::    Formatted (pictured) output
                   8702: * String Formats::              How Forth stores strings in memory
1.67      anton    8703: * Displaying characters and strings::  Other stuff
1.26      crook    8704: * Input::                       Input
                   8705: @end menu
                   8706: 
                   8707: @node Simple numeric output, Formatted numeric output, Other I/O, Other I/O
                   8708: @subsection Simple numeric output
1.28      crook    8709: @cindex numeric output - simple/free-format
1.5       anton    8710: 
1.26      crook    8711: The simplest output functions are those that display numbers from the
                   8712: data or floating-point stacks. Floating-point output is always displayed
                   8713: using base 10. Numbers displayed from the data stack use the value stored
                   8714: in @code{base}.
1.5       anton    8715: 
1.44      crook    8716: 
1.26      crook    8717: doc-.
                   8718: doc-dec.
                   8719: doc-hex.
                   8720: doc-u.
                   8721: doc-.r
                   8722: doc-u.r
                   8723: doc-d.
                   8724: doc-ud.
                   8725: doc-d.r
                   8726: doc-ud.r
                   8727: doc-f.
                   8728: doc-fe.
                   8729: doc-fs.
1.5       anton    8730: 
1.44      crook    8731: 
1.26      crook    8732: Examples of printing the number 1234.5678E23 in the different floating-point output
                   8733: formats are shown below:
1.5       anton    8734: 
                   8735: @example
1.26      crook    8736: f. 123456779999999000000000000.
                   8737: fe. 123.456779999999E24
                   8738: fs. 1.23456779999999E26
1.5       anton    8739: @end example
                   8740: 
                   8741: 
1.26      crook    8742: @node Formatted numeric output, String Formats, Simple numeric output, Other I/O
                   8743: @subsection Formatted numeric output
1.28      crook    8744: @cindex formatted numeric output
1.26      crook    8745: @cindex pictured numeric output
1.28      crook    8746: @cindex numeric output - formatted
1.26      crook    8747: 
1.29      crook    8748: Forth traditionally uses a technique called @dfn{pictured numeric
1.26      crook    8749: output} for formatted printing of integers.  In this technique, digits
                   8750: are extracted from the number (using the current output radix defined by
                   8751: @code{base}), converted to ASCII codes and appended to a string that is
                   8752: built in a scratch-pad area of memory (@pxref{core-idef,
                   8753: Implementation-defined options, Implementation-defined
                   8754: options}). Arbitrary characters can be appended to the string during the
                   8755: extraction process. The completed string is specified by an address
                   8756: and length and can be manipulated (@code{TYPE}ed, copied, modified)
                   8757: under program control.
1.5       anton    8758: 
1.75      anton    8759: All of the integer output words described in the previous section
                   8760: (@pxref{Simple numeric output}) are implemented in Gforth using pictured
                   8761: numeric output.
1.5       anton    8762: 
1.47      crook    8763: Three important things to remember about pictured numeric output:
1.5       anton    8764: 
1.26      crook    8765: @itemize @bullet
                   8766: @item
1.28      crook    8767: It always operates on double-precision numbers; to display a
1.49      anton    8768: single-precision number, convert it first (for ways of doing this
                   8769: @pxref{Double precision}).
1.26      crook    8770: @item
1.28      crook    8771: It always treats the double-precision number as though it were
                   8772: unsigned. The examples below show ways of printing signed numbers.
1.26      crook    8773: @item
                   8774: The string is built up from right to left; least significant digit first.
                   8775: @end itemize
1.5       anton    8776: 
1.44      crook    8777: 
1.26      crook    8778: doc-<#
1.47      crook    8779: doc-<<#
1.26      crook    8780: doc-#
                   8781: doc-#s
                   8782: doc-hold
                   8783: doc-sign
                   8784: doc-#>
1.47      crook    8785: doc-#>>
1.5       anton    8786: 
1.26      crook    8787: doc-represent
1.5       anton    8788: 
1.44      crook    8789: 
                   8790: @noindent
1.26      crook    8791: Here are some examples of using pictured numeric output:
1.5       anton    8792: 
                   8793: @example
1.26      crook    8794: : my-u. ( u -- )
                   8795:   \ Simplest use of pns.. behaves like Standard u. 
                   8796:   0              \ convert to unsigned double
1.75      anton    8797:   <<#            \ start conversion
1.26      crook    8798:   #s             \ convert all digits
                   8799:   #>             \ complete conversion
1.75      anton    8800:   TYPE SPACE     \ display, with trailing space
                   8801:   #>> ;          \ release hold area
1.5       anton    8802: 
1.26      crook    8803: : cents-only ( u -- )
                   8804:   0              \ convert to unsigned double
1.75      anton    8805:   <<#            \ start conversion
1.26      crook    8806:   # #            \ convert two least-significant digits
                   8807:   #>             \ complete conversion, discard other digits
1.75      anton    8808:   TYPE SPACE     \ display, with trailing space
                   8809:   #>> ;          \ release hold area
1.5       anton    8810: 
1.26      crook    8811: : dollars-and-cents ( u -- )
                   8812:   0              \ convert to unsigned double
1.75      anton    8813:   <<#            \ start conversion
1.26      crook    8814:   # #            \ convert two least-significant digits
                   8815:   [char] . hold  \ insert decimal point
                   8816:   #s             \ convert remaining digits
                   8817:   [char] $ hold  \ append currency symbol
                   8818:   #>             \ complete conversion
1.75      anton    8819:   TYPE SPACE     \ display, with trailing space
                   8820:   #>> ;          \ release hold area
1.5       anton    8821: 
1.26      crook    8822: : my-. ( n -- )
                   8823:   \ handling negatives.. behaves like Standard .
                   8824:   s>d            \ convert to signed double
                   8825:   swap over dabs \ leave sign byte followed by unsigned double
1.75      anton    8826:   <<#            \ start conversion
1.26      crook    8827:   #s             \ convert all digits
                   8828:   rot sign       \ get at sign byte, append "-" if needed
                   8829:   #>             \ complete conversion
1.75      anton    8830:   TYPE SPACE     \ display, with trailing space
                   8831:   #>> ;          \ release hold area
1.5       anton    8832: 
1.26      crook    8833: : account. ( n -- )
1.75      anton    8834:   \ accountants don't like minus signs, they use parentheses
1.26      crook    8835:   \ for negative numbers
                   8836:   s>d            \ convert to signed double
                   8837:   swap over dabs \ leave sign byte followed by unsigned double
1.75      anton    8838:   <<#            \ start conversion
1.26      crook    8839:   2 pick         \ get copy of sign byte
                   8840:   0< IF [char] ) hold THEN \ right-most character of output
                   8841:   #s             \ convert all digits
                   8842:   rot            \ get at sign byte
                   8843:   0< IF [char] ( hold THEN
                   8844:   #>             \ complete conversion
1.75      anton    8845:   TYPE SPACE     \ display, with trailing space
                   8846:   #>> ;          \ release hold area
                   8847: 
1.5       anton    8848: @end example
                   8849: 
1.26      crook    8850: Here are some examples of using these words:
1.5       anton    8851: 
                   8852: @example
1.26      crook    8853: 1 my-u. 1
                   8854: hex -1 my-u. decimal FFFFFFFF
                   8855: 1 cents-only 01
                   8856: 1234 cents-only 34
                   8857: 2 dollars-and-cents $0.02
                   8858: 1234 dollars-and-cents $12.34
                   8859: 123 my-. 123
                   8860: -123 my. -123
                   8861: 123 account. 123
                   8862: -456 account. (456)
1.5       anton    8863: @end example
                   8864: 
                   8865: 
1.26      crook    8866: @node String Formats, Displaying characters and strings, Formatted numeric output, Other I/O
                   8867: @subsection String Formats
1.27      crook    8868: @cindex strings - see character strings
                   8869: @cindex character strings - formats
1.28      crook    8870: @cindex I/O - see character strings
1.75      anton    8871: @cindex counted strings
                   8872: 
                   8873: @c anton: this does not really belong here; maybe the memory section,
                   8874: @c  or the principles chapter
1.26      crook    8875: 
1.27      crook    8876: Forth commonly uses two different methods for representing character
                   8877: strings:
1.26      crook    8878: 
                   8879: @itemize @bullet
                   8880: @item
                   8881: @cindex address of counted string
1.45      crook    8882: @cindex counted string
1.29      crook    8883: As a @dfn{counted string}, represented by a @i{c-addr}. The char
                   8884: addressed by @i{c-addr} contains a character-count, @i{n}, of the
                   8885: string and the string occupies the subsequent @i{n} char addresses in
1.26      crook    8886: memory.
                   8887: @item
1.29      crook    8888: As cell pair on the stack; @i{c-addr u}, where @i{u} is the length
                   8889: of the string in characters, and @i{c-addr} is the address of the
1.26      crook    8890: first byte of the string.
                   8891: @end itemize
                   8892: 
                   8893: ANS Forth encourages the use of the second format when representing
1.75      anton    8894: strings.
1.26      crook    8895: 
1.44      crook    8896: 
1.26      crook    8897: doc-count
                   8898: 
1.44      crook    8899: 
1.49      anton    8900: For words that move, copy and search for strings see @ref{Memory
                   8901: Blocks}. For words that display characters and strings see
                   8902: @ref{Displaying characters and strings}.
1.26      crook    8903: 
                   8904: @node Displaying characters and strings, Input, String Formats, Other I/O
                   8905: @subsection Displaying characters and strings
1.27      crook    8906: @cindex characters - compiling and displaying
                   8907: @cindex character strings - compiling and displaying
1.26      crook    8908: 
                   8909: This section starts with a glossary of Forth words and ends with a set
                   8910: of examples.
                   8911: 
1.44      crook    8912: 
1.26      crook    8913: doc-bl
                   8914: doc-space
                   8915: doc-spaces
                   8916: doc-emit
                   8917: doc-toupper
                   8918: doc-."
                   8919: doc-.(
                   8920: doc-type
1.44      crook    8921: doc-typewhite
1.26      crook    8922: doc-cr
1.27      crook    8923: @cindex cursor control
1.26      crook    8924: doc-at-xy
                   8925: doc-page
                   8926: doc-s"
                   8927: doc-c"
                   8928: doc-char
                   8929: doc-[char]
                   8930: 
1.44      crook    8931: 
                   8932: @noindent
1.26      crook    8933: As an example, consider the following text, stored in a file @file{test.fs}:
1.5       anton    8934: 
                   8935: @example
1.26      crook    8936: .( text-1)
                   8937: : my-word
                   8938:   ." text-2" cr
                   8939:   .( text-3)
                   8940: ;
                   8941: 
                   8942: ." text-4"
                   8943: 
                   8944: : my-char
                   8945:   [char] ALPHABET emit
                   8946:   char emit
                   8947: ;
1.5       anton    8948: @end example
                   8949: 
1.26      crook    8950: When you load this code into Gforth, the following output is generated:
1.5       anton    8951: 
1.26      crook    8952: @example
1.30      anton    8953: @kbd{include test.fs @key{RET}} text-1text-3text-4 ok
1.26      crook    8954: @end example
1.5       anton    8955: 
1.26      crook    8956: @itemize @bullet
                   8957: @item
                   8958: Messages @code{text-1} and @code{text-3} are displayed because @code{.(} 
                   8959: is an immediate word; it behaves in the same way whether it is used inside
                   8960: or outside a colon definition.
                   8961: @item
                   8962: Message @code{text-4} is displayed because of Gforth's added interpretation
                   8963: semantics for @code{."}.
                   8964: @item
1.29      crook    8965: Message @code{text-2} is @i{not} displayed, because the text interpreter
1.26      crook    8966: performs the compilation semantics for @code{."} within the definition of
                   8967: @code{my-word}.
                   8968: @end itemize
1.5       anton    8969: 
1.26      crook    8970: Here are some examples of executing @code{my-word} and @code{my-char}:
1.5       anton    8971: 
1.26      crook    8972: @example
1.30      anton    8973: @kbd{my-word @key{RET}} text-2
1.26      crook    8974:  ok
1.30      anton    8975: @kbd{my-char fred @key{RET}} Af ok
                   8976: @kbd{my-char jim @key{RET}} Aj ok
1.26      crook    8977: @end example
1.5       anton    8978: 
                   8979: @itemize @bullet
                   8980: @item
1.26      crook    8981: Message @code{text-2} is displayed because of the run-time behaviour of
                   8982: @code{."}.
                   8983: @item
                   8984: @code{[char]} compiles the ``A'' from ``ALPHABET'' and puts its display code
                   8985: on the stack at run-time. @code{emit} always displays the character
                   8986: when @code{my-char} is executed.
                   8987: @item
                   8988: @code{char} parses a string at run-time and the second @code{emit} displays
                   8989: the first character of the string.
1.5       anton    8990: @item
1.26      crook    8991: If you type @code{see my-char} you can see that @code{[char]} discarded
                   8992: the text ``LPHABET'' and only compiled the display code for ``A'' into the
                   8993: definition of @code{my-char}.
1.5       anton    8994: @end itemize
                   8995: 
                   8996: 
                   8997: 
1.48      anton    8998: @node Input,  , Displaying characters and strings, Other I/O
1.26      crook    8999: @subsection Input
                   9000: @cindex input
1.28      crook    9001: @cindex I/O - see input
                   9002: @cindex parsing a string
1.5       anton    9003: 
1.49      anton    9004: For ways of storing character strings in memory see @ref{String Formats}.
1.5       anton    9005: 
1.27      crook    9006: @comment TODO examples for >number >float accept key key? pad parse word refill
1.29      crook    9007: @comment then index them
1.27      crook    9008: 
1.44      crook    9009: 
1.27      crook    9010: doc-key
                   9011: doc-key?
1.45      crook    9012: doc-ekey
                   9013: doc-ekey?
                   9014: doc-ekey>char
1.26      crook    9015: doc->number
                   9016: doc->float
                   9017: doc-accept
1.27      crook    9018: doc-pad
1.75      anton    9019: @c anton: these belong in the input stream section
1.27      crook    9020: doc-parse
                   9021: doc-word
                   9022: doc-sword
1.75      anton    9023: doc-name
1.27      crook    9024: doc-refill
                   9025: @comment obsolescent words..
                   9026: doc-convert
1.26      crook    9027: doc-query
                   9028: doc-expect
1.27      crook    9029: doc-span
1.5       anton    9030: 
                   9031: 
1.78      anton    9032: @c -------------------------------------------------------------
                   9033: @node Locals, Structures, Other I/O, Words
                   9034: @section Locals
                   9035: @cindex locals
                   9036: 
                   9037: Local variables can make Forth programming more enjoyable and Forth
                   9038: programs easier to read. Unfortunately, the locals of ANS Forth are
                   9039: laden with restrictions. Therefore, we provide not only the ANS Forth
                   9040: locals wordset, but also our own, more powerful locals wordset (we
                   9041: implemented the ANS Forth locals wordset through our locals wordset).
1.44      crook    9042: 
1.78      anton    9043: The ideas in this section have also been published in M. Anton Ertl,
                   9044: @cite{@uref{http://www.complang.tuwien.ac.at/papers/ertl94l.ps.gz,
                   9045: Automatic Scoping of Local Variables}}, EuroForth '94.
1.12      anton    9046: 
                   9047: @menu
1.78      anton    9048: * Gforth locals::               
                   9049: * ANS Forth locals::            
1.5       anton    9050: @end menu
                   9051: 
1.78      anton    9052: @node Gforth locals, ANS Forth locals, Locals, Locals
                   9053: @subsection Gforth locals
                   9054: @cindex Gforth locals
                   9055: @cindex locals, Gforth style
1.5       anton    9056: 
1.78      anton    9057: Locals can be defined with
1.44      crook    9058: 
1.78      anton    9059: @example
                   9060: @{ local1 local2 ... -- comment @}
                   9061: @end example
                   9062: or
                   9063: @example
                   9064: @{ local1 local2 ... @}
                   9065: @end example
1.5       anton    9066: 
1.78      anton    9067: E.g.,
                   9068: @example
                   9069: : max @{ n1 n2 -- n3 @}
                   9070:  n1 n2 > if
                   9071:    n1
                   9072:  else
                   9073:    n2
                   9074:  endif ;
                   9075: @end example
1.44      crook    9076: 
1.78      anton    9077: The similarity of locals definitions with stack comments is intended. A
                   9078: locals definition often replaces the stack comment of a word. The order
                   9079: of the locals corresponds to the order in a stack comment and everything
                   9080: after the @code{--} is really a comment.
1.77      anton    9081: 
1.78      anton    9082: This similarity has one disadvantage: It is too easy to confuse locals
                   9083: declarations with stack comments, causing bugs and making them hard to
                   9084: find. However, this problem can be avoided by appropriate coding
                   9085: conventions: Do not use both notations in the same program. If you do,
                   9086: they should be distinguished using additional means, e.g. by position.
1.77      anton    9087: 
1.78      anton    9088: @cindex types of locals
                   9089: @cindex locals types
                   9090: The name of the local may be preceded by a type specifier, e.g.,
                   9091: @code{F:} for a floating point value:
1.5       anton    9092: 
1.78      anton    9093: @example
                   9094: : CX* @{ F: Ar F: Ai F: Br F: Bi -- Cr Ci @}
                   9095: \ complex multiplication
                   9096:  Ar Br f* Ai Bi f* f-
                   9097:  Ar Bi f* Ai Br f* f+ ;
                   9098: @end example
1.44      crook    9099: 
1.78      anton    9100: @cindex flavours of locals
                   9101: @cindex locals flavours
                   9102: @cindex value-flavoured locals
                   9103: @cindex variable-flavoured locals
                   9104: Gforth currently supports cells (@code{W:}, @code{W^}), doubles
                   9105: (@code{D:}, @code{D^}), floats (@code{F:}, @code{F^}) and characters
                   9106: (@code{C:}, @code{C^}) in two flavours: a value-flavoured local (defined
                   9107: with @code{W:}, @code{D:} etc.) produces its value and can be changed
                   9108: with @code{TO}. A variable-flavoured local (defined with @code{W^} etc.)
                   9109: produces its address (which becomes invalid when the variable's scope is
                   9110: left). E.g., the standard word @code{emit} can be defined in terms of
                   9111: @code{type} like this:
1.5       anton    9112: 
1.78      anton    9113: @example
                   9114: : emit @{ C^ char* -- @}
                   9115:     char* 1 type ;
                   9116: @end example
1.5       anton    9117: 
1.78      anton    9118: @cindex default type of locals
                   9119: @cindex locals, default type
                   9120: A local without type specifier is a @code{W:} local. Both flavours of
                   9121: locals are initialized with values from the data or FP stack.
1.44      crook    9122: 
1.78      anton    9123: Currently there is no way to define locals with user-defined data
                   9124: structures, but we are working on it.
1.5       anton    9125: 
1.78      anton    9126: Gforth allows defining locals everywhere in a colon definition. This
                   9127: poses the following questions:
1.5       anton    9128: 
1.78      anton    9129: @menu
                   9130: * Where are locals visible by name?::  
                   9131: * How long do locals live?::    
                   9132: * Locals programming style::    
                   9133: * Locals implementation::       
                   9134: @end menu
1.44      crook    9135: 
1.78      anton    9136: @node Where are locals visible by name?, How long do locals live?, Gforth locals, Gforth locals
                   9137: @subsubsection Where are locals visible by name?
                   9138: @cindex locals visibility
                   9139: @cindex visibility of locals
                   9140: @cindex scope of locals
1.5       anton    9141: 
1.78      anton    9142: Basically, the answer is that locals are visible where you would expect
                   9143: it in block-structured languages, and sometimes a little longer. If you
                   9144: want to restrict the scope of a local, enclose its definition in
                   9145: @code{SCOPE}...@code{ENDSCOPE}.
1.5       anton    9146: 
                   9147: 
1.78      anton    9148: doc-scope
                   9149: doc-endscope
1.5       anton    9150: 
                   9151: 
1.78      anton    9152: These words behave like control structure words, so you can use them
                   9153: with @code{CS-PICK} and @code{CS-ROLL} to restrict the scope in
                   9154: arbitrary ways.
1.77      anton    9155: 
1.78      anton    9156: If you want a more exact answer to the visibility question, here's the
                   9157: basic principle: A local is visible in all places that can only be
                   9158: reached through the definition of the local@footnote{In compiler
                   9159: construction terminology, all places dominated by the definition of the
                   9160: local.}. In other words, it is not visible in places that can be reached
                   9161: without going through the definition of the local. E.g., locals defined
                   9162: in @code{IF}...@code{ENDIF} are visible until the @code{ENDIF}, locals
                   9163: defined in @code{BEGIN}...@code{UNTIL} are visible after the
                   9164: @code{UNTIL} (until, e.g., a subsequent @code{ENDSCOPE}).
1.77      anton    9165: 
1.78      anton    9166: The reasoning behind this solution is: We want to have the locals
                   9167: visible as long as it is meaningful. The user can always make the
                   9168: visibility shorter by using explicit scoping. In a place that can
                   9169: only be reached through the definition of a local, the meaning of a
                   9170: local name is clear. In other places it is not: How is the local
                   9171: initialized at the control flow path that does not contain the
                   9172: definition? Which local is meant, if the same name is defined twice in
                   9173: two independent control flow paths?
1.77      anton    9174: 
1.78      anton    9175: This should be enough detail for nearly all users, so you can skip the
                   9176: rest of this section. If you really must know all the gory details and
                   9177: options, read on.
1.77      anton    9178: 
1.78      anton    9179: In order to implement this rule, the compiler has to know which places
                   9180: are unreachable. It knows this automatically after @code{AHEAD},
                   9181: @code{AGAIN}, @code{EXIT} and @code{LEAVE}; in other cases (e.g., after
                   9182: most @code{THROW}s), you can use the word @code{UNREACHABLE} to tell the
                   9183: compiler that the control flow never reaches that place. If
                   9184: @code{UNREACHABLE} is not used where it could, the only consequence is
                   9185: that the visibility of some locals is more limited than the rule above
                   9186: says. If @code{UNREACHABLE} is used where it should not (i.e., if you
                   9187: lie to the compiler), buggy code will be produced.
1.77      anton    9188: 
1.5       anton    9189: 
1.78      anton    9190: doc-unreachable
1.5       anton    9191: 
1.23      crook    9192: 
1.78      anton    9193: Another problem with this rule is that at @code{BEGIN}, the compiler
                   9194: does not know which locals will be visible on the incoming
                   9195: back-edge. All problems discussed in the following are due to this
                   9196: ignorance of the compiler (we discuss the problems using @code{BEGIN}
                   9197: loops as examples; the discussion also applies to @code{?DO} and other
                   9198: loops). Perhaps the most insidious example is:
1.26      crook    9199: @example
1.78      anton    9200: AHEAD
                   9201: BEGIN
                   9202:   x
                   9203: [ 1 CS-ROLL ] THEN
                   9204:   @{ x @}
                   9205:   ...
                   9206: UNTIL
1.26      crook    9207: @end example
1.23      crook    9208: 
1.78      anton    9209: This should be legal according to the visibility rule. The use of
                   9210: @code{x} can only be reached through the definition; but that appears
                   9211: textually below the use.
                   9212: 
                   9213: From this example it is clear that the visibility rules cannot be fully
                   9214: implemented without major headaches. Our implementation treats common
                   9215: cases as advertised and the exceptions are treated in a safe way: The
                   9216: compiler makes a reasonable guess about the locals visible after a
                   9217: @code{BEGIN}; if it is too pessimistic, the
                   9218: user will get a spurious error about the local not being defined; if the
                   9219: compiler is too optimistic, it will notice this later and issue a
                   9220: warning. In the case above the compiler would complain about @code{x}
                   9221: being undefined at its use. You can see from the obscure examples in
                   9222: this section that it takes quite unusual control structures to get the
                   9223: compiler into trouble, and even then it will often do fine.
1.23      crook    9224: 
1.78      anton    9225: If the @code{BEGIN} is reachable from above, the most optimistic guess
                   9226: is that all locals visible before the @code{BEGIN} will also be
                   9227: visible after the @code{BEGIN}. This guess is valid for all loops that
                   9228: are entered only through the @code{BEGIN}, in particular, for normal
                   9229: @code{BEGIN}...@code{WHILE}...@code{REPEAT} and
                   9230: @code{BEGIN}...@code{UNTIL} loops and it is implemented in our
                   9231: compiler. When the branch to the @code{BEGIN} is finally generated by
                   9232: @code{AGAIN} or @code{UNTIL}, the compiler checks the guess and
                   9233: warns the user if it was too optimistic:
1.26      crook    9234: @example
1.78      anton    9235: IF
                   9236:   @{ x @}
                   9237: BEGIN
                   9238:   \ x ? 
                   9239: [ 1 cs-roll ] THEN
                   9240:   ...
                   9241: UNTIL
1.26      crook    9242: @end example
1.23      crook    9243: 
1.78      anton    9244: Here, @code{x} lives only until the @code{BEGIN}, but the compiler
                   9245: optimistically assumes that it lives until the @code{THEN}. It notices
                   9246: this difference when it compiles the @code{UNTIL} and issues a
                   9247: warning. The user can avoid the warning, and make sure that @code{x}
                   9248: is not used in the wrong area by using explicit scoping:
                   9249: @example
                   9250: IF
                   9251:   SCOPE
                   9252:   @{ x @}
                   9253:   ENDSCOPE
                   9254: BEGIN
                   9255: [ 1 cs-roll ] THEN
                   9256:   ...
                   9257: UNTIL
                   9258: @end example
1.23      crook    9259: 
1.78      anton    9260: Since the guess is optimistic, there will be no spurious error messages
                   9261: about undefined locals.
1.44      crook    9262: 
1.78      anton    9263: If the @code{BEGIN} is not reachable from above (e.g., after
                   9264: @code{AHEAD} or @code{EXIT}), the compiler cannot even make an
                   9265: optimistic guess, as the locals visible after the @code{BEGIN} may be
                   9266: defined later. Therefore, the compiler assumes that no locals are
                   9267: visible after the @code{BEGIN}. However, the user can use
                   9268: @code{ASSUME-LIVE} to make the compiler assume that the same locals are
                   9269: visible at the BEGIN as at the point where the top control-flow stack
                   9270: item was created.
1.23      crook    9271: 
1.44      crook    9272: 
1.78      anton    9273: doc-assume-live
1.26      crook    9274: 
1.23      crook    9275: 
1.78      anton    9276: @noindent
                   9277: E.g.,
                   9278: @example
                   9279: @{ x @}
                   9280: AHEAD
                   9281: ASSUME-LIVE
                   9282: BEGIN
                   9283:   x
                   9284: [ 1 CS-ROLL ] THEN
                   9285:   ...
                   9286: UNTIL
                   9287: @end example
1.44      crook    9288: 
1.78      anton    9289: Other cases where the locals are defined before the @code{BEGIN} can be
                   9290: handled by inserting an appropriate @code{CS-ROLL} before the
                   9291: @code{ASSUME-LIVE} (and changing the control-flow stack manipulation
                   9292: behind the @code{ASSUME-LIVE}).
1.23      crook    9293: 
1.78      anton    9294: Cases where locals are defined after the @code{BEGIN} (but should be
                   9295: visible immediately after the @code{BEGIN}) can only be handled by
                   9296: rearranging the loop. E.g., the ``most insidious'' example above can be
                   9297: arranged into:
                   9298: @example
                   9299: BEGIN
                   9300:   @{ x @}
                   9301:   ... 0=
                   9302: WHILE
                   9303:   x
                   9304: REPEAT
                   9305: @end example
1.44      crook    9306: 
1.78      anton    9307: @node How long do locals live?, Locals programming style, Where are locals visible by name?, Gforth locals
                   9308: @subsubsection How long do locals live?
                   9309: @cindex locals lifetime
                   9310: @cindex lifetime of locals
1.23      crook    9311: 
1.78      anton    9312: The right answer for the lifetime question would be: A local lives at
                   9313: least as long as it can be accessed. For a value-flavoured local this
                   9314: means: until the end of its visibility. However, a variable-flavoured
                   9315: local could be accessed through its address far beyond its visibility
                   9316: scope. Ultimately, this would mean that such locals would have to be
                   9317: garbage collected. Since this entails un-Forth-like implementation
                   9318: complexities, I adopted the same cowardly solution as some other
                   9319: languages (e.g., C): The local lives only as long as it is visible;
                   9320: afterwards its address is invalid (and programs that access it
                   9321: afterwards are erroneous).
1.23      crook    9322: 
1.78      anton    9323: @node Locals programming style, Locals implementation, How long do locals live?, Gforth locals
                   9324: @subsubsection Locals programming style
                   9325: @cindex locals programming style
                   9326: @cindex programming style, locals
1.23      crook    9327: 
1.78      anton    9328: The freedom to define locals anywhere has the potential to change
                   9329: programming styles dramatically. In particular, the need to use the
                   9330: return stack for intermediate storage vanishes. Moreover, all stack
                   9331: manipulations (except @code{PICK}s and @code{ROLL}s with run-time
                   9332: determined arguments) can be eliminated: If the stack items are in the
                   9333: wrong order, just write a locals definition for all of them; then
                   9334: write the items in the order you want.
1.23      crook    9335: 
1.78      anton    9336: This seems a little far-fetched and eliminating stack manipulations is
                   9337: unlikely to become a conscious programming objective. Still, the number
                   9338: of stack manipulations will be reduced dramatically if local variables
                   9339: are used liberally (e.g., compare @code{max} (@pxref{Gforth locals}) with
                   9340: a traditional implementation of @code{max}).
1.23      crook    9341: 
1.78      anton    9342: This shows one potential benefit of locals: making Forth programs more
                   9343: readable. Of course, this benefit will only be realized if the
                   9344: programmers continue to honour the principle of factoring instead of
                   9345: using the added latitude to make the words longer.
1.23      crook    9346: 
1.78      anton    9347: @cindex single-assignment style for locals
                   9348: Using @code{TO} can and should be avoided.  Without @code{TO},
                   9349: every value-flavoured local has only a single assignment and many
                   9350: advantages of functional languages apply to Forth. I.e., programs are
                   9351: easier to analyse, to optimize and to read: It is clear from the
                   9352: definition what the local stands for, it does not turn into something
                   9353: different later.
1.23      crook    9354: 
1.78      anton    9355: E.g., a definition using @code{TO} might look like this:
                   9356: @example
                   9357: : strcmp @{ addr1 u1 addr2 u2 -- n @}
                   9358:  u1 u2 min 0
                   9359:  ?do
                   9360:    addr1 c@@ addr2 c@@ -
                   9361:    ?dup-if
                   9362:      unloop exit
                   9363:    then
                   9364:    addr1 char+ TO addr1
                   9365:    addr2 char+ TO addr2
                   9366:  loop
                   9367:  u1 u2 - ;
1.26      crook    9368: @end example
1.78      anton    9369: Here, @code{TO} is used to update @code{addr1} and @code{addr2} at
                   9370: every loop iteration. @code{strcmp} is a typical example of the
                   9371: readability problems of using @code{TO}. When you start reading
                   9372: @code{strcmp}, you think that @code{addr1} refers to the start of the
                   9373: string. Only near the end of the loop you realize that it is something
                   9374: else.
1.23      crook    9375: 
1.78      anton    9376: This can be avoided by defining two locals at the start of the loop that
                   9377: are initialized with the right value for the current iteration.
                   9378: @example
                   9379: : strcmp @{ addr1 u1 addr2 u2 -- n @}
                   9380:  addr1 addr2
                   9381:  u1 u2 min 0 
                   9382:  ?do @{ s1 s2 @}
                   9383:    s1 c@@ s2 c@@ -
                   9384:    ?dup-if
                   9385:      unloop exit
                   9386:    then
                   9387:    s1 char+ s2 char+
                   9388:  loop
                   9389:  2drop
                   9390:  u1 u2 - ;
                   9391: @end example
                   9392: Here it is clear from the start that @code{s1} has a different value
                   9393: in every loop iteration.
1.23      crook    9394: 
1.78      anton    9395: @node Locals implementation,  , Locals programming style, Gforth locals
                   9396: @subsubsection Locals implementation
                   9397: @cindex locals implementation
                   9398: @cindex implementation of locals
1.23      crook    9399: 
1.78      anton    9400: @cindex locals stack
                   9401: Gforth uses an extra locals stack. The most compelling reason for
                   9402: this is that the return stack is not float-aligned; using an extra stack
                   9403: also eliminates the problems and restrictions of using the return stack
                   9404: as locals stack. Like the other stacks, the locals stack grows toward
                   9405: lower addresses. A few primitives allow an efficient implementation:
                   9406: 
                   9407: 
                   9408: doc-@local#
                   9409: doc-f@local#
                   9410: doc-laddr#
                   9411: doc-lp+!#
                   9412: doc-lp!
                   9413: doc->l
                   9414: doc-f>l
                   9415: 
                   9416: 
                   9417: In addition to these primitives, some specializations of these
                   9418: primitives for commonly occurring inline arguments are provided for
                   9419: efficiency reasons, e.g., @code{@@local0} as specialization of
                   9420: @code{@@local#} for the inline argument 0. The following compiling words
                   9421: compile the right specialized version, or the general version, as
                   9422: appropriate:
1.23      crook    9423: 
1.5       anton    9424: 
1.78      anton    9425: doc-compile-@local
                   9426: doc-compile-f@local
                   9427: doc-compile-lp+!
1.5       anton    9428: 
                   9429: 
1.78      anton    9430: Combinations of conditional branches and @code{lp+!#} like
                   9431: @code{?branch-lp+!#} (the locals pointer is only changed if the branch
                   9432: is taken) are provided for efficiency and correctness in loops.
1.5       anton    9433: 
1.78      anton    9434: A special area in the dictionary space is reserved for keeping the
                   9435: local variable names. @code{@{} switches the dictionary pointer to this
                   9436: area and @code{@}} switches it back and generates the locals
                   9437: initializing code. @code{W:} etc.@ are normal defining words. This
                   9438: special area is cleared at the start of every colon definition.
1.5       anton    9439: 
1.78      anton    9440: @cindex word list for defining locals
                   9441: A special feature of Gforth's dictionary is used to implement the
                   9442: definition of locals without type specifiers: every word list (aka
                   9443: vocabulary) has its own methods for searching
                   9444: etc. (@pxref{Word Lists}). For the present purpose we defined a word list
                   9445: with a special search method: When it is searched for a word, it
                   9446: actually creates that word using @code{W:}. @code{@{} changes the search
                   9447: order to first search the word list containing @code{@}}, @code{W:} etc.,
                   9448: and then the word list for defining locals without type specifiers.
1.5       anton    9449: 
1.78      anton    9450: The lifetime rules support a stack discipline within a colon
                   9451: definition: The lifetime of a local is either nested with other locals
                   9452: lifetimes or it does not overlap them.
1.23      crook    9453: 
1.78      anton    9454: At @code{BEGIN}, @code{IF}, and @code{AHEAD} no code for locals stack
                   9455: pointer manipulation is generated. Between control structure words
                   9456: locals definitions can push locals onto the locals stack. @code{AGAIN}
                   9457: is the simplest of the other three control flow words. It has to
                   9458: restore the locals stack depth of the corresponding @code{BEGIN}
                   9459: before branching. The code looks like this:
                   9460: @format
                   9461: @code{lp+!#} current-locals-size @minus{} dest-locals-size
                   9462: @code{branch} <begin>
                   9463: @end format
1.26      crook    9464: 
1.78      anton    9465: @code{UNTIL} is a little more complicated: If it branches back, it
                   9466: must adjust the stack just like @code{AGAIN}. But if it falls through,
                   9467: the locals stack must not be changed. The compiler generates the
                   9468: following code:
                   9469: @format
                   9470: @code{?branch-lp+!#} <begin> current-locals-size @minus{} dest-locals-size
                   9471: @end format
                   9472: The locals stack pointer is only adjusted if the branch is taken.
1.44      crook    9473: 
1.78      anton    9474: @code{THEN} can produce somewhat inefficient code:
                   9475: @format
                   9476: @code{lp+!#} current-locals-size @minus{} orig-locals-size
                   9477: <orig target>:
                   9478: @code{lp+!#} orig-locals-size @minus{} new-locals-size
                   9479: @end format
                   9480: The second @code{lp+!#} adjusts the locals stack pointer from the
                   9481: level at the @i{orig} point to the level after the @code{THEN}. The
                   9482: first @code{lp+!#} adjusts the locals stack pointer from the current
                   9483: level to the level at the orig point, so the complete effect is an
                   9484: adjustment from the current level to the right level after the
                   9485: @code{THEN}.
1.26      crook    9486: 
1.78      anton    9487: @cindex locals information on the control-flow stack
                   9488: @cindex control-flow stack items, locals information
                   9489: In a conventional Forth implementation a dest control-flow stack entry
                   9490: is just the target address and an orig entry is just the address to be
                   9491: patched. Our locals implementation adds a word list to every orig or dest
                   9492: item. It is the list of locals visible (or assumed visible) at the point
                   9493: described by the entry. Our implementation also adds a tag to identify
                   9494: the kind of entry, in particular to differentiate between live and dead
                   9495: (reachable and unreachable) orig entries.
1.26      crook    9496: 
1.78      anton    9497: A few unusual operations have to be performed on locals word lists:
1.44      crook    9498: 
1.5       anton    9499: 
1.78      anton    9500: doc-common-list
                   9501: doc-sub-list?
                   9502: doc-list-size
1.52      anton    9503: 
                   9504: 
1.78      anton    9505: Several features of our locals word list implementation make these
                   9506: operations easy to implement: The locals word lists are organised as
                   9507: linked lists; the tails of these lists are shared, if the lists
                   9508: contain some of the same locals; and the address of a name is greater
                   9509: than the address of the names behind it in the list.
1.5       anton    9510: 
1.78      anton    9511: Another important implementation detail is the variable
                   9512: @code{dead-code}. It is used by @code{BEGIN} and @code{THEN} to
                   9513: determine if they can be reached directly or only through the branch
                   9514: that they resolve. @code{dead-code} is set by @code{UNREACHABLE},
                   9515: @code{AHEAD}, @code{EXIT} etc., and cleared at the start of a colon
                   9516: definition, by @code{BEGIN} and usually by @code{THEN}.
1.5       anton    9517: 
1.78      anton    9518: Counted loops are similar to other loops in most respects, but
                   9519: @code{LEAVE} requires special attention: It performs basically the same
                   9520: service as @code{AHEAD}, but it does not create a control-flow stack
                   9521: entry. Therefore the information has to be stored elsewhere;
                   9522: traditionally, the information was stored in the target fields of the
                   9523: branches created by the @code{LEAVE}s, by organizing these fields into a
                   9524: linked list. Unfortunately, this clever trick does not provide enough
                   9525: space for storing our extended control flow information. Therefore, we
                   9526: introduce another stack, the leave stack. It contains the control-flow
                   9527: stack entries for all unresolved @code{LEAVE}s.
1.44      crook    9528: 
1.78      anton    9529: Local names are kept until the end of the colon definition, even if
                   9530: they are no longer visible in any control-flow path. In a few cases
                   9531: this may lead to increased space needs for the locals name area, but
                   9532: usually less than reclaiming this space would cost in code size.
1.5       anton    9533: 
1.44      crook    9534: 
1.78      anton    9535: @node ANS Forth locals,  , Gforth locals, Locals
                   9536: @subsection ANS Forth locals
                   9537: @cindex locals, ANS Forth style
1.5       anton    9538: 
1.78      anton    9539: The ANS Forth locals wordset does not define a syntax for locals, but
                   9540: words that make it possible to define various syntaxes. One of the
                   9541: possible syntaxes is a subset of the syntax we used in the Gforth locals
                   9542: wordset, i.e.:
1.29      crook    9543: 
                   9544: @example
1.78      anton    9545: @{ local1 local2 ... -- comment @}
                   9546: @end example
                   9547: @noindent
                   9548: or
                   9549: @example
                   9550: @{ local1 local2 ... @}
1.29      crook    9551: @end example
                   9552: 
1.78      anton    9553: The order of the locals corresponds to the order in a stack comment. The
                   9554: restrictions are:
1.5       anton    9555: 
1.78      anton    9556: @itemize @bullet
                   9557: @item
                   9558: Locals can only be cell-sized values (no type specifiers are allowed).
                   9559: @item
                   9560: Locals can be defined only outside control structures.
                   9561: @item
                   9562: Locals can interfere with explicit usage of the return stack. For the
                   9563: exact (and long) rules, see the standard. If you don't use return stack
                   9564: accessing words in a definition using locals, you will be all right. The
                   9565: purpose of this rule is to make locals implementation on the return
                   9566: stack easier.
                   9567: @item
                   9568: The whole definition must be in one line.
                   9569: @end itemize
1.5       anton    9570: 
1.78      anton    9571: Locals defined in ANS Forth behave like @code{VALUE}s
                   9572: (@pxref{Values}). I.e., they are initialized from the stack. Using their
                   9573: name produces their value. Their value can be changed using @code{TO}.
1.77      anton    9574: 
1.78      anton    9575: Since the syntax above is supported by Gforth directly, you need not do
                   9576: anything to use it. If you want to port a program using this syntax to
                   9577: another ANS Forth system, use @file{compat/anslocal.fs} to implement the
                   9578: syntax on the other system.
1.5       anton    9579: 
1.78      anton    9580: Note that a syntax shown in the standard, section A.13 looks
                   9581: similar, but is quite different in having the order of locals
                   9582: reversed. Beware!
1.5       anton    9583: 
1.78      anton    9584: The ANS Forth locals wordset itself consists of one word:
1.5       anton    9585: 
1.78      anton    9586: doc-(local)
1.5       anton    9587: 
1.78      anton    9588: The ANS Forth locals extension wordset defines a syntax using
                   9589: @code{locals|}, but it is so awful that we strongly recommend not to use
                   9590: it. We have implemented this syntax to make porting to Gforth easy, but
                   9591: do not document it here. The problem with this syntax is that the locals
                   9592: are defined in an order reversed with respect to the standard stack
                   9593: comment notation, making programs harder to read, and easier to misread
                   9594: and miswrite. The only merit of this syntax is that it is easy to
                   9595: implement using the ANS Forth locals wordset.
1.53      anton    9596: 
                   9597: 
1.78      anton    9598: @c ----------------------------------------------------------
                   9599: @node Structures, Object-oriented Forth, Locals, Words
                   9600: @section  Structures
                   9601: @cindex structures
                   9602: @cindex records
1.53      anton    9603: 
1.78      anton    9604: This section presents the structure package that comes with Gforth. A
                   9605: version of the package implemented in ANS Forth is available in
                   9606: @file{compat/struct.fs}. This package was inspired by a posting on
                   9607: comp.lang.forth in 1989 (unfortunately I don't remember, by whom;
                   9608: possibly John Hayes). A version of this section has been published in
                   9609: M. Anton Ertl,
                   9610: @uref{http://www.complang.tuwien.ac.at/forth/objects/structs.html, Yet
                   9611: Another Forth Structures Package}, Forth Dimensions 19(3), pages
                   9612: 13--16. Marcel Hendrix provided helpful comments.
1.53      anton    9613: 
1.78      anton    9614: @menu
                   9615: * Why explicit structure support?::  
                   9616: * Structure Usage::             
                   9617: * Structure Naming Convention::  
                   9618: * Structure Implementation::    
                   9619: * Structure Glossary::          
                   9620: @end menu
1.55      anton    9621: 
1.78      anton    9622: @node Why explicit structure support?, Structure Usage, Structures, Structures
                   9623: @subsection Why explicit structure support?
1.53      anton    9624: 
1.78      anton    9625: @cindex address arithmetic for structures
                   9626: @cindex structures using address arithmetic
                   9627: If we want to use a structure containing several fields, we could simply
                   9628: reserve memory for it, and access the fields using address arithmetic
                   9629: (@pxref{Address arithmetic}). As an example, consider a structure with
                   9630: the following fields
1.57      anton    9631: 
1.78      anton    9632: @table @code
                   9633: @item a
                   9634: is a float
                   9635: @item b
                   9636: is a cell
                   9637: @item c
                   9638: is a float
                   9639: @end table
1.57      anton    9640: 
1.78      anton    9641: Given the (float-aligned) base address of the structure we get the
                   9642: address of the field
1.52      anton    9643: 
1.78      anton    9644: @table @code
                   9645: @item a
                   9646: without doing anything further.
                   9647: @item b
                   9648: with @code{float+}
                   9649: @item c
                   9650: with @code{float+ cell+ faligned}
                   9651: @end table
1.52      anton    9652: 
1.78      anton    9653: It is easy to see that this can become quite tiring. 
1.52      anton    9654: 
1.78      anton    9655: Moreover, it is not very readable, because seeing a
                   9656: @code{cell+} tells us neither which kind of structure is
                   9657: accessed nor what field is accessed; we have to somehow infer the kind
                   9658: of structure, and then look up in the documentation, which field of
                   9659: that structure corresponds to that offset.
1.53      anton    9660: 
1.78      anton    9661: Finally, this kind of address arithmetic also causes maintenance
                   9662: troubles: If you add or delete a field somewhere in the middle of the
                   9663: structure, you have to find and change all computations for the fields
                   9664: afterwards.
1.52      anton    9665: 
1.78      anton    9666: So, instead of using @code{cell+} and friends directly, how
                   9667: about storing the offsets in constants:
1.52      anton    9668: 
1.78      anton    9669: @example
                   9670: 0 constant a-offset
                   9671: 0 float+ constant b-offset
                   9672: 0 float+ cell+ faligned c-offset
                   9673: @end example
1.64      pazsan   9674: 
1.78      anton    9675: Now we can get the address of field @code{x} with @code{x-offset
                   9676: +}. This is much better in all respects. Of course, you still
                   9677: have to change all later offset definitions if you add a field. You can
                   9678: fix this by declaring the offsets in the following way:
1.57      anton    9679: 
1.78      anton    9680: @example
                   9681: 0 constant a-offset
                   9682: a-offset float+ constant b-offset
                   9683: b-offset cell+ faligned constant c-offset
                   9684: @end example
1.57      anton    9685: 
1.78      anton    9686: Since we always use the offsets with @code{+}, we could use a defining
                   9687: word @code{cfield} that includes the @code{+} in the action of the
                   9688: defined word:
1.64      pazsan   9689: 
1.78      anton    9690: @example
                   9691: : cfield ( n "name" -- )
                   9692:     create ,
                   9693: does> ( name execution: addr1 -- addr2 )
                   9694:     @@ + ;
1.64      pazsan   9695: 
1.78      anton    9696: 0 cfield a
                   9697: 0 a float+ cfield b
                   9698: 0 b cell+ faligned cfield c
                   9699: @end example
1.64      pazsan   9700: 
1.78      anton    9701: Instead of @code{x-offset +}, we now simply write @code{x}.
1.64      pazsan   9702: 
1.78      anton    9703: The structure field words now can be used quite nicely. However,
                   9704: their definition is still a bit cumbersome: We have to repeat the
                   9705: name, the information about size and alignment is distributed before
                   9706: and after the field definitions etc.  The structure package presented
                   9707: here addresses these problems.
1.64      pazsan   9708: 
1.78      anton    9709: @node Structure Usage, Structure Naming Convention, Why explicit structure support?, Structures
                   9710: @subsection Structure Usage
                   9711: @cindex structure usage
1.57      anton    9712: 
1.78      anton    9713: @cindex @code{field} usage
                   9714: @cindex @code{struct} usage
                   9715: @cindex @code{end-struct} usage
                   9716: You can define a structure for a (data-less) linked list with:
1.57      anton    9717: @example
1.78      anton    9718: struct
                   9719:     cell% field list-next
                   9720: end-struct list%
1.57      anton    9721: @end example
                   9722: 
1.78      anton    9723: With the address of the list node on the stack, you can compute the
                   9724: address of the field that contains the address of the next node with
                   9725: @code{list-next}. E.g., you can determine the length of a list
                   9726: with:
1.57      anton    9727: 
                   9728: @example
1.78      anton    9729: : list-length ( list -- n )
                   9730: \ "list" is a pointer to the first element of a linked list
                   9731: \ "n" is the length of the list
                   9732:     0 BEGIN ( list1 n1 )
                   9733:         over
                   9734:     WHILE ( list1 n1 )
                   9735:         1+ swap list-next @@ swap
                   9736:     REPEAT
                   9737:     nip ;
1.57      anton    9738: @end example
                   9739: 
1.78      anton    9740: You can reserve memory for a list node in the dictionary with
                   9741: @code{list% %allot}, which leaves the address of the list node on the
                   9742: stack. For the equivalent allocation on the heap you can use @code{list%
                   9743: %alloc} (or, for an @code{allocate}-like stack effect (i.e., with ior),
                   9744: use @code{list% %allocate}). You can get the the size of a list
                   9745: node with @code{list% %size} and its alignment with @code{list%
                   9746: %alignment}.
                   9747: 
                   9748: Note that in ANS Forth the body of a @code{create}d word is
                   9749: @code{aligned} but not necessarily @code{faligned};
                   9750: therefore, if you do a:
1.57      anton    9751: 
                   9752: @example
1.78      anton    9753: create @emph{name} foo% %allot drop
1.57      anton    9754: @end example
                   9755: 
1.78      anton    9756: @noindent
                   9757: then the memory alloted for @code{foo%} is guaranteed to start at the
                   9758: body of @code{@emph{name}} only if @code{foo%} contains only character,
                   9759: cell and double fields.  Therefore, if your structure contains floats,
                   9760: better use
1.57      anton    9761: 
                   9762: @example
1.78      anton    9763: foo% %allot constant @emph{name}
1.57      anton    9764: @end example
                   9765: 
1.78      anton    9766: @cindex structures containing structures
                   9767: You can include a structure @code{foo%} as a field of
                   9768: another structure, like this:
1.65      anton    9769: @example
1.78      anton    9770: struct
                   9771: ...
                   9772:     foo% field ...
                   9773: ...
                   9774: end-struct ...
1.65      anton    9775: @end example
1.52      anton    9776: 
1.78      anton    9777: @cindex structure extension
                   9778: @cindex extended records
                   9779: Instead of starting with an empty structure, you can extend an
                   9780: existing structure. E.g., a plain linked list without data, as defined
                   9781: above, is hardly useful; You can extend it to a linked list of integers,
                   9782: like this:@footnote{This feature is also known as @emph{extended
                   9783: records}. It is the main innovation in the Oberon language; in other
                   9784: words, adding this feature to Modula-2 led Wirth to create a new
                   9785: language, write a new compiler etc.  Adding this feature to Forth just
                   9786: required a few lines of code.}
1.52      anton    9787: 
1.78      anton    9788: @example
                   9789: list%
                   9790:     cell% field intlist-int
                   9791: end-struct intlist%
                   9792: @end example
1.55      anton    9793: 
1.78      anton    9794: @code{intlist%} is a structure with two fields:
                   9795: @code{list-next} and @code{intlist-int}.
1.55      anton    9796: 
1.78      anton    9797: @cindex structures containing arrays
                   9798: You can specify an array type containing @emph{n} elements of
                   9799: type @code{foo%} like this:
1.55      anton    9800: 
                   9801: @example
1.78      anton    9802: foo% @emph{n} *
1.56      anton    9803: @end example
1.55      anton    9804: 
1.78      anton    9805: You can use this array type in any place where you can use a normal
                   9806: type, e.g., when defining a @code{field}, or with
                   9807: @code{%allot}.
                   9808: 
                   9809: @cindex first field optimization
                   9810: The first field is at the base address of a structure and the word for
                   9811: this field (e.g., @code{list-next}) actually does not change the address
                   9812: on the stack. You may be tempted to leave it away in the interest of
                   9813: run-time and space efficiency. This is not necessary, because the
                   9814: structure package optimizes this case: If you compile a first-field
                   9815: words, no code is generated. So, in the interest of readability and
                   9816: maintainability you should include the word for the field when accessing
                   9817: the field.
1.52      anton    9818: 
                   9819: 
1.78      anton    9820: @node Structure Naming Convention, Structure Implementation, Structure Usage, Structures
                   9821: @subsection Structure Naming Convention
                   9822: @cindex structure naming convention
1.52      anton    9823: 
1.78      anton    9824: The field names that come to (my) mind are often quite generic, and,
                   9825: if used, would cause frequent name clashes. E.g., many structures
                   9826: probably contain a @code{counter} field. The structure names
                   9827: that come to (my) mind are often also the logical choice for the names
                   9828: of words that create such a structure.
1.52      anton    9829: 
1.78      anton    9830: Therefore, I have adopted the following naming conventions: 
1.52      anton    9831: 
1.78      anton    9832: @itemize @bullet
                   9833: @cindex field naming convention
                   9834: @item
                   9835: The names of fields are of the form
                   9836: @code{@emph{struct}-@emph{field}}, where
                   9837: @code{@emph{struct}} is the basic name of the structure, and
                   9838: @code{@emph{field}} is the basic name of the field. You can
                   9839: think of field words as converting the (address of the)
                   9840: structure into the (address of the) field.
1.52      anton    9841: 
1.78      anton    9842: @cindex structure naming convention
                   9843: @item
                   9844: The names of structures are of the form
                   9845: @code{@emph{struct}%}, where
                   9846: @code{@emph{struct}} is the basic name of the structure.
                   9847: @end itemize
1.52      anton    9848: 
1.78      anton    9849: This naming convention does not work that well for fields of extended
                   9850: structures; e.g., the integer list structure has a field
                   9851: @code{intlist-int}, but has @code{list-next}, not
                   9852: @code{intlist-next}.
1.53      anton    9853: 
1.78      anton    9854: @node Structure Implementation, Structure Glossary, Structure Naming Convention, Structures
                   9855: @subsection Structure Implementation
                   9856: @cindex structure implementation
                   9857: @cindex implementation of structures
1.52      anton    9858: 
1.78      anton    9859: The central idea in the implementation is to pass the data about the
                   9860: structure being built on the stack, not in some global
                   9861: variable. Everything else falls into place naturally once this design
                   9862: decision is made.
1.53      anton    9863: 
1.78      anton    9864: The type description on the stack is of the form @emph{align
                   9865: size}. Keeping the size on the top-of-stack makes dealing with arrays
                   9866: very simple.
1.53      anton    9867: 
1.78      anton    9868: @code{field} is a defining word that uses @code{Create}
                   9869: and @code{DOES>}. The body of the field contains the offset
                   9870: of the field, and the normal @code{DOES>} action is simply:
1.53      anton    9871: 
                   9872: @example
1.78      anton    9873: @@ +
1.53      anton    9874: @end example
                   9875: 
1.78      anton    9876: @noindent
                   9877: i.e., add the offset to the address, giving the stack effect
                   9878: @i{addr1 -- addr2} for a field.
                   9879: 
                   9880: @cindex first field optimization, implementation
                   9881: This simple structure is slightly complicated by the optimization
                   9882: for fields with offset 0, which requires a different
                   9883: @code{DOES>}-part (because we cannot rely on there being
                   9884: something on the stack if such a field is invoked during
                   9885: compilation). Therefore, we put the different @code{DOES>}-parts
                   9886: in separate words, and decide which one to invoke based on the
                   9887: offset. For a zero offset, the field is basically a noop; it is
                   9888: immediate, and therefore no code is generated when it is compiled.
1.53      anton    9889: 
1.78      anton    9890: @node Structure Glossary,  , Structure Implementation, Structures
                   9891: @subsection Structure Glossary
                   9892: @cindex structure glossary
1.53      anton    9893: 
1.5       anton    9894: 
1.78      anton    9895: doc-%align
                   9896: doc-%alignment
                   9897: doc-%alloc
                   9898: doc-%allocate
                   9899: doc-%allot
                   9900: doc-cell%
                   9901: doc-char%
                   9902: doc-dfloat%
                   9903: doc-double%
                   9904: doc-end-struct
                   9905: doc-field
                   9906: doc-float%
                   9907: doc-naligned
                   9908: doc-sfloat%
                   9909: doc-%size
                   9910: doc-struct
1.54      anton    9911: 
                   9912: 
1.26      crook    9913: @c -------------------------------------------------------------
1.78      anton    9914: @node Object-oriented Forth, Programming Tools, Structures, Words
                   9915: @section Object-oriented Forth
                   9916: 
                   9917: Gforth comes with three packages for object-oriented programming:
                   9918: @file{objects.fs}, @file{oof.fs}, and @file{mini-oof.fs}; none of them
                   9919: is preloaded, so you have to @code{include} them before use. The most
                   9920: important differences between these packages (and others) are discussed
                   9921: in @ref{Comparison with other object models}. All packages are written
                   9922: in ANS Forth and can be used with any other ANS Forth.
1.5       anton    9923: 
1.78      anton    9924: @menu
                   9925: * Why object-oriented programming?::  
                   9926: * Object-Oriented Terminology::  
                   9927: * Objects::                     
                   9928: * OOF::                         
                   9929: * Mini-OOF::                    
                   9930: * Comparison with other object models::  
                   9931: @end menu
1.5       anton    9932: 
1.78      anton    9933: @c ----------------------------------------------------------------
                   9934: @node Why object-oriented programming?, Object-Oriented Terminology, Object-oriented Forth, Object-oriented Forth
                   9935: @subsection Why object-oriented programming?
                   9936: @cindex object-oriented programming motivation
                   9937: @cindex motivation for object-oriented programming
1.44      crook    9938: 
1.78      anton    9939: Often we have to deal with several data structures (@emph{objects}),
                   9940: that have to be treated similarly in some respects, but differently in
                   9941: others. Graphical objects are the textbook example: circles, triangles,
                   9942: dinosaurs, icons, and others, and we may want to add more during program
                   9943: development. We want to apply some operations to any graphical object,
                   9944: e.g., @code{draw} for displaying it on the screen. However, @code{draw}
                   9945: has to do something different for every kind of object.
                   9946: @comment TODO add some other operations eg perimeter, area
                   9947: @comment and tie in to concrete examples later..
1.5       anton    9948: 
1.78      anton    9949: We could implement @code{draw} as a big @code{CASE}
                   9950: control structure that executes the appropriate code depending on the
                   9951: kind of object to be drawn. This would be not be very elegant, and,
                   9952: moreover, we would have to change @code{draw} every time we add
                   9953: a new kind of graphical object (say, a spaceship).
1.44      crook    9954: 
1.78      anton    9955: What we would rather do is: When defining spaceships, we would tell
                   9956: the system: ``Here's how you @code{draw} a spaceship; you figure
                   9957: out the rest''.
1.5       anton    9958: 
1.78      anton    9959: This is the problem that all systems solve that (rightfully) call
                   9960: themselves object-oriented; the object-oriented packages presented here
                   9961: solve this problem (and not much else).
                   9962: @comment TODO ?list properties of oo systems.. oo vs o-based?
1.44      crook    9963: 
1.78      anton    9964: @c ------------------------------------------------------------------------
                   9965: @node Object-Oriented Terminology, Objects, Why object-oriented programming?, Object-oriented Forth
                   9966: @subsection Object-Oriented Terminology
                   9967: @cindex object-oriented terminology
                   9968: @cindex terminology for object-oriented programming
1.5       anton    9969: 
1.78      anton    9970: This section is mainly for reference, so you don't have to understand
                   9971: all of it right away.  The terminology is mainly Smalltalk-inspired.  In
                   9972: short:
1.44      crook    9973: 
1.78      anton    9974: @table @emph
                   9975: @cindex class
                   9976: @item class
                   9977: a data structure definition with some extras.
1.5       anton    9978: 
1.78      anton    9979: @cindex object
                   9980: @item object
                   9981: an instance of the data structure described by the class definition.
1.5       anton    9982: 
1.78      anton    9983: @cindex instance variables
                   9984: @item instance variables
                   9985: fields of the data structure.
1.5       anton    9986: 
1.78      anton    9987: @cindex selector
                   9988: @cindex method selector
                   9989: @cindex virtual function
                   9990: @item selector
                   9991: (or @emph{method selector}) a word (e.g.,
                   9992: @code{draw}) that performs an operation on a variety of data
                   9993: structures (classes). A selector describes @emph{what} operation to
                   9994: perform. In C++ terminology: a (pure) virtual function.
1.5       anton    9995: 
1.78      anton    9996: @cindex method
                   9997: @item method
                   9998: the concrete definition that performs the operation
                   9999: described by the selector for a specific class. A method specifies
                   10000: @emph{how} the operation is performed for a specific class.
1.5       anton    10001: 
1.78      anton    10002: @cindex selector invocation
                   10003: @cindex message send
                   10004: @cindex invoking a selector
                   10005: @item selector invocation
                   10006: a call of a selector. One argument of the call (the TOS (top-of-stack))
                   10007: is used for determining which method is used. In Smalltalk terminology:
                   10008: a message (consisting of the selector and the other arguments) is sent
                   10009: to the object.
1.5       anton    10010: 
1.78      anton    10011: @cindex receiving object
                   10012: @item receiving object
                   10013: the object used for determining the method executed by a selector
                   10014: invocation. In the @file{objects.fs} model, it is the object that is on
                   10015: the TOS when the selector is invoked. (@emph{Receiving} comes from
                   10016: the Smalltalk @emph{message} terminology.)
1.5       anton    10017: 
1.78      anton    10018: @cindex child class
                   10019: @cindex parent class
                   10020: @cindex inheritance
                   10021: @item child class
                   10022: a class that has (@emph{inherits}) all properties (instance variables,
                   10023: selectors, methods) from a @emph{parent class}. In Smalltalk
                   10024: terminology: The subclass inherits from the superclass. In C++
                   10025: terminology: The derived class inherits from the base class.
1.5       anton    10026: 
1.78      anton    10027: @end table
1.5       anton    10028: 
1.78      anton    10029: @c If you wonder about the message sending terminology, it comes from
                   10030: @c a time when each object had it's own task and objects communicated via
                   10031: @c message passing; eventually the Smalltalk developers realized that
                   10032: @c they can do most things through simple (indirect) calls. They kept the
                   10033: @c terminology.
1.5       anton    10034: 
1.78      anton    10035: @c --------------------------------------------------------------
                   10036: @node Objects, OOF, Object-Oriented Terminology, Object-oriented Forth
                   10037: @subsection The @file{objects.fs} model
                   10038: @cindex objects
                   10039: @cindex object-oriented programming
1.26      crook    10040: 
1.78      anton    10041: @cindex @file{objects.fs}
                   10042: @cindex @file{oof.fs}
1.26      crook    10043: 
1.78      anton    10044: This section describes the @file{objects.fs} package. This material also
                   10045: has been published in M. Anton Ertl,
                   10046: @cite{@uref{http://www.complang.tuwien.ac.at/forth/objects/objects.html,
                   10047: Yet Another Forth Objects Package}}, Forth Dimensions 19(2), pages
                   10048: 37--43.
                   10049: @c McKewan's and Zsoter's packages
1.26      crook    10050: 
1.78      anton    10051: This section assumes that you have read @ref{Structures}.
1.5       anton    10052: 
1.78      anton    10053: The techniques on which this model is based have been used to implement
                   10054: the parser generator, Gray, and have also been used in Gforth for
                   10055: implementing the various flavours of word lists (hashed or not,
                   10056: case-sensitive or not, special-purpose word lists for locals etc.).
1.5       anton    10057: 
                   10058: 
1.26      crook    10059: @menu
1.78      anton    10060: * Properties of the Objects model::  
                   10061: * Basic Objects Usage::         
                   10062: * The Objects base class::      
                   10063: * Creating objects::            
                   10064: * Object-Oriented Programming Style::  
                   10065: * Class Binding::               
                   10066: * Method conveniences::         
                   10067: * Classes and Scoping::         
                   10068: * Dividing classes::            
                   10069: * Object Interfaces::           
                   10070: * Objects Implementation::      
                   10071: * Objects Glossary::            
1.26      crook    10072: @end menu
1.5       anton    10073: 
1.78      anton    10074: Marcel Hendrix provided helpful comments on this section.
1.5       anton    10075: 
1.78      anton    10076: @node Properties of the Objects model, Basic Objects Usage, Objects, Objects
                   10077: @subsubsection Properties of the @file{objects.fs} model
                   10078: @cindex @file{objects.fs} properties
1.5       anton    10079: 
1.78      anton    10080: @itemize @bullet
                   10081: @item
                   10082: It is straightforward to pass objects on the stack. Passing
                   10083: selectors on the stack is a little less convenient, but possible.
1.44      crook    10084: 
1.78      anton    10085: @item
                   10086: Objects are just data structures in memory, and are referenced by their
                   10087: address. You can create words for objects with normal defining words
                   10088: like @code{constant}. Likewise, there is no difference between instance
                   10089: variables that contain objects and those that contain other data.
1.5       anton    10090: 
1.78      anton    10091: @item
                   10092: Late binding is efficient and easy to use.
1.44      crook    10093: 
1.78      anton    10094: @item
                   10095: It avoids parsing, and thus avoids problems with state-smartness
                   10096: and reduced extensibility; for convenience there are a few parsing
                   10097: words, but they have non-parsing counterparts. There are also a few
                   10098: defining words that parse. This is hard to avoid, because all standard
                   10099: defining words parse (except @code{:noname}); however, such
                   10100: words are not as bad as many other parsing words, because they are not
                   10101: state-smart.
1.5       anton    10102: 
1.78      anton    10103: @item
                   10104: It does not try to incorporate everything. It does a few things and does
                   10105: them well (IMO). In particular, this model was not designed to support
                   10106: information hiding (although it has features that may help); you can use
                   10107: a separate package for achieving this.
1.5       anton    10108: 
1.78      anton    10109: @item
                   10110: It is layered; you don't have to learn and use all features to use this
                   10111: model. Only a few features are necessary (@pxref{Basic Objects Usage},
                   10112: @pxref{The Objects base class}, @pxref{Creating objects}.), the others
                   10113: are optional and independent of each other.
1.5       anton    10114: 
1.78      anton    10115: @item
                   10116: An implementation in ANS Forth is available.
1.5       anton    10117: 
1.78      anton    10118: @end itemize
1.5       anton    10119: 
1.44      crook    10120: 
1.78      anton    10121: @node Basic Objects Usage, The Objects base class, Properties of the Objects model, Objects
                   10122: @subsubsection Basic @file{objects.fs} Usage
                   10123: @cindex basic objects usage
                   10124: @cindex objects, basic usage
1.5       anton    10125: 
1.78      anton    10126: You can define a class for graphical objects like this:
1.44      crook    10127: 
1.78      anton    10128: @cindex @code{class} usage
                   10129: @cindex @code{end-class} usage
                   10130: @cindex @code{selector} usage
1.5       anton    10131: @example
1.78      anton    10132: object class \ "object" is the parent class
                   10133:   selector draw ( x y graphical -- )
                   10134: end-class graphical
                   10135: @end example
                   10136: 
                   10137: This code defines a class @code{graphical} with an
                   10138: operation @code{draw}.  We can perform the operation
                   10139: @code{draw} on any @code{graphical} object, e.g.:
                   10140: 
                   10141: @example
                   10142: 100 100 t-rex draw
1.26      crook    10143: @end example
1.5       anton    10144: 
1.78      anton    10145: @noindent
                   10146: where @code{t-rex} is a word (say, a constant) that produces a
                   10147: graphical object.
                   10148: 
                   10149: @comment TODO add a 2nd operation eg perimeter.. and use for
                   10150: @comment a concrete example
1.5       anton    10151: 
1.78      anton    10152: @cindex abstract class
                   10153: How do we create a graphical object? With the present definitions,
                   10154: we cannot create a useful graphical object. The class
                   10155: @code{graphical} describes graphical objects in general, but not
                   10156: any concrete graphical object type (C++ users would call it an
                   10157: @emph{abstract class}); e.g., there is no method for the selector
                   10158: @code{draw} in the class @code{graphical}.
1.5       anton    10159: 
1.78      anton    10160: For concrete graphical objects, we define child classes of the
                   10161: class @code{graphical}, e.g.:
1.5       anton    10162: 
1.78      anton    10163: @cindex @code{overrides} usage
                   10164: @cindex @code{field} usage in class definition
1.26      crook    10165: @example
1.78      anton    10166: graphical class \ "graphical" is the parent class
                   10167:   cell% field circle-radius
1.5       anton    10168: 
1.78      anton    10169: :noname ( x y circle -- )
                   10170:   circle-radius @@ draw-circle ;
                   10171: overrides draw
1.5       anton    10172: 
1.78      anton    10173: :noname ( n-radius circle -- )
                   10174:   circle-radius ! ;
                   10175: overrides construct
1.5       anton    10176: 
1.78      anton    10177: end-class circle
                   10178: @end example
1.44      crook    10179: 
1.78      anton    10180: Here we define a class @code{circle} as a child of @code{graphical},
                   10181: with field @code{circle-radius} (which behaves just like a field
                   10182: (@pxref{Structures}); it defines (using @code{overrides}) new methods
                   10183: for the selectors @code{draw} and @code{construct} (@code{construct} is
                   10184: defined in @code{object}, the parent class of @code{graphical}).
1.5       anton    10185: 
1.78      anton    10186: Now we can create a circle on the heap (i.e.,
                   10187: @code{allocate}d memory) with:
1.44      crook    10188: 
1.78      anton    10189: @cindex @code{heap-new} usage
1.5       anton    10190: @example
1.78      anton    10191: 50 circle heap-new constant my-circle
1.5       anton    10192: @end example
                   10193: 
1.78      anton    10194: @noindent
                   10195: @code{heap-new} invokes @code{construct}, thus
                   10196: initializing the field @code{circle-radius} with 50. We can draw
                   10197: this new circle at (100,100) with:
1.5       anton    10198: 
                   10199: @example
1.78      anton    10200: 100 100 my-circle draw
1.5       anton    10201: @end example
                   10202: 
1.78      anton    10203: @cindex selector invocation, restrictions
                   10204: @cindex class definition, restrictions
                   10205: Note: You can only invoke a selector if the object on the TOS
                   10206: (the receiving object) belongs to the class where the selector was
                   10207: defined or one of its descendents; e.g., you can invoke
                   10208: @code{draw} only for objects belonging to @code{graphical}
                   10209: or its descendents (e.g., @code{circle}).  Immediately before
                   10210: @code{end-class}, the search order has to be the same as
                   10211: immediately after @code{class}.
                   10212: 
                   10213: @node The Objects base class, Creating objects, Basic Objects Usage, Objects
                   10214: @subsubsection The @file{object.fs} base class
                   10215: @cindex @code{object} class
                   10216: 
                   10217: When you define a class, you have to specify a parent class.  So how do
                   10218: you start defining classes? There is one class available from the start:
                   10219: @code{object}. It is ancestor for all classes and so is the
                   10220: only class that has no parent. It has two selectors: @code{construct}
                   10221: and @code{print}.
                   10222: 
                   10223: @node Creating objects, Object-Oriented Programming Style, The Objects base class, Objects
                   10224: @subsubsection Creating objects
                   10225: @cindex creating objects
                   10226: @cindex object creation
                   10227: @cindex object allocation options
                   10228: 
                   10229: @cindex @code{heap-new} discussion
                   10230: @cindex @code{dict-new} discussion
                   10231: @cindex @code{construct} discussion
                   10232: You can create and initialize an object of a class on the heap with
                   10233: @code{heap-new} ( ... class -- object ) and in the dictionary
                   10234: (allocation with @code{allot}) with @code{dict-new} (
                   10235: ... class -- object ). Both words invoke @code{construct}, which
                   10236: consumes the stack items indicated by "..." above.
                   10237: 
                   10238: @cindex @code{init-object} discussion
                   10239: @cindex @code{class-inst-size} discussion
                   10240: If you want to allocate memory for an object yourself, you can get its
                   10241: alignment and size with @code{class-inst-size 2@@} ( class --
                   10242: align size ). Once you have memory for an object, you can initialize
                   10243: it with @code{init-object} ( ... class object -- );
                   10244: @code{construct} does only a part of the necessary work.
                   10245: 
                   10246: @node Object-Oriented Programming Style, Class Binding, Creating objects, Objects
                   10247: @subsubsection Object-Oriented Programming Style
                   10248: @cindex object-oriented programming style
                   10249: @cindex programming style, object-oriented
1.5       anton    10250: 
1.78      anton    10251: This section is not exhaustive.
1.5       anton    10252: 
1.78      anton    10253: @cindex stack effects of selectors
                   10254: @cindex selectors and stack effects
                   10255: In general, it is a good idea to ensure that all methods for the
                   10256: same selector have the same stack effect: when you invoke a selector,
                   10257: you often have no idea which method will be invoked, so, unless all
                   10258: methods have the same stack effect, you will not know the stack effect
                   10259: of the selector invocation.
1.5       anton    10260: 
1.78      anton    10261: One exception to this rule is methods for the selector
                   10262: @code{construct}. We know which method is invoked, because we
                   10263: specify the class to be constructed at the same place. Actually, I
                   10264: defined @code{construct} as a selector only to give the users a
                   10265: convenient way to specify initialization. The way it is used, a
                   10266: mechanism different from selector invocation would be more natural
                   10267: (but probably would take more code and more space to explain).
1.5       anton    10268: 
1.78      anton    10269: @node Class Binding, Method conveniences, Object-Oriented Programming Style, Objects
                   10270: @subsubsection Class Binding
                   10271: @cindex class binding
                   10272: @cindex early binding
1.5       anton    10273: 
1.78      anton    10274: @cindex late binding
                   10275: Normal selector invocations determine the method at run-time depending
                   10276: on the class of the receiving object. This run-time selection is called
                   10277: @i{late binding}.
1.5       anton    10278: 
1.78      anton    10279: Sometimes it's preferable to invoke a different method. For example,
                   10280: you might want to use the simple method for @code{print}ing
                   10281: @code{object}s instead of the possibly long-winded @code{print} method
                   10282: of the receiver class. You can achieve this by replacing the invocation
                   10283: of @code{print} with:
1.5       anton    10284: 
1.78      anton    10285: @cindex @code{[bind]} usage
1.5       anton    10286: @example
1.78      anton    10287: [bind] object print
1.5       anton    10288: @end example
                   10289: 
1.78      anton    10290: @noindent
                   10291: in compiled code or:
                   10292: 
                   10293: @cindex @code{bind} usage
1.5       anton    10294: @example
1.78      anton    10295: bind object print
1.5       anton    10296: @end example
                   10297: 
1.78      anton    10298: @cindex class binding, alternative to
                   10299: @noindent
                   10300: in interpreted code. Alternatively, you can define the method with a
                   10301: name (e.g., @code{print-object}), and then invoke it through the
                   10302: name. Class binding is just a (often more convenient) way to achieve
                   10303: the same effect; it avoids name clutter and allows you to invoke
                   10304: methods directly without naming them first.
1.5       anton    10305: 
1.78      anton    10306: @cindex superclass binding
                   10307: @cindex parent class binding
                   10308: A frequent use of class binding is this: When we define a method
                   10309: for a selector, we often want the method to do what the selector does
                   10310: in the parent class, and a little more. There is a special word for
                   10311: this purpose: @code{[parent]}; @code{[parent]
                   10312: @emph{selector}} is equivalent to @code{[bind] @emph{parent
                   10313: selector}}, where @code{@emph{parent}} is the parent
                   10314: class of the current class. E.g., a method definition might look like:
1.44      crook    10315: 
1.78      anton    10316: @cindex @code{[parent]} usage
                   10317: @example
                   10318: :noname
                   10319:   dup [parent] foo \ do parent's foo on the receiving object
                   10320:   ... \ do some more
                   10321: ; overrides foo
                   10322: @end example
1.6       pazsan   10323: 
1.78      anton    10324: @cindex class binding as optimization
                   10325: In @cite{Object-oriented programming in ANS Forth} (Forth Dimensions,
                   10326: March 1997), Andrew McKewan presents class binding as an optimization
                   10327: technique. I recommend not using it for this purpose unless you are in
                   10328: an emergency. Late binding is pretty fast with this model anyway, so the
                   10329: benefit of using class binding is small; the cost of using class binding
                   10330: where it is not appropriate is reduced maintainability.
1.44      crook    10331: 
1.78      anton    10332: While we are at programming style questions: You should bind
                   10333: selectors only to ancestor classes of the receiving object. E.g., say,
                   10334: you know that the receiving object is of class @code{foo} or its
                   10335: descendents; then you should bind only to @code{foo} and its
                   10336: ancestors.
1.12      anton    10337: 
1.78      anton    10338: @node Method conveniences, Classes and Scoping, Class Binding, Objects
                   10339: @subsubsection Method conveniences
                   10340: @cindex method conveniences
1.44      crook    10341: 
1.78      anton    10342: In a method you usually access the receiving object pretty often.  If
                   10343: you define the method as a plain colon definition (e.g., with
                   10344: @code{:noname}), you may have to do a lot of stack
                   10345: gymnastics. To avoid this, you can define the method with @code{m:
                   10346: ... ;m}. E.g., you could define the method for
                   10347: @code{draw}ing a @code{circle} with
1.6       pazsan   10348: 
1.78      anton    10349: @cindex @code{this} usage
                   10350: @cindex @code{m:} usage
                   10351: @cindex @code{;m} usage
                   10352: @example
                   10353: m: ( x y circle -- )
                   10354:   ( x y ) this circle-radius @@ draw-circle ;m
                   10355: @end example
1.6       pazsan   10356: 
1.78      anton    10357: @cindex @code{exit} in @code{m: ... ;m}
                   10358: @cindex @code{exitm} discussion
                   10359: @cindex @code{catch} in @code{m: ... ;m}
                   10360: When this method is executed, the receiver object is removed from the
                   10361: stack; you can access it with @code{this} (admittedly, in this
                   10362: example the use of @code{m: ... ;m} offers no advantage). Note
                   10363: that I specify the stack effect for the whole method (i.e. including
                   10364: the receiver object), not just for the code between @code{m:}
                   10365: and @code{;m}. You cannot use @code{exit} in
                   10366: @code{m:...;m}; instead, use
                   10367: @code{exitm}.@footnote{Moreover, for any word that calls
                   10368: @code{catch} and was defined before loading
                   10369: @code{objects.fs}, you have to redefine it like I redefined
                   10370: @code{catch}: @code{: catch this >r catch r> to-this ;}}
1.12      anton    10371: 
1.78      anton    10372: @cindex @code{inst-var} usage
                   10373: You will frequently use sequences of the form @code{this
                   10374: @emph{field}} (in the example above: @code{this
                   10375: circle-radius}). If you use the field only in this way, you can
                   10376: define it with @code{inst-var} and eliminate the
                   10377: @code{this} before the field name. E.g., the @code{circle}
                   10378: class above could also be defined with:
1.6       pazsan   10379: 
1.78      anton    10380: @example
                   10381: graphical class
                   10382:   cell% inst-var radius
1.6       pazsan   10383: 
1.78      anton    10384: m: ( x y circle -- )
                   10385:   radius @@ draw-circle ;m
                   10386: overrides draw
1.6       pazsan   10387: 
1.78      anton    10388: m: ( n-radius circle -- )
                   10389:   radius ! ;m
                   10390: overrides construct
1.6       pazsan   10391: 
1.78      anton    10392: end-class circle
                   10393: @end example
1.6       pazsan   10394: 
1.78      anton    10395: @code{radius} can only be used in @code{circle} and its
                   10396: descendent classes and inside @code{m:...;m}.
1.6       pazsan   10397: 
1.78      anton    10398: @cindex @code{inst-value} usage
                   10399: You can also define fields with @code{inst-value}, which is
                   10400: to @code{inst-var} what @code{value} is to
                   10401: @code{variable}.  You can change the value of such a field with
                   10402: @code{[to-inst]}.  E.g., we could also define the class
                   10403: @code{circle} like this:
1.44      crook    10404: 
1.78      anton    10405: @example
                   10406: graphical class
                   10407:   inst-value radius
1.6       pazsan   10408: 
1.78      anton    10409: m: ( x y circle -- )
                   10410:   radius draw-circle ;m
                   10411: overrides draw
1.44      crook    10412: 
1.78      anton    10413: m: ( n-radius circle -- )
                   10414:   [to-inst] radius ;m
                   10415: overrides construct
1.6       pazsan   10416: 
1.78      anton    10417: end-class circle
                   10418: @end example
1.6       pazsan   10419: 
1.78      anton    10420: @c !! :m is easy to confuse with m:.  Another name would be better.
1.6       pazsan   10421: 
1.78      anton    10422: @c Finally, you can define named methods with @code{:m}.  One use of this
                   10423: @c feature is the definition of words that occur only in one class and are
                   10424: @c not intended to be overridden, but which still need method context
                   10425: @c (e.g., for accessing @code{inst-var}s).  Another use is for methods that
                   10426: @c would be bound frequently, if defined anonymously.
1.6       pazsan   10427: 
                   10428: 
1.78      anton    10429: @node Classes and Scoping, Dividing classes, Method conveniences, Objects
                   10430: @subsubsection Classes and Scoping
                   10431: @cindex classes and scoping
                   10432: @cindex scoping and classes
1.6       pazsan   10433: 
1.78      anton    10434: Inheritance is frequent, unlike structure extension. This exacerbates
                   10435: the problem with the field name convention (@pxref{Structure Naming
                   10436: Convention}): One always has to remember in which class the field was
                   10437: originally defined; changing a part of the class structure would require
                   10438: changes for renaming in otherwise unaffected code.
1.6       pazsan   10439: 
1.78      anton    10440: @cindex @code{inst-var} visibility
                   10441: @cindex @code{inst-value} visibility
                   10442: To solve this problem, I added a scoping mechanism (which was not in my
                   10443: original charter): A field defined with @code{inst-var} (or
                   10444: @code{inst-value}) is visible only in the class where it is defined and in
                   10445: the descendent classes of this class.  Using such fields only makes
                   10446: sense in @code{m:}-defined methods in these classes anyway.
1.6       pazsan   10447: 
1.78      anton    10448: This scoping mechanism allows us to use the unadorned field name,
                   10449: because name clashes with unrelated words become much less likely.
1.6       pazsan   10450: 
1.78      anton    10451: @cindex @code{protected} discussion
                   10452: @cindex @code{private} discussion
                   10453: Once we have this mechanism, we can also use it for controlling the
                   10454: visibility of other words: All words defined after
                   10455: @code{protected} are visible only in the current class and its
                   10456: descendents. @code{public} restores the compilation
                   10457: (i.e. @code{current}) word list that was in effect before. If you
                   10458: have several @code{protected}s without an intervening
                   10459: @code{public} or @code{set-current}, @code{public}
                   10460: will restore the compilation word list in effect before the first of
                   10461: these @code{protected}s.
1.6       pazsan   10462: 
1.78      anton    10463: @node Dividing classes, Object Interfaces, Classes and Scoping, Objects
                   10464: @subsubsection Dividing classes
                   10465: @cindex Dividing classes
                   10466: @cindex @code{methods}...@code{end-methods}
1.6       pazsan   10467: 
1.78      anton    10468: You may want to do the definition of methods separate from the
                   10469: definition of the class, its selectors, fields, and instance variables,
                   10470: i.e., separate the implementation from the definition.  You can do this
                   10471: in the following way:
1.6       pazsan   10472: 
1.78      anton    10473: @example
                   10474: graphical class
                   10475:   inst-value radius
                   10476: end-class circle
1.6       pazsan   10477: 
1.78      anton    10478: ... \ do some other stuff
1.6       pazsan   10479: 
1.78      anton    10480: circle methods \ now we are ready
1.44      crook    10481: 
1.78      anton    10482: m: ( x y circle -- )
                   10483:   radius draw-circle ;m
                   10484: overrides draw
1.6       pazsan   10485: 
1.78      anton    10486: m: ( n-radius circle -- )
                   10487:   [to-inst] radius ;m
                   10488: overrides construct
1.44      crook    10489: 
1.78      anton    10490: end-methods
                   10491: @end example
1.7       pazsan   10492: 
1.78      anton    10493: You can use several @code{methods}...@code{end-methods} sections.  The
                   10494: only things you can do to the class in these sections are: defining
                   10495: methods, and overriding the class's selectors.  You must not define new
                   10496: selectors or fields.
1.7       pazsan   10497: 
1.78      anton    10498: Note that you often have to override a selector before using it.  In
                   10499: particular, you usually have to override @code{construct} with a new
                   10500: method before you can invoke @code{heap-new} and friends.  E.g., you
                   10501: must not create a circle before the @code{overrides construct} sequence
                   10502: in the example above.
1.7       pazsan   10503: 
1.78      anton    10504: @node Object Interfaces, Objects Implementation, Dividing classes, Objects
                   10505: @subsubsection Object Interfaces
                   10506: @cindex object interfaces
                   10507: @cindex interfaces for objects
1.7       pazsan   10508: 
1.78      anton    10509: In this model you can only call selectors defined in the class of the
                   10510: receiving objects or in one of its ancestors. If you call a selector
                   10511: with a receiving object that is not in one of these classes, the
                   10512: result is undefined; if you are lucky, the program crashes
                   10513: immediately.
1.7       pazsan   10514: 
1.78      anton    10515: @cindex selectors common to hardly-related classes
                   10516: Now consider the case when you want to have a selector (or several)
                   10517: available in two classes: You would have to add the selector to a
                   10518: common ancestor class, in the worst case to @code{object}. You
                   10519: may not want to do this, e.g., because someone else is responsible for
                   10520: this ancestor class.
1.7       pazsan   10521: 
1.78      anton    10522: The solution for this problem is interfaces. An interface is a
                   10523: collection of selectors. If a class implements an interface, the
                   10524: selectors become available to the class and its descendents. A class
                   10525: can implement an unlimited number of interfaces. For the problem
                   10526: discussed above, we would define an interface for the selector(s), and
                   10527: both classes would implement the interface.
1.7       pazsan   10528: 
1.78      anton    10529: As an example, consider an interface @code{storage} for
                   10530: writing objects to disk and getting them back, and a class
                   10531: @code{foo} that implements it. The code would look like this:
1.7       pazsan   10532: 
1.78      anton    10533: @cindex @code{interface} usage
                   10534: @cindex @code{end-interface} usage
                   10535: @cindex @code{implementation} usage
                   10536: @example
                   10537: interface
                   10538:   selector write ( file object -- )
                   10539:   selector read1 ( file object -- )
                   10540: end-interface storage
1.13      pazsan   10541: 
1.78      anton    10542: bar class
                   10543:   storage implementation
1.13      pazsan   10544: 
1.78      anton    10545: ... overrides write
                   10546: ... overrides read1
                   10547: ...
                   10548: end-class foo
                   10549: @end example
1.13      pazsan   10550: 
1.78      anton    10551: @noindent
                   10552: (I would add a word @code{read} @i{( file -- object )} that uses
                   10553: @code{read1} internally, but that's beyond the point illustrated
                   10554: here.)
1.13      pazsan   10555: 
1.78      anton    10556: Note that you cannot use @code{protected} in an interface; and
                   10557: of course you cannot define fields.
1.13      pazsan   10558: 
1.78      anton    10559: In the Neon model, all selectors are available for all classes;
                   10560: therefore it does not need interfaces. The price you pay in this model
                   10561: is slower late binding, and therefore, added complexity to avoid late
                   10562: binding.
1.13      pazsan   10563: 
1.78      anton    10564: @node Objects Implementation, Objects Glossary, Object Interfaces, Objects
                   10565: @subsubsection @file{objects.fs} Implementation
                   10566: @cindex @file{objects.fs} implementation
1.13      pazsan   10567: 
1.78      anton    10568: @cindex @code{object-map} discussion
                   10569: An object is a piece of memory, like one of the data structures
                   10570: described with @code{struct...end-struct}. It has a field
                   10571: @code{object-map} that points to the method map for the object's
                   10572: class.
1.13      pazsan   10573: 
1.78      anton    10574: @cindex method map
                   10575: @cindex virtual function table
                   10576: The @emph{method map}@footnote{This is Self terminology; in C++
                   10577: terminology: virtual function table.} is an array that contains the
                   10578: execution tokens (@i{xt}s) of the methods for the object's class. Each
                   10579: selector contains an offset into a method map.
1.13      pazsan   10580: 
1.78      anton    10581: @cindex @code{selector} implementation, class
                   10582: @code{selector} is a defining word that uses
                   10583: @code{CREATE} and @code{DOES>}. The body of the
                   10584: selector contains the offset; the @code{DOES>} action for a
                   10585: class selector is, basically:
1.8       pazsan   10586: 
                   10587: @example
1.78      anton    10588: ( object addr ) @@ over object-map @@ + @@ execute
1.13      pazsan   10589: @end example
                   10590: 
1.78      anton    10591: Since @code{object-map} is the first field of the object, it
                   10592: does not generate any code. As you can see, calling a selector has a
                   10593: small, constant cost.
1.26      crook    10594: 
1.78      anton    10595: @cindex @code{current-interface} discussion
                   10596: @cindex class implementation and representation
                   10597: A class is basically a @code{struct} combined with a method
                   10598: map. During the class definition the alignment and size of the class
                   10599: are passed on the stack, just as with @code{struct}s, so
                   10600: @code{field} can also be used for defining class
                   10601: fields. However, passing more items on the stack would be
                   10602: inconvenient, so @code{class} builds a data structure in memory,
                   10603: which is accessed through the variable
                   10604: @code{current-interface}. After its definition is complete, the
                   10605: class is represented on the stack by a pointer (e.g., as parameter for
                   10606: a child class definition).
1.26      crook    10607: 
1.78      anton    10608: A new class starts off with the alignment and size of its parent,
                   10609: and a copy of the parent's method map. Defining new fields extends the
                   10610: size and alignment; likewise, defining new selectors extends the
                   10611: method map. @code{overrides} just stores a new @i{xt} in the method
                   10612: map at the offset given by the selector.
1.13      pazsan   10613: 
1.78      anton    10614: @cindex class binding, implementation
                   10615: Class binding just gets the @i{xt} at the offset given by the selector
                   10616: from the class's method map and @code{compile,}s (in the case of
                   10617: @code{[bind]}) it.
1.13      pazsan   10618: 
1.78      anton    10619: @cindex @code{this} implementation
                   10620: @cindex @code{catch} and @code{this}
                   10621: @cindex @code{this} and @code{catch}
                   10622: I implemented @code{this} as a @code{value}. At the
                   10623: start of an @code{m:...;m} method the old @code{this} is
                   10624: stored to the return stack and restored at the end; and the object on
                   10625: the TOS is stored @code{TO this}. This technique has one
                   10626: disadvantage: If the user does not leave the method via
                   10627: @code{;m}, but via @code{throw} or @code{exit},
                   10628: @code{this} is not restored (and @code{exit} may
                   10629: crash). To deal with the @code{throw} problem, I have redefined
                   10630: @code{catch} to save and restore @code{this}; the same
                   10631: should be done with any word that can catch an exception. As for
                   10632: @code{exit}, I simply forbid it (as a replacement, there is
                   10633: @code{exitm}).
1.13      pazsan   10634: 
1.78      anton    10635: @cindex @code{inst-var} implementation
                   10636: @code{inst-var} is just the same as @code{field}, with
                   10637: a different @code{DOES>} action:
1.13      pazsan   10638: @example
1.78      anton    10639: @@ this +
1.8       pazsan   10640: @end example
1.78      anton    10641: Similar for @code{inst-value}.
1.8       pazsan   10642: 
1.78      anton    10643: @cindex class scoping implementation
                   10644: Each class also has a word list that contains the words defined with
                   10645: @code{inst-var} and @code{inst-value}, and its protected
                   10646: words. It also has a pointer to its parent. @code{class} pushes
                   10647: the word lists of the class and all its ancestors onto the search order stack,
                   10648: and @code{end-class} drops them.
1.20      pazsan   10649: 
1.78      anton    10650: @cindex interface implementation
                   10651: An interface is like a class without fields, parent and protected
                   10652: words; i.e., it just has a method map. If a class implements an
                   10653: interface, its method map contains a pointer to the method map of the
                   10654: interface. The positive offsets in the map are reserved for class
                   10655: methods, therefore interface map pointers have negative
                   10656: offsets. Interfaces have offsets that are unique throughout the
                   10657: system, unlike class selectors, whose offsets are only unique for the
                   10658: classes where the selector is available (invokable).
1.20      pazsan   10659: 
1.78      anton    10660: This structure means that interface selectors have to perform one
                   10661: indirection more than class selectors to find their method. Their body
                   10662: contains the interface map pointer offset in the class method map, and
                   10663: the method offset in the interface method map. The
                   10664: @code{does>} action for an interface selector is, basically:
1.20      pazsan   10665: 
                   10666: @example
1.78      anton    10667: ( object selector-body )
                   10668: 2dup selector-interface @@ ( object selector-body object interface-offset )
                   10669: swap object-map @@ + @@ ( object selector-body map )
                   10670: swap selector-offset @@ + @@ execute
1.20      pazsan   10671: @end example
                   10672: 
1.78      anton    10673: where @code{object-map} and @code{selector-offset} are
                   10674: first fields and generate no code.
1.20      pazsan   10675: 
1.78      anton    10676: As a concrete example, consider the following code:
1.20      pazsan   10677: 
                   10678: @example
1.78      anton    10679: interface
                   10680:   selector if1sel1
                   10681:   selector if1sel2
                   10682: end-interface if1
1.20      pazsan   10683: 
1.78      anton    10684: object class
                   10685:   if1 implementation
                   10686:   selector cl1sel1
                   10687:   cell% inst-var cl1iv1
1.20      pazsan   10688: 
1.78      anton    10689: ' m1 overrides construct
                   10690: ' m2 overrides if1sel1
                   10691: ' m3 overrides if1sel2
                   10692: ' m4 overrides cl1sel2
                   10693: end-class cl1
1.20      pazsan   10694: 
1.78      anton    10695: create obj1 object dict-new drop
                   10696: create obj2 cl1    dict-new drop
                   10697: @end example
1.20      pazsan   10698: 
1.78      anton    10699: The data structure created by this code (including the data structure
                   10700: for @code{object}) is shown in the
                   10701: @uref{objects-implementation.eps,figure}, assuming a cell size of 4.
                   10702: @comment TODO add this diagram..
1.20      pazsan   10703: 
1.78      anton    10704: @node Objects Glossary,  , Objects Implementation, Objects
                   10705: @subsubsection @file{objects.fs} Glossary
                   10706: @cindex @file{objects.fs} Glossary
1.20      pazsan   10707: 
                   10708: 
1.78      anton    10709: doc---objects-bind
                   10710: doc---objects-<bind>
                   10711: doc---objects-bind'
                   10712: doc---objects-[bind]
                   10713: doc---objects-class
                   10714: doc---objects-class->map
                   10715: doc---objects-class-inst-size
                   10716: doc---objects-class-override!
1.79      anton    10717: doc---objects-class-previous
                   10718: doc---objects-class>order
1.78      anton    10719: doc---objects-construct
                   10720: doc---objects-current'
                   10721: doc---objects-[current]
                   10722: doc---objects-current-interface
                   10723: doc---objects-dict-new
                   10724: doc---objects-end-class
                   10725: doc---objects-end-class-noname
                   10726: doc---objects-end-interface
                   10727: doc---objects-end-interface-noname
                   10728: doc---objects-end-methods
                   10729: doc---objects-exitm
                   10730: doc---objects-heap-new
                   10731: doc---objects-implementation
                   10732: doc---objects-init-object
                   10733: doc---objects-inst-value
                   10734: doc---objects-inst-var
                   10735: doc---objects-interface
                   10736: doc---objects-m:
                   10737: doc---objects-:m
                   10738: doc---objects-;m
                   10739: doc---objects-method
                   10740: doc---objects-methods
                   10741: doc---objects-object
                   10742: doc---objects-overrides
                   10743: doc---objects-[parent]
                   10744: doc---objects-print
                   10745: doc---objects-protected
                   10746: doc---objects-public
                   10747: doc---objects-selector
                   10748: doc---objects-this
                   10749: doc---objects-<to-inst>
                   10750: doc---objects-[to-inst]
                   10751: doc---objects-to-this
                   10752: doc---objects-xt-new
1.20      pazsan   10753: 
                   10754: 
1.78      anton    10755: @c -------------------------------------------------------------
                   10756: @node OOF, Mini-OOF, Objects, Object-oriented Forth
                   10757: @subsection The @file{oof.fs} model
                   10758: @cindex oof
                   10759: @cindex object-oriented programming
1.20      pazsan   10760: 
1.78      anton    10761: @cindex @file{objects.fs}
                   10762: @cindex @file{oof.fs}
1.20      pazsan   10763: 
1.78      anton    10764: This section describes the @file{oof.fs} package.
1.20      pazsan   10765: 
1.78      anton    10766: The package described in this section has been used in bigFORTH since 1991, and
                   10767: used for two large applications: a chromatographic system used to
                   10768: create new medicaments, and a graphic user interface library (MINOS).
1.20      pazsan   10769: 
1.78      anton    10770: You can find a description (in German) of @file{oof.fs} in @cite{Object
                   10771: oriented bigFORTH} by Bernd Paysan, published in @cite{Vierte Dimension}
                   10772: 10(2), 1994.
1.20      pazsan   10773: 
1.78      anton    10774: @menu
                   10775: * Properties of the OOF model::  
                   10776: * Basic OOF Usage::             
                   10777: * The OOF base class::          
                   10778: * Class Declaration::           
                   10779: * Class Implementation::        
                   10780: @end menu
1.20      pazsan   10781: 
1.78      anton    10782: @node Properties of the OOF model, Basic OOF Usage, OOF, OOF
                   10783: @subsubsection Properties of the @file{oof.fs} model
                   10784: @cindex @file{oof.fs} properties
1.20      pazsan   10785: 
1.78      anton    10786: @itemize @bullet
                   10787: @item
                   10788: This model combines object oriented programming with information
                   10789: hiding. It helps you writing large application, where scoping is
                   10790: necessary, because it provides class-oriented scoping.
1.20      pazsan   10791: 
1.78      anton    10792: @item
                   10793: Named objects, object pointers, and object arrays can be created,
                   10794: selector invocation uses the ``object selector'' syntax. Selector invocation
                   10795: to objects and/or selectors on the stack is a bit less convenient, but
                   10796: possible.
1.44      crook    10797: 
1.78      anton    10798: @item
                   10799: Selector invocation and instance variable usage of the active object is
                   10800: straightforward, since both make use of the active object.
1.44      crook    10801: 
1.78      anton    10802: @item
                   10803: Late binding is efficient and easy to use.
1.20      pazsan   10804: 
1.78      anton    10805: @item
                   10806: State-smart objects parse selectors. However, extensibility is provided
                   10807: using a (parsing) selector @code{postpone} and a selector @code{'}.
1.20      pazsan   10808: 
1.78      anton    10809: @item
                   10810: An implementation in ANS Forth is available.
1.20      pazsan   10811: 
1.78      anton    10812: @end itemize
1.23      crook    10813: 
                   10814: 
1.78      anton    10815: @node Basic OOF Usage, The OOF base class, Properties of the OOF model, OOF
                   10816: @subsubsection Basic @file{oof.fs} Usage
                   10817: @cindex @file{oof.fs} usage
1.23      crook    10818: 
1.78      anton    10819: This section uses the same example as for @code{objects} (@pxref{Basic Objects Usage}).
1.23      crook    10820: 
1.78      anton    10821: You can define a class for graphical objects like this:
1.23      crook    10822: 
1.78      anton    10823: @cindex @code{class} usage
                   10824: @cindex @code{class;} usage
                   10825: @cindex @code{method} usage
                   10826: @example
                   10827: object class graphical \ "object" is the parent class
                   10828:   method draw ( x y graphical -- )
                   10829: class;
                   10830: @end example
1.23      crook    10831: 
1.78      anton    10832: This code defines a class @code{graphical} with an
                   10833: operation @code{draw}.  We can perform the operation
                   10834: @code{draw} on any @code{graphical} object, e.g.:
1.23      crook    10835: 
1.78      anton    10836: @example
                   10837: 100 100 t-rex draw
                   10838: @end example
1.23      crook    10839: 
1.78      anton    10840: @noindent
                   10841: where @code{t-rex} is an object or object pointer, created with e.g.
                   10842: @code{graphical : t-rex}.
1.23      crook    10843: 
1.78      anton    10844: @cindex abstract class
                   10845: How do we create a graphical object? With the present definitions,
                   10846: we cannot create a useful graphical object. The class
                   10847: @code{graphical} describes graphical objects in general, but not
                   10848: any concrete graphical object type (C++ users would call it an
                   10849: @emph{abstract class}); e.g., there is no method for the selector
                   10850: @code{draw} in the class @code{graphical}.
1.23      crook    10851: 
1.78      anton    10852: For concrete graphical objects, we define child classes of the
                   10853: class @code{graphical}, e.g.:
1.23      crook    10854: 
1.78      anton    10855: @example
                   10856: graphical class circle \ "graphical" is the parent class
                   10857:   cell var circle-radius
                   10858: how:
                   10859:   : draw ( x y -- )
                   10860:     circle-radius @@ draw-circle ;
1.23      crook    10861: 
1.78      anton    10862:   : init ( n-radius -- (
                   10863:     circle-radius ! ;
                   10864: class;
                   10865: @end example
1.1       anton    10866: 
1.78      anton    10867: Here we define a class @code{circle} as a child of @code{graphical},
                   10868: with a field @code{circle-radius}; it defines new methods for the
                   10869: selectors @code{draw} and @code{init} (@code{init} is defined in
                   10870: @code{object}, the parent class of @code{graphical}).
1.1       anton    10871: 
1.78      anton    10872: Now we can create a circle in the dictionary with:
1.1       anton    10873: 
1.78      anton    10874: @example
                   10875: 50 circle : my-circle
                   10876: @end example
1.21      crook    10877: 
1.78      anton    10878: @noindent
                   10879: @code{:} invokes @code{init}, thus initializing the field
                   10880: @code{circle-radius} with 50. We can draw this new circle at (100,100)
                   10881: with:
1.1       anton    10882: 
1.78      anton    10883: @example
                   10884: 100 100 my-circle draw
                   10885: @end example
1.1       anton    10886: 
1.78      anton    10887: @cindex selector invocation, restrictions
                   10888: @cindex class definition, restrictions
                   10889: Note: You can only invoke a selector if the receiving object belongs to
                   10890: the class where the selector was defined or one of its descendents;
                   10891: e.g., you can invoke @code{draw} only for objects belonging to
                   10892: @code{graphical} or its descendents (e.g., @code{circle}). The scoping
                   10893: mechanism will check if you try to invoke a selector that is not
                   10894: defined in this class hierarchy, so you'll get an error at compilation
                   10895: time.
1.1       anton    10896: 
                   10897: 
1.78      anton    10898: @node The OOF base class, Class Declaration, Basic OOF Usage, OOF
                   10899: @subsubsection The @file{oof.fs} base class
                   10900: @cindex @file{oof.fs} base class
1.1       anton    10901: 
1.78      anton    10902: When you define a class, you have to specify a parent class.  So how do
                   10903: you start defining classes? There is one class available from the start:
                   10904: @code{object}. You have to use it as ancestor for all classes. It is the
                   10905: only class that has no parent. Classes are also objects, except that
                   10906: they don't have instance variables; class manipulation such as
                   10907: inheritance or changing definitions of a class is handled through
                   10908: selectors of the class @code{object}.
1.1       anton    10909: 
1.78      anton    10910: @code{object} provides a number of selectors:
1.1       anton    10911: 
1.78      anton    10912: @itemize @bullet
                   10913: @item
                   10914: @code{class} for subclassing, @code{definitions} to add definitions
                   10915: later on, and @code{class?} to get type informations (is the class a
                   10916: subclass of the class passed on the stack?).
1.1       anton    10917: 
1.78      anton    10918: doc---object-class
                   10919: doc---object-definitions
                   10920: doc---object-class?
1.1       anton    10921: 
                   10922: 
1.26      crook    10923: @item
1.78      anton    10924: @code{init} and @code{dispose} as constructor and destructor of the
                   10925: object. @code{init} is invocated after the object's memory is allocated,
                   10926: while @code{dispose} also handles deallocation. Thus if you redefine
                   10927: @code{dispose}, you have to call the parent's dispose with @code{super
                   10928: dispose}, too.
                   10929: 
                   10930: doc---object-init
                   10931: doc---object-dispose
                   10932: 
1.1       anton    10933: 
1.26      crook    10934: @item
1.78      anton    10935: @code{new}, @code{new[]}, @code{:}, @code{ptr}, @code{asptr}, and
                   10936: @code{[]} to create named and unnamed objects and object arrays or
                   10937: object pointers.
                   10938: 
                   10939: doc---object-new
                   10940: doc---object-new[]
                   10941: doc---object-:
                   10942: doc---object-ptr
                   10943: doc---object-asptr
                   10944: doc---object-[]
                   10945: 
1.1       anton    10946: 
1.26      crook    10947: @item
1.78      anton    10948: @code{::} and @code{super} for explicit scoping. You should use explicit
                   10949: scoping only for super classes or classes with the same set of instance
                   10950: variables. Explicitly-scoped selectors use early binding.
1.21      crook    10951: 
1.78      anton    10952: doc---object-::
                   10953: doc---object-super
1.21      crook    10954: 
                   10955: 
1.26      crook    10956: @item
1.78      anton    10957: @code{self} to get the address of the object
1.21      crook    10958: 
1.78      anton    10959: doc---object-self
1.21      crook    10960: 
                   10961: 
1.78      anton    10962: @item
                   10963: @code{bind}, @code{bound}, @code{link}, and @code{is} to assign object
                   10964: pointers and instance defers.
1.21      crook    10965: 
1.78      anton    10966: doc---object-bind
                   10967: doc---object-bound
                   10968: doc---object-link
                   10969: doc---object-is
1.21      crook    10970: 
                   10971: 
1.78      anton    10972: @item
                   10973: @code{'} to obtain selector tokens, @code{send} to invocate selectors
                   10974: form the stack, and @code{postpone} to generate selector invocation code.
1.21      crook    10975: 
1.78      anton    10976: doc---object-'
                   10977: doc---object-postpone
1.21      crook    10978: 
                   10979: 
1.78      anton    10980: @item
                   10981: @code{with} and @code{endwith} to select the active object from the
                   10982: stack, and enable its scope. Using @code{with} and @code{endwith}
                   10983: also allows you to create code using selector @code{postpone} without being
                   10984: trapped by the state-smart objects.
1.21      crook    10985: 
1.78      anton    10986: doc---object-with
                   10987: doc---object-endwith
1.21      crook    10988: 
                   10989: 
1.78      anton    10990: @end itemize
1.21      crook    10991: 
1.78      anton    10992: @node Class Declaration, Class Implementation, The OOF base class, OOF
                   10993: @subsubsection Class Declaration
                   10994: @cindex class declaration
1.21      crook    10995: 
1.78      anton    10996: @itemize @bullet
                   10997: @item
                   10998: Instance variables
1.21      crook    10999: 
1.78      anton    11000: doc---oof-var
1.21      crook    11001: 
                   11002: 
1.78      anton    11003: @item
                   11004: Object pointers
1.21      crook    11005: 
1.78      anton    11006: doc---oof-ptr
                   11007: doc---oof-asptr
1.21      crook    11008: 
                   11009: 
1.78      anton    11010: @item
                   11011: Instance defers
1.21      crook    11012: 
1.78      anton    11013: doc---oof-defer
1.21      crook    11014: 
                   11015: 
1.78      anton    11016: @item
                   11017: Method selectors
1.21      crook    11018: 
1.78      anton    11019: doc---oof-early
                   11020: doc---oof-method
1.21      crook    11021: 
                   11022: 
1.78      anton    11023: @item
                   11024: Class-wide variables
1.21      crook    11025: 
1.78      anton    11026: doc---oof-static
1.21      crook    11027: 
                   11028: 
1.78      anton    11029: @item
                   11030: End declaration
1.1       anton    11031: 
1.78      anton    11032: doc---oof-how:
                   11033: doc---oof-class;
1.21      crook    11034: 
                   11035: 
1.78      anton    11036: @end itemize
1.21      crook    11037: 
1.78      anton    11038: @c -------------------------------------------------------------
                   11039: @node Class Implementation,  , Class Declaration, OOF
                   11040: @subsubsection Class Implementation
                   11041: @cindex class implementation
1.21      crook    11042: 
1.78      anton    11043: @c -------------------------------------------------------------
                   11044: @node Mini-OOF, Comparison with other object models, OOF, Object-oriented Forth
                   11045: @subsection The @file{mini-oof.fs} model
                   11046: @cindex mini-oof
1.21      crook    11047: 
1.78      anton    11048: Gforth's third object oriented Forth package is a 12-liner. It uses a
1.79      anton    11049: mixture of the @file{objects.fs} and the @file{oof.fs} syntax,
1.78      anton    11050: and reduces to the bare minimum of features. This is based on a posting
                   11051: of Bernd Paysan in comp.lang.forth.
1.21      crook    11052: 
1.78      anton    11053: @menu
                   11054: * Basic Mini-OOF Usage::        
                   11055: * Mini-OOF Example::            
                   11056: * Mini-OOF Implementation::     
                   11057: @end menu
1.21      crook    11058: 
1.78      anton    11059: @c -------------------------------------------------------------
                   11060: @node Basic Mini-OOF Usage, Mini-OOF Example, Mini-OOF, Mini-OOF
                   11061: @subsubsection Basic @file{mini-oof.fs} Usage
                   11062: @cindex mini-oof usage
1.21      crook    11063: 
1.78      anton    11064: There is a base class (@code{class}, which allocates one cell for the
                   11065: object pointer) plus seven other words: to define a method, a variable,
                   11066: a class; to end a class, to resolve binding, to allocate an object and
                   11067: to compile a class method.
                   11068: @comment TODO better description of the last one
1.26      crook    11069: 
1.21      crook    11070: 
1.78      anton    11071: doc-object
                   11072: doc-method
                   11073: doc-var
                   11074: doc-class
                   11075: doc-end-class
                   11076: doc-defines
                   11077: doc-new
                   11078: doc-::
1.21      crook    11079: 
                   11080: 
                   11081: 
1.78      anton    11082: @c -------------------------------------------------------------
                   11083: @node Mini-OOF Example, Mini-OOF Implementation, Basic Mini-OOF Usage, Mini-OOF
                   11084: @subsubsection Mini-OOF Example
                   11085: @cindex mini-oof example
1.1       anton    11086: 
1.78      anton    11087: A short example shows how to use this package. This example, in slightly
                   11088: extended form, is supplied as @file{moof-exm.fs}
                   11089: @comment TODO could flesh this out with some comments from the Forthwrite article
1.20      pazsan   11090: 
1.26      crook    11091: @example
1.78      anton    11092: object class
                   11093:   method init
                   11094:   method draw
                   11095: end-class graphical
1.26      crook    11096: @end example
1.20      pazsan   11097: 
1.78      anton    11098: This code defines a class @code{graphical} with an
                   11099: operation @code{draw}.  We can perform the operation
                   11100: @code{draw} on any @code{graphical} object, e.g.:
1.20      pazsan   11101: 
1.26      crook    11102: @example
1.78      anton    11103: 100 100 t-rex draw
1.26      crook    11104: @end example
1.12      anton    11105: 
1.78      anton    11106: where @code{t-rex} is an object or object pointer, created with e.g.
                   11107: @code{graphical new Constant t-rex}.
1.12      anton    11108: 
1.78      anton    11109: For concrete graphical objects, we define child classes of the
                   11110: class @code{graphical}, e.g.:
1.12      anton    11111: 
1.26      crook    11112: @example
                   11113: graphical class
1.78      anton    11114:   cell var circle-radius
                   11115: end-class circle \ "graphical" is the parent class
1.12      anton    11116: 
1.78      anton    11117: :noname ( x y -- )
                   11118:   circle-radius @@ draw-circle ; circle defines draw
                   11119: :noname ( r -- )
                   11120:   circle-radius ! ; circle defines init
                   11121: @end example
1.12      anton    11122: 
1.78      anton    11123: There is no implicit init method, so we have to define one. The creation
                   11124: code of the object now has to call init explicitely.
1.21      crook    11125: 
1.78      anton    11126: @example
                   11127: circle new Constant my-circle
                   11128: 50 my-circle init
1.12      anton    11129: @end example
                   11130: 
1.78      anton    11131: It is also possible to add a function to create named objects with
                   11132: automatic call of @code{init}, given that all objects have @code{init}
                   11133: on the same place:
1.38      anton    11134: 
1.78      anton    11135: @example
                   11136: : new: ( .. o "name" -- )
                   11137:     new dup Constant init ;
                   11138: 80 circle new: large-circle
                   11139: @end example
1.12      anton    11140: 
1.78      anton    11141: We can draw this new circle at (100,100) with:
1.12      anton    11142: 
1.78      anton    11143: @example
                   11144: 100 100 my-circle draw
                   11145: @end example
1.12      anton    11146: 
1.78      anton    11147: @node Mini-OOF Implementation,  , Mini-OOF Example, Mini-OOF
                   11148: @subsubsection @file{mini-oof.fs} Implementation
1.12      anton    11149: 
1.78      anton    11150: Object-oriented systems with late binding typically use a
                   11151: ``vtable''-approach: the first variable in each object is a pointer to a
                   11152: table, which contains the methods as function pointers. The vtable
                   11153: may also contain other information.
1.12      anton    11154: 
1.79      anton    11155: So first, let's declare selectors:
1.37      anton    11156: 
                   11157: @example
1.79      anton    11158: : method ( m v "name" -- m' v ) Create  over , swap cell+ swap
1.78      anton    11159:   DOES> ( ... o -- ... ) @@ over @@ + @@ execute ;
                   11160: @end example
1.37      anton    11161: 
1.79      anton    11162: During selector declaration, the number of selectors and instance
                   11163: variables is on the stack (in address units). @code{method} creates one
                   11164: selector and increments the selector number. To execute a selector, it
1.78      anton    11165: takes the object, fetches the vtable pointer, adds the offset, and
1.79      anton    11166: executes the method @i{xt} stored there. Each selector takes the object
                   11167: it is invoked with as top of stack parameter; it passes the parameters
                   11168: (including the object) unchanged to the appropriate method which should
1.78      anton    11169: consume that object.
1.37      anton    11170: 
1.78      anton    11171: Now, we also have to declare instance variables
1.37      anton    11172: 
1.78      anton    11173: @example
1.79      anton    11174: : var ( m v size "name" -- m v' ) Create  over , +
1.78      anton    11175:   DOES> ( o -- addr ) @@ + ;
1.37      anton    11176: @end example
                   11177: 
1.78      anton    11178: As before, a word is created with the current offset. Instance
                   11179: variables can have different sizes (cells, floats, doubles, chars), so
                   11180: all we do is take the size and add it to the offset. If your machine
                   11181: has alignment restrictions, put the proper @code{aligned} or
                   11182: @code{faligned} before the variable, to adjust the variable
                   11183: offset. That's why it is on the top of stack.
1.37      anton    11184: 
1.78      anton    11185: We need a starting point (the base object) and some syntactic sugar:
1.37      anton    11186: 
1.78      anton    11187: @example
                   11188: Create object  1 cells , 2 cells ,
1.79      anton    11189: : class ( class -- class selectors vars ) dup 2@@ ;
1.78      anton    11190: @end example
1.12      anton    11191: 
1.78      anton    11192: For inheritance, the vtable of the parent object has to be
                   11193: copied when a new, derived class is declared. This gives all the
                   11194: methods of the parent class, which can be overridden, though.
1.12      anton    11195: 
1.78      anton    11196: @example
1.79      anton    11197: : end-class  ( class selectors vars "name" -- )
1.78      anton    11198:   Create  here >r , dup , 2 cells ?DO ['] noop , 1 cells +LOOP
                   11199:   cell+ dup cell+ r> rot @@ 2 cells /string move ;
                   11200: @end example
1.12      anton    11201: 
1.78      anton    11202: The first line creates the vtable, initialized with
                   11203: @code{noop}s. The second line is the inheritance mechanism, it
                   11204: copies the xts from the parent vtable.
1.12      anton    11205: 
1.78      anton    11206: We still have no way to define new methods, let's do that now:
1.12      anton    11207: 
1.26      crook    11208: @example
1.79      anton    11209: : defines ( xt class "name" -- ) ' >body @@ + ! ;
1.78      anton    11210: @end example
1.12      anton    11211: 
1.78      anton    11212: To allocate a new object, we need a word, too:
1.12      anton    11213: 
1.78      anton    11214: @example
                   11215: : new ( class -- o )  here over @@ allot swap over ! ;
1.12      anton    11216: @end example
                   11217: 
1.78      anton    11218: Sometimes derived classes want to access the method of the
                   11219: parent object. There are two ways to achieve this with Mini-OOF:
                   11220: first, you could use named words, and second, you could look up the
                   11221: vtable of the parent object.
1.12      anton    11222: 
1.78      anton    11223: @example
                   11224: : :: ( class "name" -- ) ' >body @@ + @@ compile, ;
                   11225: @end example
1.12      anton    11226: 
                   11227: 
1.78      anton    11228: Nothing can be more confusing than a good example, so here is
                   11229: one. First let's declare a text object (called
                   11230: @code{button}), that stores text and position:
1.12      anton    11231: 
1.78      anton    11232: @example
                   11233: object class
                   11234:   cell var text
                   11235:   cell var len
                   11236:   cell var x
                   11237:   cell var y
                   11238:   method init
                   11239:   method draw
                   11240: end-class button
                   11241: @end example
1.12      anton    11242: 
1.78      anton    11243: @noindent
                   11244: Now, implement the two methods, @code{draw} and @code{init}:
1.21      crook    11245: 
1.26      crook    11246: @example
1.78      anton    11247: :noname ( o -- )
                   11248:  >r r@@ x @@ r@@ y @@ at-xy  r@@ text @@ r> len @@ type ;
                   11249:  button defines draw
                   11250: :noname ( addr u o -- )
                   11251:  >r 0 r@@ x ! 0 r@@ y ! r@@ len ! r> text ! ;
                   11252:  button defines init
1.26      crook    11253: @end example
1.12      anton    11254: 
1.78      anton    11255: @noindent
                   11256: To demonstrate inheritance, we define a class @code{bold-button}, with no
1.79      anton    11257: new data and no new selectors:
1.78      anton    11258: 
                   11259: @example
                   11260: button class
                   11261: end-class bold-button
1.12      anton    11262: 
1.78      anton    11263: : bold   27 emit ." [1m" ;
                   11264: : normal 27 emit ." [0m" ;
                   11265: @end example
1.1       anton    11266: 
1.78      anton    11267: @noindent
                   11268: The class @code{bold-button} has a different draw method to
                   11269: @code{button}, but the new method is defined in terms of the draw method
                   11270: for @code{button}:
1.20      pazsan   11271: 
1.78      anton    11272: @example
                   11273: :noname bold [ button :: draw ] normal ; bold-button defines draw
                   11274: @end example
1.21      crook    11275: 
1.78      anton    11276: @noindent
1.79      anton    11277: Finally, create two objects and apply selectors:
1.21      crook    11278: 
1.26      crook    11279: @example
1.78      anton    11280: button new Constant foo
                   11281: s" thin foo" foo init
                   11282: page
                   11283: foo draw
                   11284: bold-button new Constant bar
                   11285: s" fat bar" bar init
                   11286: 1 bar y !
                   11287: bar draw
1.26      crook    11288: @end example
1.21      crook    11289: 
                   11290: 
1.78      anton    11291: @node Comparison with other object models,  , Mini-OOF, Object-oriented Forth
                   11292: @subsection Comparison with other object models
                   11293: @cindex comparison of object models
                   11294: @cindex object models, comparison
                   11295: 
                   11296: Many object-oriented Forth extensions have been proposed (@cite{A survey
                   11297: of object-oriented Forths} (SIGPLAN Notices, April 1996) by Bradford
                   11298: J. Rodriguez and W. F. S. Poehlman lists 17). This section discusses the
                   11299: relation of the object models described here to two well-known and two
                   11300: closely-related (by the use of method maps) models.  Andras Zsoter
                   11301: helped us with this section.
                   11302: 
                   11303: @cindex Neon model
                   11304: The most popular model currently seems to be the Neon model (see
                   11305: @cite{Object-oriented programming in ANS Forth} (Forth Dimensions, March
                   11306: 1997) by Andrew McKewan) but this model has a number of limitations
                   11307: @footnote{A longer version of this critique can be
                   11308: found in @cite{On Standardizing Object-Oriented Forth Extensions} (Forth
                   11309: Dimensions, May 1997) by Anton Ertl.}:
                   11310: 
                   11311: @itemize @bullet
                   11312: @item
                   11313: It uses a @code{@emph{selector object}} syntax, which makes it unnatural
                   11314: to pass objects on the stack.
1.21      crook    11315: 
1.78      anton    11316: @item
                   11317: It requires that the selector parses the input stream (at
1.79      anton    11318: compile time); this leads to reduced extensibility and to bugs that are
1.78      anton    11319: hard to find.
1.21      crook    11320: 
1.78      anton    11321: @item
1.79      anton    11322: It allows using every selector on every object; this eliminates the
                   11323: need for interfaces, but makes it harder to create efficient
                   11324: implementations.
1.78      anton    11325: @end itemize
1.21      crook    11326: 
1.78      anton    11327: @cindex Pountain's object-oriented model
                   11328: Another well-known publication is @cite{Object-Oriented Forth} (Academic
                   11329: Press, London, 1987) by Dick Pountain. However, it is not really about
                   11330: object-oriented programming, because it hardly deals with late
                   11331: binding. Instead, it focuses on features like information hiding and
                   11332: overloading that are characteristic of modular languages like Ada (83).
1.26      crook    11333: 
1.78      anton    11334: @cindex Zsoter's object-oriented model
1.79      anton    11335: In @uref{http://www.forth.org/oopf.html, Does late binding have to be
                   11336: slow?} (Forth Dimensions 18(1) 1996, pages 31-35) Andras Zsoter
                   11337: describes a model that makes heavy use of an active object (like
                   11338: @code{this} in @file{objects.fs}): The active object is not only used
                   11339: for accessing all fields, but also specifies the receiving object of
                   11340: every selector invocation; you have to change the active object
                   11341: explicitly with @code{@{ ... @}}, whereas in @file{objects.fs} it
                   11342: changes more or less implicitly at @code{m: ... ;m}. Such a change at
                   11343: the method entry point is unnecessary with Zsoter's model, because the
                   11344: receiving object is the active object already. On the other hand, the
                   11345: explicit change is absolutely necessary in that model, because otherwise
                   11346: no one could ever change the active object. An ANS Forth implementation
                   11347: of this model is available through
                   11348: @uref{http://www.forth.org/oopf.html}.
1.21      crook    11349: 
1.78      anton    11350: @cindex @file{oof.fs}, differences to other models
                   11351: The @file{oof.fs} model combines information hiding and overloading
                   11352: resolution (by keeping names in various word lists) with object-oriented
                   11353: programming. It sets the active object implicitly on method entry, but
                   11354: also allows explicit changing (with @code{>o...o>} or with
                   11355: @code{with...endwith}). It uses parsing and state-smart objects and
                   11356: classes for resolving overloading and for early binding: the object or
                   11357: class parses the selector and determines the method from this. If the
                   11358: selector is not parsed by an object or class, it performs a call to the
                   11359: selector for the active object (late binding), like Zsoter's model.
                   11360: Fields are always accessed through the active object. The big
                   11361: disadvantage of this model is the parsing and the state-smartness, which
                   11362: reduces extensibility and increases the opportunities for subtle bugs;
                   11363: essentially, you are only safe if you never tick or @code{postpone} an
                   11364: object or class (Bernd disagrees, but I (Anton) am not convinced).
1.21      crook    11365: 
1.78      anton    11366: @cindex @file{mini-oof.fs}, differences to other models
                   11367: The @file{mini-oof.fs} model is quite similar to a very stripped-down
                   11368: version of the @file{objects.fs} model, but syntactically it is a
                   11369: mixture of the @file{objects.fs} and @file{oof.fs} models.
1.21      crook    11370: 
                   11371: 
1.78      anton    11372: @c -------------------------------------------------------------
                   11373: @node Programming Tools, Assembler and Code Words, Object-oriented Forth, Words
                   11374: @section Programming Tools
                   11375: @cindex programming tools
1.21      crook    11376: 
1.78      anton    11377: @c !! move this and assembler down below OO stuff.
1.21      crook    11378: 
1.78      anton    11379: @menu
                   11380: * Examining::                   
                   11381: * Forgetting words::            
                   11382: * Debugging::                   Simple and quick.
                   11383: * Assertions::                  Making your programs self-checking.
                   11384: * Singlestep Debugger::         Executing your program word by word.
                   11385: @end menu
1.21      crook    11386: 
1.78      anton    11387: @node Examining, Forgetting words, Programming Tools, Programming Tools
                   11388: @subsection Examining data and code
                   11389: @cindex examining data and code
                   11390: @cindex data examination
                   11391: @cindex code examination
1.44      crook    11392: 
1.78      anton    11393: The following words inspect the stack non-destructively:
1.21      crook    11394: 
1.78      anton    11395: doc-.s
                   11396: doc-f.s
1.44      crook    11397: 
1.78      anton    11398: There is a word @code{.r} but it does @i{not} display the return stack!
                   11399: It is used for formatted numeric output (@pxref{Simple numeric output}).
1.21      crook    11400: 
1.78      anton    11401: doc-depth
                   11402: doc-fdepth
                   11403: doc-clearstack
1.21      crook    11404: 
1.78      anton    11405: The following words inspect memory.
1.21      crook    11406: 
1.78      anton    11407: doc-?
                   11408: doc-dump
1.21      crook    11409: 
1.78      anton    11410: And finally, @code{see} allows to inspect code:
1.21      crook    11411: 
1.78      anton    11412: doc-see
                   11413: doc-xt-see
1.21      crook    11414: 
1.78      anton    11415: @node Forgetting words, Debugging, Examining, Programming Tools
                   11416: @subsection Forgetting words
                   11417: @cindex words, forgetting
                   11418: @cindex forgeting words
1.21      crook    11419: 
1.78      anton    11420: @c  anton: other, maybe better places for this subsection: Defining Words;
                   11421: @c  Dictionary allocation.  At least a reference should be there.
1.21      crook    11422: 
1.78      anton    11423: Forth allows you to forget words (and everything that was alloted in the
                   11424: dictonary after them) in a LIFO manner.
1.21      crook    11425: 
1.78      anton    11426: doc-marker
1.21      crook    11427: 
1.78      anton    11428: The most common use of this feature is during progam development: when
                   11429: you change a source file, forget all the words it defined and load it
                   11430: again (since you also forget everything defined after the source file
                   11431: was loaded, you have to reload that, too).  Note that effects like
                   11432: storing to variables and destroyed system words are not undone when you
                   11433: forget words.  With a system like Gforth, that is fast enough at
                   11434: starting up and compiling, I find it more convenient to exit and restart
                   11435: Gforth, as this gives me a clean slate.
1.21      crook    11436: 
1.78      anton    11437: Here's an example of using @code{marker} at the start of a source file
                   11438: that you are debugging; it ensures that you only ever have one copy of
                   11439: the file's definitions compiled at any time:
1.21      crook    11440: 
1.78      anton    11441: @example
                   11442: [IFDEF] my-code
                   11443:     my-code
                   11444: [ENDIF]
1.26      crook    11445: 
1.78      anton    11446: marker my-code
                   11447: init-included-files
1.21      crook    11448: 
1.78      anton    11449: \ .. definitions start here
                   11450: \ .
                   11451: \ .
                   11452: \ end
                   11453: @end example
1.21      crook    11454: 
1.26      crook    11455: 
1.78      anton    11456: @node Debugging, Assertions, Forgetting words, Programming Tools
                   11457: @subsection Debugging
                   11458: @cindex debugging
1.21      crook    11459: 
1.78      anton    11460: Languages with a slow edit/compile/link/test development loop tend to
                   11461: require sophisticated tracing/stepping debuggers to facilate debugging.
1.21      crook    11462: 
1.78      anton    11463: A much better (faster) way in fast-compiling languages is to add
                   11464: printing code at well-selected places, let the program run, look at
                   11465: the output, see where things went wrong, add more printing code, etc.,
                   11466: until the bug is found.
1.21      crook    11467: 
1.78      anton    11468: The simple debugging aids provided in @file{debugs.fs}
                   11469: are meant to support this style of debugging.
1.21      crook    11470: 
1.78      anton    11471: The word @code{~~} prints debugging information (by default the source
                   11472: location and the stack contents). It is easy to insert. If you use Emacs
                   11473: it is also easy to remove (@kbd{C-x ~} in the Emacs Forth mode to
                   11474: query-replace them with nothing). The deferred words
                   11475: @code{printdebugdata} and @code{printdebugline} control the output of
                   11476: @code{~~}. The default source location output format works well with
                   11477: Emacs' compilation mode, so you can step through the program at the
                   11478: source level using @kbd{C-x `} (the advantage over a stepping debugger
                   11479: is that you can step in any direction and you know where the crash has
                   11480: happened or where the strange data has occurred).
1.21      crook    11481: 
1.78      anton    11482: doc-~~
                   11483: doc-printdebugdata
                   11484: doc-printdebugline
1.21      crook    11485: 
1.78      anton    11486: @node Assertions, Singlestep Debugger, Debugging, Programming Tools
                   11487: @subsection Assertions
                   11488: @cindex assertions
1.21      crook    11489: 
1.78      anton    11490: It is a good idea to make your programs self-checking, especially if you
                   11491: make an assumption that may become invalid during maintenance (for
                   11492: example, that a certain field of a data structure is never zero). Gforth
                   11493: supports @dfn{assertions} for this purpose. They are used like this:
1.21      crook    11494: 
                   11495: @example
1.78      anton    11496: assert( @i{flag} )
1.26      crook    11497: @end example
                   11498: 
1.78      anton    11499: The code between @code{assert(} and @code{)} should compute a flag, that
                   11500: should be true if everything is alright and false otherwise. It should
                   11501: not change anything else on the stack. The overall stack effect of the
                   11502: assertion is @code{( -- )}. E.g.
1.21      crook    11503: 
1.26      crook    11504: @example
1.78      anton    11505: assert( 1 1 + 2 = ) \ what we learn in school
                   11506: assert( dup 0<> ) \ assert that the top of stack is not zero
                   11507: assert( false ) \ this code should not be reached
1.21      crook    11508: @end example
                   11509: 
1.78      anton    11510: The need for assertions is different at different times. During
                   11511: debugging, we want more checking, in production we sometimes care more
                   11512: for speed. Therefore, assertions can be turned off, i.e., the assertion
                   11513: becomes a comment. Depending on the importance of an assertion and the
                   11514: time it takes to check it, you may want to turn off some assertions and
                   11515: keep others turned on. Gforth provides several levels of assertions for
                   11516: this purpose:
                   11517: 
                   11518: 
                   11519: doc-assert0(
                   11520: doc-assert1(
                   11521: doc-assert2(
                   11522: doc-assert3(
                   11523: doc-assert(
                   11524: doc-)
1.21      crook    11525: 
                   11526: 
1.78      anton    11527: The variable @code{assert-level} specifies the highest assertions that
                   11528: are turned on. I.e., at the default @code{assert-level} of one,
                   11529: @code{assert0(} and @code{assert1(} assertions perform checking, while
                   11530: @code{assert2(} and @code{assert3(} assertions are treated as comments.
1.26      crook    11531: 
1.78      anton    11532: The value of @code{assert-level} is evaluated at compile-time, not at
                   11533: run-time. Therefore you cannot turn assertions on or off at run-time;
                   11534: you have to set the @code{assert-level} appropriately before compiling a
                   11535: piece of code. You can compile different pieces of code at different
                   11536: @code{assert-level}s (e.g., a trusted library at level 1 and
                   11537: newly-written code at level 3).
1.26      crook    11538: 
                   11539: 
1.78      anton    11540: doc-assert-level
1.26      crook    11541: 
                   11542: 
1.78      anton    11543: If an assertion fails, a message compatible with Emacs' compilation mode
                   11544: is produced and the execution is aborted (currently with @code{ABORT"}.
                   11545: If there is interest, we will introduce a special throw code. But if you
                   11546: intend to @code{catch} a specific condition, using @code{throw} is
                   11547: probably more appropriate than an assertion).
1.44      crook    11548: 
1.78      anton    11549: Definitions in ANS Forth for these assertion words are provided
                   11550: in @file{compat/assert.fs}.
1.26      crook    11551: 
1.44      crook    11552: 
1.78      anton    11553: @node Singlestep Debugger,  , Assertions, Programming Tools
                   11554: @subsection Singlestep Debugger
                   11555: @cindex singlestep Debugger
                   11556: @cindex debugging Singlestep
1.44      crook    11557: 
1.78      anton    11558: When you create a new word there's often the need to check whether it
                   11559: behaves correctly or not. You can do this by typing @code{dbg
                   11560: badword}. A debug session might look like this:
1.26      crook    11561: 
1.78      anton    11562: @example
                   11563: : badword 0 DO i . LOOP ;  ok
                   11564: 2 dbg badword 
                   11565: : badword  
                   11566: Scanning code...
1.44      crook    11567: 
1.78      anton    11568: Nesting debugger ready!
1.44      crook    11569: 
1.78      anton    11570: 400D4738  8049BC4 0              -> [ 2 ] 00002 00000 
                   11571: 400D4740  8049F68 DO             -> [ 0 ] 
                   11572: 400D4744  804A0C8 i              -> [ 1 ] 00000 
                   11573: 400D4748 400C5E60 .              -> 0 [ 0 ] 
                   11574: 400D474C  8049D0C LOOP           -> [ 0 ] 
                   11575: 400D4744  804A0C8 i              -> [ 1 ] 00001 
                   11576: 400D4748 400C5E60 .              -> 1 [ 0 ] 
                   11577: 400D474C  8049D0C LOOP           -> [ 0 ] 
                   11578: 400D4758  804B384 ;              ->  ok
                   11579: @end example
1.21      crook    11580: 
1.78      anton    11581: Each line displayed is one step. You always have to hit return to
                   11582: execute the next word that is displayed. If you don't want to execute
                   11583: the next word in a whole, you have to type @kbd{n} for @code{nest}. Here is
                   11584: an overview what keys are available:
1.44      crook    11585: 
1.78      anton    11586: @table @i
1.44      crook    11587: 
1.78      anton    11588: @item @key{RET}
                   11589: Next; Execute the next word.
1.21      crook    11590: 
1.78      anton    11591: @item n
                   11592: Nest; Single step through next word.
1.44      crook    11593: 
1.78      anton    11594: @item u
                   11595: Unnest; Stop debugging and execute rest of word. If we got to this word
                   11596: with nest, continue debugging with the calling word.
1.44      crook    11597: 
1.78      anton    11598: @item d
                   11599: Done; Stop debugging and execute rest.
1.21      crook    11600: 
1.78      anton    11601: @item s
                   11602: Stop; Abort immediately.
1.44      crook    11603: 
1.78      anton    11604: @end table
1.44      crook    11605: 
1.78      anton    11606: Debugging large application with this mechanism is very difficult, because
                   11607: you have to nest very deeply into the program before the interesting part
                   11608: begins. This takes a lot of time. 
1.26      crook    11609: 
1.78      anton    11610: To do it more directly put a @code{BREAK:} command into your source code.
                   11611: When program execution reaches @code{BREAK:} the single step debugger is
                   11612: invoked and you have all the features described above.
1.44      crook    11613: 
1.78      anton    11614: If you have more than one part to debug it is useful to know where the
                   11615: program has stopped at the moment. You can do this by the 
                   11616: @code{BREAK" string"} command. This behaves like @code{BREAK:} except that
                   11617: string is typed out when the ``breakpoint'' is reached.
1.44      crook    11618: 
1.26      crook    11619: 
1.78      anton    11620: doc-dbg
                   11621: doc-break:
                   11622: doc-break"
1.44      crook    11623: 
                   11624: 
1.26      crook    11625: 
1.78      anton    11626: @c -------------------------------------------------------------
                   11627: @node Assembler and Code Words, Threading Words, Programming Tools, Words
                   11628: @section Assembler and Code Words
                   11629: @cindex assembler
                   11630: @cindex code words
1.44      crook    11631: 
1.78      anton    11632: @menu
                   11633: * Code and ;code::              
                   11634: * Common Assembler::            Assembler Syntax
                   11635: * Common Disassembler::         
                   11636: * 386 Assembler::               Deviations and special cases
                   11637: * Alpha Assembler::             Deviations and special cases
                   11638: * MIPS assembler::              Deviations and special cases
                   11639: * Other assemblers::            How to write them
                   11640: @end menu
1.21      crook    11641: 
1.78      anton    11642: @node Code and ;code, Common Assembler, Assembler and Code Words, Assembler and Code Words
                   11643: @subsection @code{Code} and @code{;code}
1.26      crook    11644: 
1.78      anton    11645: Gforth provides some words for defining primitives (words written in
                   11646: machine code), and for defining the machine-code equivalent of
                   11647: @code{DOES>}-based defining words. However, the machine-independent
                   11648: nature of Gforth poses a few problems: First of all, Gforth runs on
                   11649: several architectures, so it can provide no standard assembler. What's
                   11650: worse is that the register allocation not only depends on the processor,
                   11651: but also on the @code{gcc} version and options used.
1.44      crook    11652: 
1.78      anton    11653: The words that Gforth offers encapsulate some system dependences (e.g.,
                   11654: the header structure), so a system-independent assembler may be used in
                   11655: Gforth. If you do not have an assembler, you can compile machine code
                   11656: directly with @code{,} and @code{c,}@footnote{This isn't portable,
                   11657: because these words emit stuff in @i{data} space; it works because
                   11658: Gforth has unified code/data spaces. Assembler isn't likely to be
                   11659: portable anyway.}.
1.21      crook    11660: 
1.44      crook    11661: 
1.78      anton    11662: doc-assembler
                   11663: doc-init-asm
                   11664: doc-code
                   11665: doc-end-code
                   11666: doc-;code
                   11667: doc-flush-icache
1.44      crook    11668: 
1.21      crook    11669: 
1.78      anton    11670: If @code{flush-icache} does not work correctly, @code{code} words
                   11671: etc. will not work (reliably), either.
1.44      crook    11672: 
1.78      anton    11673: The typical usage of these @code{code} words can be shown most easily by
                   11674: analogy to the equivalent high-level defining words:
1.44      crook    11675: 
1.78      anton    11676: @example
                   11677: : foo                              code foo
                   11678:    <high-level Forth words>              <assembler>
                   11679: ;                                  end-code
                   11680:                                 
                   11681: : bar                              : bar
                   11682:    <high-level Forth words>           <high-level Forth words>
                   11683:    CREATE                             CREATE
                   11684:       <high-level Forth words>           <high-level Forth words>
                   11685:    DOES>                              ;code
                   11686:       <high-level Forth words>           <assembler>
                   11687: ;                                  end-code
                   11688: @end example
1.21      crook    11689: 
1.78      anton    11690: @c anton: the following stuff is also in "Common Assembler", in less detail.
1.44      crook    11691: 
1.78      anton    11692: @cindex registers of the inner interpreter
                   11693: In the assembly code you will want to refer to the inner interpreter's
                   11694: registers (e.g., the data stack pointer) and you may want to use other
                   11695: registers for temporary storage. Unfortunately, the register allocation
                   11696: is installation-dependent.
1.44      crook    11697: 
1.78      anton    11698: In particular, @code{ip} (Forth instruction pointer) and @code{rp}
                   11699: (return stack pointer) are in different places in @code{gforth} and
                   11700: @code{gforth-fast}.  This means that you cannot write a @code{NEXT}
                   11701: routine that works on both versions; so for doing @code{NEXT}, I
                   11702: recomment jumping to @code{' noop >code-address}, which contains nothing
                   11703: but a @code{NEXT}.
1.21      crook    11704: 
1.78      anton    11705: For general accesses to the inner interpreter's registers, the easiest
                   11706: solution is to use explicit register declarations (@pxref{Explicit Reg
                   11707: Vars, , Variables in Specified Registers, gcc.info, GNU C Manual}) for
                   11708: all of the inner interpreter's registers: You have to compile Gforth
                   11709: with @code{-DFORCE_REG} (configure option @code{--enable-force-reg}) and
                   11710: the appropriate declarations must be present in the @code{machine.h}
                   11711: file (see @code{mips.h} for an example; you can find a full list of all
                   11712: declarable register symbols with @code{grep register engine.c}). If you
                   11713: give explicit registers to all variables that are declared at the
                   11714: beginning of @code{engine()}, you should be able to use the other
                   11715: caller-saved registers for temporary storage. Alternatively, you can use
                   11716: the @code{gcc} option @code{-ffixed-REG} (@pxref{Code Gen Options, ,
                   11717: Options for Code Generation Conventions, gcc.info, GNU C Manual}) to
                   11718: reserve a register (however, this restriction on register allocation may
                   11719: slow Gforth significantly).
1.44      crook    11720: 
1.78      anton    11721: If this solution is not viable (e.g., because @code{gcc} does not allow
                   11722: you to explicitly declare all the registers you need), you have to find
                   11723: out by looking at the code where the inner interpreter's registers
                   11724: reside and which registers can be used for temporary storage. You can
                   11725: get an assembly listing of the engine's code with @code{make engine.s}.
1.44      crook    11726: 
1.78      anton    11727: In any case, it is good practice to abstract your assembly code from the
                   11728: actual register allocation. E.g., if the data stack pointer resides in
                   11729: register @code{$17}, create an alias for this register called @code{sp},
                   11730: and use that in your assembly code.
1.21      crook    11731: 
1.78      anton    11732: @cindex code words, portable
                   11733: Another option for implementing normal and defining words efficiently
                   11734: is to add the desired functionality to the source of Gforth. For normal
                   11735: words you just have to edit @file{primitives} (@pxref{Automatic
                   11736: Generation}). Defining words (equivalent to @code{;CODE} words, for fast
                   11737: defined words) may require changes in @file{engine.c}, @file{kernel.fs},
                   11738: @file{prims2x.fs}, and possibly @file{cross.fs}.
1.44      crook    11739: 
1.78      anton    11740: @node Common Assembler, Common Disassembler, Code and ;code, Assembler and Code Words
                   11741: @subsection Common Assembler
1.44      crook    11742: 
1.78      anton    11743: The assemblers in Gforth generally use a postfix syntax, i.e., the
                   11744: instruction name follows the operands.
1.21      crook    11745: 
1.78      anton    11746: The operands are passed in the usual order (the same that is used in the
                   11747: manual of the architecture).  Since they all are Forth words, they have
                   11748: to be separated by spaces; you can also use Forth words to compute the
                   11749: operands.
1.44      crook    11750: 
1.78      anton    11751: The instruction names usually end with a @code{,}.  This makes it easier
                   11752: to visually separate instructions if you put several of them on one
                   11753: line; it also avoids shadowing other Forth words (e.g., @code{and}).
1.21      crook    11754: 
1.78      anton    11755: Registers are usually specified by number; e.g., (decimal) @code{11}
                   11756: specifies registers R11 and F11 on the Alpha architecture (which one,
                   11757: depends on the instruction).  The usual names are also available, e.g.,
                   11758: @code{s2} for R11 on Alpha.
1.21      crook    11759: 
1.78      anton    11760: Control flow is specified similar to normal Forth code (@pxref{Arbitrary
                   11761: control structures}), with @code{if,}, @code{ahead,}, @code{then,},
                   11762: @code{begin,}, @code{until,}, @code{again,}, @code{cs-roll},
                   11763: @code{cs-pick}, @code{else,}, @code{while,}, and @code{repeat,}.  The
                   11764: conditions are specified in a way specific to each assembler.
1.1       anton    11765: 
1.78      anton    11766: Note that the register assignments of the Gforth engine can change
                   11767: between Gforth versions, or even between different compilations of the
                   11768: same Gforth version (e.g., if you use a different GCC version).  So if
                   11769: you want to refer to Gforth's registers (e.g., the stack pointer or
                   11770: TOS), I recommend defining your own words for refering to these
                   11771: registers, and using them later on; then you can easily adapt to a
                   11772: changed register assignment.  The stability of the register assignment
                   11773: is usually better if you build Gforth with @code{--enable-force-reg}.
1.1       anton    11774: 
1.78      anton    11775: In particular, the return stack pointer and the instruction pointer are
                   11776: in memory in @code{gforth}, and usually in registers in
                   11777: @code{gforth-fast}.  The most common use of these registers is to
                   11778: dispatch to the next word (the @code{next} routine).  A portable way to
                   11779: do this is to jump to @code{' noop >code-address} (of course, this is
                   11780: less efficient than integrating the @code{next} code and scheduling it
                   11781: well).
1.1       anton    11782: 
1.78      anton    11783: @node  Common Disassembler, 386 Assembler, Common Assembler, Assembler and Code Words
                   11784: @subsection Common Disassembler
1.1       anton    11785: 
1.78      anton    11786: You can disassemble a @code{code} word with @code{see}
                   11787: (@pxref{Debugging}).  You can disassemble a section of memory with
1.1       anton    11788: 
1.78      anton    11789: doc-disasm
1.44      crook    11790: 
1.78      anton    11791: The disassembler generally produces output that can be fed into the
                   11792: assembler (i.e., same syntax, etc.).  It also includes additional
                   11793: information in comments.  In particular, the address of the instruction
                   11794: is given in a comment before the instruction.
1.1       anton    11795: 
1.78      anton    11796: @code{See} may display more or less than the actual code of the word,
                   11797: because the recognition of the end of the code is unreliable.  You can
                   11798: use @code{disasm} if it did not display enough.  It may display more, if
                   11799: the code word is not immediately followed by a named word.  If you have
                   11800: something else there, you can follow the word with @code{align last @ ,}
                   11801: to ensure that the end is recognized.
1.21      crook    11802: 
1.78      anton    11803: @node 386 Assembler, Alpha Assembler, Common Disassembler, Assembler and Code Words
                   11804: @subsection 386 Assembler
1.44      crook    11805: 
1.78      anton    11806: The 386 assembler included in Gforth was written by Bernd Paysan, it's
                   11807: available under GPL, and originally part of bigFORTH.
1.21      crook    11808: 
1.78      anton    11809: The 386 disassembler included in Gforth was written by Andrew McKewan
                   11810: and is in the public domain.
1.21      crook    11811: 
1.78      anton    11812: The disassembler displays code in prefix Intel syntax.
1.21      crook    11813: 
1.78      anton    11814: The assembler uses a postfix syntax with reversed parameters.
1.1       anton    11815: 
1.78      anton    11816: The assembler includes all instruction of the Athlon, i.e. 486 core
                   11817: instructions, Pentium and PPro extensions, floating point, MMX, 3Dnow!,
                   11818: but not ISSE. It's an integrated 16- and 32-bit assembler. Default is 32
                   11819: bit, you can switch to 16 bit with .86 and back to 32 bit with .386.
1.1       anton    11820: 
1.78      anton    11821: There are several prefixes to switch between different operation sizes,
                   11822: @code{.b} for byte accesses, @code{.w} for word accesses, @code{.d} for
                   11823: double-word accesses. Addressing modes can be switched with @code{.wa}
                   11824: for 16 bit addresses, and @code{.da} for 32 bit addresses. You don't
                   11825: need a prefix for byte register names (@code{AL} et al).
1.1       anton    11826: 
1.78      anton    11827: For floating point operations, the prefixes are @code{.fs} (IEEE
                   11828: single), @code{.fl} (IEEE double), @code{.fx} (extended), @code{.fw}
                   11829: (word), @code{.fd} (double-word), and @code{.fq} (quad-word).
1.21      crook    11830: 
1.78      anton    11831: The MMX opcodes don't have size prefixes, they are spelled out like in
                   11832: the Intel assembler. Instead of move from and to memory, there are
                   11833: PLDQ/PLDD and PSTQ/PSTD.
1.21      crook    11834: 
1.78      anton    11835: The registers lack the 'e' prefix; even in 32 bit mode, eax is called
                   11836: ax.  Immediate values are indicated by postfixing them with @code{#},
                   11837: e.g., @code{3 #}.  Here are some examples of addressing modes:
1.21      crook    11838: 
1.26      crook    11839: @example
1.78      anton    11840: 3 #          \ immediate
1.85      anton    11841: 1000 #)      \ absolute
1.78      anton    11842: ax           \ register
                   11843: 100 di d)    \ 100[edi]
                   11844: 4 bx cx di)  \ 4[ebx][ecx]
                   11845: di ax *4 i)  \ [edi][eax*4]
                   11846: 20 ax *4 i#) \ 20[eax*4]
1.26      crook    11847: @end example
1.21      crook    11848: 
1.78