Annotation of gforth/doc/gforth.ds, revision 1.173
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
1.113 anton 14: @c
15: @c Karl Berry writes:
16: @c If they don't like the all-caps for @var Info output, all I can say is
17: @c that it's always been that way, and the usage of all-caps for
18: @c metavariables has a long tradition. I think it's best to just let it be
19: @c what it is, for the sake of consistency among manuals.
20: @c
1.29 crook 21: @comment .. would be useful to have a word that identified all deferred words
22: @comment should semantics stuff in intro be moved to another section
23:
1.66 anton 24: @c POSTPONE, COMPILE, [COMPILE], LITERAL should have their own section
1.28 crook 25:
1.1 anton 26: @comment %**start of header (This is for running Texinfo on a region.)
27: @setfilename gforth.info
1.113 anton 28: @include version.texi
1.1 anton 29: @settitle Gforth Manual
1.113 anton 30: @c @syncodeindex pg cp
1.49 anton 31:
1.12 anton 32: @macro progstyle {}
33: Programming style note:
1.3 anton 34: @end macro
1.48 anton 35:
36: @macro assignment {}
37: @table @i
38: @item Assignment:
39: @end macro
40: @macro endassignment {}
41: @end table
42: @end macro
43:
1.29 crook 44: @comment macros for beautifying glossary entries
45: @macro GLOSS-START {}
46: @iftex
47: @ninerm
48: @end iftex
49: @end macro
50:
51: @macro GLOSS-END {}
52: @iftex
53: @rm
54: @end iftex
55: @end macro
56:
1.113 anton 57: @comment %**end of header (This is for running Texinfo on a region.)
58: @copying
1.125 anton 59: This manual is for Gforth (version @value{VERSION}, @value{UPDATED}),
60: a fast and portable implementation of the ANS Forth language. It
61: serves as reference manual, but it also contains an introduction to
62: Forth and a Forth tutorial.
1.29 crook 63:
1.169 anton 64: Copyright @copyright{} 1995, 1996, 1997, 1998, 2000, 2003, 2004,2005,2006 Free Software Foundation, Inc.
1.29 crook 65:
1.113 anton 66: @quotation
67: Permission is granted to copy, distribute and/or modify this document
68: under the terms of the GNU Free Documentation License, Version 1.1 or
69: any later version published by the Free Software Foundation; with no
70: Invariant Sections, with the Front-Cover texts being ``A GNU Manual,''
71: and with the Back-Cover Texts as in (a) below. A copy of the
72: license is included in the section entitled ``GNU Free Documentation
73: License.''
74:
75: (a) The FSF's Back-Cover Text is: ``You have freedom to copy and modify
76: this GNU Manual, like GNU software. Copies published by the Free
77: Software Foundation raise funds for GNU development.''
78: @end quotation
79: @end copying
1.10 anton 80:
1.113 anton 81: @dircategory Software development
82: @direntry
83: * Gforth: (gforth). A fast interpreter for the Forth language.
84: @end direntry
85: @c The Texinfo manual also recommends doing this, but for Gforth it may
86: @c not make much sense
87: @c @dircategory Individual utilities
88: @c @direntry
89: @c * Gforth: (gforth)Invoking Gforth. gforth, gforth-fast, gforthmi
90: @c @end direntry
1.1 anton 91:
92: @titlepage
1.113 anton 93: @title Gforth
94: @subtitle for version @value{VERSION}, @value{UPDATED}
95: @author Neal Crook
96: @author Anton Ertl
1.114 anton 97: @author David Kuehling
1.113 anton 98: @author Bernd Paysan
99: @author Jens Wilke
1.1 anton 100: @page
101: @vskip 0pt plus 1filll
1.113 anton 102: @insertcopying
103: @end titlepage
1.1 anton 104:
1.113 anton 105: @contents
1.1 anton 106:
1.113 anton 107: @ifnottex
108: @node Top, Goals, (dir), (dir)
109: @top Gforth
1.1 anton 110:
1.113 anton 111: @insertcopying
1.49 anton 112: @end ifnottex
1.1 anton 113:
114: @menu
1.26 crook 115: * Goals:: About the Gforth Project
1.29 crook 116: * Gforth Environment:: Starting (and exiting) Gforth
1.48 anton 117: * Tutorial:: Hands-on Forth Tutorial
1.21 crook 118: * Introduction:: An introduction to ANS Forth
1.1 anton 119: * Words:: Forth words available in Gforth
1.24 anton 120: * Error messages:: How to interpret them
1.1 anton 121: * Tools:: Programming tools
122: * ANS conformance:: Implementation-defined options etc.
1.65 anton 123: * Standard vs Extensions:: Should I use extensions?
1.1 anton 124: * Model:: The abstract machine of Gforth
125: * Integrating Gforth:: Forth as scripting language for applications
126: * Emacs and Gforth:: The Gforth Mode
127: * Image Files:: @code{.fi} files contain compiled code
128: * Engine:: The inner interpreter and the primitives
1.13 pazsan 129: * Cross Compiler:: The Cross Compiler
1.1 anton 130: * Bugs:: How to report them
131: * Origin:: Authors and ancestors of Gforth
1.21 crook 132: * Forth-related information:: Books and places to look on the WWW
1.113 anton 133: * Licenses::
1.1 anton 134: * Word Index:: An item for each Forth word
135: * Concept Index:: A menu covering many topics
1.12 anton 136:
1.91 anton 137: @detailmenu
138: --- The Detailed Node Listing ---
1.12 anton 139:
1.29 crook 140: Gforth Environment
141:
1.32 anton 142: * Invoking Gforth:: Getting in
143: * Leaving Gforth:: Getting out
144: * Command-line editing::
1.48 anton 145: * Environment variables:: that affect how Gforth starts up
1.32 anton 146: * Gforth Files:: What gets installed and where
1.112 anton 147: * Gforth in pipes::
1.48 anton 148: * Startup speed:: When 35ms is not fast enough ...
149:
150: Forth Tutorial
151:
152: * Starting Gforth Tutorial::
153: * Syntax Tutorial::
154: * Crash Course Tutorial::
155: * Stack Tutorial::
156: * Arithmetics Tutorial::
157: * Stack Manipulation Tutorial::
158: * Using files for Forth code Tutorial::
159: * Comments Tutorial::
160: * Colon Definitions Tutorial::
161: * Decompilation Tutorial::
162: * Stack-Effect Comments Tutorial::
163: * Types Tutorial::
164: * Factoring Tutorial::
165: * Designing the stack effect Tutorial::
166: * Local Variables Tutorial::
167: * Conditional execution Tutorial::
168: * Flags and Comparisons Tutorial::
169: * General Loops Tutorial::
170: * Counted loops Tutorial::
171: * Recursion Tutorial::
172: * Leaving definitions or loops Tutorial::
173: * Return Stack Tutorial::
174: * Memory Tutorial::
175: * Characters and Strings Tutorial::
176: * Alignment Tutorial::
1.87 anton 177: * Files Tutorial::
1.48 anton 178: * Interpretation and Compilation Semantics and Immediacy Tutorial::
179: * Execution Tokens Tutorial::
180: * Exceptions Tutorial::
181: * Defining Words Tutorial::
182: * Arrays and Records Tutorial::
183: * POSTPONE Tutorial::
184: * Literal Tutorial::
185: * Advanced macros Tutorial::
186: * Compilation Tokens Tutorial::
187: * Wordlists and Search Order Tutorial::
1.29 crook 188:
1.24 anton 189: An Introduction to ANS Forth
190:
1.67 anton 191: * Introducing the Text Interpreter::
192: * Stacks and Postfix notation::
193: * Your first definition::
194: * How does that work?::
195: * Forth is written in Forth::
196: * Review - elements of a Forth system::
197: * Where to go next::
198: * Exercises::
1.24 anton 199:
1.12 anton 200: Forth Words
201:
202: * Notation::
1.65 anton 203: * Case insensitivity::
204: * Comments::
205: * Boolean Flags::
1.12 anton 206: * Arithmetic::
207: * Stack Manipulation::
208: * Memory::
209: * Control Structures::
210: * Defining Words::
1.65 anton 211: * Interpretation and Compilation Semantics::
1.47 crook 212: * Tokens for Words::
1.81 anton 213: * Compiling words::
1.65 anton 214: * The Text Interpreter::
1.111 anton 215: * The Input Stream::
1.65 anton 216: * Word Lists::
217: * Environmental Queries::
1.12 anton 218: * Files::
219: * Blocks::
220: * Other I/O::
1.121 anton 221: * OS command line arguments::
1.78 anton 222: * Locals::
223: * Structures::
224: * Object-oriented Forth::
1.12 anton 225: * Programming Tools::
1.150 anton 226: * C Interface::
1.12 anton 227: * Assembler and Code Words::
228: * Threading Words::
1.65 anton 229: * Passing Commands to the OS::
230: * Keeping track of Time::
231: * Miscellaneous Words::
1.12 anton 232:
233: Arithmetic
234:
235: * Single precision::
1.67 anton 236: * Double precision:: Double-cell integer arithmetic
1.12 anton 237: * Bitwise operations::
1.67 anton 238: * Numeric comparison::
1.32 anton 239: * Mixed precision:: Operations with single and double-cell integers
1.12 anton 240: * Floating Point::
241:
242: Stack Manipulation
243:
244: * Data stack::
245: * Floating point stack::
246: * Return stack::
247: * Locals stack::
248: * Stack pointer manipulation::
249:
250: Memory
251:
1.32 anton 252: * Memory model::
253: * Dictionary allocation::
254: * Heap Allocation::
255: * Memory Access::
256: * Address arithmetic::
257: * Memory Blocks::
1.12 anton 258:
259: Control Structures
260:
1.41 anton 261: * Selection:: IF ... ELSE ... ENDIF
262: * Simple Loops:: BEGIN ...
1.32 anton 263: * Counted Loops:: DO
1.67 anton 264: * Arbitrary control structures::
265: * Calls and returns::
1.12 anton 266: * Exception Handling::
267:
268: Defining Words
269:
1.67 anton 270: * CREATE::
1.44 crook 271: * Variables:: Variables and user variables
1.67 anton 272: * Constants::
1.44 crook 273: * Values:: Initialised variables
1.67 anton 274: * Colon Definitions::
1.44 crook 275: * Anonymous Definitions:: Definitions without names
1.71 anton 276: * Supplying names:: Passing definition names as strings
1.67 anton 277: * User-defined Defining Words::
1.170 pazsan 278: * Deferred Words:: Allow forward references
1.67 anton 279: * Aliases::
1.47 crook 280:
1.63 anton 281: User-defined Defining Words
282:
283: * CREATE..DOES> applications::
284: * CREATE..DOES> details::
285: * Advanced does> usage example::
1.155 anton 286: * Const-does>::
1.63 anton 287:
1.47 crook 288: Interpretation and Compilation Semantics
289:
1.67 anton 290: * Combined words::
1.12 anton 291:
1.71 anton 292: Tokens for Words
293:
294: * Execution token:: represents execution/interpretation semantics
295: * Compilation token:: represents compilation semantics
296: * Name token:: represents named words
297:
1.82 anton 298: Compiling words
299:
300: * Literals:: Compiling data values
301: * Macros:: Compiling words
302:
1.21 crook 303: The Text Interpreter
304:
1.67 anton 305: * Input Sources::
306: * Number Conversion::
307: * Interpret/Compile states::
308: * Interpreter Directives::
1.21 crook 309:
1.26 crook 310: Word Lists
311:
1.75 anton 312: * Vocabularies::
1.67 anton 313: * Why use word lists?::
1.75 anton 314: * Word list example::
1.26 crook 315:
316: Files
317:
1.48 anton 318: * Forth source files::
319: * General files::
1.167 anton 320: * Redirection::
1.48 anton 321: * Search Paths::
322:
323: Search Paths
324:
1.75 anton 325: * Source Search Paths::
1.26 crook 326: * General Search Paths::
327:
328: Other I/O
329:
1.32 anton 330: * Simple numeric output:: Predefined formats
331: * Formatted numeric output:: Formatted (pictured) output
332: * String Formats:: How Forth stores strings in memory
1.67 anton 333: * Displaying characters and strings:: Other stuff
1.32 anton 334: * Input:: Input
1.112 anton 335: * Pipes:: How to create your own pipes
1.149 pazsan 336: * Xchars and Unicode:: Non-ASCII characters
1.26 crook 337:
338: Locals
339:
340: * Gforth locals::
341: * ANS Forth locals::
342:
343: Gforth locals
344:
345: * Where are locals visible by name?::
346: * How long do locals live?::
1.78 anton 347: * Locals programming style::
348: * Locals implementation::
1.26 crook 349:
1.12 anton 350: Structures
351:
352: * Why explicit structure support?::
353: * Structure Usage::
354: * Structure Naming Convention::
355: * Structure Implementation::
356: * Structure Glossary::
357:
358: Object-oriented Forth
359:
1.48 anton 360: * Why object-oriented programming?::
361: * Object-Oriented Terminology::
362: * Objects::
363: * OOF::
364: * Mini-OOF::
1.23 crook 365: * Comparison with other object models::
1.12 anton 366:
1.24 anton 367: The @file{objects.fs} model
1.12 anton 368:
369: * Properties of the Objects model::
370: * Basic Objects Usage::
1.41 anton 371: * The Objects base class::
1.12 anton 372: * Creating objects::
373: * Object-Oriented Programming Style::
374: * Class Binding::
375: * Method conveniences::
376: * Classes and Scoping::
1.41 anton 377: * Dividing classes::
1.12 anton 378: * Object Interfaces::
379: * Objects Implementation::
380: * Objects Glossary::
381:
1.24 anton 382: The @file{oof.fs} model
1.12 anton 383:
1.67 anton 384: * Properties of the OOF model::
385: * Basic OOF Usage::
386: * The OOF base class::
387: * Class Declaration::
388: * Class Implementation::
1.12 anton 389:
1.24 anton 390: The @file{mini-oof.fs} model
1.23 crook 391:
1.48 anton 392: * Basic Mini-OOF Usage::
393: * Mini-OOF Example::
394: * Mini-OOF Implementation::
1.23 crook 395:
1.78 anton 396: Programming Tools
397:
1.150 anton 398: * Examining:: Data and Code.
399: * Forgetting words:: Usually before reloading.
1.78 anton 400: * Debugging:: Simple and quick.
401: * Assertions:: Making your programs self-checking.
402: * Singlestep Debugger:: Executing your program word by word.
403:
1.155 anton 404: C Interface
405:
406: * Calling C Functions::
407: * Declaring C Functions::
408: * Callbacks::
409: * Low-Level C Interface Words::
410:
1.78 anton 411: Assembler and Code Words
412:
413: * Code and ;code::
414: * Common Assembler:: Assembler Syntax
415: * Common Disassembler::
416: * 386 Assembler:: Deviations and special cases
417: * Alpha Assembler:: Deviations and special cases
418: * MIPS assembler:: Deviations and special cases
1.167 anton 419: * PowerPC assembler:: Deviations and special cases
1.78 anton 420: * Other assemblers:: How to write them
421:
1.12 anton 422: Tools
423:
424: * ANS Report:: Report the words used, sorted by wordset.
1.127 anton 425: * Stack depth changes:: Where does this stack item come from?
1.12 anton 426:
427: ANS conformance
428:
429: * The Core Words::
430: * The optional Block word set::
431: * The optional Double Number word set::
432: * The optional Exception word set::
433: * The optional Facility word set::
434: * The optional File-Access word set::
435: * The optional Floating-Point word set::
436: * The optional Locals word set::
437: * The optional Memory-Allocation word set::
438: * The optional Programming-Tools word set::
439: * The optional Search-Order word set::
440:
441: The Core Words
442:
443: * core-idef:: Implementation Defined Options
444: * core-ambcond:: Ambiguous Conditions
445: * core-other:: Other System Documentation
446:
447: The optional Block word set
448:
449: * block-idef:: Implementation Defined Options
450: * block-ambcond:: Ambiguous Conditions
451: * block-other:: Other System Documentation
452:
453: The optional Double Number word set
454:
455: * double-ambcond:: Ambiguous Conditions
456:
457: The optional Exception word set
458:
459: * exception-idef:: Implementation Defined Options
460:
461: The optional Facility word set
462:
463: * facility-idef:: Implementation Defined Options
464: * facility-ambcond:: Ambiguous Conditions
465:
466: The optional File-Access word set
467:
468: * file-idef:: Implementation Defined Options
469: * file-ambcond:: Ambiguous Conditions
470:
471: The optional Floating-Point word set
472:
473: * floating-idef:: Implementation Defined Options
474: * floating-ambcond:: Ambiguous Conditions
475:
476: The optional Locals word set
477:
478: * locals-idef:: Implementation Defined Options
479: * locals-ambcond:: Ambiguous Conditions
480:
481: The optional Memory-Allocation word set
482:
483: * memory-idef:: Implementation Defined Options
484:
485: The optional Programming-Tools word set
486:
487: * programming-idef:: Implementation Defined Options
488: * programming-ambcond:: Ambiguous Conditions
489:
490: The optional Search-Order word set
491:
492: * search-idef:: Implementation Defined Options
493: * search-ambcond:: Ambiguous Conditions
494:
1.109 anton 495: Emacs and Gforth
496:
497: * Installing gforth.el:: Making Emacs aware of Forth.
498: * Emacs Tags:: Viewing the source of a word in Emacs.
499: * Hilighting:: Making Forth code look prettier.
500: * Auto-Indentation:: Customizing auto-indentation.
501: * Blocks Files:: Reading and writing blocks files.
502:
1.12 anton 503: Image Files
504:
1.24 anton 505: * Image Licensing Issues:: Distribution terms for images.
506: * Image File Background:: Why have image files?
1.67 anton 507: * Non-Relocatable Image Files:: don't always work.
1.24 anton 508: * Data-Relocatable Image Files:: are better.
1.67 anton 509: * Fully Relocatable Image Files:: better yet.
1.24 anton 510: * Stack and Dictionary Sizes:: Setting the default sizes for an image.
1.32 anton 511: * Running Image Files:: @code{gforth -i @i{file}} or @i{file}.
1.24 anton 512: * Modifying the Startup Sequence:: and turnkey applications.
1.12 anton 513:
514: Fully Relocatable Image Files
515:
1.27 crook 516: * gforthmi:: The normal way
1.12 anton 517: * cross.fs:: The hard way
518:
519: Engine
520:
521: * Portability::
522: * Threading::
523: * Primitives::
524: * Performance::
525:
526: Threading
527:
528: * Scheduling::
529: * Direct or Indirect Threaded?::
1.109 anton 530: * Dynamic Superinstructions::
1.12 anton 531: * DOES>::
532:
533: Primitives
534:
535: * Automatic Generation::
536: * TOS Optimization::
537: * Produced code::
1.13 pazsan 538:
539: Cross Compiler
540:
1.67 anton 541: * Using the Cross Compiler::
542: * How the Cross Compiler Works::
1.13 pazsan 543:
1.113 anton 544: Licenses
545:
546: * GNU Free Documentation License:: License for copying this manual.
547: * Copying:: GPL (for copying this software).
548:
1.24 anton 549: @end detailmenu
1.1 anton 550: @end menu
551:
1.113 anton 552: @c ----------------------------------------------------------
1.1 anton 553: @iftex
554: @unnumbered Preface
555: @cindex Preface
1.21 crook 556: This manual documents Gforth. Some introductory material is provided for
557: readers who are unfamiliar with Forth or who are migrating to Gforth
558: from other Forth compilers. However, this manual is primarily a
559: reference manual.
1.1 anton 560: @end iftex
561:
1.28 crook 562: @comment TODO much more blurb here.
1.26 crook 563:
564: @c ******************************************************************
1.113 anton 565: @node Goals, Gforth Environment, Top, Top
1.26 crook 566: @comment node-name, next, previous, up
567: @chapter Goals of Gforth
568: @cindex goals of the Gforth project
569: The goal of the Gforth Project is to develop a standard model for
570: ANS Forth. This can be split into several subgoals:
571:
572: @itemize @bullet
573: @item
574: Gforth should conform to the ANS Forth Standard.
575: @item
576: It should be a model, i.e. it should define all the
577: implementation-dependent things.
578: @item
579: It should become standard, i.e. widely accepted and used. This goal
580: is the most difficult one.
581: @end itemize
582:
583: To achieve these goals Gforth should be
584: @itemize @bullet
585: @item
586: Similar to previous models (fig-Forth, F83)
587: @item
588: Powerful. It should provide for all the things that are considered
589: necessary today and even some that are not yet considered necessary.
590: @item
591: Efficient. It should not get the reputation of being exceptionally
592: slow.
593: @item
594: Free.
595: @item
596: Available on many machines/easy to port.
597: @end itemize
598:
599: Have we achieved these goals? Gforth conforms to the ANS Forth
600: standard. It may be considered a model, but we have not yet documented
601: which parts of the model are stable and which parts we are likely to
602: change. It certainly has not yet become a de facto standard, but it
603: appears to be quite popular. It has some similarities to and some
604: differences from previous models. It has some powerful features, but not
605: yet everything that we envisioned. We certainly have achieved our
1.65 anton 606: execution speed goals (@pxref{Performance})@footnote{However, in 1998
607: the bar was raised when the major commercial Forth vendors switched to
608: native code compilers.}. It is free and available on many machines.
1.29 crook 609:
1.26 crook 610: @c ******************************************************************
1.48 anton 611: @node Gforth Environment, Tutorial, Goals, Top
1.29 crook 612: @chapter Gforth Environment
613: @cindex Gforth environment
1.21 crook 614:
1.45 crook 615: Note: ultimately, the Gforth man page will be auto-generated from the
1.29 crook 616: material in this chapter.
1.21 crook 617:
618: @menu
1.29 crook 619: * Invoking Gforth:: Getting in
620: * Leaving Gforth:: Getting out
621: * Command-line editing::
1.48 anton 622: * Environment variables:: that affect how Gforth starts up
1.29 crook 623: * Gforth Files:: What gets installed and where
1.112 anton 624: * Gforth in pipes::
1.48 anton 625: * Startup speed:: When 35ms is not fast enough ...
1.21 crook 626: @end menu
627:
1.49 anton 628: For related information about the creation of images see @ref{Image Files}.
1.29 crook 629:
1.21 crook 630: @comment ----------------------------------------------
1.48 anton 631: @node Invoking Gforth, Leaving Gforth, Gforth Environment, Gforth Environment
1.29 crook 632: @section Invoking Gforth
633: @cindex invoking Gforth
634: @cindex running Gforth
635: @cindex command-line options
636: @cindex options on the command line
637: @cindex flags on the command line
1.21 crook 638:
1.30 anton 639: Gforth is made up of two parts; an executable ``engine'' (named
1.109 anton 640: @command{gforth} or @command{gforth-fast}) and an image file. To start it, you
1.30 anton 641: will usually just say @code{gforth} -- this automatically loads the
642: default image file @file{gforth.fi}. In many other cases the default
643: Gforth image will be invoked like this:
1.21 crook 644: @example
1.30 anton 645: gforth [file | -e forth-code] ...
1.21 crook 646: @end example
1.29 crook 647: @noindent
648: This interprets the contents of the files and the Forth code in the order they
649: are given.
1.21 crook 650:
1.109 anton 651: In addition to the @command{gforth} engine, there is also an engine
652: called @command{gforth-fast}, which is faster, but gives less
653: informative error messages (@pxref{Error messages}) and may catch some
1.166 anton 654: errors (in particular, stack underflows and integer division errors)
655: later or not at all. You should use it for debugged,
1.109 anton 656: performance-critical programs.
657:
658: Moreover, there is an engine called @command{gforth-itc}, which is
659: useful in some backwards-compatibility situations (@pxref{Direct or
660: Indirect Threaded?}).
1.30 anton 661:
1.29 crook 662: In general, the command line looks like this:
1.21 crook 663:
664: @example
1.30 anton 665: gforth[-fast] [engine options] [image options]
1.21 crook 666: @end example
667:
1.30 anton 668: The engine options must come before the rest of the command
1.29 crook 669: line. They are:
1.26 crook 670:
1.29 crook 671: @table @code
672: @cindex -i, command-line option
673: @cindex --image-file, command-line option
674: @item --image-file @i{file}
675: @itemx -i @i{file}
676: Loads the Forth image @i{file} instead of the default
677: @file{gforth.fi} (@pxref{Image Files}).
1.21 crook 678:
1.39 anton 679: @cindex --appl-image, command-line option
680: @item --appl-image @i{file}
681: Loads the image @i{file} and leaves all further command-line arguments
1.65 anton 682: to the image (instead of processing them as engine options). This is
683: useful for building executable application images on Unix, built with
1.39 anton 684: @code{gforthmi --application ...}.
685:
1.29 crook 686: @cindex --path, command-line option
687: @cindex -p, command-line option
688: @item --path @i{path}
689: @itemx -p @i{path}
690: Uses @i{path} for searching the image file and Forth source code files
691: instead of the default in the environment variable @code{GFORTHPATH} or
692: the path specified at installation time (e.g.,
693: @file{/usr/local/share/gforth/0.2.0:.}). A path is given as a list of
694: directories, separated by @samp{:} (on Unix) or @samp{;} (on other OSs).
1.21 crook 695:
1.29 crook 696: @cindex --dictionary-size, command-line option
697: @cindex -m, command-line option
698: @cindex @i{size} parameters for command-line options
699: @cindex size of the dictionary and the stacks
700: @item --dictionary-size @i{size}
701: @itemx -m @i{size}
702: Allocate @i{size} space for the Forth dictionary space instead of
703: using the default specified in the image (typically 256K). The
704: @i{size} specification for this and subsequent options consists of
705: an integer and a unit (e.g.,
706: @code{4M}). The unit can be one of @code{b} (bytes), @code{e} (element
707: size, in this case Cells), @code{k} (kilobytes), @code{M} (Megabytes),
708: @code{G} (Gigabytes), and @code{T} (Terabytes). If no unit is specified,
709: @code{e} is used.
1.21 crook 710:
1.29 crook 711: @cindex --data-stack-size, command-line option
712: @cindex -d, command-line option
713: @item --data-stack-size @i{size}
714: @itemx -d @i{size}
715: Allocate @i{size} space for the data stack instead of using the
716: default specified in the image (typically 16K).
1.21 crook 717:
1.29 crook 718: @cindex --return-stack-size, command-line option
719: @cindex -r, command-line option
720: @item --return-stack-size @i{size}
721: @itemx -r @i{size}
722: Allocate @i{size} space for the return stack instead of using the
723: default specified in the image (typically 15K).
1.21 crook 724:
1.29 crook 725: @cindex --fp-stack-size, command-line option
726: @cindex -f, command-line option
727: @item --fp-stack-size @i{size}
728: @itemx -f @i{size}
729: Allocate @i{size} space for the floating point stack instead of
730: using the default specified in the image (typically 15.5K). In this case
731: the unit specifier @code{e} refers to floating point numbers.
1.21 crook 732:
1.48 anton 733: @cindex --locals-stack-size, command-line option
734: @cindex -l, command-line option
735: @item --locals-stack-size @i{size}
736: @itemx -l @i{size}
737: Allocate @i{size} space for the locals stack instead of using the
738: default specified in the image (typically 14.5K).
739:
740: @cindex -h, command-line option
741: @cindex --help, command-line option
742: @item --help
743: @itemx -h
744: Print a message about the command-line options
745:
746: @cindex -v, command-line option
747: @cindex --version, command-line option
748: @item --version
749: @itemx -v
750: Print version and exit
751:
752: @cindex --debug, command-line option
753: @item --debug
754: Print some information useful for debugging on startup.
755:
756: @cindex --offset-image, command-line option
757: @item --offset-image
758: Start the dictionary at a slightly different position than would be used
759: otherwise (useful for creating data-relocatable images,
760: @pxref{Data-Relocatable Image Files}).
761:
762: @cindex --no-offset-im, command-line option
763: @item --no-offset-im
764: Start the dictionary at the normal position.
765:
766: @cindex --clear-dictionary, command-line option
767: @item --clear-dictionary
768: Initialize all bytes in the dictionary to 0 before loading the image
769: (@pxref{Data-Relocatable Image Files}).
770:
771: @cindex --die-on-signal, command-line-option
772: @item --die-on-signal
773: Normally Gforth handles most signals (e.g., the user interrupt SIGINT,
774: or the segmentation violation SIGSEGV) by translating it into a Forth
775: @code{THROW}. With this option, Gforth exits if it receives such a
776: signal. This option is useful when the engine and/or the image might be
777: severely broken (such that it causes another signal before recovering
778: from the first); this option avoids endless loops in such cases.
1.109 anton 779:
1.119 anton 780: @cindex --no-dynamic, command-line option
781: @cindex --dynamic, command-line option
1.109 anton 782: @item --no-dynamic
783: @item --dynamic
784: Disable or enable dynamic superinstructions with replication
785: (@pxref{Dynamic Superinstructions}).
786:
1.119 anton 787: @cindex --no-super, command-line option
1.109 anton 788: @item --no-super
1.110 anton 789: Disable dynamic superinstructions, use just dynamic replication; this is
790: useful if you want to patch threaded code (@pxref{Dynamic
791: Superinstructions}).
1.119 anton 792:
793: @cindex --ss-number, command-line option
794: @item --ss-number=@var{N}
795: Use only the first @var{N} static superinstructions compiled into the
796: engine (default: use them all; note that only @code{gforth-fast} has
797: any). This option is useful for measuring the performance impact of
798: static superinstructions.
799:
800: @cindex --ss-min-..., command-line options
801: @item --ss-min-codesize
802: @item --ss-min-ls
803: @item --ss-min-lsu
804: @item --ss-min-nexts
805: Use specified metric for determining the cost of a primitive or static
806: superinstruction for static superinstruction selection. @code{Codesize}
807: is the native code size of the primive or static superinstruction,
808: @code{ls} is the number of loads and stores, @code{lsu} is the number of
809: loads, stores, and updates, and @code{nexts} is the number of dispatches
810: (not taking dynamic superinstructions into account), i.e. every
811: primitive or static superinstruction has cost 1. Default:
812: @code{codesize} if you use dynamic code generation, otherwise
813: @code{nexts}.
814:
815: @cindex --ss-greedy, command-line option
816: @item --ss-greedy
817: This option is useful for measuring the performance impact of static
818: superinstructions. By default, an optimal shortest-path algorithm is
819: used for selecting static superinstructions. With @option{--ss-greedy}
820: this algorithm is modified to assume that anything after the static
821: superinstruction currently under consideration is not combined into
822: static superinstructions. With @option{--ss-min-nexts} this produces
823: the same result as a greedy algorithm that always selects the longest
824: superinstruction available at the moment. E.g., if there are
825: superinstructions AB and BCD, then for the sequence A B C D the optimal
826: algorithm will select A BCD and the greedy algorithm will select AB C D.
827:
828: @cindex --print-metrics, command-line option
829: @item --print-metrics
830: Prints some metrics used during static superinstruction selection:
831: @code{code size} is the actual size of the dynamically generated code.
832: @code{Metric codesize} is the sum of the codesize metrics as seen by
833: static superinstruction selection; there is a difference from @code{code
834: size}, because not all primitives and static superinstructions are
835: compiled into dynamically generated code, and because of markers. The
836: other metrics correspond to the @option{ss-min-...} options. This
837: option is useful for evaluating the effects of the @option{--ss-...}
838: options.
1.109 anton 839:
1.48 anton 840: @end table
841:
842: @cindex loading files at startup
843: @cindex executing code on startup
844: @cindex batch processing with Gforth
845: As explained above, the image-specific command-line arguments for the
846: default image @file{gforth.fi} consist of a sequence of filenames and
847: @code{-e @var{forth-code}} options that are interpreted in the sequence
848: in which they are given. The @code{-e @var{forth-code}} or
1.121 anton 849: @code{--evaluate @var{forth-code}} option evaluates the Forth code. This
850: option takes only one argument; if you want to evaluate more Forth
851: words, you have to quote them or use @code{-e} several times. To exit
1.48 anton 852: after processing the command line (instead of entering interactive mode)
1.121 anton 853: append @code{-e bye} to the command line. You can also process the
854: command-line arguments with a Forth program (@pxref{OS command line
855: arguments}).
1.48 anton 856:
857: @cindex versions, invoking other versions of Gforth
858: If you have several versions of Gforth installed, @code{gforth} will
859: invoke the version that was installed last. @code{gforth-@i{version}}
860: invokes a specific version. If your environment contains the variable
861: @code{GFORTHPATH}, you may want to override it by using the
862: @code{--path} option.
863:
864: Not yet implemented:
865: On startup the system first executes the system initialization file
866: (unless the option @code{--no-init-file} is given; note that the system
867: resulting from using this option may not be ANS Forth conformant). Then
868: the user initialization file @file{.gforth.fs} is executed, unless the
1.62 crook 869: option @code{--no-rc} is given; this file is searched for in @file{.},
1.48 anton 870: then in @file{~}, then in the normal path (see above).
871:
872:
873:
874: @comment ----------------------------------------------
875: @node Leaving Gforth, Command-line editing, Invoking Gforth, Gforth Environment
876: @section Leaving Gforth
877: @cindex Gforth - leaving
878: @cindex leaving Gforth
879:
880: You can leave Gforth by typing @code{bye} or @kbd{Ctrl-d} (at the start
881: of a line) or (if you invoked Gforth with the @code{--die-on-signal}
882: option) @kbd{Ctrl-c}. When you leave Gforth, all of your definitions and
1.49 anton 883: data are discarded. For ways of saving the state of the system before
884: leaving Gforth see @ref{Image Files}.
1.48 anton 885:
886: doc-bye
887:
888:
889: @comment ----------------------------------------------
1.65 anton 890: @node Command-line editing, Environment variables, Leaving Gforth, Gforth Environment
1.48 anton 891: @section Command-line editing
892: @cindex command-line editing
893:
894: Gforth maintains a history file that records every line that you type to
895: the text interpreter. This file is preserved between sessions, and is
896: used to provide a command-line recall facility; if you type @kbd{Ctrl-P}
897: repeatedly you can recall successively older commands from this (or
898: previous) session(s). The full list of command-line editing facilities is:
899:
900: @itemize @bullet
901: @item
902: @kbd{Ctrl-p} (``previous'') (or up-arrow) to recall successively older
903: commands from the history buffer.
904: @item
905: @kbd{Ctrl-n} (``next'') (or down-arrow) to recall successively newer commands
906: from the history buffer.
907: @item
908: @kbd{Ctrl-f} (or right-arrow) to move the cursor right, non-destructively.
909: @item
910: @kbd{Ctrl-b} (or left-arrow) to move the cursor left, non-destructively.
911: @item
912: @kbd{Ctrl-h} (backspace) to delete the character to the left of the cursor,
913: closing up the line.
914: @item
915: @kbd{Ctrl-k} to delete (``kill'') from the cursor to the end of the line.
916: @item
917: @kbd{Ctrl-a} to move the cursor to the start of the line.
918: @item
919: @kbd{Ctrl-e} to move the cursor to the end of the line.
920: @item
921: @key{RET} (@kbd{Ctrl-m}) or @key{LFD} (@kbd{Ctrl-j}) to submit the current
922: line.
923: @item
924: @key{TAB} to step through all possible full-word completions of the word
925: currently being typed.
926: @item
1.65 anton 927: @kbd{Ctrl-d} on an empty line line to terminate Gforth (gracefully,
928: using @code{bye}).
929: @item
930: @kbd{Ctrl-x} (or @code{Ctrl-d} on a non-empty line) to delete the
931: character under the cursor.
1.48 anton 932: @end itemize
933:
934: When editing, displayable characters are inserted to the left of the
935: cursor position; the line is always in ``insert'' (as opposed to
936: ``overstrike'') mode.
937:
938: @cindex history file
939: @cindex @file{.gforth-history}
940: On Unix systems, the history file is @file{~/.gforth-history} by
941: default@footnote{i.e. it is stored in the user's home directory.}. You
942: can find out the name and location of your history file using:
943:
944: @example
945: history-file type \ Unix-class systems
946:
947: history-file type \ Other systems
948: history-dir type
949: @end example
950:
951: If you enter long definitions by hand, you can use a text editor to
952: paste them out of the history file into a Forth source file for reuse at
953: a later time.
954:
955: Gforth never trims the size of the history file, so you should do this
956: periodically, if necessary.
957:
958: @comment this is all defined in history.fs
959: @comment NAC TODO the ctrl-D behaviour can either do a bye or a beep.. how is that option
960: @comment chosen?
961:
962:
963: @comment ----------------------------------------------
1.65 anton 964: @node Environment variables, Gforth Files, Command-line editing, Gforth Environment
1.48 anton 965: @section Environment variables
966: @cindex environment variables
967:
968: Gforth uses these environment variables:
969:
970: @itemize @bullet
971: @item
972: @cindex @code{GFORTHHIST} -- environment variable
973: @code{GFORTHHIST} -- (Unix systems only) specifies the directory in which to
974: open/create the history file, @file{.gforth-history}. Default:
975: @code{$HOME}.
976:
977: @item
978: @cindex @code{GFORTHPATH} -- environment variable
979: @code{GFORTHPATH} -- specifies the path used when searching for the gforth image file and
980: for Forth source-code files.
981:
982: @item
1.147 anton 983: @cindex @code{LANG} -- environment variable
984: @code{LANG} -- see @code{LC_CTYPE}
985:
986: @item
987: @cindex @code{LC_ALL} -- environment variable
988: @code{LC_ALL} -- see @code{LC_CTYPE}
989:
990: @item
991: @cindex @code{LC_CTYPE} -- environment variable
992: @code{LC_CTYPE} -- If this variable contains ``UTF-8'' on Gforth
993: startup, Gforth uses the UTF-8 encoding for strings internally and
994: expects its input and produces its output in UTF-8 encoding, otherwise
995: the encoding is 8bit (see @pxref{Xchars and Unicode}). If this
996: environment variable is unset, Gforth looks in @code{LC_ALL}, and if
997: that is unset, in @code{LANG}.
998:
999: @item
1.129 anton 1000: @cindex @code{GFORTHSYSTEMPREFIX} -- environment variable
1001:
1002: @code{GFORTHSYSTEMPREFIX} -- specifies what to prepend to the argument
1003: of @code{system} before passing it to C's @code{system()}. Default:
1.130 anton 1004: @code{"./$COMSPEC /c "} on Windows, @code{""} on other OSs. The prefix
1.129 anton 1005: and the command are directly concatenated, so if a space between them is
1006: necessary, append it to the prefix.
1007:
1008: @item
1.48 anton 1009: @cindex @code{GFORTH} -- environment variable
1.49 anton 1010: @code{GFORTH} -- used by @file{gforthmi}, @xref{gforthmi}.
1.48 anton 1011:
1012: @item
1013: @cindex @code{GFORTHD} -- environment variable
1.62 crook 1014: @code{GFORTHD} -- used by @file{gforthmi}, @xref{gforthmi}.
1.48 anton 1015:
1016: @item
1017: @cindex @code{TMP}, @code{TEMP} - environment variable
1018: @code{TMP}, @code{TEMP} - (non-Unix systems only) used as a potential
1019: location for the history file.
1020: @end itemize
1021:
1022: @comment also POSIXELY_CORRECT LINES COLUMNS HOME but no interest in
1023: @comment mentioning these.
1024:
1025: All the Gforth environment variables default to sensible values if they
1026: are not set.
1027:
1028:
1029: @comment ----------------------------------------------
1.112 anton 1030: @node Gforth Files, Gforth in pipes, Environment variables, Gforth Environment
1.48 anton 1031: @section Gforth files
1032: @cindex Gforth files
1033:
1034: When you install Gforth on a Unix system, it installs files in these
1035: locations by default:
1036:
1037: @itemize @bullet
1038: @item
1039: @file{/usr/local/bin/gforth}
1040: @item
1041: @file{/usr/local/bin/gforthmi}
1042: @item
1043: @file{/usr/local/man/man1/gforth.1} - man page.
1044: @item
1045: @file{/usr/local/info} - the Info version of this manual.
1046: @item
1047: @file{/usr/local/lib/gforth/<version>/...} - Gforth @file{.fi} files.
1048: @item
1049: @file{/usr/local/share/gforth/<version>/TAGS} - Emacs TAGS file.
1050: @item
1051: @file{/usr/local/share/gforth/<version>/...} - Gforth source files.
1052: @item
1053: @file{.../emacs/site-lisp/gforth.el} - Emacs gforth mode.
1054: @end itemize
1055:
1056: You can select different places for installation by using
1057: @code{configure} options (listed with @code{configure --help}).
1058:
1059: @comment ----------------------------------------------
1.112 anton 1060: @node Gforth in pipes, Startup speed, Gforth Files, Gforth Environment
1061: @section Gforth in pipes
1062: @cindex pipes, Gforth as part of
1063:
1064: Gforth can be used in pipes created elsewhere (described here). It can
1065: also create pipes on its own (@pxref{Pipes}).
1066:
1067: @cindex input from pipes
1068: If you pipe into Gforth, your program should read with @code{read-file}
1069: or @code{read-line} from @code{stdin} (@pxref{General files}).
1070: @code{Key} does not recognize the end of input. Words like
1071: @code{accept} echo the input and are therefore usually not useful for
1072: reading from a pipe. You have to invoke the Forth program with an OS
1073: command-line option, as you have no chance to use the Forth command line
1074: (the text interpreter would try to interpret the pipe input).
1075:
1076: @cindex output in pipes
1077: You can output to a pipe with @code{type}, @code{emit}, @code{cr} etc.
1078:
1079: @cindex silent exiting from Gforth
1080: When you write to a pipe that has been closed at the other end, Gforth
1081: receives a SIGPIPE signal (``pipe broken''). Gforth translates this
1082: into the exception @code{broken-pipe-error}. If your application does
1083: not catch that exception, the system catches it and exits, usually
1084: silently (unless you were working on the Forth command line; then it
1085: prints an error message and exits). This is usually the desired
1086: behaviour.
1087:
1088: If you do not like this behaviour, you have to catch the exception
1089: yourself, and react to it.
1090:
1091: Here's an example of an invocation of Gforth that is usable in a pipe:
1092:
1093: @example
1094: gforth -e ": foo begin pad dup 10 stdin read-file throw dup while \
1095: type repeat ; foo bye"
1096: @end example
1097:
1098: This example just copies the input verbatim to the output. A very
1099: simple pipe containing this example looks like this:
1100:
1101: @example
1102: cat startup.fs |
1103: gforth -e ": foo begin pad dup 80 stdin read-file throw dup while \
1104: type repeat ; foo bye"|
1105: head
1106: @end example
1107:
1108: @cindex stderr and pipes
1109: Pipes involving Gforth's @code{stderr} output do not work.
1110:
1111: @comment ----------------------------------------------
1112: @node Startup speed, , Gforth in pipes, Gforth Environment
1.48 anton 1113: @section Startup speed
1114: @cindex Startup speed
1115: @cindex speed, startup
1116:
1117: If Gforth is used for CGI scripts or in shell scripts, its startup
1118: speed may become a problem. On a 300MHz 21064a under Linux-2.2.13 with
1119: glibc-2.0.7, @code{gforth -e bye} takes about 24.6ms user and 11.3ms
1120: system time.
1121:
1122: If startup speed is a problem, you may consider the following ways to
1123: improve it; or you may consider ways to reduce the number of startups
1.62 crook 1124: (for example, by using Fast-CGI).
1.48 anton 1125:
1.112 anton 1126: An easy step that influences Gforth startup speed is the use of the
1127: @option{--no-dynamic} option; this decreases image loading speed, but
1128: increases compile-time and run-time.
1129:
1130: Another step to improve startup speed is to statically link Gforth, by
1.48 anton 1131: building it with @code{XLDFLAGS=-static}. This requires more memory for
1132: the code and will therefore slow down the first invocation, but
1133: subsequent invocations avoid the dynamic linking overhead. Another
1134: disadvantage is that Gforth won't profit from library upgrades. As a
1135: result, @code{gforth-static -e bye} takes about 17.1ms user and
1136: 8.2ms system time.
1137:
1138: The next step to improve startup speed is to use a non-relocatable image
1.65 anton 1139: (@pxref{Non-Relocatable Image Files}). You can create this image with
1.48 anton 1140: @code{gforth -e "savesystem gforthnr.fi bye"} and later use it with
1141: @code{gforth -i gforthnr.fi ...}. This avoids the relocation overhead
1142: and a part of the copy-on-write overhead. The disadvantage is that the
1.62 crook 1143: non-relocatable image does not work if the OS gives Gforth a different
1.48 anton 1144: address for the dictionary, for whatever reason; so you better provide a
1145: fallback on a relocatable image. @code{gforth-static -i gforthnr.fi -e
1146: bye} takes about 15.3ms user and 7.5ms system time.
1147:
1148: The final step is to disable dictionary hashing in Gforth. Gforth
1149: builds the hash table on startup, which takes much of the startup
1150: overhead. You can do this by commenting out the @code{include hash.fs}
1151: in @file{startup.fs} and everything that requires @file{hash.fs} (at the
1152: moment @file{table.fs} and @file{ekey.fs}) and then doing @code{make}.
1153: The disadvantages are that functionality like @code{table} and
1154: @code{ekey} is missing and that text interpretation (e.g., compiling)
1155: now takes much longer. So, you should only use this method if there is
1156: no significant text interpretation to perform (the script should be
1.62 crook 1157: compiled into the image, amongst other things). @code{gforth-static -i
1.48 anton 1158: gforthnrnh.fi -e bye} takes about 2.1ms user and 6.1ms system time.
1159:
1160: @c ******************************************************************
1161: @node Tutorial, Introduction, Gforth Environment, Top
1162: @chapter Forth Tutorial
1163: @cindex Tutorial
1164: @cindex Forth Tutorial
1165:
1.67 anton 1166: @c Topics from nac's Introduction that could be mentioned:
1167: @c press <ret> after each line
1168: @c Prompt
1169: @c numbers vs. words in dictionary on text interpretation
1170: @c what happens on redefinition
1171: @c parsing words (in particular, defining words)
1172:
1.83 anton 1173: The difference of this chapter from the Introduction
1174: (@pxref{Introduction}) is that this tutorial is more fast-paced, should
1175: be used while sitting in front of a computer, and covers much more
1176: material, but does not explain how the Forth system works.
1177:
1.62 crook 1178: This tutorial can be used with any ANS-compliant Forth; any
1179: Gforth-specific features are marked as such and you can skip them if you
1180: work with another Forth. This tutorial does not explain all features of
1181: Forth, just enough to get you started and give you some ideas about the
1182: facilities available in Forth. Read the rest of the manual and the
1183: standard when you are through this.
1.48 anton 1184:
1185: The intended way to use this tutorial is that you work through it while
1186: sitting in front of the console, take a look at the examples and predict
1187: what they will do, then try them out; if the outcome is not as expected,
1188: find out why (e.g., by trying out variations of the example), so you
1189: understand what's going on. There are also some assignments that you
1190: should solve.
1191:
1192: This tutorial assumes that you have programmed before and know what,
1193: e.g., a loop is.
1194:
1195: @c !! explain compat library
1196:
1197: @menu
1198: * Starting Gforth Tutorial::
1199: * Syntax Tutorial::
1200: * Crash Course Tutorial::
1201: * Stack Tutorial::
1202: * Arithmetics Tutorial::
1203: * Stack Manipulation Tutorial::
1204: * Using files for Forth code Tutorial::
1205: * Comments Tutorial::
1206: * Colon Definitions Tutorial::
1207: * Decompilation Tutorial::
1208: * Stack-Effect Comments Tutorial::
1209: * Types Tutorial::
1210: * Factoring Tutorial::
1211: * Designing the stack effect Tutorial::
1212: * Local Variables Tutorial::
1213: * Conditional execution Tutorial::
1214: * Flags and Comparisons Tutorial::
1215: * General Loops Tutorial::
1216: * Counted loops Tutorial::
1217: * Recursion Tutorial::
1218: * Leaving definitions or loops Tutorial::
1219: * Return Stack Tutorial::
1220: * Memory Tutorial::
1221: * Characters and Strings Tutorial::
1222: * Alignment Tutorial::
1.87 anton 1223: * Files Tutorial::
1.48 anton 1224: * Interpretation and Compilation Semantics and Immediacy Tutorial::
1225: * Execution Tokens Tutorial::
1226: * Exceptions Tutorial::
1227: * Defining Words Tutorial::
1228: * Arrays and Records Tutorial::
1229: * POSTPONE Tutorial::
1230: * Literal Tutorial::
1231: * Advanced macros Tutorial::
1232: * Compilation Tokens Tutorial::
1233: * Wordlists and Search Order Tutorial::
1234: @end menu
1235:
1236: @node Starting Gforth Tutorial, Syntax Tutorial, Tutorial, Tutorial
1237: @section Starting Gforth
1.66 anton 1238: @cindex starting Gforth tutorial
1.48 anton 1239: You can start Gforth by typing its name:
1240:
1241: @example
1242: gforth
1243: @end example
1244:
1245: That puts you into interactive mode; you can leave Gforth by typing
1246: @code{bye}. While in Gforth, you can edit the command line and access
1247: the command line history with cursor keys, similar to bash.
1248:
1249:
1250: @node Syntax Tutorial, Crash Course Tutorial, Starting Gforth Tutorial, Tutorial
1251: @section Syntax
1.66 anton 1252: @cindex syntax tutorial
1.48 anton 1253:
1.171 anton 1254: A @dfn{word} is a sequence of arbitrary characters (except white
1.48 anton 1255: space). Words are separated by white space. E.g., each of the
1256: following lines contains exactly one word:
1257:
1258: @example
1259: word
1260: !@@#$%^&*()
1261: 1234567890
1262: 5!a
1263: @end example
1264:
1265: A frequent beginner's error is to leave away necessary white space,
1266: resulting in an error like @samp{Undefined word}; so if you see such an
1267: error, check if you have put spaces wherever necessary.
1268:
1269: @example
1270: ." hello, world" \ correct
1271: ."hello, world" \ gives an "Undefined word" error
1272: @end example
1273:
1.65 anton 1274: Gforth and most other Forth systems ignore differences in case (they are
1.48 anton 1275: case-insensitive), i.e., @samp{word} is the same as @samp{Word}. If
1276: your system is case-sensitive, you may have to type all the examples
1277: given here in upper case.
1278:
1279:
1280: @node Crash Course Tutorial, Stack Tutorial, Syntax Tutorial, Tutorial
1281: @section Crash Course
1282:
1283: Type
1284:
1285: @example
1286: 0 0 !
1287: here execute
1288: ' catch >body 20 erase abort
1289: ' (quit) >body 20 erase
1290: @end example
1291:
1292: The last two examples are guaranteed to destroy parts of Gforth (and
1293: most other systems), so you better leave Gforth afterwards (if it has
1294: not finished by itself). On some systems you may have to kill gforth
1295: from outside (e.g., in Unix with @code{kill}).
1296:
1297: Now that you know how to produce crashes (and that there's not much to
1298: them), let's learn how to produce meaningful programs.
1299:
1300:
1301: @node Stack Tutorial, Arithmetics Tutorial, Crash Course Tutorial, Tutorial
1302: @section Stack
1.66 anton 1303: @cindex stack tutorial
1.48 anton 1304:
1305: The most obvious feature of Forth is the stack. When you type in a
1306: number, it is pushed on the stack. You can display the content of the
1307: stack with @code{.s}.
1308:
1309: @example
1310: 1 2 .s
1311: 3 .s
1312: @end example
1313:
1314: @code{.s} displays the top-of-stack to the right, i.e., the numbers
1315: appear in @code{.s} output as they appeared in the input.
1316:
1317: You can print the top of stack element with @code{.}.
1318:
1319: @example
1320: 1 2 3 . . .
1321: @end example
1322:
1323: In general, words consume their stack arguments (@code{.s} is an
1324: exception).
1325:
1.141 anton 1326: @quotation Assignment
1.48 anton 1327: What does the stack contain after @code{5 6 7 .}?
1.141 anton 1328: @end quotation
1.48 anton 1329:
1330:
1331: @node Arithmetics Tutorial, Stack Manipulation Tutorial, Stack Tutorial, Tutorial
1332: @section Arithmetics
1.66 anton 1333: @cindex arithmetics tutorial
1.48 anton 1334:
1335: The words @code{+}, @code{-}, @code{*}, @code{/}, and @code{mod} always
1336: operate on the top two stack items:
1337:
1338: @example
1.67 anton 1339: 2 2 .s
1340: + .s
1341: .
1.48 anton 1342: 2 1 - .
1343: 7 3 mod .
1344: @end example
1345:
1346: The operands of @code{-}, @code{/}, and @code{mod} are in the same order
1347: as in the corresponding infix expression (this is generally the case in
1348: Forth).
1349:
1350: Parentheses are superfluous (and not available), because the order of
1351: the words unambiguously determines the order of evaluation and the
1352: operands:
1353:
1354: @example
1355: 3 4 + 5 * .
1356: 3 4 5 * + .
1357: @end example
1358:
1.141 anton 1359: @quotation Assignment
1.48 anton 1360: What are the infix expressions corresponding to the Forth code above?
1361: Write @code{6-7*8+9} in Forth notation@footnote{This notation is also
1362: known as Postfix or RPN (Reverse Polish Notation).}.
1.141 anton 1363: @end quotation
1.48 anton 1364:
1365: To change the sign, use @code{negate}:
1366:
1367: @example
1368: 2 negate .
1369: @end example
1370:
1.141 anton 1371: @quotation Assignment
1.48 anton 1372: Convert -(-3)*4-5 to Forth.
1.141 anton 1373: @end quotation
1.48 anton 1374:
1375: @code{/mod} performs both @code{/} and @code{mod}.
1376:
1377: @example
1378: 7 3 /mod . .
1379: @end example
1380:
1.66 anton 1381: Reference: @ref{Arithmetic}.
1382:
1383:
1.48 anton 1384: @node Stack Manipulation Tutorial, Using files for Forth code Tutorial, Arithmetics Tutorial, Tutorial
1385: @section Stack Manipulation
1.66 anton 1386: @cindex stack manipulation tutorial
1.48 anton 1387:
1388: Stack manipulation words rearrange the data on the stack.
1389:
1390: @example
1391: 1 .s drop .s
1392: 1 .s dup .s drop drop .s
1393: 1 2 .s over .s drop drop drop
1394: 1 2 .s swap .s drop drop
1395: 1 2 3 .s rot .s drop drop drop
1396: @end example
1397:
1398: These are the most important stack manipulation words. There are also
1399: variants that manipulate twice as many stack items:
1400:
1401: @example
1402: 1 2 3 4 .s 2swap .s 2drop 2drop
1403: @end example
1404:
1405: Two more stack manipulation words are:
1406:
1407: @example
1408: 1 2 .s nip .s drop
1409: 1 2 .s tuck .s 2drop drop
1410: @end example
1411:
1.141 anton 1412: @quotation Assignment
1.48 anton 1413: Replace @code{nip} and @code{tuck} with combinations of other stack
1414: manipulation words.
1415:
1416: @example
1417: Given: How do you get:
1418: 1 2 3 3 2 1
1419: 1 2 3 1 2 3 2
1420: 1 2 3 1 2 3 3
1421: 1 2 3 1 3 3
1422: 1 2 3 2 1 3
1423: 1 2 3 4 4 3 2 1
1424: 1 2 3 1 2 3 1 2 3
1425: 1 2 3 4 1 2 3 4 1 2
1426: 1 2 3
1427: 1 2 3 1 2 3 4
1428: 1 2 3 1 3
1429: @end example
1.141 anton 1430: @end quotation
1.48 anton 1431:
1432: @example
1433: 5 dup * .
1434: @end example
1435:
1.141 anton 1436: @quotation Assignment
1.48 anton 1437: Write 17^3 and 17^4 in Forth, without writing @code{17} more than once.
1438: Write a piece of Forth code that expects two numbers on the stack
1439: (@var{a} and @var{b}, with @var{b} on top) and computes
1440: @code{(a-b)(a+1)}.
1.141 anton 1441: @end quotation
1.48 anton 1442:
1.66 anton 1443: Reference: @ref{Stack Manipulation}.
1444:
1445:
1.48 anton 1446: @node Using files for Forth code Tutorial, Comments Tutorial, Stack Manipulation Tutorial, Tutorial
1447: @section Using files for Forth code
1.66 anton 1448: @cindex loading Forth code, tutorial
1449: @cindex files containing Forth code, tutorial
1.48 anton 1450:
1451: While working at the Forth command line is convenient for one-line
1452: examples and short one-off code, you probably want to store your source
1453: code in files for convenient editing and persistence. You can use your
1454: favourite editor (Gforth includes Emacs support, @pxref{Emacs and
1.102 anton 1455: Gforth}) to create @var{file.fs} and use
1.48 anton 1456:
1457: @example
1.102 anton 1458: s" @var{file.fs}" included
1.48 anton 1459: @end example
1460:
1461: to load it into your Forth system. The file name extension I use for
1462: Forth files is @samp{.fs}.
1463:
1464: You can easily start Gforth with some files loaded like this:
1465:
1466: @example
1.102 anton 1467: gforth @var{file1.fs} @var{file2.fs}
1.48 anton 1468: @end example
1469:
1470: If an error occurs during loading these files, Gforth terminates,
1471: whereas an error during @code{INCLUDED} within Gforth usually gives you
1472: a Gforth command line. Starting the Forth system every time gives you a
1473: clean start every time, without interference from the results of earlier
1474: tries.
1475:
1476: I often put all the tests in a file, then load the code and run the
1477: tests with
1478:
1479: @example
1.102 anton 1480: gforth @var{code.fs} @var{tests.fs} -e bye
1.48 anton 1481: @end example
1482:
1483: (often by performing this command with @kbd{C-x C-e} in Emacs). The
1484: @code{-e bye} ensures that Gforth terminates afterwards so that I can
1485: restart this command without ado.
1486:
1487: The advantage of this approach is that the tests can be repeated easily
1488: every time the program ist changed, making it easy to catch bugs
1489: introduced by the change.
1490:
1.66 anton 1491: Reference: @ref{Forth source files}.
1492:
1.48 anton 1493:
1494: @node Comments Tutorial, Colon Definitions Tutorial, Using files for Forth code Tutorial, Tutorial
1495: @section Comments
1.66 anton 1496: @cindex comments tutorial
1.48 anton 1497:
1498: @example
1499: \ That's a comment; it ends at the end of the line
1500: ( Another comment; it ends here: ) .s
1501: @end example
1502:
1503: @code{\} and @code{(} are ordinary Forth words and therefore have to be
1504: separated with white space from the following text.
1505:
1506: @example
1507: \This gives an "Undefined word" error
1508: @end example
1509:
1510: The first @code{)} ends a comment started with @code{(}, so you cannot
1511: nest @code{(}-comments; and you cannot comment out text containing a
1512: @code{)} with @code{( ... )}@footnote{therefore it's a good idea to
1513: avoid @code{)} in word names.}.
1514:
1515: I use @code{\}-comments for descriptive text and for commenting out code
1516: of one or more line; I use @code{(}-comments for describing the stack
1517: effect, the stack contents, or for commenting out sub-line pieces of
1518: code.
1519:
1520: The Emacs mode @file{gforth.el} (@pxref{Emacs and Gforth}) supports
1521: these uses by commenting out a region with @kbd{C-x \}, uncommenting a
1522: region with @kbd{C-u C-x \}, and filling a @code{\}-commented region
1523: with @kbd{M-q}.
1524:
1.66 anton 1525: Reference: @ref{Comments}.
1526:
1.48 anton 1527:
1528: @node Colon Definitions Tutorial, Decompilation Tutorial, Comments Tutorial, Tutorial
1529: @section Colon Definitions
1.66 anton 1530: @cindex colon definitions, tutorial
1531: @cindex definitions, tutorial
1532: @cindex procedures, tutorial
1533: @cindex functions, tutorial
1.48 anton 1534:
1535: are similar to procedures and functions in other programming languages.
1536:
1537: @example
1538: : squared ( n -- n^2 )
1539: dup * ;
1540: 5 squared .
1541: 7 squared .
1542: @end example
1543:
1544: @code{:} starts the colon definition; its name is @code{squared}. The
1545: following comment describes its stack effect. The words @code{dup *}
1546: are not executed, but compiled into the definition. @code{;} ends the
1547: colon definition.
1548:
1549: The newly-defined word can be used like any other word, including using
1550: it in other definitions:
1551:
1552: @example
1553: : cubed ( n -- n^3 )
1554: dup squared * ;
1555: -5 cubed .
1556: : fourth-power ( n -- n^4 )
1557: squared squared ;
1558: 3 fourth-power .
1559: @end example
1560:
1.141 anton 1561: @quotation Assignment
1.48 anton 1562: Write colon definitions for @code{nip}, @code{tuck}, @code{negate}, and
1563: @code{/mod} in terms of other Forth words, and check if they work (hint:
1564: test your tests on the originals first). Don't let the
1565: @samp{redefined}-Messages spook you, they are just warnings.
1.141 anton 1566: @end quotation
1.48 anton 1567:
1.66 anton 1568: Reference: @ref{Colon Definitions}.
1569:
1.48 anton 1570:
1571: @node Decompilation Tutorial, Stack-Effect Comments Tutorial, Colon Definitions Tutorial, Tutorial
1572: @section Decompilation
1.66 anton 1573: @cindex decompilation tutorial
1574: @cindex see tutorial
1.48 anton 1575:
1576: You can decompile colon definitions with @code{see}:
1577:
1578: @example
1579: see squared
1580: see cubed
1581: @end example
1582:
1583: In Gforth @code{see} shows you a reconstruction of the source code from
1584: the executable code. Informations that were present in the source, but
1585: not in the executable code, are lost (e.g., comments).
1586:
1.65 anton 1587: You can also decompile the predefined words:
1588:
1589: @example
1590: see .
1591: see +
1592: @end example
1593:
1594:
1.48 anton 1595: @node Stack-Effect Comments Tutorial, Types Tutorial, Decompilation Tutorial, Tutorial
1596: @section Stack-Effect Comments
1.66 anton 1597: @cindex stack-effect comments, tutorial
1598: @cindex --, tutorial
1.48 anton 1599: By convention the comment after the name of a definition describes the
1.171 anton 1600: stack effect: The part in front of the @samp{--} describes the state of
1.48 anton 1601: the stack before the execution of the definition, i.e., the parameters
1602: that are passed into the colon definition; the part behind the @samp{--}
1603: is the state of the stack after the execution of the definition, i.e.,
1604: the results of the definition. The stack comment only shows the top
1605: stack items that the definition accesses and/or changes.
1606:
1607: You should put a correct stack effect on every definition, even if it is
1608: just @code{( -- )}. You should also add some descriptive comment to
1609: more complicated words (I usually do this in the lines following
1610: @code{:}). If you don't do this, your code becomes unreadable (because
1.117 anton 1611: you have to work through every definition before you can understand
1.48 anton 1612: any).
1613:
1.141 anton 1614: @quotation Assignment
1.48 anton 1615: The stack effect of @code{swap} can be written like this: @code{x1 x2 --
1616: x2 x1}. Describe the stack effect of @code{-}, @code{drop}, @code{dup},
1617: @code{over}, @code{rot}, @code{nip}, and @code{tuck}. Hint: When you
1.65 anton 1618: are done, you can compare your stack effects to those in this manual
1.48 anton 1619: (@pxref{Word Index}).
1.141 anton 1620: @end quotation
1.48 anton 1621:
1622: Sometimes programmers put comments at various places in colon
1623: definitions that describe the contents of the stack at that place (stack
1624: comments); i.e., they are like the first part of a stack-effect
1625: comment. E.g.,
1626:
1627: @example
1628: : cubed ( n -- n^3 )
1629: dup squared ( n n^2 ) * ;
1630: @end example
1631:
1632: In this case the stack comment is pretty superfluous, because the word
1633: is simple enough. If you think it would be a good idea to add such a
1634: comment to increase readability, you should also consider factoring the
1635: word into several simpler words (@pxref{Factoring Tutorial,,
1.60 anton 1636: Factoring}), which typically eliminates the need for the stack comment;
1.48 anton 1637: however, if you decide not to refactor it, then having such a comment is
1638: better than not having it.
1639:
1640: The names of the stack items in stack-effect and stack comments in the
1641: standard, in this manual, and in many programs specify the type through
1642: a type prefix, similar to Fortran and Hungarian notation. The most
1643: frequent prefixes are:
1644:
1645: @table @code
1646: @item n
1647: signed integer
1648: @item u
1649: unsigned integer
1650: @item c
1651: character
1652: @item f
1653: Boolean flags, i.e. @code{false} or @code{true}.
1654: @item a-addr,a-
1655: Cell-aligned address
1656: @item c-addr,c-
1657: Char-aligned address (note that a Char may have two bytes in Windows NT)
1658: @item xt
1659: Execution token, same size as Cell
1660: @item w,x
1661: Cell, can contain an integer or an address. It usually takes 32, 64 or
1662: 16 bits (depending on your platform and Forth system). A cell is more
1663: commonly known as machine word, but the term @emph{word} already means
1664: something different in Forth.
1665: @item d
1666: signed double-cell integer
1667: @item ud
1668: unsigned double-cell integer
1669: @item r
1670: Float (on the FP stack)
1671: @end table
1672:
1673: You can find a more complete list in @ref{Notation}.
1674:
1.141 anton 1675: @quotation Assignment
1.48 anton 1676: Write stack-effect comments for all definitions you have written up to
1677: now.
1.141 anton 1678: @end quotation
1.48 anton 1679:
1680:
1681: @node Types Tutorial, Factoring Tutorial, Stack-Effect Comments Tutorial, Tutorial
1682: @section Types
1.66 anton 1683: @cindex types tutorial
1.48 anton 1684:
1685: In Forth the names of the operations are not overloaded; so similar
1686: operations on different types need different names; e.g., @code{+} adds
1687: integers, and you have to use @code{f+} to add floating-point numbers.
1688: The following prefixes are often used for related operations on
1689: different types:
1690:
1691: @table @code
1692: @item (none)
1693: signed integer
1694: @item u
1695: unsigned integer
1696: @item c
1697: character
1698: @item d
1699: signed double-cell integer
1700: @item ud, du
1701: unsigned double-cell integer
1702: @item 2
1703: two cells (not-necessarily double-cell numbers)
1704: @item m, um
1705: mixed single-cell and double-cell operations
1706: @item f
1707: floating-point (note that in stack comments @samp{f} represents flags,
1.66 anton 1708: and @samp{r} represents FP numbers).
1.48 anton 1709: @end table
1710:
1711: If there are no differences between the signed and the unsigned variant
1712: (e.g., for @code{+}), there is only the prefix-less variant.
1713:
1714: Forth does not perform type checking, neither at compile time, nor at
1715: run time. If you use the wrong oeration, the data are interpreted
1716: incorrectly:
1717:
1718: @example
1719: -1 u.
1720: @end example
1721:
1722: If you have only experience with type-checked languages until now, and
1723: have heard how important type-checking is, don't panic! In my
1724: experience (and that of other Forthers), type errors in Forth code are
1725: usually easy to find (once you get used to it), the increased vigilance
1726: of the programmer tends to catch some harder errors in addition to most
1727: type errors, and you never have to work around the type system, so in
1728: most situations the lack of type-checking seems to be a win (projects to
1729: add type checking to Forth have not caught on).
1730:
1731:
1732: @node Factoring Tutorial, Designing the stack effect Tutorial, Types Tutorial, Tutorial
1733: @section Factoring
1.66 anton 1734: @cindex factoring tutorial
1.48 anton 1735:
1736: If you try to write longer definitions, you will soon find it hard to
1737: keep track of the stack contents. Therefore, good Forth programmers
1738: tend to write only short definitions (e.g., three lines). The art of
1739: finding meaningful short definitions is known as factoring (as in
1740: factoring polynomials).
1741:
1742: Well-factored programs offer additional advantages: smaller, more
1743: general words, are easier to test and debug and can be reused more and
1744: better than larger, specialized words.
1745:
1746: So, if you run into difficulties with stack management, when writing
1747: code, try to define meaningful factors for the word, and define the word
1748: in terms of those. Even if a factor contains only two words, it is
1749: often helpful.
1750:
1.65 anton 1751: Good factoring is not easy, and it takes some practice to get the knack
1752: for it; but even experienced Forth programmers often don't find the
1753: right solution right away, but only when rewriting the program. So, if
1754: you don't come up with a good solution immediately, keep trying, don't
1755: despair.
1.48 anton 1756:
1757: @c example !!
1758:
1759:
1760: @node Designing the stack effect Tutorial, Local Variables Tutorial, Factoring Tutorial, Tutorial
1761: @section Designing the stack effect
1.66 anton 1762: @cindex Stack effect design, tutorial
1763: @cindex design of stack effects, tutorial
1.48 anton 1764:
1765: In other languages you can use an arbitrary order of parameters for a
1.65 anton 1766: function; and since there is only one result, you don't have to deal with
1.48 anton 1767: the order of results, either.
1768:
1.117 anton 1769: In Forth (and other stack-based languages, e.g., PostScript) the
1.48 anton 1770: parameter and result order of a definition is important and should be
1771: designed well. The general guideline is to design the stack effect such
1772: that the word is simple to use in most cases, even if that complicates
1773: the implementation of the word. Some concrete rules are:
1774:
1775: @itemize @bullet
1776:
1777: @item
1778: Words consume all of their parameters (e.g., @code{.}).
1779:
1780: @item
1781: If there is a convention on the order of parameters (e.g., from
1782: mathematics or another programming language), stick with it (e.g.,
1783: @code{-}).
1784:
1785: @item
1786: If one parameter usually requires only a short computation (e.g., it is
1787: a constant), pass it on the top of the stack. Conversely, parameters
1788: that usually require a long sequence of code to compute should be passed
1789: as the bottom (i.e., first) parameter. This makes the code easier to
1.171 anton 1790: read, because the reader does not need to keep track of the bottom item
1.48 anton 1791: through a long sequence of code (or, alternatively, through stack
1.49 anton 1792: manipulations). E.g., @code{!} (store, @pxref{Memory}) expects the
1.48 anton 1793: address on top of the stack because it is usually simpler to compute
1794: than the stored value (often the address is just a variable).
1795:
1796: @item
1797: Similarly, results that are usually consumed quickly should be returned
1798: on the top of stack, whereas a result that is often used in long
1799: computations should be passed as bottom result. E.g., the file words
1800: like @code{open-file} return the error code on the top of stack, because
1801: it is usually consumed quickly by @code{throw}; moreover, the error code
1802: has to be checked before doing anything with the other results.
1803:
1804: @end itemize
1805:
1806: These rules are just general guidelines, don't lose sight of the overall
1807: goal to make the words easy to use. E.g., if the convention rule
1808: conflicts with the computation-length rule, you might decide in favour
1809: of the convention if the word will be used rarely, and in favour of the
1810: computation-length rule if the word will be used frequently (because
1811: with frequent use the cost of breaking the computation-length rule would
1812: be quite high, and frequent use makes it easier to remember an
1813: unconventional order).
1814:
1815: @c example !! structure package
1816:
1.65 anton 1817:
1.48 anton 1818: @node Local Variables Tutorial, Conditional execution Tutorial, Designing the stack effect Tutorial, Tutorial
1819: @section Local Variables
1.66 anton 1820: @cindex local variables, tutorial
1.48 anton 1821:
1822: You can define local variables (@emph{locals}) in a colon definition:
1823:
1824: @example
1825: : swap @{ a b -- b a @}
1826: b a ;
1827: 1 2 swap .s 2drop
1828: @end example
1829:
1830: (If your Forth system does not support this syntax, include
1831: @file{compat/anslocals.fs} first).
1832:
1833: In this example @code{@{ a b -- b a @}} is the locals definition; it
1834: takes two cells from the stack, puts the top of stack in @code{b} and
1835: the next stack element in @code{a}. @code{--} starts a comment ending
1836: with @code{@}}. After the locals definition, using the name of the
1837: local will push its value on the stack. You can leave the comment
1838: part (@code{-- b a}) away:
1839:
1840: @example
1841: : swap ( x1 x2 -- x2 x1 )
1842: @{ a b @} b a ;
1843: @end example
1844:
1845: In Gforth you can have several locals definitions, anywhere in a colon
1846: definition; in contrast, in a standard program you can have only one
1847: locals definition per colon definition, and that locals definition must
1.163 anton 1848: be outside any control structure.
1.48 anton 1849:
1850: With locals you can write slightly longer definitions without running
1851: into stack trouble. However, I recommend trying to write colon
1852: definitions without locals for exercise purposes to help you gain the
1853: essential factoring skills.
1854:
1.141 anton 1855: @quotation Assignment
1.48 anton 1856: Rewrite your definitions until now with locals
1.141 anton 1857: @end quotation
1.48 anton 1858:
1.66 anton 1859: Reference: @ref{Locals}.
1860:
1.48 anton 1861:
1862: @node Conditional execution Tutorial, Flags and Comparisons Tutorial, Local Variables Tutorial, Tutorial
1863: @section Conditional execution
1.66 anton 1864: @cindex conditionals, tutorial
1865: @cindex if, tutorial
1.48 anton 1866:
1867: In Forth you can use control structures only inside colon definitions.
1868: An @code{if}-structure looks like this:
1869:
1870: @example
1871: : abs ( n1 -- +n2 )
1872: dup 0 < if
1873: negate
1874: endif ;
1875: 5 abs .
1876: -5 abs .
1877: @end example
1878:
1879: @code{if} takes a flag from the stack. If the flag is non-zero (true),
1880: the following code is performed, otherwise execution continues after the
1.51 pazsan 1881: @code{endif} (or @code{else}). @code{<} compares the top two stack
1.171 anton 1882: elements and produces a flag:
1.48 anton 1883:
1884: @example
1885: 1 2 < .
1886: 2 1 < .
1887: 1 1 < .
1888: @end example
1889:
1890: Actually the standard name for @code{endif} is @code{then}. This
1891: tutorial presents the examples using @code{endif}, because this is often
1892: less confusing for people familiar with other programming languages
1893: where @code{then} has a different meaning. If your system does not have
1894: @code{endif}, define it with
1895:
1896: @example
1897: : endif postpone then ; immediate
1898: @end example
1899:
1900: You can optionally use an @code{else}-part:
1901:
1902: @example
1903: : min ( n1 n2 -- n )
1904: 2dup < if
1905: drop
1906: else
1907: nip
1908: endif ;
1909: 2 3 min .
1910: 3 2 min .
1911: @end example
1912:
1.141 anton 1913: @quotation Assignment
1.48 anton 1914: Write @code{min} without @code{else}-part (hint: what's the definition
1915: of @code{nip}?).
1.141 anton 1916: @end quotation
1.48 anton 1917:
1.66 anton 1918: Reference: @ref{Selection}.
1919:
1.48 anton 1920:
1921: @node Flags and Comparisons Tutorial, General Loops Tutorial, Conditional execution Tutorial, Tutorial
1922: @section Flags and Comparisons
1.66 anton 1923: @cindex flags tutorial
1924: @cindex comparison tutorial
1.48 anton 1925:
1926: In a false-flag all bits are clear (0 when interpreted as integer). In
1927: a canonical true-flag all bits are set (-1 as a twos-complement signed
1928: integer); in many contexts (e.g., @code{if}) any non-zero value is
1929: treated as true flag.
1930:
1931: @example
1932: false .
1933: true .
1934: true hex u. decimal
1935: @end example
1936:
1937: Comparison words produce canonical flags:
1938:
1939: @example
1940: 1 1 = .
1941: 1 0= .
1942: 0 1 < .
1943: 0 0 < .
1944: -1 1 u< . \ type error, u< interprets -1 as large unsigned number
1945: -1 1 < .
1946: @end example
1947:
1.66 anton 1948: Gforth supports all combinations of the prefixes @code{0 u d d0 du f f0}
1949: (or none) and the comparisons @code{= <> < > <= >=}. Only a part of
1950: these combinations are standard (for details see the standard,
1951: @ref{Numeric comparison}, @ref{Floating Point} or @ref{Word Index}).
1.48 anton 1952:
1.171 anton 1953: You can use @code{and or xor invert} as operations on canonical flags.
1954: Actually they are bitwise operations:
1.48 anton 1955:
1956: @example
1957: 1 2 and .
1958: 1 2 or .
1959: 1 3 xor .
1960: 1 invert .
1961: @end example
1962:
1963: You can convert a zero/non-zero flag into a canonical flag with
1964: @code{0<>} (and complement it on the way with @code{0=}).
1965:
1966: @example
1967: 1 0= .
1968: 1 0<> .
1969: @end example
1970:
1.65 anton 1971: You can use the all-bits-set feature of canonical flags and the bitwise
1.48 anton 1972: operation of the Boolean operations to avoid @code{if}s:
1973:
1974: @example
1975: : foo ( n1 -- n2 )
1976: 0= if
1977: 14
1978: else
1979: 0
1980: endif ;
1981: 0 foo .
1982: 1 foo .
1983:
1984: : foo ( n1 -- n2 )
1985: 0= 14 and ;
1986: 0 foo .
1987: 1 foo .
1988: @end example
1989:
1.141 anton 1990: @quotation Assignment
1.48 anton 1991: Write @code{min} without @code{if}.
1.141 anton 1992: @end quotation
1.48 anton 1993:
1.66 anton 1994: For reference, see @ref{Boolean Flags}, @ref{Numeric comparison}, and
1995: @ref{Bitwise operations}.
1996:
1.48 anton 1997:
1998: @node General Loops Tutorial, Counted loops Tutorial, Flags and Comparisons Tutorial, Tutorial
1999: @section General Loops
1.66 anton 2000: @cindex loops, indefinite, tutorial
1.48 anton 2001:
2002: The endless loop is the most simple one:
2003:
2004: @example
2005: : endless ( -- )
2006: 0 begin
2007: dup . 1+
2008: again ;
2009: endless
2010: @end example
2011:
2012: Terminate this loop by pressing @kbd{Ctrl-C} (in Gforth). @code{begin}
2013: does nothing at run-time, @code{again} jumps back to @code{begin}.
2014:
2015: A loop with one exit at any place looks like this:
2016:
2017: @example
2018: : log2 ( +n1 -- n2 )
2019: \ logarithmus dualis of n1>0, rounded down to the next integer
2020: assert( dup 0> )
2021: 2/ 0 begin
2022: over 0> while
2023: 1+ swap 2/ swap
2024: repeat
2025: nip ;
2026: 7 log2 .
2027: 8 log2 .
2028: @end example
2029:
2030: At run-time @code{while} consumes a flag; if it is 0, execution
1.51 pazsan 2031: continues behind the @code{repeat}; if the flag is non-zero, execution
1.48 anton 2032: continues behind the @code{while}. @code{Repeat} jumps back to
2033: @code{begin}, just like @code{again}.
2034:
2035: In Forth there are many combinations/abbreviations, like @code{1+}.
1.90 anton 2036: However, @code{2/} is not one of them; it shifts its argument right by
1.48 anton 2037: one bit (arithmetic shift right):
2038:
2039: @example
2040: -5 2 / .
2041: -5 2/ .
2042: @end example
2043:
2044: @code{assert(} is no standard word, but you can get it on systems other
2045: then Gforth by including @file{compat/assert.fs}. You can see what it
2046: does by trying
2047:
2048: @example
2049: 0 log2 .
2050: @end example
2051:
2052: Here's a loop with an exit at the end:
2053:
2054: @example
2055: : log2 ( +n1 -- n2 )
2056: \ logarithmus dualis of n1>0, rounded down to the next integer
2057: assert( dup 0 > )
2058: -1 begin
2059: 1+ swap 2/ swap
2060: over 0 <=
2061: until
2062: nip ;
2063: @end example
2064:
2065: @code{Until} consumes a flag; if it is non-zero, execution continues at
2066: the @code{begin}, otherwise after the @code{until}.
2067:
1.141 anton 2068: @quotation Assignment
1.48 anton 2069: Write a definition for computing the greatest common divisor.
1.141 anton 2070: @end quotation
1.48 anton 2071:
1.66 anton 2072: Reference: @ref{Simple Loops}.
2073:
1.48 anton 2074:
2075: @node Counted loops Tutorial, Recursion Tutorial, General Loops Tutorial, Tutorial
2076: @section Counted loops
1.66 anton 2077: @cindex loops, counted, tutorial
1.48 anton 2078:
2079: @example
2080: : ^ ( n1 u -- n )
1.171 anton 2081: \ n = the uth power of n1
1.48 anton 2082: 1 swap 0 u+do
2083: over *
2084: loop
2085: nip ;
2086: 3 2 ^ .
2087: 4 3 ^ .
2088: @end example
2089:
2090: @code{U+do} (from @file{compat/loops.fs}, if your Forth system doesn't
2091: have it) takes two numbers of the stack @code{( u3 u4 -- )}, and then
2092: performs the code between @code{u+do} and @code{loop} for @code{u3-u4}
2093: times (or not at all, if @code{u3-u4<0}).
2094:
2095: You can see the stack effect design rules at work in the stack effect of
2096: the loop start words: Since the start value of the loop is more
2097: frequently constant than the end value, the start value is passed on
2098: the top-of-stack.
2099:
2100: You can access the counter of a counted loop with @code{i}:
2101:
2102: @example
2103: : fac ( u -- u! )
2104: 1 swap 1+ 1 u+do
2105: i *
2106: loop ;
2107: 5 fac .
2108: 7 fac .
2109: @end example
2110:
2111: There is also @code{+do}, which expects signed numbers (important for
2112: deciding whether to enter the loop).
2113:
1.141 anton 2114: @quotation Assignment
1.48 anton 2115: Write a definition for computing the nth Fibonacci number.
1.141 anton 2116: @end quotation
1.48 anton 2117:
1.65 anton 2118: You can also use increments other than 1:
2119:
2120: @example
2121: : up2 ( n1 n2 -- )
2122: +do
2123: i .
2124: 2 +loop ;
2125: 10 0 up2
2126:
2127: : down2 ( n1 n2 -- )
2128: -do
2129: i .
2130: 2 -loop ;
2131: 0 10 down2
2132: @end example
1.48 anton 2133:
1.66 anton 2134: Reference: @ref{Counted Loops}.
2135:
1.48 anton 2136:
2137: @node Recursion Tutorial, Leaving definitions or loops Tutorial, Counted loops Tutorial, Tutorial
2138: @section Recursion
1.66 anton 2139: @cindex recursion tutorial
1.48 anton 2140:
2141: Usually the name of a definition is not visible in the definition; but
2142: earlier definitions are usually visible:
2143:
2144: @example
1.166 anton 2145: 1 0 / . \ "Floating-point unidentified fault" in Gforth on some platforms
1.48 anton 2146: : / ( n1 n2 -- n )
2147: dup 0= if
2148: -10 throw \ report division by zero
2149: endif
2150: / \ old version
2151: ;
2152: 1 0 /
2153: @end example
2154:
2155: For recursive definitions you can use @code{recursive} (non-standard) or
2156: @code{recurse}:
2157:
2158: @example
2159: : fac1 ( n -- n! ) recursive
2160: dup 0> if
2161: dup 1- fac1 *
2162: else
2163: drop 1
2164: endif ;
2165: 7 fac1 .
2166:
2167: : fac2 ( n -- n! )
2168: dup 0> if
2169: dup 1- recurse *
2170: else
2171: drop 1
2172: endif ;
2173: 8 fac2 .
2174: @end example
2175:
1.141 anton 2176: @quotation Assignment
1.48 anton 2177: Write a recursive definition for computing the nth Fibonacci number.
1.141 anton 2178: @end quotation
1.48 anton 2179:
1.66 anton 2180: Reference (including indirect recursion): @xref{Calls and returns}.
2181:
1.48 anton 2182:
2183: @node Leaving definitions or loops Tutorial, Return Stack Tutorial, Recursion Tutorial, Tutorial
2184: @section Leaving definitions or loops
1.66 anton 2185: @cindex leaving definitions, tutorial
2186: @cindex leaving loops, tutorial
1.48 anton 2187:
2188: @code{EXIT} exits the current definition right away. For every counted
2189: loop that is left in this way, an @code{UNLOOP} has to be performed
2190: before the @code{EXIT}:
2191:
2192: @c !! real examples
2193: @example
2194: : ...
2195: ... u+do
2196: ... if
2197: ... unloop exit
2198: endif
2199: ...
2200: loop
2201: ... ;
2202: @end example
2203:
2204: @code{LEAVE} leaves the innermost counted loop right away:
2205:
2206: @example
2207: : ...
2208: ... u+do
2209: ... if
2210: ... leave
2211: endif
2212: ...
2213: loop
2214: ... ;
2215: @end example
2216:
1.65 anton 2217: @c !! example
1.48 anton 2218:
1.66 anton 2219: Reference: @ref{Calls and returns}, @ref{Counted Loops}.
2220:
2221:
1.48 anton 2222: @node Return Stack Tutorial, Memory Tutorial, Leaving definitions or loops Tutorial, Tutorial
2223: @section Return Stack
1.66 anton 2224: @cindex return stack tutorial
1.48 anton 2225:
2226: In addition to the data stack Forth also has a second stack, the return
2227: stack; most Forth systems store the return addresses of procedure calls
2228: there (thus its name). Programmers can also use this stack:
2229:
2230: @example
2231: : foo ( n1 n2 -- )
2232: .s
2233: >r .s
1.50 anton 2234: r@@ .
1.48 anton 2235: >r .s
1.50 anton 2236: r@@ .
1.48 anton 2237: r> .
1.50 anton 2238: r@@ .
1.48 anton 2239: r> . ;
2240: 1 2 foo
2241: @end example
2242:
2243: @code{>r} takes an element from the data stack and pushes it onto the
2244: return stack; conversely, @code{r>} moves an elementm from the return to
2245: the data stack; @code{r@@} pushes a copy of the top of the return stack
1.148 anton 2246: on the data stack.
1.48 anton 2247:
2248: Forth programmers usually use the return stack for storing data
2249: temporarily, if using the data stack alone would be too complex, and
2250: factoring and locals are not an option:
2251:
2252: @example
2253: : 2swap ( x1 x2 x3 x4 -- x3 x4 x1 x2 )
2254: rot >r rot r> ;
2255: @end example
2256:
2257: The return address of the definition and the loop control parameters of
2258: counted loops usually reside on the return stack, so you have to take
2259: all items, that you have pushed on the return stack in a colon
2260: definition or counted loop, from the return stack before the definition
2261: or loop ends. You cannot access items that you pushed on the return
2262: stack outside some definition or loop within the definition of loop.
2263:
2264: If you miscount the return stack items, this usually ends in a crash:
2265:
2266: @example
2267: : crash ( n -- )
2268: >r ;
2269: 5 crash
2270: @end example
2271:
2272: You cannot mix using locals and using the return stack (according to the
2273: standard; Gforth has no problem). However, they solve the same
2274: problems, so this shouldn't be an issue.
2275:
1.141 anton 2276: @quotation Assignment
1.48 anton 2277: Can you rewrite any of the definitions you wrote until now in a better
2278: way using the return stack?
1.141 anton 2279: @end quotation
1.48 anton 2280:
1.66 anton 2281: Reference: @ref{Return stack}.
2282:
1.48 anton 2283:
2284: @node Memory Tutorial, Characters and Strings Tutorial, Return Stack Tutorial, Tutorial
2285: @section Memory
1.66 anton 2286: @cindex memory access/allocation tutorial
1.48 anton 2287:
2288: You can create a global variable @code{v} with
2289:
2290: @example
2291: variable v ( -- addr )
2292: @end example
2293:
2294: @code{v} pushes the address of a cell in memory on the stack. This cell
2295: was reserved by @code{variable}. You can use @code{!} (store) to store
2296: values into this cell and @code{@@} (fetch) to load the value from the
2297: stack into memory:
2298:
2299: @example
2300: v .
2301: 5 v ! .s
1.50 anton 2302: v @@ .
1.48 anton 2303: @end example
2304:
1.65 anton 2305: You can see a raw dump of memory with @code{dump}:
2306:
2307: @example
2308: v 1 cells .s dump
2309: @end example
2310:
2311: @code{Cells ( n1 -- n2 )} gives you the number of bytes (or, more
2312: generally, address units (aus)) that @code{n1 cells} occupy. You can
2313: also reserve more memory:
1.48 anton 2314:
2315: @example
2316: create v2 20 cells allot
1.65 anton 2317: v2 20 cells dump
1.48 anton 2318: @end example
2319:
1.65 anton 2320: creates a word @code{v2} and reserves 20 uninitialized cells; the
2321: address pushed by @code{v2} points to the start of these 20 cells. You
2322: can use address arithmetic to access these cells:
1.48 anton 2323:
2324: @example
2325: 3 v2 5 cells + !
1.65 anton 2326: v2 20 cells dump
1.48 anton 2327: @end example
2328:
2329: You can reserve and initialize memory with @code{,}:
2330:
2331: @example
2332: create v3
2333: 5 , 4 , 3 , 2 , 1 ,
1.50 anton 2334: v3 @@ .
2335: v3 cell+ @@ .
2336: v3 2 cells + @@ .
1.65 anton 2337: v3 5 cells dump
1.48 anton 2338: @end example
2339:
1.141 anton 2340: @quotation Assignment
1.48 anton 2341: Write a definition @code{vsum ( addr u -- n )} that computes the sum of
2342: @code{u} cells, with the first of these cells at @code{addr}, the next
2343: one at @code{addr cell+} etc.
1.141 anton 2344: @end quotation
1.48 anton 2345:
2346: You can also reserve memory without creating a new word:
2347:
2348: @example
1.60 anton 2349: here 10 cells allot .
2350: here .
1.48 anton 2351: @end example
2352:
2353: @code{Here} pushes the start address of the memory area. You should
2354: store it somewhere, or you will have a hard time finding the memory area
2355: again.
2356:
2357: @code{Allot} manages dictionary memory. The dictionary memory contains
2358: the system's data structures for words etc. on Gforth and most other
2359: Forth systems. It is managed like a stack: You can free the memory that
2360: you have just @code{allot}ed with
2361:
2362: @example
2363: -10 cells allot
1.60 anton 2364: here .
1.48 anton 2365: @end example
2366:
2367: Note that you cannot do this if you have created a new word in the
2368: meantime (because then your @code{allot}ed memory is no longer on the
2369: top of the dictionary ``stack'').
2370:
2371: Alternatively, you can use @code{allocate} and @code{free} which allow
2372: freeing memory in any order:
2373:
2374: @example
2375: 10 cells allocate throw .s
2376: 20 cells allocate throw .s
2377: swap
2378: free throw
2379: free throw
2380: @end example
2381:
2382: The @code{throw}s deal with errors (e.g., out of memory).
2383:
1.65 anton 2384: And there is also a
2385: @uref{http://www.complang.tuwien.ac.at/forth/garbage-collection.zip,
2386: garbage collector}, which eliminates the need to @code{free} memory
2387: explicitly.
1.48 anton 2388:
1.66 anton 2389: Reference: @ref{Memory}.
2390:
1.48 anton 2391:
2392: @node Characters and Strings Tutorial, Alignment Tutorial, Memory Tutorial, Tutorial
2393: @section Characters and Strings
1.66 anton 2394: @cindex strings tutorial
2395: @cindex characters tutorial
1.48 anton 2396:
2397: On the stack characters take up a cell, like numbers. In memory they
2398: have their own size (one 8-bit byte on most systems), and therefore
2399: require their own words for memory access:
2400:
2401: @example
2402: create v4
2403: 104 c, 97 c, 108 c, 108 c, 111 c,
1.50 anton 2404: v4 4 chars + c@@ .
1.65 anton 2405: v4 5 chars dump
1.48 anton 2406: @end example
2407:
2408: The preferred representation of strings on the stack is @code{addr
2409: u-count}, where @code{addr} is the address of the first character and
2410: @code{u-count} is the number of characters in the string.
2411:
2412: @example
2413: v4 5 type
2414: @end example
2415:
2416: You get a string constant with
2417:
2418: @example
2419: s" hello, world" .s
2420: type
2421: @end example
2422:
2423: Make sure you have a space between @code{s"} and the string; @code{s"}
2424: is a normal Forth word and must be delimited with white space (try what
2425: happens when you remove the space).
2426:
2427: However, this interpretive use of @code{s"} is quite restricted: the
2428: string exists only until the next call of @code{s"} (some Forth systems
2429: keep more than one of these strings, but usually they still have a
1.62 crook 2430: limited lifetime).
1.48 anton 2431:
2432: @example
2433: s" hello," s" world" .s
2434: type
2435: type
2436: @end example
2437:
1.62 crook 2438: You can also use @code{s"} in a definition, and the resulting
2439: strings then live forever (well, for as long as the definition):
1.48 anton 2440:
2441: @example
2442: : foo s" hello," s" world" ;
2443: foo .s
2444: type
2445: type
2446: @end example
2447:
1.141 anton 2448: @quotation Assignment
1.48 anton 2449: @code{Emit ( c -- )} types @code{c} as character (not a number).
2450: Implement @code{type ( addr u -- )}.
1.141 anton 2451: @end quotation
1.48 anton 2452:
1.66 anton 2453: Reference: @ref{Memory Blocks}.
2454:
2455:
1.84 pazsan 2456: @node Alignment Tutorial, Files Tutorial, Characters and Strings Tutorial, Tutorial
1.48 anton 2457: @section Alignment
1.66 anton 2458: @cindex alignment tutorial
2459: @cindex memory alignment tutorial
1.48 anton 2460:
2461: On many processors cells have to be aligned in memory, if you want to
2462: access them with @code{@@} and @code{!} (and even if the processor does
1.62 crook 2463: not require alignment, access to aligned cells is faster).
1.48 anton 2464:
2465: @code{Create} aligns @code{here} (i.e., the place where the next
2466: allocation will occur, and that the @code{create}d word points to).
2467: Likewise, the memory produced by @code{allocate} starts at an aligned
2468: address. Adding a number of @code{cells} to an aligned address produces
2469: another aligned address.
2470:
2471: However, address arithmetic involving @code{char+} and @code{chars} can
2472: create an address that is not cell-aligned. @code{Aligned ( addr --
2473: a-addr )} produces the next aligned address:
2474:
2475: @example
1.50 anton 2476: v3 char+ aligned .s @@ .
2477: v3 char+ .s @@ .
1.48 anton 2478: @end example
2479:
2480: Similarly, @code{align} advances @code{here} to the next aligned
2481: address:
2482:
2483: @example
2484: create v5 97 c,
2485: here .
2486: align here .
2487: 1000 ,
2488: @end example
2489:
2490: Note that you should use aligned addresses even if your processor does
2491: not require them, if you want your program to be portable.
2492:
1.66 anton 2493: Reference: @ref{Address arithmetic}.
2494:
1.48 anton 2495:
1.84 pazsan 2496: @node Files Tutorial, Interpretation and Compilation Semantics and Immediacy Tutorial, Alignment Tutorial, Tutorial
2497: @section Files
2498: @cindex files tutorial
2499:
2500: This section gives a short introduction into how to use files inside
2501: Forth. It's broken up into five easy steps:
2502:
2503: @enumerate 1
2504: @item Opened an ASCII text file for input
2505: @item Opened a file for output
2506: @item Read input file until string matched (or some other condition matched)
2507: @item Wrote some lines from input ( modified or not) to output
2508: @item Closed the files.
2509: @end enumerate
2510:
1.153 anton 2511: Reference: @ref{General files}.
2512:
1.84 pazsan 2513: @subsection Open file for input
2514:
2515: @example
2516: s" foo.in" r/o open-file throw Value fd-in
2517: @end example
2518:
2519: @subsection Create file for output
2520:
2521: @example
2522: s" foo.out" w/o create-file throw Value fd-out
2523: @end example
2524:
2525: The available file modes are r/o for read-only access, r/w for
2526: read-write access, and w/o for write-only access. You could open both
2527: files with r/w, too, if you like. All file words return error codes; for
2528: most applications, it's best to pass there error codes with @code{throw}
2529: to the outer error handler.
2530:
2531: If you want words for opening and assigning, define them as follows:
2532:
2533: @example
2534: 0 Value fd-in
2535: 0 Value fd-out
2536: : open-input ( addr u -- ) r/o open-file throw to fd-in ;
2537: : open-output ( addr u -- ) w/o create-file throw to fd-out ;
2538: @end example
2539:
2540: Usage example:
2541:
2542: @example
2543: s" foo.in" open-input
2544: s" foo.out" open-output
2545: @end example
2546:
2547: @subsection Scan file for a particular line
2548:
2549: @example
2550: 256 Constant max-line
2551: Create line-buffer max-line 2 + allot
2552:
2553: : scan-file ( addr u -- )
2554: begin
2555: line-buffer max-line fd-in read-line throw
2556: while
2557: >r 2dup line-buffer r> compare 0=
2558: until
2559: else
2560: drop
2561: then
2562: 2drop ;
2563: @end example
2564:
2565: @code{read-line ( addr u1 fd -- u2 flag ior )} reads up to u1 bytes into
1.94 anton 2566: the buffer at addr, and returns the number of bytes read, a flag that is
2567: false when the end of file is reached, and an error code.
1.84 pazsan 2568:
2569: @code{compare ( addr1 u1 addr2 u2 -- n )} compares two strings and
2570: returns zero if both strings are equal. It returns a positive number if
2571: the first string is lexically greater, a negative if the second string
2572: is lexically greater.
2573:
2574: We haven't seen this loop here; it has two exits. Since the @code{while}
2575: exits with the number of bytes read on the stack, we have to clean up
2576: that separately; that's after the @code{else}.
2577:
2578: Usage example:
2579:
2580: @example
2581: s" The text I search is here" scan-file
2582: @end example
2583:
2584: @subsection Copy input to output
2585:
2586: @example
2587: : copy-file ( -- )
2588: begin
2589: line-buffer max-line fd-in read-line throw
2590: while
2591: line-buffer swap fd-out write-file throw
2592: repeat ;
2593: @end example
2594:
2595: @subsection Close files
2596:
2597: @example
2598: fd-in close-file throw
2599: fd-out close-file throw
2600: @end example
2601:
2602: Likewise, you can put that into definitions, too:
2603:
2604: @example
2605: : close-input ( -- ) fd-in close-file throw ;
2606: : close-output ( -- ) fd-out close-file throw ;
2607: @end example
2608:
1.141 anton 2609: @quotation Assignment
1.84 pazsan 2610: How could you modify @code{copy-file} so that it copies until a second line is
2611: matched? Can you write a program that extracts a section of a text file,
2612: given the line that starts and the line that terminates that section?
1.141 anton 2613: @end quotation
1.84 pazsan 2614:
2615: @node Interpretation and Compilation Semantics and Immediacy Tutorial, Execution Tokens Tutorial, Files Tutorial, Tutorial
1.48 anton 2616: @section Interpretation and Compilation Semantics and Immediacy
1.66 anton 2617: @cindex semantics tutorial
2618: @cindex interpretation semantics tutorial
2619: @cindex compilation semantics tutorial
2620: @cindex immediate, tutorial
1.48 anton 2621:
2622: When a word is compiled, it behaves differently from being interpreted.
2623: E.g., consider @code{+}:
2624:
2625: @example
2626: 1 2 + .
2627: : foo + ;
2628: @end example
2629:
2630: These two behaviours are known as compilation and interpretation
2631: semantics. For normal words (e.g., @code{+}), the compilation semantics
2632: is to append the interpretation semantics to the currently defined word
2633: (@code{foo} in the example above). I.e., when @code{foo} is executed
2634: later, the interpretation semantics of @code{+} (i.e., adding two
2635: numbers) will be performed.
2636:
2637: However, there are words with non-default compilation semantics, e.g.,
2638: the control-flow words like @code{if}. You can use @code{immediate} to
2639: change the compilation semantics of the last defined word to be equal to
2640: the interpretation semantics:
2641:
2642: @example
2643: : [FOO] ( -- )
2644: 5 . ; immediate
2645:
2646: [FOO]
2647: : bar ( -- )
2648: [FOO] ;
2649: bar
2650: see bar
2651: @end example
2652:
2653: Two conventions to mark words with non-default compilation semnatics are
2654: names with brackets (more frequently used) and to write them all in
2655: upper case (less frequently used).
2656:
2657: In Gforth (and many other systems) you can also remove the
2658: interpretation semantics with @code{compile-only} (the compilation
2659: semantics is derived from the original interpretation semantics):
2660:
2661: @example
2662: : flip ( -- )
2663: 6 . ; compile-only \ but not immediate
2664: flip
2665:
2666: : flop ( -- )
2667: flip ;
2668: flop
2669: @end example
2670:
2671: In this example the interpretation semantics of @code{flop} is equal to
2672: the original interpretation semantics of @code{flip}.
2673:
2674: The text interpreter has two states: in interpret state, it performs the
2675: interpretation semantics of words it encounters; in compile state, it
2676: performs the compilation semantics of these words.
2677:
2678: Among other things, @code{:} switches into compile state, and @code{;}
2679: switches back to interpret state. They contain the factors @code{]}
2680: (switch to compile state) and @code{[} (switch to interpret state), that
2681: do nothing but switch the state.
2682:
2683: @example
2684: : xxx ( -- )
2685: [ 5 . ]
2686: ;
2687:
2688: xxx
2689: see xxx
2690: @end example
2691:
2692: These brackets are also the source of the naming convention mentioned
2693: above.
2694:
1.66 anton 2695: Reference: @ref{Interpretation and Compilation Semantics}.
2696:
1.48 anton 2697:
2698: @node Execution Tokens Tutorial, Exceptions Tutorial, Interpretation and Compilation Semantics and Immediacy Tutorial, Tutorial
2699: @section Execution Tokens
1.66 anton 2700: @cindex execution tokens tutorial
2701: @cindex XT tutorial
1.48 anton 2702:
2703: @code{' word} gives you the execution token (XT) of a word. The XT is a
2704: cell representing the interpretation semantics of a word. You can
2705: execute this semantics with @code{execute}:
2706:
2707: @example
2708: ' + .s
2709: 1 2 rot execute .
2710: @end example
2711:
2712: The XT is similar to a function pointer in C. However, parameter
2713: passing through the stack makes it a little more flexible:
2714:
2715: @example
2716: : map-array ( ... addr u xt -- ... )
1.50 anton 2717: \ executes xt ( ... x -- ... ) for every element of the array starting
2718: \ at addr and containing u elements
1.48 anton 2719: @{ xt @}
2720: cells over + swap ?do
1.50 anton 2721: i @@ xt execute
1.48 anton 2722: 1 cells +loop ;
2723:
2724: create a 3 , 4 , 2 , -1 , 4 ,
2725: a 5 ' . map-array .s
2726: 0 a 5 ' + map-array .
2727: s" max-n" environment? drop .s
2728: a 5 ' min map-array .
2729: @end example
2730:
2731: You can use map-array with the XTs of words that consume one element
2732: more than they produce. In theory you can also use it with other XTs,
2733: but the stack effect then depends on the size of the array, which is
2734: hard to understand.
2735:
1.51 pazsan 2736: Since XTs are cell-sized, you can store them in memory and manipulate
2737: them on the stack like other cells. You can also compile the XT into a
1.48 anton 2738: word with @code{compile,}:
2739:
2740: @example
2741: : foo1 ( n1 n2 -- n )
2742: [ ' + compile, ] ;
2743: see foo
2744: @end example
2745:
2746: This is non-standard, because @code{compile,} has no compilation
2747: semantics in the standard, but it works in good Forth systems. For the
2748: broken ones, use
2749:
2750: @example
2751: : [compile,] compile, ; immediate
2752:
2753: : foo1 ( n1 n2 -- n )
2754: [ ' + ] [compile,] ;
2755: see foo
2756: @end example
2757:
2758: @code{'} is a word with default compilation semantics; it parses the
2759: next word when its interpretation semantics are executed, not during
2760: compilation:
2761:
2762: @example
2763: : foo ( -- xt )
2764: ' ;
2765: see foo
2766: : bar ( ... "word" -- ... )
2767: ' execute ;
2768: see bar
1.60 anton 2769: 1 2 bar + .
1.48 anton 2770: @end example
2771:
2772: You often want to parse a word during compilation and compile its XT so
2773: it will be pushed on the stack at run-time. @code{[']} does this:
2774:
2775: @example
2776: : xt-+ ( -- xt )
2777: ['] + ;
2778: see xt-+
2779: 1 2 xt-+ execute .
2780: @end example
2781:
2782: Many programmers tend to see @code{'} and the word it parses as one
2783: unit, and expect it to behave like @code{[']} when compiled, and are
2784: confused by the actual behaviour. If you are, just remember that the
2785: Forth system just takes @code{'} as one unit and has no idea that it is
2786: a parsing word (attempts to convenience programmers in this issue have
2787: usually resulted in even worse pitfalls, see
1.66 anton 2788: @uref{http://www.complang.tuwien.ac.at/papers/ertl98.ps.gz,
2789: @code{State}-smartness---Why it is evil and How to Exorcise it}).
1.48 anton 2790:
2791: Note that the state of the interpreter does not come into play when
1.51 pazsan 2792: creating and executing XTs. I.e., even when you execute @code{'} in
1.48 anton 2793: compile state, it still gives you the interpretation semantics. And
2794: whatever that state is, @code{execute} performs the semantics
1.66 anton 2795: represented by the XT (i.e., for XTs produced with @code{'} the
2796: interpretation semantics).
2797:
2798: Reference: @ref{Tokens for Words}.
1.48 anton 2799:
2800:
2801: @node Exceptions Tutorial, Defining Words Tutorial, Execution Tokens Tutorial, Tutorial
2802: @section Exceptions
1.66 anton 2803: @cindex exceptions tutorial
1.48 anton 2804:
2805: @code{throw ( n -- )} causes an exception unless n is zero.
2806:
2807: @example
2808: 100 throw .s
2809: 0 throw .s
2810: @end example
2811:
2812: @code{catch ( ... xt -- ... n )} behaves similar to @code{execute}, but
2813: it catches exceptions and pushes the number of the exception on the
2814: stack (or 0, if the xt executed without exception). If there was an
2815: exception, the stacks have the same depth as when entering @code{catch}:
2816:
2817: @example
2818: .s
2819: 3 0 ' / catch .s
2820: 3 2 ' / catch .s
2821: @end example
2822:
1.141 anton 2823: @quotation Assignment
1.48 anton 2824: Try the same with @code{execute} instead of @code{catch}.
1.141 anton 2825: @end quotation
1.48 anton 2826:
2827: @code{Throw} always jumps to the dynamically next enclosing
2828: @code{catch}, even if it has to leave several call levels to achieve
2829: this:
2830:
2831: @example
2832: : foo 100 throw ;
2833: : foo1 foo ." after foo" ;
1.51 pazsan 2834: : bar ['] foo1 catch ;
1.60 anton 2835: bar .
1.48 anton 2836: @end example
2837:
2838: It is often important to restore a value upon leaving a definition, even
2839: if the definition is left through an exception. You can ensure this
2840: like this:
2841:
2842: @example
2843: : ...
2844: save-x
1.51 pazsan 2845: ['] word-changing-x catch ( ... n )
1.48 anton 2846: restore-x
2847: ( ... n ) throw ;
2848: @end example
2849:
1.172 anton 2850: However, this is still not safe against, e.g., the user pressing
2851: @kbd{Ctrl-C} when execution is between the @code{catch} and
2852: @code{restore-x}.
2853:
2854: Gforth provides an alternative exception handling syntax that is safe
2855: against such cases: @code{try ... restore ... endtry}. If the code
2856: between @code{try} and @code{endtry} has an exception, the stack
2857: depths are restored, the exception number is pushed on the stack, and
2858: the execution continues right after @code{restore}.
1.48 anton 2859:
1.172 anton 2860: The safer equivalent to the restoration code above is
1.48 anton 2861:
2862: @example
2863: : ...
2864: save-x
2865: try
1.92 anton 2866: word-changing-x 0
1.172 anton 2867: restore
2868: restore-x
2869: endtry
1.48 anton 2870: throw ;
2871: @end example
2872:
1.66 anton 2873: Reference: @ref{Exception Handling}.
2874:
1.48 anton 2875:
2876: @node Defining Words Tutorial, Arrays and Records Tutorial, Exceptions Tutorial, Tutorial
2877: @section Defining Words
1.66 anton 2878: @cindex defining words tutorial
2879: @cindex does> tutorial
2880: @cindex create...does> tutorial
2881:
2882: @c before semantics?
1.48 anton 2883:
2884: @code{:}, @code{create}, and @code{variable} are definition words: They
2885: define other words. @code{Constant} is another definition word:
2886:
2887: @example
2888: 5 constant foo
2889: foo .
2890: @end example
2891:
2892: You can also use the prefixes @code{2} (double-cell) and @code{f}
2893: (floating point) with @code{variable} and @code{constant}.
2894:
2895: You can also define your own defining words. E.g.:
2896:
2897: @example
2898: : variable ( "name" -- )
2899: create 0 , ;
2900: @end example
2901:
2902: You can also define defining words that create words that do something
2903: other than just producing their address:
2904:
2905: @example
2906: : constant ( n "name" -- )
2907: create ,
2908: does> ( -- n )
1.50 anton 2909: ( addr ) @@ ;
1.48 anton 2910:
2911: 5 constant foo
2912: foo .
2913: @end example
2914:
2915: The definition of @code{constant} above ends at the @code{does>}; i.e.,
2916: @code{does>} replaces @code{;}, but it also does something else: It
2917: changes the last defined word such that it pushes the address of the
2918: body of the word and then performs the code after the @code{does>}
2919: whenever it is called.
2920:
2921: In the example above, @code{constant} uses @code{,} to store 5 into the
2922: body of @code{foo}. When @code{foo} executes, it pushes the address of
2923: the body onto the stack, then (in the code after the @code{does>})
2924: fetches the 5 from there.
2925:
2926: The stack comment near the @code{does>} reflects the stack effect of the
2927: defined word, not the stack effect of the code after the @code{does>}
2928: (the difference is that the code expects the address of the body that
2929: the stack comment does not show).
2930:
2931: You can use these definition words to do factoring in cases that involve
2932: (other) definition words. E.g., a field offset is always added to an
2933: address. Instead of defining
2934:
2935: @example
2936: 2 cells constant offset-field1
2937: @end example
2938:
2939: and using this like
2940:
2941: @example
2942: ( addr ) offset-field1 +
2943: @end example
2944:
2945: you can define a definition word
2946:
2947: @example
2948: : simple-field ( n "name" -- )
2949: create ,
2950: does> ( n1 -- n1+n )
1.50 anton 2951: ( addr ) @@ + ;
1.48 anton 2952: @end example
1.21 crook 2953:
1.48 anton 2954: Definition and use of field offsets now look like this:
1.21 crook 2955:
1.48 anton 2956: @example
2957: 2 cells simple-field field1
1.60 anton 2958: create mystruct 4 cells allot
2959: mystruct .s field1 .s drop
1.48 anton 2960: @end example
1.21 crook 2961:
1.48 anton 2962: If you want to do something with the word without performing the code
2963: after the @code{does>}, you can access the body of a @code{create}d word
2964: with @code{>body ( xt -- addr )}:
1.21 crook 2965:
1.48 anton 2966: @example
2967: : value ( n "name" -- )
2968: create ,
2969: does> ( -- n1 )
1.50 anton 2970: @@ ;
1.48 anton 2971: : to ( n "name" -- )
2972: ' >body ! ;
1.21 crook 2973:
1.48 anton 2974: 5 value foo
2975: foo .
2976: 7 to foo
2977: foo .
2978: @end example
1.21 crook 2979:
1.141 anton 2980: @quotation Assignment
1.48 anton 2981: Define @code{defer ( "name" -- )}, which creates a word that stores an
2982: XT (at the start the XT of @code{abort}), and upon execution
2983: @code{execute}s the XT. Define @code{is ( xt "name" -- )} that stores
2984: @code{xt} into @code{name}, a word defined with @code{defer}. Indirect
2985: recursion is one application of @code{defer}.
1.141 anton 2986: @end quotation
1.29 crook 2987:
1.66 anton 2988: Reference: @ref{User-defined Defining Words}.
2989:
2990:
1.48 anton 2991: @node Arrays and Records Tutorial, POSTPONE Tutorial, Defining Words Tutorial, Tutorial
2992: @section Arrays and Records
1.66 anton 2993: @cindex arrays tutorial
2994: @cindex records tutorial
2995: @cindex structs tutorial
1.29 crook 2996:
1.48 anton 2997: Forth has no standard words for defining data structures such as arrays
2998: and records (structs in C terminology), but you can build them yourself
2999: based on address arithmetic. You can also define words for defining
3000: arrays and records (@pxref{Defining Words Tutorial,, Defining Words}).
1.29 crook 3001:
1.48 anton 3002: One of the first projects a Forth newcomer sets out upon when learning
3003: about defining words is an array defining word (possibly for
3004: n-dimensional arrays). Go ahead and do it, I did it, too; you will
3005: learn something from it. However, don't be disappointed when you later
3006: learn that you have little use for these words (inappropriate use would
3007: be even worse). I have not yet found a set of useful array words yet;
3008: the needs are just too diverse, and named, global arrays (the result of
3009: naive use of defining words) are often not flexible enough (e.g.,
1.66 anton 3010: consider how to pass them as parameters). Another such project is a set
3011: of words to help dealing with strings.
1.29 crook 3012:
1.48 anton 3013: On the other hand, there is a useful set of record words, and it has
3014: been defined in @file{compat/struct.fs}; these words are predefined in
3015: Gforth. They are explained in depth elsewhere in this manual (see
3016: @pxref{Structures}). The @code{simple-field} example above is
3017: simplified variant of fields in this package.
1.21 crook 3018:
3019:
1.48 anton 3020: @node POSTPONE Tutorial, Literal Tutorial, Arrays and Records Tutorial, Tutorial
3021: @section @code{POSTPONE}
1.66 anton 3022: @cindex postpone tutorial
1.21 crook 3023:
1.48 anton 3024: You can compile the compilation semantics (instead of compiling the
3025: interpretation semantics) of a word with @code{POSTPONE}:
1.21 crook 3026:
1.48 anton 3027: @example
3028: : MY-+ ( Compilation: -- ; Run-time of compiled code: n1 n2 -- n )
1.51 pazsan 3029: POSTPONE + ; immediate
1.48 anton 3030: : foo ( n1 n2 -- n )
3031: MY-+ ;
3032: 1 2 foo .
3033: see foo
3034: @end example
1.21 crook 3035:
1.48 anton 3036: During the definition of @code{foo} the text interpreter performs the
3037: compilation semantics of @code{MY-+}, which performs the compilation
3038: semantics of @code{+}, i.e., it compiles @code{+} into @code{foo}.
3039:
3040: This example also displays separate stack comments for the compilation
3041: semantics and for the stack effect of the compiled code. For words with
3042: default compilation semantics these stack effects are usually not
3043: displayed; the stack effect of the compilation semantics is always
3044: @code{( -- )} for these words, the stack effect for the compiled code is
3045: the stack effect of the interpretation semantics.
3046:
3047: Note that the state of the interpreter does not come into play when
3048: performing the compilation semantics in this way. You can also perform
3049: it interpretively, e.g.:
3050:
3051: @example
3052: : foo2 ( n1 n2 -- n )
3053: [ MY-+ ] ;
3054: 1 2 foo .
3055: see foo
3056: @end example
1.21 crook 3057:
1.48 anton 3058: However, there are some broken Forth systems where this does not always
1.62 crook 3059: work, and therefore this practice was been declared non-standard in
1.48 anton 3060: 1999.
3061: @c !! repair.fs
3062:
3063: Here is another example for using @code{POSTPONE}:
1.44 crook 3064:
1.48 anton 3065: @example
3066: : MY-- ( Compilation: -- ; Run-time of compiled code: n1 n2 -- n )
3067: POSTPONE negate POSTPONE + ; immediate compile-only
3068: : bar ( n1 n2 -- n )
3069: MY-- ;
3070: 2 1 bar .
3071: see bar
3072: @end example
1.21 crook 3073:
1.48 anton 3074: You can define @code{ENDIF} in this way:
1.21 crook 3075:
1.48 anton 3076: @example
3077: : ENDIF ( Compilation: orig -- )
3078: POSTPONE then ; immediate
3079: @end example
1.21 crook 3080:
1.141 anton 3081: @quotation Assignment
1.48 anton 3082: Write @code{MY-2DUP} that has compilation semantics equivalent to
3083: @code{2dup}, but compiles @code{over over}.
1.141 anton 3084: @end quotation
1.29 crook 3085:
1.66 anton 3086: @c !! @xref{Macros} for reference
3087:
3088:
1.48 anton 3089: @node Literal Tutorial, Advanced macros Tutorial, POSTPONE Tutorial, Tutorial
3090: @section @code{Literal}
1.66 anton 3091: @cindex literal tutorial
1.29 crook 3092:
1.48 anton 3093: You cannot @code{POSTPONE} numbers:
1.21 crook 3094:
1.48 anton 3095: @example
3096: : [FOO] POSTPONE 500 ; immediate
1.21 crook 3097: @end example
3098:
1.48 anton 3099: Instead, you can use @code{LITERAL (compilation: n --; run-time: -- n )}:
1.29 crook 3100:
1.48 anton 3101: @example
3102: : [FOO] ( compilation: --; run-time: -- n )
3103: 500 POSTPONE literal ; immediate
1.29 crook 3104:
1.60 anton 3105: : flip [FOO] ;
1.48 anton 3106: flip .
3107: see flip
3108: @end example
1.29 crook 3109:
1.48 anton 3110: @code{LITERAL} consumes a number at compile-time (when it's compilation
3111: semantics are executed) and pushes it at run-time (when the code it
3112: compiled is executed). A frequent use of @code{LITERAL} is to compile a
3113: number computed at compile time into the current word:
1.29 crook 3114:
1.48 anton 3115: @example
3116: : bar ( -- n )
3117: [ 2 2 + ] literal ;
3118: see bar
3119: @end example
1.29 crook 3120:
1.141 anton 3121: @quotation Assignment
1.48 anton 3122: Write @code{]L} which allows writing the example above as @code{: bar (
3123: -- n ) [ 2 2 + ]L ;}
1.141 anton 3124: @end quotation
1.48 anton 3125:
1.66 anton 3126: @c !! @xref{Macros} for reference
3127:
1.48 anton 3128:
3129: @node Advanced macros Tutorial, Compilation Tokens Tutorial, Literal Tutorial, Tutorial
3130: @section Advanced macros
1.66 anton 3131: @cindex macros, advanced tutorial
3132: @cindex run-time code generation, tutorial
1.48 anton 3133:
1.66 anton 3134: Reconsider @code{map-array} from @ref{Execution Tokens Tutorial,,
3135: Execution Tokens}. It frequently performs @code{execute}, a relatively
3136: expensive operation in some Forth implementations. You can use
1.48 anton 3137: @code{compile,} and @code{POSTPONE} to eliminate these @code{execute}s
3138: and produce a word that contains the word to be performed directly:
3139:
3140: @c use ]] ... [[
3141: @example
3142: : compile-map-array ( compilation: xt -- ; run-time: ... addr u -- ... )
3143: \ at run-time, execute xt ( ... x -- ... ) for each element of the
3144: \ array beginning at addr and containing u elements
3145: @{ xt @}
3146: POSTPONE cells POSTPONE over POSTPONE + POSTPONE swap POSTPONE ?do
1.50 anton 3147: POSTPONE i POSTPONE @@ xt compile,
1.48 anton 3148: 1 cells POSTPONE literal POSTPONE +loop ;
3149:
3150: : sum-array ( addr u -- n )
3151: 0 rot rot [ ' + compile-map-array ] ;
3152: see sum-array
3153: a 5 sum-array .
3154: @end example
3155:
3156: You can use the full power of Forth for generating the code; here's an
3157: example where the code is generated in a loop:
3158:
3159: @example
3160: : compile-vmul-step ( compilation: n --; run-time: n1 addr1 -- n2 addr2 )
3161: \ n2=n1+(addr1)*n, addr2=addr1+cell
1.50 anton 3162: POSTPONE tuck POSTPONE @@
1.48 anton 3163: POSTPONE literal POSTPONE * POSTPONE +
3164: POSTPONE swap POSTPONE cell+ ;
3165:
3166: : compile-vmul ( compilation: addr1 u -- ; run-time: addr2 -- n )
1.51 pazsan 3167: \ n=v1*v2 (inner product), where the v_i are represented as addr_i u
1.48 anton 3168: 0 postpone literal postpone swap
3169: [ ' compile-vmul-step compile-map-array ]
3170: postpone drop ;
3171: see compile-vmul
3172:
3173: : a-vmul ( addr -- n )
1.51 pazsan 3174: \ n=a*v, where v is a vector that's as long as a and starts at addr
1.48 anton 3175: [ a 5 compile-vmul ] ;
3176: see a-vmul
3177: a a-vmul .
3178: @end example
3179:
3180: This example uses @code{compile-map-array} to show off, but you could
1.66 anton 3181: also use @code{map-array} instead (try it now!).
1.48 anton 3182:
3183: You can use this technique for efficient multiplication of large
3184: matrices. In matrix multiplication, you multiply every line of one
3185: matrix with every column of the other matrix. You can generate the code
3186: for one line once, and use it for every column. The only downside of
3187: this technique is that it is cumbersome to recover the memory consumed
3188: by the generated code when you are done (and in more complicated cases
3189: it is not possible portably).
3190:
1.66 anton 3191: @c !! @xref{Macros} for reference
3192:
3193:
1.48 anton 3194: @node Compilation Tokens Tutorial, Wordlists and Search Order Tutorial, Advanced macros Tutorial, Tutorial
3195: @section Compilation Tokens
1.66 anton 3196: @cindex compilation tokens, tutorial
3197: @cindex CT, tutorial
1.48 anton 3198:
3199: This section is Gforth-specific. You can skip it.
3200:
3201: @code{' word compile,} compiles the interpretation semantics. For words
3202: with default compilation semantics this is the same as performing the
3203: compilation semantics. To represent the compilation semantics of other
3204: words (e.g., words like @code{if} that have no interpretation
3205: semantics), Gforth has the concept of a compilation token (CT,
3206: consisting of two cells), and words @code{comp'} and @code{[comp']}.
3207: You can perform the compilation semantics represented by a CT with
3208: @code{execute}:
1.29 crook 3209:
1.48 anton 3210: @example
3211: : foo2 ( n1 n2 -- n )
3212: [ comp' + execute ] ;
3213: see foo
3214: @end example
1.29 crook 3215:
1.48 anton 3216: You can compile the compilation semantics represented by a CT with
3217: @code{postpone,}:
1.30 anton 3218:
1.48 anton 3219: @example
3220: : foo3 ( -- )
3221: [ comp' + postpone, ] ;
3222: see foo3
3223: @end example
1.30 anton 3224:
1.51 pazsan 3225: @code{[ comp' word postpone, ]} is equivalent to @code{POSTPONE word}.
1.48 anton 3226: @code{comp'} is particularly useful for words that have no
3227: interpretation semantics:
1.29 crook 3228:
1.30 anton 3229: @example
1.48 anton 3230: ' if
1.60 anton 3231: comp' if .s 2drop
1.30 anton 3232: @end example
3233:
1.66 anton 3234: Reference: @ref{Tokens for Words}.
3235:
1.29 crook 3236:
1.48 anton 3237: @node Wordlists and Search Order Tutorial, , Compilation Tokens Tutorial, Tutorial
3238: @section Wordlists and Search Order
1.66 anton 3239: @cindex wordlists tutorial
3240: @cindex search order, tutorial
1.48 anton 3241:
3242: The dictionary is not just a memory area that allows you to allocate
3243: memory with @code{allot}, it also contains the Forth words, arranged in
3244: several wordlists. When searching for a word in a wordlist,
3245: conceptually you start searching at the youngest and proceed towards
3246: older words (in reality most systems nowadays use hash-tables); i.e., if
3247: you define a word with the same name as an older word, the new word
3248: shadows the older word.
3249:
3250: Which wordlists are searched in which order is determined by the search
3251: order. You can display the search order with @code{order}. It displays
3252: first the search order, starting with the wordlist searched first, then
3253: it displays the wordlist that will contain newly defined words.
1.21 crook 3254:
1.48 anton 3255: You can create a new, empty wordlist with @code{wordlist ( -- wid )}:
1.21 crook 3256:
1.48 anton 3257: @example
3258: wordlist constant mywords
3259: @end example
1.21 crook 3260:
1.48 anton 3261: @code{Set-current ( wid -- )} sets the wordlist that will contain newly
3262: defined words (the @emph{current} wordlist):
1.21 crook 3263:
1.48 anton 3264: @example
3265: mywords set-current
3266: order
3267: @end example
1.26 crook 3268:
1.48 anton 3269: Gforth does not display a name for the wordlist in @code{mywords}
3270: because this wordlist was created anonymously with @code{wordlist}.
1.21 crook 3271:
1.48 anton 3272: You can get the current wordlist with @code{get-current ( -- wid)}. If
3273: you want to put something into a specific wordlist without overall
3274: effect on the current wordlist, this typically looks like this:
1.21 crook 3275:
1.48 anton 3276: @example
3277: get-current mywords set-current ( wid )
3278: create someword
3279: ( wid ) set-current
3280: @end example
1.21 crook 3281:
1.48 anton 3282: You can write the search order with @code{set-order ( wid1 .. widn n --
3283: )} and read it with @code{get-order ( -- wid1 .. widn n )}. The first
3284: searched wordlist is topmost.
1.21 crook 3285:
1.48 anton 3286: @example
3287: get-order mywords swap 1+ set-order
3288: order
3289: @end example
1.21 crook 3290:
1.48 anton 3291: Yes, the order of wordlists in the output of @code{order} is reversed
3292: from stack comments and the output of @code{.s} and thus unintuitive.
1.21 crook 3293:
1.141 anton 3294: @quotation Assignment
1.48 anton 3295: Define @code{>order ( wid -- )} with adds @code{wid} as first searched
3296: wordlist to the search order. Define @code{previous ( -- )}, which
3297: removes the first searched wordlist from the search order. Experiment
3298: with boundary conditions (you will see some crashes or situations that
3299: are hard or impossible to leave).
1.141 anton 3300: @end quotation
1.21 crook 3301:
1.48 anton 3302: The search order is a powerful foundation for providing features similar
3303: to Modula-2 modules and C++ namespaces. However, trying to modularize
3304: programs in this way has disadvantages for debugging and reuse/factoring
3305: that overcome the advantages in my experience (I don't do huge projects,
1.55 anton 3306: though). These disadvantages are not so clear in other
1.82 anton 3307: languages/programming environments, because these languages are not so
1.48 anton 3308: strong in debugging and reuse.
1.21 crook 3309:
1.66 anton 3310: @c !! example
3311:
3312: Reference: @ref{Word Lists}.
1.21 crook 3313:
1.29 crook 3314: @c ******************************************************************
1.48 anton 3315: @node Introduction, Words, Tutorial, Top
1.29 crook 3316: @comment node-name, next, previous, up
3317: @chapter An Introduction to ANS Forth
3318: @cindex Forth - an introduction
1.21 crook 3319:
1.83 anton 3320: The difference of this chapter from the Tutorial (@pxref{Tutorial}) is
3321: that it is slower-paced in its examples, but uses them to dive deep into
3322: explaining Forth internals (not covered by the Tutorial). Apart from
3323: that, this chapter covers far less material. It is suitable for reading
3324: without using a computer.
3325:
1.29 crook 3326: The primary purpose of this manual is to document Gforth. However, since
3327: Forth is not a widely-known language and there is a lack of up-to-date
3328: teaching material, it seems worthwhile to provide some introductory
1.49 anton 3329: material. For other sources of Forth-related
3330: information, see @ref{Forth-related information}.
1.21 crook 3331:
1.29 crook 3332: The examples in this section should work on any ANS Forth; the
3333: output shown was produced using Gforth. Each example attempts to
3334: reproduce the exact output that Gforth produces. If you try out the
3335: examples (and you should), what you should type is shown @kbd{like this}
3336: and Gforth's response is shown @code{like this}. The single exception is
1.30 anton 3337: that, where the example shows @key{RET} it means that you should
1.29 crook 3338: press the ``carriage return'' key. Unfortunately, some output formats for
3339: this manual cannot show the difference between @kbd{this} and
3340: @code{this} which will make trying out the examples harder (but not
3341: impossible).
1.21 crook 3342:
1.29 crook 3343: Forth is an unusual language. It provides an interactive development
3344: environment which includes both an interpreter and compiler. Forth
3345: programming style encourages you to break a problem down into many
3346: @cindex factoring
3347: small fragments (@dfn{factoring}), and then to develop and test each
3348: fragment interactively. Forth advocates assert that breaking the
3349: edit-compile-test cycle used by conventional programming languages can
3350: lead to great productivity improvements.
1.21 crook 3351:
1.29 crook 3352: @menu
1.67 anton 3353: * Introducing the Text Interpreter::
3354: * Stacks and Postfix notation::
3355: * Your first definition::
3356: * How does that work?::
3357: * Forth is written in Forth::
3358: * Review - elements of a Forth system::
3359: * Where to go next::
3360: * Exercises::
1.29 crook 3361: @end menu
1.21 crook 3362:
1.29 crook 3363: @comment ----------------------------------------------
3364: @node Introducing the Text Interpreter, Stacks and Postfix notation, Introduction, Introduction
3365: @section Introducing the Text Interpreter
3366: @cindex text interpreter
3367: @cindex outer interpreter
1.21 crook 3368:
1.30 anton 3369: @c IMO this is too detailed and the pace is too slow for
3370: @c an introduction. If you know German, take a look at
3371: @c http://www.complang.tuwien.ac.at/anton/lvas/skriptum-stack.html
3372: @c to see how I do it - anton
3373:
1.44 crook 3374: @c nac-> Where I have accepted your comments 100% and modified the text
3375: @c accordingly, I have deleted your comments. Elsewhere I have added a
3376: @c response like this to attempt to rationalise what I have done. Of
3377: @c course, this is a very clumsy mechanism for something that would be
3378: @c done far more efficiently over a beer. Please delete any dialogue
3379: @c you consider closed.
3380:
1.29 crook 3381: When you invoke the Forth image, you will see a startup banner printed
3382: and nothing else (if you have Gforth installed on your system, try
1.30 anton 3383: invoking it now, by typing @kbd{gforth@key{RET}}). Forth is now running
1.29 crook 3384: its command line interpreter, which is called the @dfn{Text Interpreter}
3385: (also known as the @dfn{Outer Interpreter}). (You will learn a lot
1.49 anton 3386: about the text interpreter as you read through this chapter, for more
3387: detail @pxref{The Text Interpreter}).
1.21 crook 3388:
1.29 crook 3389: Although it's not obvious, Forth is actually waiting for your
1.30 anton 3390: input. Type a number and press the @key{RET} key:
1.21 crook 3391:
1.26 crook 3392: @example
1.30 anton 3393: @kbd{45@key{RET}} ok
1.26 crook 3394: @end example
1.21 crook 3395:
1.29 crook 3396: Rather than give you a prompt to invite you to input something, the text
3397: interpreter prints a status message @i{after} it has processed a line
3398: of input. The status message in this case (``@code{ ok}'' followed by
3399: carriage-return) indicates that the text interpreter was able to process
3400: all of your input successfully. Now type something illegal:
3401:
3402: @example
1.30 anton 3403: @kbd{qwer341@key{RET}}
1.134 anton 3404: *the terminal*:2: Undefined word
3405: >>>qwer341<<<
3406: Backtrace:
3407: $2A95B42A20 throw
3408: $2A95B57FB8 no.extensions
1.29 crook 3409: @end example
1.23 crook 3410:
1.134 anton 3411: The exact text, other than the ``Undefined word'' may differ slightly
3412: on your system, but the effect is the same; when the text interpreter
1.29 crook 3413: detects an error, it discards any remaining text on a line, resets
1.134 anton 3414: certain internal state and prints an error message. For a detailed
3415: description of error messages see @ref{Error messages}.
1.23 crook 3416:
1.29 crook 3417: The text interpreter waits for you to press carriage-return, and then
3418: processes your input line. Starting at the beginning of the line, it
3419: breaks the line into groups of characters separated by spaces. For each
3420: group of characters in turn, it makes two attempts to do something:
1.23 crook 3421:
1.29 crook 3422: @itemize @bullet
3423: @item
1.44 crook 3424: @cindex name dictionary
1.29 crook 3425: It tries to treat it as a command. It does this by searching a @dfn{name
3426: dictionary}. If the group of characters matches an entry in the name
3427: dictionary, the name dictionary provides the text interpreter with
3428: information that allows the text interpreter perform some actions. In
3429: Forth jargon, we say that the group
3430: @cindex word
3431: @cindex definition
3432: @cindex execution token
3433: @cindex xt
3434: of characters names a @dfn{word}, that the dictionary search returns an
3435: @dfn{execution token (xt)} corresponding to the @dfn{definition} of the
3436: word, and that the text interpreter executes the xt. Often, the terms
3437: @dfn{word} and @dfn{definition} are used interchangeably.
3438: @item
3439: If the text interpreter fails to find a match in the name dictionary, it
3440: tries to treat the group of characters as a number in the current number
3441: base (when you start up Forth, the current number base is base 10). If
3442: the group of characters legitimately represents a number, the text
3443: interpreter pushes the number onto a stack (we'll learn more about that
3444: in the next section).
3445: @end itemize
1.23 crook 3446:
1.29 crook 3447: If the text interpreter is unable to do either of these things with any
3448: group of characters, it discards the group of characters and the rest of
3449: the line, then prints an error message. If the text interpreter reaches
3450: the end of the line without error, it prints the status message ``@code{ ok}''
3451: followed by carriage-return.
1.21 crook 3452:
1.29 crook 3453: This is the simplest command we can give to the text interpreter:
1.23 crook 3454:
3455: @example
1.30 anton 3456: @key{RET} ok
1.23 crook 3457: @end example
1.21 crook 3458:
1.29 crook 3459: The text interpreter did everything we asked it to do (nothing) without
3460: an error, so it said that everything is ``@code{ ok}''. Try a slightly longer
3461: command:
1.21 crook 3462:
1.23 crook 3463: @example
1.30 anton 3464: @kbd{12 dup fred dup@key{RET}}
1.134 anton 3465: *the terminal*:3: Undefined word
3466: 12 dup >>>fred<<< dup
3467: Backtrace:
3468: $2A95B42A20 throw
3469: $2A95B57FB8 no.extensions
1.23 crook 3470: @end example
1.21 crook 3471:
1.29 crook 3472: When you press the carriage-return key, the text interpreter starts to
3473: work its way along the line:
1.21 crook 3474:
1.29 crook 3475: @itemize @bullet
3476: @item
3477: When it gets to the space after the @code{2}, it takes the group of
3478: characters @code{12} and looks them up in the name
3479: dictionary@footnote{We can't tell if it found them or not, but assume
3480: for now that it did not}. There is no match for this group of characters
3481: in the name dictionary, so it tries to treat them as a number. It is
3482: able to do this successfully, so it puts the number, 12, ``on the stack''
3483: (whatever that means).
3484: @item
3485: The text interpreter resumes scanning the line and gets the next group
3486: of characters, @code{dup}. It looks it up in the name dictionary and
3487: (you'll have to take my word for this) finds it, and executes the word
3488: @code{dup} (whatever that means).
3489: @item
3490: Once again, the text interpreter resumes scanning the line and gets the
3491: group of characters @code{fred}. It looks them up in the name
3492: dictionary, but can't find them. It tries to treat them as a number, but
3493: they don't represent any legal number.
3494: @end itemize
1.21 crook 3495:
1.29 crook 3496: At this point, the text interpreter gives up and prints an error
3497: message. The error message shows exactly how far the text interpreter
3498: got in processing the line. In particular, it shows that the text
3499: interpreter made no attempt to do anything with the final character
3500: group, @code{dup}, even though we have good reason to believe that the
3501: text interpreter would have no problem looking that word up and
3502: executing it a second time.
1.21 crook 3503:
3504:
1.29 crook 3505: @comment ----------------------------------------------
3506: @node Stacks and Postfix notation, Your first definition, Introducing the Text Interpreter, Introduction
3507: @section Stacks, postfix notation and parameter passing
3508: @cindex text interpreter
3509: @cindex outer interpreter
1.21 crook 3510:
1.29 crook 3511: In procedural programming languages (like C and Pascal), the
3512: building-block of programs is the @dfn{function} or @dfn{procedure}. These
3513: functions or procedures are called with @dfn{explicit parameters}. For
3514: example, in C we might write:
1.21 crook 3515:
1.23 crook 3516: @example
1.29 crook 3517: total = total + new_volume(length,height,depth);
1.23 crook 3518: @end example
1.21 crook 3519:
1.23 crook 3520: @noindent
1.29 crook 3521: where new_volume is a function-call to another piece of code, and total,
3522: length, height and depth are all variables. length, height and depth are
3523: parameters to the function-call.
1.21 crook 3524:
1.29 crook 3525: In Forth, the equivalent of the function or procedure is the
3526: @dfn{definition} and parameters are implicitly passed between
3527: definitions using a shared stack that is visible to the
3528: programmer. Although Forth does support variables, the existence of the
3529: stack means that they are used far less often than in most other
3530: programming languages. When the text interpreter encounters a number, it
3531: will place (@dfn{push}) it on the stack. There are several stacks (the
1.30 anton 3532: actual number is implementation-dependent ...) and the particular stack
1.29 crook 3533: used for any operation is implied unambiguously by the operation being
3534: performed. The stack used for all integer operations is called the @dfn{data
3535: stack} and, since this is the stack used most commonly, references to
3536: ``the data stack'' are often abbreviated to ``the stack''.
1.21 crook 3537:
1.29 crook 3538: The stacks have a last-in, first-out (LIFO) organisation. If you type:
1.21 crook 3539:
1.23 crook 3540: @example
1.30 anton 3541: @kbd{1 2 3@key{RET}} ok
1.23 crook 3542: @end example
1.21 crook 3543:
1.29 crook 3544: Then this instructs the text interpreter to placed three numbers on the
3545: (data) stack. An analogy for the behaviour of the stack is to take a
3546: pack of playing cards and deal out the ace (1), 2 and 3 into a pile on
3547: the table. The 3 was the last card onto the pile (``last-in'') and if
3548: you take a card off the pile then, unless you're prepared to fiddle a
3549: bit, the card that you take off will be the 3 (``first-out''). The
3550: number that will be first-out of the stack is called the @dfn{top of
3551: stack}, which
3552: @cindex TOS definition
3553: is often abbreviated to @dfn{TOS}.
1.21 crook 3554:
1.29 crook 3555: To understand how parameters are passed in Forth, consider the
3556: behaviour of the definition @code{+} (pronounced ``plus''). You will not
3557: be surprised to learn that this definition performs addition. More
3558: precisely, it adds two number together and produces a result. Where does
3559: it get the two numbers from? It takes the top two numbers off the
3560: stack. Where does it place the result? On the stack. You can act-out the
3561: behaviour of @code{+} with your playing cards like this:
1.21 crook 3562:
3563: @itemize @bullet
3564: @item
1.29 crook 3565: Pick up two cards from the stack on the table
1.21 crook 3566: @item
1.29 crook 3567: Stare at them intently and ask yourself ``what @i{is} the sum of these two
3568: numbers''
1.21 crook 3569: @item
1.29 crook 3570: Decide that the answer is 5
1.21 crook 3571: @item
1.29 crook 3572: Shuffle the two cards back into the pack and find a 5
1.21 crook 3573: @item
1.29 crook 3574: Put a 5 on the remaining ace that's on the table.
1.21 crook 3575: @end itemize
3576:
1.29 crook 3577: If you don't have a pack of cards handy but you do have Forth running,
3578: you can use the definition @code{.s} to show the current state of the stack,
3579: without affecting the stack. Type:
1.21 crook 3580:
3581: @example
1.124 anton 3582: @kbd{clearstacks 1 2 3@key{RET}} ok
1.30 anton 3583: @kbd{.s@key{RET}} <3> 1 2 3 ok
1.23 crook 3584: @end example
3585:
1.124 anton 3586: The text interpreter looks up the word @code{clearstacks} and executes
3587: it; it tidies up the stacks and removes any entries that may have been
1.29 crook 3588: left on it by earlier examples. The text interpreter pushes each of the
3589: three numbers in turn onto the stack. Finally, the text interpreter
3590: looks up the word @code{.s} and executes it. The effect of executing
3591: @code{.s} is to print the ``<3>'' (the total number of items on the stack)
3592: followed by a list of all the items on the stack; the item on the far
3593: right-hand side is the TOS.
1.21 crook 3594:
1.29 crook 3595: You can now type:
1.21 crook 3596:
1.29 crook 3597: @example
1.30 anton 3598: @kbd{+ .s@key{RET}} <2> 1 5 ok
1.29 crook 3599: @end example
1.21 crook 3600:
1.29 crook 3601: @noindent
3602: which is correct; there are now 2 items on the stack and the result of
3603: the addition is 5.
1.23 crook 3604:
1.29 crook 3605: If you're playing with cards, try doing a second addition: pick up the
3606: two cards, work out that their sum is 6, shuffle them into the pack,
3607: look for a 6 and place that on the table. You now have just one item on
3608: the stack. What happens if you try to do a third addition? Pick up the
3609: first card, pick up the second card -- ah! There is no second card. This
3610: is called a @dfn{stack underflow} and consitutes an error. If you try to
1.95 anton 3611: do the same thing with Forth it often reports an error (probably a Stack
1.29 crook 3612: Underflow or an Invalid Memory Address error).
1.23 crook 3613:
1.29 crook 3614: The opposite situation to a stack underflow is a @dfn{stack overflow},
3615: which simply accepts that there is a finite amount of storage space
3616: reserved for the stack. To stretch the playing card analogy, if you had
3617: enough packs of cards and you piled the cards up on the table, you would
3618: eventually be unable to add another card; you'd hit the ceiling. Gforth
3619: allows you to set the maximum size of the stacks. In general, the only
3620: time that you will get a stack overflow is because a definition has a
3621: bug in it and is generating data on the stack uncontrollably.
1.23 crook 3622:
1.29 crook 3623: There's one final use for the playing card analogy. If you model your
3624: stack using a pack of playing cards, the maximum number of items on
3625: your stack will be 52 (I assume you didn't use the Joker). The maximum
3626: @i{value} of any item on the stack is 13 (the King). In fact, the only
3627: possible numbers are positive integer numbers 1 through 13; you can't
3628: have (for example) 0 or 27 or 3.52 or -2. If you change the way you
3629: think about some of the cards, you can accommodate different
3630: numbers. For example, you could think of the Jack as representing 0,
3631: the Queen as representing -1 and the King as representing -2. Your
1.45 crook 3632: @i{range} remains unchanged (you can still only represent a total of 13
1.29 crook 3633: numbers) but the numbers that you can represent are -2 through 10.
1.28 crook 3634:
1.29 crook 3635: In that analogy, the limit was the amount of information that a single
3636: stack entry could hold, and Forth has a similar limit. In Forth, the
3637: size of a stack entry is called a @dfn{cell}. The actual size of a cell is
3638: implementation dependent and affects the maximum value that a stack
3639: entry can hold. A Standard Forth provides a cell size of at least
3640: 16-bits, and most desktop systems use a cell size of 32-bits.
1.21 crook 3641:
1.29 crook 3642: Forth does not do any type checking for you, so you are free to
3643: manipulate and combine stack items in any way you wish. A convenient way
3644: of treating stack items is as 2's complement signed integers, and that
3645: is what Standard words like @code{+} do. Therefore you can type:
1.21 crook 3646:
1.29 crook 3647: @example
1.30 anton 3648: @kbd{-5 12 + .s@key{RET}} <1> 7 ok
1.29 crook 3649: @end example
1.21 crook 3650:
1.29 crook 3651: If you use numbers and definitions like @code{+} in order to turn Forth
3652: into a great big pocket calculator, you will realise that it's rather
3653: different from a normal calculator. Rather than typing 2 + 3 = you had
3654: to type 2 3 + (ignore the fact that you had to use @code{.s} to see the
3655: result). The terminology used to describe this difference is to say that
3656: your calculator uses @dfn{Infix Notation} (parameters and operators are
3657: mixed) whilst Forth uses @dfn{Postfix Notation} (parameters and
3658: operators are separate), also called @dfn{Reverse Polish Notation}.
1.21 crook 3659:
1.29 crook 3660: Whilst postfix notation might look confusing to begin with, it has
3661: several important advantages:
1.21 crook 3662:
1.23 crook 3663: @itemize @bullet
3664: @item
1.29 crook 3665: it is unambiguous
1.23 crook 3666: @item
1.29 crook 3667: it is more concise
1.23 crook 3668: @item
1.29 crook 3669: it fits naturally with a stack-based system
1.23 crook 3670: @end itemize
1.21 crook 3671:
1.29 crook 3672: To examine these claims in more detail, consider these sums:
1.21 crook 3673:
1.29 crook 3674: @example
3675: 6 + 5 * 4 =
3676: 4 * 5 + 6 =
3677: @end example
1.21 crook 3678:
1.29 crook 3679: If you're just learning maths or your maths is very rusty, you will
3680: probably come up with the answer 44 for the first and 26 for the
3681: second. If you are a bit of a whizz at maths you will remember the
3682: @i{convention} that multiplication takes precendence over addition, and
3683: you'd come up with the answer 26 both times. To explain the answer 26
3684: to someone who got the answer 44, you'd probably rewrite the first sum
3685: like this:
1.21 crook 3686:
1.29 crook 3687: @example
3688: 6 + (5 * 4) =
3689: @end example
1.21 crook 3690:
1.29 crook 3691: If what you really wanted was to perform the addition before the
3692: multiplication, you would have to use parentheses to force it.
1.21 crook 3693:
1.29 crook 3694: If you did the first two sums on a pocket calculator you would probably
3695: get the right answers, unless you were very cautious and entered them using
3696: these keystroke sequences:
1.21 crook 3697:
1.29 crook 3698: 6 + 5 = * 4 =
3699: 4 * 5 = + 6 =
1.21 crook 3700:
1.29 crook 3701: Postfix notation is unambiguous because the order that the operators
3702: are applied is always explicit; that also means that parentheses are
3703: never required. The operators are @i{active} (the act of quoting the
3704: operator makes the operation occur) which removes the need for ``=''.
1.28 crook 3705:
1.29 crook 3706: The sum 6 + 5 * 4 can be written (in postfix notation) in two
3707: equivalent ways:
1.26 crook 3708:
3709: @example
1.29 crook 3710: 6 5 4 * + or:
3711: 5 4 * 6 +
1.26 crook 3712: @end example
1.23 crook 3713:
1.29 crook 3714: An important thing that you should notice about this notation is that
3715: the @i{order} of the numbers does not change; if you want to subtract
3716: 2 from 10 you type @code{10 2 -}.
1.1 anton 3717:
1.29 crook 3718: The reason that Forth uses postfix notation is very simple to explain: it
3719: makes the implementation extremely simple, and it follows naturally from
3720: using the stack as a mechanism for passing parameters. Another way of
3721: thinking about this is to realise that all Forth definitions are
3722: @i{active}; they execute as they are encountered by the text
3723: interpreter. The result of this is that the syntax of Forth is trivially
3724: simple.
1.1 anton 3725:
3726:
3727:
1.29 crook 3728: @comment ----------------------------------------------
3729: @node Your first definition, How does that work?, Stacks and Postfix notation, Introduction
3730: @section Your first Forth definition
3731: @cindex first definition
1.1 anton 3732:
1.29 crook 3733: Until now, the examples we've seen have been trivial; we've just been
3734: using Forth as a bigger-than-pocket calculator. Also, each calculation
3735: we've shown has been a ``one-off'' -- to repeat it we'd need to type it in
3736: again@footnote{That's not quite true. If you press the up-arrow key on
3737: your keyboard you should be able to scroll back to any earlier command,
3738: edit it and re-enter it.} In this section we'll see how to add new
3739: words to Forth's vocabulary.
1.1 anton 3740:
1.29 crook 3741: The easiest way to create a new word is to use a @dfn{colon
3742: definition}. We'll define a few and try them out before worrying too
3743: much about how they work. Try typing in these examples; be careful to
3744: copy the spaces accurately:
1.1 anton 3745:
1.29 crook 3746: @example
3747: : add-two 2 + . ;
3748: : greet ." Hello and welcome" ;
3749: : demo 5 add-two ;
3750: @end example
1.1 anton 3751:
1.29 crook 3752: @noindent
3753: Now try them out:
1.1 anton 3754:
1.29 crook 3755: @example
1.30 anton 3756: @kbd{greet@key{RET}} Hello and welcome ok
3757: @kbd{greet greet@key{RET}} Hello and welcomeHello and welcome ok
3758: @kbd{4 add-two@key{RET}} 6 ok
3759: @kbd{demo@key{RET}} 7 ok
3760: @kbd{9 greet demo add-two@key{RET}} Hello and welcome7 11 ok
1.29 crook 3761: @end example
1.1 anton 3762:
1.29 crook 3763: The first new thing that we've introduced here is the pair of words
3764: @code{:} and @code{;}. These are used to start and terminate a new
3765: definition, respectively. The first word after the @code{:} is the name
3766: for the new definition.
1.1 anton 3767:
1.29 crook 3768: As you can see from the examples, a definition is built up of words that
3769: have already been defined; Forth makes no distinction between
3770: definitions that existed when you started the system up, and those that
3771: you define yourself.
1.1 anton 3772:
1.29 crook 3773: The examples also introduce the words @code{.} (dot), @code{."}
3774: (dot-quote) and @code{dup} (dewp). Dot takes the value from the top of
3775: the stack and displays it. It's like @code{.s} except that it only
3776: displays the top item of the stack and it is destructive; after it has
3777: executed, the number is no longer on the stack. There is always one
3778: space printed after the number, and no spaces before it. Dot-quote
3779: defines a string (a sequence of characters) that will be printed when
3780: the word is executed. The string can contain any printable characters
3781: except @code{"}. A @code{"} has a special function; it is not a Forth
3782: word but it acts as a delimiter (the way that delimiters work is
3783: described in the next section). Finally, @code{dup} duplicates the value
3784: at the top of the stack. Try typing @code{5 dup .s} to see what it does.
1.1 anton 3785:
1.29 crook 3786: We already know that the text interpreter searches through the
3787: dictionary to locate names. If you've followed the examples earlier, you
3788: will already have a definition called @code{add-two}. Lets try modifying
3789: it by typing in a new definition:
1.1 anton 3790:
1.29 crook 3791: @example
1.30 anton 3792: @kbd{: add-two dup . ." + 2 =" 2 + . ;@key{RET}} redefined add-two ok
1.29 crook 3793: @end example
1.5 anton 3794:
1.29 crook 3795: Forth recognised that we were defining a word that already exists, and
3796: printed a message to warn us of that fact. Let's try out the new
3797: definition:
1.5 anton 3798:
1.29 crook 3799: @example
1.30 anton 3800: @kbd{9 add-two@key{RET}} 9 + 2 =11 ok
1.29 crook 3801: @end example
1.1 anton 3802:
1.29 crook 3803: @noindent
3804: All that we've actually done here, though, is to create a new
3805: definition, with a particular name. The fact that there was already a
3806: definition with the same name did not make any difference to the way
3807: that the new definition was created (except that Forth printed a warning
3808: message). The old definition of add-two still exists (try @code{demo}
3809: again to see that this is true). Any new definition will use the new
3810: definition of @code{add-two}, but old definitions continue to use the
3811: version that already existed at the time that they were @code{compiled}.
1.1 anton 3812:
1.29 crook 3813: Before you go on to the next section, try defining and redefining some
3814: words of your own.
1.1 anton 3815:
1.29 crook 3816: @comment ----------------------------------------------
3817: @node How does that work?, Forth is written in Forth, Your first definition, Introduction
3818: @section How does that work?
3819: @cindex parsing words
1.1 anton 3820:
1.30 anton 3821: @c That's pretty deep (IMO way too deep) for an introduction. - anton
3822:
3823: @c Is it a good idea to talk about the interpretation semantics of a
3824: @c number? We don't have an xt to go along with it. - anton
3825:
3826: @c Now that I have eliminated execution semantics, I wonder if it would not
3827: @c be better to keep them (or add run-time semantics), to make it easier to
3828: @c explain what compilation semantics usually does. - anton
3829:
1.44 crook 3830: @c nac-> I removed the term ``default compilation sematics'' from the
3831: @c introductory chapter. Removing ``execution semantics'' was making
3832: @c everything simpler to explain, then I think the use of this term made
3833: @c everything more complex again. I replaced it with ``default
3834: @c semantics'' (which is used elsewhere in the manual) by which I mean
3835: @c ``a definition that has neither the immediate nor the compile-only
1.83 anton 3836: @c flag set''.
3837:
3838: @c anton: I have eliminated default semantics (except in one place where it
3839: @c means "default interpretation and compilation semantics"), because it
3840: @c makes no sense in the presence of combined words. I reverted to
3841: @c "execution semantics" where necessary.
3842:
3843: @c nac-> I reworded big chunks of the ``how does that work''
1.44 crook 3844: @c section (and, unusually for me, I think I even made it shorter!). See
3845: @c what you think -- I know I have not addressed your primary concern
3846: @c that it is too heavy-going for an introduction. From what I understood
3847: @c of your course notes it looks as though they might be a good framework.
3848: @c Things that I've tried to capture here are some things that came as a
3849: @c great revelation here when I first understood them. Also, I like the
3850: @c fact that a very simple code example shows up almost all of the issues
3851: @c that you need to understand to see how Forth works. That's unique and
3852: @c worthwhile to emphasise.
3853:
1.83 anton 3854: @c anton: I think it's a good idea to present the details, especially those
3855: @c that you found to be a revelation, and probably the tutorial tries to be
3856: @c too superficial and does not get some of the things across that make
3857: @c Forth special. I do believe that most of the time these things should
3858: @c be discussed at the end of a section or in separate sections instead of
3859: @c in the middle of a section (e.g., the stuff you added in "User-defined
3860: @c defining words" leads in a completely different direction from the rest
3861: @c of the section).
3862:
1.29 crook 3863: Now we're going to take another look at the definition of @code{add-two}
3864: from the previous section. From our knowledge of the way that the text
3865: interpreter works, we would have expected this result when we tried to
3866: define @code{add-two}:
1.21 crook 3867:
1.29 crook 3868: @example
1.44 crook 3869: @kbd{: add-two 2 + . ;@key{RET}}
1.134 anton 3870: *the terminal*:4: Undefined word
3871: : >>>add-two<<< 2 + . ;
1.29 crook 3872: @end example
1.28 crook 3873:
1.29 crook 3874: The reason that this didn't happen is bound up in the way that @code{:}
3875: works. The word @code{:} does two special things. The first special
3876: thing that it does prevents the text interpreter from ever seeing the
3877: characters @code{add-two}. The text interpreter uses a variable called
3878: @cindex modifying >IN
1.44 crook 3879: @code{>IN} (pronounced ``to-in'') to keep track of where it is in the
1.29 crook 3880: input line. When it encounters the word @code{:} it behaves in exactly
3881: the same way as it does for any other word; it looks it up in the name
3882: dictionary, finds its xt and executes it. When @code{:} executes, it
3883: looks at the input buffer, finds the word @code{add-two} and advances the
3884: value of @code{>IN} to point past it. It then does some other stuff
3885: associated with creating the new definition (including creating an entry
3886: for @code{add-two} in the name dictionary). When the execution of @code{:}
3887: completes, control returns to the text interpreter, which is oblivious
3888: to the fact that it has been tricked into ignoring part of the input
3889: line.
1.21 crook 3890:
1.29 crook 3891: @cindex parsing words
3892: Words like @code{:} -- words that advance the value of @code{>IN} and so
3893: prevent the text interpreter from acting on the whole of the input line
3894: -- are called @dfn{parsing words}.
1.21 crook 3895:
1.29 crook 3896: @cindex @code{state} - effect on the text interpreter
3897: @cindex text interpreter - effect of state
3898: The second special thing that @code{:} does is change the value of a
3899: variable called @code{state}, which affects the way that the text
3900: interpreter behaves. When Gforth starts up, @code{state} has the value
3901: 0, and the text interpreter is said to be @dfn{interpreting}. During a
3902: colon definition (started with @code{:}), @code{state} is set to -1 and
1.44 crook 3903: the text interpreter is said to be @dfn{compiling}.
3904:
3905: In this example, the text interpreter is compiling when it processes the
3906: string ``@code{2 + . ;}''. It still breaks the string down into
3907: character sequences in the same way. However, instead of pushing the
3908: number @code{2} onto the stack, it lays down (@dfn{compiles}) some magic
3909: into the definition of @code{add-two} that will make the number @code{2} get
3910: pushed onto the stack when @code{add-two} is @dfn{executed}. Similarly,
3911: the behaviours of @code{+} and @code{.} are also compiled into the
3912: definition.
3913:
3914: One category of words don't get compiled. These so-called @dfn{immediate
3915: words} get executed (performed @i{now}) regardless of whether the text
3916: interpreter is interpreting or compiling. The word @code{;} is an
3917: immediate word. Rather than being compiled into the definition, it
3918: executes. Its effect is to terminate the current definition, which
3919: includes changing the value of @code{state} back to 0.
3920:
3921: When you execute @code{add-two}, it has a @dfn{run-time effect} that is
3922: exactly the same as if you had typed @code{2 + . @key{RET}} outside of a
3923: definition.
1.28 crook 3924:
1.30 anton 3925: In Forth, every word or number can be described in terms of two
1.29 crook 3926: properties:
1.28 crook 3927:
3928: @itemize @bullet
3929: @item
1.29 crook 3930: @cindex interpretation semantics
1.44 crook 3931: Its @dfn{interpretation semantics} describe how it will behave when the
3932: text interpreter encounters it in @dfn{interpret} state. The
3933: interpretation semantics of a word are represented by an @dfn{execution
3934: token}.
1.28 crook 3935: @item
1.29 crook 3936: @cindex compilation semantics
1.44 crook 3937: Its @dfn{compilation semantics} describe how it will behave when the
3938: text interpreter encounters it in @dfn{compile} state. The compilation
3939: semantics of a word are represented in an implementation-dependent way;
3940: Gforth uses a @dfn{compilation token}.
1.29 crook 3941: @end itemize
3942:
3943: @noindent
3944: Numbers are always treated in a fixed way:
3945:
3946: @itemize @bullet
1.28 crook 3947: @item
1.44 crook 3948: When the number is @dfn{interpreted}, its behaviour is to push the
3949: number onto the stack.
1.28 crook 3950: @item
1.30 anton 3951: When the number is @dfn{compiled}, a piece of code is appended to the
3952: current definition that pushes the number when it runs. (In other words,
3953: the compilation semantics of a number are to postpone its interpretation
3954: semantics until the run-time of the definition that it is being compiled
3955: into.)
1.29 crook 3956: @end itemize
3957:
1.44 crook 3958: Words don't behave in such a regular way, but most have @i{default
3959: semantics} which means that they behave like this:
1.29 crook 3960:
3961: @itemize @bullet
1.28 crook 3962: @item
1.30 anton 3963: The @dfn{interpretation semantics} of the word are to do something useful.
3964: @item
1.29 crook 3965: The @dfn{compilation semantics} of the word are to append its
1.30 anton 3966: @dfn{interpretation semantics} to the current definition (so that its
3967: run-time behaviour is to do something useful).
1.28 crook 3968: @end itemize
3969:
1.30 anton 3970: @cindex immediate words
1.44 crook 3971: The actual behaviour of any particular word can be controlled by using
3972: the words @code{immediate} and @code{compile-only} when the word is
3973: defined. These words set flags in the name dictionary entry of the most
3974: recently defined word, and these flags are retrieved by the text
3975: interpreter when it finds the word in the name dictionary.
3976:
3977: A word that is marked as @dfn{immediate} has compilation semantics that
3978: are identical to its interpretation semantics. In other words, it
3979: behaves like this:
1.29 crook 3980:
3981: @itemize @bullet
3982: @item
1.30 anton 3983: The @dfn{interpretation semantics} of the word are to do something useful.
1.29 crook 3984: @item
1.30 anton 3985: The @dfn{compilation semantics} of the word are to do something useful
3986: (and actually the same thing); i.e., it is executed during compilation.
1.29 crook 3987: @end itemize
1.28 crook 3988:
1.44 crook 3989: Marking a word as @dfn{compile-only} prohibits the text interpreter from
3990: performing the interpretation semantics of the word directly; an attempt
3991: to do so will generate an error. It is never necessary to use
3992: @code{compile-only} (and it is not even part of ANS Forth, though it is
3993: provided by many implementations) but it is good etiquette to apply it
3994: to a word that will not behave correctly (and might have unexpected
3995: side-effects) in interpret state. For example, it is only legal to use
3996: the conditional word @code{IF} within a definition. If you forget this
3997: and try to use it elsewhere, the fact that (in Gforth) it is marked as
3998: @code{compile-only} allows the text interpreter to generate a helpful
3999: error message rather than subjecting you to the consequences of your
4000: folly.
4001:
1.29 crook 4002: This example shows the difference between an immediate and a
4003: non-immediate word:
1.28 crook 4004:
1.29 crook 4005: @example
4006: : show-state state @@ . ;
4007: : show-state-now show-state ; immediate
4008: : word1 show-state ;
4009: : word2 show-state-now ;
1.28 crook 4010: @end example
1.23 crook 4011:
1.29 crook 4012: The word @code{immediate} after the definition of @code{show-state-now}
4013: makes that word an immediate word. These definitions introduce a new
4014: word: @code{@@} (pronounced ``fetch''). This word fetches the value of a
4015: variable, and leaves it on the stack. Therefore, the behaviour of
4016: @code{show-state} is to print a number that represents the current value
4017: of @code{state}.
1.28 crook 4018:
1.29 crook 4019: When you execute @code{word1}, it prints the number 0, indicating that
4020: the system is interpreting. When the text interpreter compiled the
4021: definition of @code{word1}, it encountered @code{show-state} whose
1.30 anton 4022: compilation semantics are to append its interpretation semantics to the
1.29 crook 4023: current definition. When you execute @code{word1}, it performs the
1.30 anton 4024: interpretation semantics of @code{show-state}. At the time that @code{word1}
1.29 crook 4025: (and therefore @code{show-state}) are executed, the system is
4026: interpreting.
1.28 crook 4027:
1.30 anton 4028: When you pressed @key{RET} after entering the definition of @code{word2},
1.29 crook 4029: you should have seen the number -1 printed, followed by ``@code{
4030: ok}''. When the text interpreter compiled the definition of
4031: @code{word2}, it encountered @code{show-state-now}, an immediate word,
1.30 anton 4032: whose compilation semantics are therefore to perform its interpretation
1.29 crook 4033: semantics. It is executed straight away (even before the text
4034: interpreter has moved on to process another group of characters; the
4035: @code{;} in this example). The effect of executing it are to display the
4036: value of @code{state} @i{at the time that the definition of}
4037: @code{word2} @i{is being defined}. Printing -1 demonstrates that the
4038: system is compiling at this time. If you execute @code{word2} it does
4039: nothing at all.
1.28 crook 4040:
1.29 crook 4041: @cindex @code{."}, how it works
4042: Before leaving the subject of immediate words, consider the behaviour of
4043: @code{."} in the definition of @code{greet}, in the previous
4044: section. This word is both a parsing word and an immediate word. Notice
4045: that there is a space between @code{."} and the start of the text
4046: @code{Hello and welcome}, but that there is no space between the last
4047: letter of @code{welcome} and the @code{"} character. The reason for this
4048: is that @code{."} is a Forth word; it must have a space after it so that
4049: the text interpreter can identify it. The @code{"} is not a Forth word;
4050: it is a @dfn{delimiter}. The examples earlier show that, when the string
4051: is displayed, there is neither a space before the @code{H} nor after the
4052: @code{e}. Since @code{."} is an immediate word, it executes at the time
4053: that @code{greet} is defined. When it executes, its behaviour is to
4054: search forward in the input line looking for the delimiter. When it
4055: finds the delimiter, it updates @code{>IN} to point past the
4056: delimiter. It also compiles some magic code into the definition of
4057: @code{greet}; the xt of a run-time routine that prints a text string. It
4058: compiles the string @code{Hello and welcome} into memory so that it is
4059: available to be printed later. When the text interpreter gains control,
4060: the next word it finds in the input stream is @code{;} and so it
4061: terminates the definition of @code{greet}.
1.28 crook 4062:
4063:
4064: @comment ----------------------------------------------
1.29 crook 4065: @node Forth is written in Forth, Review - elements of a Forth system, How does that work?, Introduction
4066: @section Forth is written in Forth
4067: @cindex structure of Forth programs
4068:
4069: When you start up a Forth compiler, a large number of definitions
4070: already exist. In Forth, you develop a new application using bottom-up
4071: programming techniques to create new definitions that are defined in
4072: terms of existing definitions. As you create each definition you can
4073: test and debug it interactively.
4074:
4075: If you have tried out the examples in this section, you will probably
4076: have typed them in by hand; when you leave Gforth, your definitions will
4077: be lost. You can avoid this by using a text editor to enter Forth source
4078: code into a file, and then loading code from the file using
1.49 anton 4079: @code{include} (@pxref{Forth source files}). A Forth source file is
1.29 crook 4080: processed by the text interpreter, just as though you had typed it in by
4081: hand@footnote{Actually, there are some subtle differences -- see
4082: @ref{The Text Interpreter}.}.
4083:
4084: Gforth also supports the traditional Forth alternative to using text
1.49 anton 4085: files for program entry (@pxref{Blocks}).
1.28 crook 4086:
1.29 crook 4087: In common with many, if not most, Forth compilers, most of Gforth is
4088: actually written in Forth. All of the @file{.fs} files in the
4089: installation directory@footnote{For example,
1.30 anton 4090: @file{/usr/local/share/gforth...}} are Forth source files, which you can
1.29 crook 4091: study to see examples of Forth programming.
1.28 crook 4092:
1.29 crook 4093: Gforth maintains a history file that records every line that you type to
4094: the text interpreter. This file is preserved between sessions, and is
4095: used to provide a command-line recall facility. If you enter long
4096: definitions by hand, you can use a text editor to paste them out of the
4097: history file into a Forth source file for reuse at a later time
1.49 anton 4098: (for more information @pxref{Command-line editing}).
1.28 crook 4099:
4100:
4101: @comment ----------------------------------------------
1.29 crook 4102: @node Review - elements of a Forth system, Where to go next, Forth is written in Forth, Introduction
4103: @section Review - elements of a Forth system
4104: @cindex elements of a Forth system
1.28 crook 4105:
1.29 crook 4106: To summarise this chapter:
1.28 crook 4107:
4108: @itemize @bullet
4109: @item
1.29 crook 4110: Forth programs use @dfn{factoring} to break a problem down into small
4111: fragments called @dfn{words} or @dfn{definitions}.
4112: @item
4113: Forth program development is an interactive process.
4114: @item
4115: The main command loop that accepts input, and controls both
4116: interpretation and compilation, is called the @dfn{text interpreter}
4117: (also known as the @dfn{outer interpreter}).
4118: @item
4119: Forth has a very simple syntax, consisting of words and numbers
4120: separated by spaces or carriage-return characters. Any additional syntax
4121: is imposed by @dfn{parsing words}.
4122: @item
4123: Forth uses a stack to pass parameters between words. As a result, it
4124: uses postfix notation.
4125: @item
4126: To use a word that has previously been defined, the text interpreter
4127: searches for the word in the @dfn{name dictionary}.
4128: @item
1.30 anton 4129: Words have @dfn{interpretation semantics} and @dfn{compilation semantics}.
1.28 crook 4130: @item
1.29 crook 4131: The text interpreter uses the value of @code{state} to select between
4132: the use of the @dfn{interpretation semantics} and the @dfn{compilation
4133: semantics} of a word that it encounters.
1.28 crook 4134: @item
1.30 anton 4135: The relationship between the @dfn{interpretation semantics} and
4136: @dfn{compilation semantics} for a word
1.29 crook 4137: depend upon the way in which the word was defined (for example, whether
4138: it is an @dfn{immediate} word).
1.28 crook 4139: @item
1.29 crook 4140: Forth definitions can be implemented in Forth (called @dfn{high-level
4141: definitions}) or in some other way (usually a lower-level language and
4142: as a result often called @dfn{low-level definitions}, @dfn{code
4143: definitions} or @dfn{primitives}).
1.28 crook 4144: @item
1.29 crook 4145: Many Forth systems are implemented mainly in Forth.
1.28 crook 4146: @end itemize
4147:
4148:
1.29 crook 4149: @comment ----------------------------------------------
1.48 anton 4150: @node Where to go next, Exercises, Review - elements of a Forth system, Introduction
1.29 crook 4151: @section Where To Go Next
4152: @cindex where to go next
1.28 crook 4153:
1.29 crook 4154: Amazing as it may seem, if you have read (and understood) this far, you
4155: know almost all the fundamentals about the inner workings of a Forth
4156: system. You certainly know enough to be able to read and understand the
4157: rest of this manual and the ANS Forth document, to learn more about the
4158: facilities that Forth in general and Gforth in particular provide. Even
4159: scarier, you know almost enough to implement your own Forth system.
1.30 anton 4160: However, that's not a good idea just yet... better to try writing some
1.29 crook 4161: programs in Gforth.
1.28 crook 4162:
1.29 crook 4163: Forth has such a rich vocabulary that it can be hard to know where to
4164: start in learning it. This section suggests a few sets of words that are
4165: enough to write small but useful programs. Use the word index in this
4166: document to learn more about each word, then try it out and try to write
4167: small definitions using it. Start by experimenting with these words:
1.28 crook 4168:
4169: @itemize @bullet
4170: @item
1.29 crook 4171: Arithmetic: @code{+ - * / /MOD */ ABS INVERT}
4172: @item
4173: Comparison: @code{MIN MAX =}
4174: @item
4175: Logic: @code{AND OR XOR NOT}
4176: @item
4177: Stack manipulation: @code{DUP DROP SWAP OVER}
1.28 crook 4178: @item
1.29 crook 4179: Loops and decisions: @code{IF ELSE ENDIF ?DO I LOOP}
1.28 crook 4180: @item
1.29 crook 4181: Input/Output: @code{. ." EMIT CR KEY}
1.28 crook 4182: @item
1.29 crook 4183: Defining words: @code{: ; CREATE}
1.28 crook 4184: @item
1.29 crook 4185: Memory allocation words: @code{ALLOT ,}
1.28 crook 4186: @item
1.29 crook 4187: Tools: @code{SEE WORDS .S MARKER}
4188: @end itemize
4189:
4190: When you have mastered those, go on to:
4191:
4192: @itemize @bullet
1.28 crook 4193: @item
1.29 crook 4194: More defining words: @code{VARIABLE CONSTANT VALUE TO CREATE DOES>}
1.28 crook 4195: @item
1.29 crook 4196: Memory access: @code{@@ !}
1.28 crook 4197: @end itemize
1.23 crook 4198:
1.29 crook 4199: When you have mastered these, there's nothing for it but to read through
4200: the whole of this manual and find out what you've missed.
4201:
4202: @comment ----------------------------------------------
1.48 anton 4203: @node Exercises, , Where to go next, Introduction
1.29 crook 4204: @section Exercises
4205: @cindex exercises
4206:
4207: TODO: provide a set of programming excercises linked into the stuff done
4208: already and into other sections of the manual. Provide solutions to all
4209: the exercises in a .fs file in the distribution.
4210:
4211: @c Get some inspiration from Starting Forth and Kelly&Spies.
4212:
4213: @c excercises:
4214: @c 1. take inches and convert to feet and inches.
4215: @c 2. take temperature and convert from fahrenheight to celcius;
4216: @c may need to care about symmetric vs floored??
4217: @c 3. take input line and do character substitution
4218: @c to encipher or decipher
4219: @c 4. as above but work on a file for in and out
4220: @c 5. take input line and convert to pig-latin
4221: @c
4222: @c thing of sets of things to exercise then come up with
4223: @c problems that need those things.
4224:
4225:
1.26 crook 4226: @c ******************************************************************
1.29 crook 4227: @node Words, Error messages, Introduction, Top
1.1 anton 4228: @chapter Forth Words
1.26 crook 4229: @cindex words
1.1 anton 4230:
4231: @menu
4232: * Notation::
1.65 anton 4233: * Case insensitivity::
4234: * Comments::
4235: * Boolean Flags::
1.1 anton 4236: * Arithmetic::
4237: * Stack Manipulation::
1.5 anton 4238: * Memory::
1.1 anton 4239: * Control Structures::
4240: * Defining Words::
1.65 anton 4241: * Interpretation and Compilation Semantics::
1.47 crook 4242: * Tokens for Words::
1.81 anton 4243: * Compiling words::
1.65 anton 4244: * The Text Interpreter::
1.111 anton 4245: * The Input Stream::
1.65 anton 4246: * Word Lists::
4247: * Environmental Queries::
1.12 anton 4248: * Files::
4249: * Blocks::
4250: * Other I/O::
1.121 anton 4251: * OS command line arguments::
1.78 anton 4252: * Locals::
4253: * Structures::
4254: * Object-oriented Forth::
1.12 anton 4255: * Programming Tools::
1.150 anton 4256: * C Interface::
1.12 anton 4257: * Assembler and Code Words::
4258: * Threading Words::
1.65 anton 4259: * Passing Commands to the OS::
4260: * Keeping track of Time::
4261: * Miscellaneous Words::
1.1 anton 4262: @end menu
4263:
1.65 anton 4264: @node Notation, Case insensitivity, Words, Words
1.1 anton 4265: @section Notation
4266: @cindex notation of glossary entries
4267: @cindex format of glossary entries
4268: @cindex glossary notation format
4269: @cindex word glossary entry format
4270:
4271: The Forth words are described in this section in the glossary notation
1.67 anton 4272: that has become a de-facto standard for Forth texts:
1.1 anton 4273:
4274: @format
1.29 crook 4275: @i{word} @i{Stack effect} @i{wordset} @i{pronunciation}
1.1 anton 4276: @end format
1.29 crook 4277: @i{Description}
1.1 anton 4278:
4279: @table @var
4280: @item word
1.28 crook 4281: The name of the word.
1.1 anton 4282:
4283: @item Stack effect
4284: @cindex stack effect
1.29 crook 4285: The stack effect is written in the notation @code{@i{before} --
4286: @i{after}}, where @i{before} and @i{after} describe the top of
1.1 anton 4287: stack entries before and after the execution of the word. The rest of
4288: the stack is not touched by the word. The top of stack is rightmost,
4289: i.e., a stack sequence is written as it is typed in. Note that Gforth
4290: uses a separate floating point stack, but a unified stack
1.29 crook 4291: notation. Also, return stack effects are not shown in @i{stack
4292: effect}, but in @i{Description}. The name of a stack item describes
1.1 anton 4293: the type and/or the function of the item. See below for a discussion of
4294: the types.
4295:
4296: All words have two stack effects: A compile-time stack effect and a
4297: run-time stack effect. The compile-time stack-effect of most words is
1.29 crook 4298: @i{ -- }. If the compile-time stack-effect of a word deviates from
1.1 anton 4299: this standard behaviour, or the word does other unusual things at
4300: compile time, both stack effects are shown; otherwise only the run-time
4301: stack effect is shown.
4302:
4303: @cindex pronounciation of words
4304: @item pronunciation
4305: How the word is pronounced.
4306:
4307: @cindex wordset
1.67 anton 4308: @cindex environment wordset
1.1 anton 4309: @item wordset
1.21 crook 4310: The ANS Forth standard is divided into several word sets. A standard
4311: system need not support all of them. Therefore, in theory, the fewer
4312: word sets your program uses the more portable it will be. However, we
4313: suspect that most ANS Forth systems on personal machines will feature
1.26 crook 4314: all word sets. Words that are not defined in ANS Forth have
1.21 crook 4315: @code{gforth} or @code{gforth-internal} as word set. @code{gforth}
1.1 anton 4316: describes words that will work in future releases of Gforth;
4317: @code{gforth-internal} words are more volatile. Environmental query
4318: strings are also displayed like words; you can recognize them by the
1.21 crook 4319: @code{environment} in the word set field.
1.1 anton 4320:
4321: @item Description
4322: A description of the behaviour of the word.
4323: @end table
4324:
4325: @cindex types of stack items
4326: @cindex stack item types
4327: The type of a stack item is specified by the character(s) the name
4328: starts with:
4329:
4330: @table @code
4331: @item f
4332: @cindex @code{f}, stack item type
4333: Boolean flags, i.e. @code{false} or @code{true}.
4334: @item c
4335: @cindex @code{c}, stack item type
4336: Char
4337: @item w
4338: @cindex @code{w}, stack item type
4339: Cell, can contain an integer or an address
4340: @item n
4341: @cindex @code{n}, stack item type
4342: signed integer
4343: @item u
4344: @cindex @code{u}, stack item type
4345: unsigned integer
4346: @item d
4347: @cindex @code{d}, stack item type
4348: double sized signed integer
4349: @item ud
4350: @cindex @code{ud}, stack item type
4351: double sized unsigned integer
4352: @item r
4353: @cindex @code{r}, stack item type
4354: Float (on the FP stack)
1.21 crook 4355: @item a-
1.1 anton 4356: @cindex @code{a_}, stack item type
4357: Cell-aligned address
1.21 crook 4358: @item c-
1.1 anton 4359: @cindex @code{c_}, stack item type
4360: Char-aligned address (note that a Char may have two bytes in Windows NT)
1.21 crook 4361: @item f-
1.1 anton 4362: @cindex @code{f_}, stack item type
4363: Float-aligned address
1.21 crook 4364: @item df-
1.1 anton 4365: @cindex @code{df_}, stack item type
4366: Address aligned for IEEE double precision float
1.21 crook 4367: @item sf-
1.1 anton 4368: @cindex @code{sf_}, stack item type
4369: Address aligned for IEEE single precision float
4370: @item xt
4371: @cindex @code{xt}, stack item type
4372: Execution token, same size as Cell
4373: @item wid
4374: @cindex @code{wid}, stack item type
1.21 crook 4375: Word list ID, same size as Cell
1.68 anton 4376: @item ior, wior
4377: @cindex ior type description
4378: @cindex wior type description
4379: I/O result code, cell-sized. In Gforth, you can @code{throw} iors.
1.1 anton 4380: @item f83name
4381: @cindex @code{f83name}, stack item type
4382: Pointer to a name structure
4383: @item "
4384: @cindex @code{"}, stack item type
1.12 anton 4385: string in the input stream (not on the stack). The terminating character
4386: is a blank by default. If it is not a blank, it is shown in @code{<>}
1.1 anton 4387: quotes.
4388: @end table
4389:
1.65 anton 4390: @comment ----------------------------------------------
4391: @node Case insensitivity, Comments, Notation, Words
4392: @section Case insensitivity
4393: @cindex case sensitivity
4394: @cindex upper and lower case
4395:
4396: Gforth is case-insensitive; you can enter definitions and invoke
4397: Standard words using upper, lower or mixed case (however,
4398: @pxref{core-idef, Implementation-defined options, Implementation-defined
4399: options}).
4400:
4401: ANS Forth only @i{requires} implementations to recognise Standard words
4402: when they are typed entirely in upper case. Therefore, a Standard
4403: program must use upper case for all Standard words. You can use whatever
4404: case you like for words that you define, but in a Standard program you
4405: have to use the words in the same case that you defined them.
4406:
4407: Gforth supports case sensitivity through @code{table}s (case-sensitive
4408: wordlists, @pxref{Word Lists}).
4409:
4410: Two people have asked how to convert Gforth to be case-sensitive; while
4411: we think this is a bad idea, you can change all wordlists into tables
4412: like this:
4413:
4414: @example
4415: ' table-find forth-wordlist wordlist-map @ !
4416: @end example
4417:
4418: Note that you now have to type the predefined words in the same case
4419: that we defined them, which are varying. You may want to convert them
4420: to your favourite case before doing this operation (I won't explain how,
4421: because if you are even contemplating doing this, you'd better have
4422: enough knowledge of Forth systems to know this already).
4423:
4424: @node Comments, Boolean Flags, Case insensitivity, Words
1.21 crook 4425: @section Comments
1.26 crook 4426: @cindex comments
1.21 crook 4427:
1.29 crook 4428: Forth supports two styles of comment; the traditional @i{in-line} comment,
4429: @code{(} and its modern cousin, the @i{comment to end of line}; @code{\}.
1.21 crook 4430:
1.44 crook 4431:
1.23 crook 4432: doc-(
1.21 crook 4433: doc-\
1.23 crook 4434: doc-\G
1.21 crook 4435:
1.44 crook 4436:
1.21 crook 4437: @node Boolean Flags, Arithmetic, Comments, Words
4438: @section Boolean Flags
1.26 crook 4439: @cindex Boolean flags
1.21 crook 4440:
4441: A Boolean flag is cell-sized. A cell with all bits clear represents the
4442: flag @code{false} and a flag with all bits set represents the flag
1.26 crook 4443: @code{true}. Words that check a flag (for example, @code{IF}) will treat
1.29 crook 4444: a cell that has @i{any} bit set as @code{true}.
1.67 anton 4445: @c on and off to Memory?
4446: @c true and false to "Bitwise operations" or "Numeric comparison"?
1.44 crook 4447:
1.21 crook 4448: doc-true
4449: doc-false
1.29 crook 4450: doc-on
4451: doc-off
1.21 crook 4452:
1.44 crook 4453:
1.21 crook 4454: @node Arithmetic, Stack Manipulation, Boolean Flags, Words
1.1 anton 4455: @section Arithmetic
4456: @cindex arithmetic words
4457:
4458: @cindex division with potentially negative operands
4459: Forth arithmetic is not checked, i.e., you will not hear about integer
4460: overflow on addition or multiplication, you may hear about division by
4461: zero if you are lucky. The operator is written after the operands, but
4462: the operands are still in the original order. I.e., the infix @code{2-1}
4463: corresponds to @code{2 1 -}. Forth offers a variety of division
4464: operators. If you perform division with potentially negative operands,
4465: you do not want to use @code{/} or @code{/mod} with its undefined
4466: behaviour, but rather @code{fm/mod} or @code{sm/mod} (probably the
4467: former, @pxref{Mixed precision}).
1.26 crook 4468: @comment TODO discuss the different division forms and the std approach
1.1 anton 4469:
4470: @menu
4471: * Single precision::
1.67 anton 4472: * Double precision:: Double-cell integer arithmetic
1.1 anton 4473: * Bitwise operations::
1.67 anton 4474: * Numeric comparison::
1.29 crook 4475: * Mixed precision:: Operations with single and double-cell integers
1.1 anton 4476: * Floating Point::
4477: @end menu
4478:
1.67 anton 4479: @node Single precision, Double precision, Arithmetic, Arithmetic
1.1 anton 4480: @subsection Single precision
4481: @cindex single precision arithmetic words
4482:
1.67 anton 4483: @c !! cell undefined
4484:
4485: By default, numbers in Forth are single-precision integers that are one
1.26 crook 4486: cell in size. They can be signed or unsigned, depending upon how you
1.49 anton 4487: treat them. For the rules used by the text interpreter for recognising
4488: single-precision integers see @ref{Number Conversion}.
1.21 crook 4489:
1.67 anton 4490: These words are all defined for signed operands, but some of them also
4491: work for unsigned numbers: @code{+}, @code{1+}, @code{-}, @code{1-},
4492: @code{*}.
1.44 crook 4493:
1.1 anton 4494: doc-+
1.21 crook 4495: doc-1+
1.128 anton 4496: doc-under+
1.1 anton 4497: doc--
1.21 crook 4498: doc-1-
1.1 anton 4499: doc-*
4500: doc-/
4501: doc-mod
4502: doc-/mod
4503: doc-negate
4504: doc-abs
4505: doc-min
4506: doc-max
1.27 crook 4507: doc-floored
1.1 anton 4508:
1.44 crook 4509:
1.67 anton 4510: @node Double precision, Bitwise operations, Single precision, Arithmetic
1.21 crook 4511: @subsection Double precision
4512: @cindex double precision arithmetic words
4513:
1.49 anton 4514: For the rules used by the text interpreter for
4515: recognising double-precision integers, see @ref{Number Conversion}.
1.21 crook 4516:
4517: A double precision number is represented by a cell pair, with the most
1.67 anton 4518: significant cell at the TOS. It is trivial to convert an unsigned single
4519: to a double: simply push a @code{0} onto the TOS. Since numbers are
4520: represented by Gforth using 2's complement arithmetic, converting a
4521: signed single to a (signed) double requires sign-extension across the
4522: most significant cell. This can be achieved using @code{s>d}. The moral
4523: of the story is that you cannot convert a number without knowing whether
4524: it represents an unsigned or a signed number.
1.21 crook 4525:
1.67 anton 4526: These words are all defined for signed operands, but some of them also
4527: work for unsigned numbers: @code{d+}, @code{d-}.
1.44 crook 4528:
1.21 crook 4529: doc-s>d
1.67 anton 4530: doc-d>s
1.21 crook 4531: doc-d+
4532: doc-d-
4533: doc-dnegate
4534: doc-dabs
4535: doc-dmin
4536: doc-dmax
4537:
1.44 crook 4538:
1.67 anton 4539: @node Bitwise operations, Numeric comparison, Double precision, Arithmetic
4540: @subsection Bitwise operations
4541: @cindex bitwise operation words
4542:
4543:
4544: doc-and
4545: doc-or
4546: doc-xor
4547: doc-invert
4548: doc-lshift
4549: doc-rshift
4550: doc-2*
4551: doc-d2*
4552: doc-2/
4553: doc-d2/
4554:
4555:
4556: @node Numeric comparison, Mixed precision, Bitwise operations, Arithmetic
1.21 crook 4557: @subsection Numeric comparison
4558: @cindex numeric comparison words
4559:
1.67 anton 4560: Note that the words that compare for equality (@code{= <> 0= 0<> d= d<>
4561: d0= d0<>}) work for for both signed and unsigned numbers.
1.44 crook 4562:
1.28 crook 4563: doc-<
4564: doc-<=
4565: doc-<>
4566: doc-=
4567: doc->
4568: doc->=
4569:
1.21 crook 4570: doc-0<
1.23 crook 4571: doc-0<=
1.21 crook 4572: doc-0<>
4573: doc-0=
1.23 crook 4574: doc-0>
4575: doc-0>=
1.28 crook 4576:
4577: doc-u<
4578: doc-u<=
1.44 crook 4579: @c u<> and u= exist but are the same as <> and =
1.31 anton 4580: @c doc-u<>
4581: @c doc-u=
1.28 crook 4582: doc-u>
4583: doc-u>=
4584:
4585: doc-within
4586:
4587: doc-d<
4588: doc-d<=
4589: doc-d<>
4590: doc-d=
4591: doc-d>
4592: doc-d>=
1.23 crook 4593:
1.21 crook 4594: doc-d0<
1.23 crook 4595: doc-d0<=
4596: doc-d0<>
1.21 crook 4597: doc-d0=
1.23 crook 4598: doc-d0>
4599: doc-d0>=
4600:
1.21 crook 4601: doc-du<
1.28 crook 4602: doc-du<=
1.44 crook 4603: @c du<> and du= exist but are the same as d<> and d=
1.31 anton 4604: @c doc-du<>
4605: @c doc-du=
1.28 crook 4606: doc-du>
4607: doc-du>=
1.1 anton 4608:
1.44 crook 4609:
1.21 crook 4610: @node Mixed precision, Floating Point, Numeric comparison, Arithmetic
1.1 anton 4611: @subsection Mixed precision
4612: @cindex mixed precision arithmetic words
4613:
1.44 crook 4614:
1.1 anton 4615: doc-m+
4616: doc-*/
4617: doc-*/mod
4618: doc-m*
4619: doc-um*
4620: doc-m*/
4621: doc-um/mod
4622: doc-fm/mod
4623: doc-sm/rem
4624:
1.44 crook 4625:
1.21 crook 4626: @node Floating Point, , Mixed precision, Arithmetic
1.1 anton 4627: @subsection Floating Point
4628: @cindex floating point arithmetic words
4629:
1.49 anton 4630: For the rules used by the text interpreter for
4631: recognising floating-point numbers see @ref{Number Conversion}.
1.1 anton 4632:
1.67 anton 4633: Gforth has a separate floating point stack, but the documentation uses
4634: the unified notation.@footnote{It's easy to generate the separate
4635: notation from that by just separating the floating-point numbers out:
4636: e.g. @code{( n r1 u r2 -- r3 )} becomes @code{( n u -- ) ( F: r1 r2 --
4637: r3 )}.}
1.1 anton 4638:
4639: @cindex floating-point arithmetic, pitfalls
4640: Floating point numbers have a number of unpleasant surprises for the
4641: unwary (e.g., floating point addition is not associative) and even a few
4642: for the wary. You should not use them unless you know what you are doing
4643: or you don't care that the results you get are totally bogus. If you
4644: want to learn about the problems of floating point numbers (and how to
1.66 anton 4645: avoid them), you might start with @cite{David Goldberg,
4646: @uref{http://www.validgh.com/goldberg/paper.ps,What Every Computer
4647: Scientist Should Know About Floating-Point Arithmetic}, ACM Computing
4648: Surveys 23(1):5@minus{}48, March 1991}.
1.1 anton 4649:
1.44 crook 4650:
1.21 crook 4651: doc-d>f
4652: doc-f>d
1.1 anton 4653: doc-f+
4654: doc-f-
4655: doc-f*
4656: doc-f/
4657: doc-fnegate
4658: doc-fabs
4659: doc-fmax
4660: doc-fmin
4661: doc-floor
4662: doc-fround
4663: doc-f**
4664: doc-fsqrt
4665: doc-fexp
4666: doc-fexpm1
4667: doc-fln
4668: doc-flnp1
4669: doc-flog
4670: doc-falog
1.32 anton 4671: doc-f2*
4672: doc-f2/
4673: doc-1/f
4674: doc-precision
4675: doc-set-precision
4676:
4677: @cindex angles in trigonometric operations
4678: @cindex trigonometric operations
4679: Angles in floating point operations are given in radians (a full circle
4680: has 2 pi radians).
4681:
1.1 anton 4682: doc-fsin
4683: doc-fcos
4684: doc-fsincos
4685: doc-ftan
4686: doc-fasin
4687: doc-facos
4688: doc-fatan
4689: doc-fatan2
4690: doc-fsinh
4691: doc-fcosh
4692: doc-ftanh
4693: doc-fasinh
4694: doc-facosh
4695: doc-fatanh
1.21 crook 4696: doc-pi
1.28 crook 4697:
1.32 anton 4698: @cindex equality of floats
4699: @cindex floating-point comparisons
1.31 anton 4700: One particular problem with floating-point arithmetic is that comparison
4701: for equality often fails when you would expect it to succeed. For this
4702: reason approximate equality is often preferred (but you still have to
1.67 anton 4703: know what you are doing). Also note that IEEE NaNs may compare
1.68 anton 4704: differently from what you might expect. The comparison words are:
1.31 anton 4705:
4706: doc-f~rel
4707: doc-f~abs
1.68 anton 4708: doc-f~
1.31 anton 4709: doc-f=
4710: doc-f<>
4711:
4712: doc-f<
4713: doc-f<=
4714: doc-f>
4715: doc-f>=
4716:
1.21 crook 4717: doc-f0<
1.28 crook 4718: doc-f0<=
4719: doc-f0<>
1.21 crook 4720: doc-f0=
1.28 crook 4721: doc-f0>
4722: doc-f0>=
4723:
1.1 anton 4724:
4725: @node Stack Manipulation, Memory, Arithmetic, Words
4726: @section Stack Manipulation
4727: @cindex stack manipulation words
4728:
4729: @cindex floating-point stack in the standard
1.21 crook 4730: Gforth maintains a number of separate stacks:
4731:
1.29 crook 4732: @cindex data stack
4733: @cindex parameter stack
1.21 crook 4734: @itemize @bullet
4735: @item
1.29 crook 4736: A data stack (also known as the @dfn{parameter stack}) -- for
4737: characters, cells, addresses, and double cells.
1.21 crook 4738:
1.29 crook 4739: @cindex floating-point stack
1.21 crook 4740: @item
1.44 crook 4741: A floating point stack -- for holding floating point (FP) numbers.
1.21 crook 4742:
1.29 crook 4743: @cindex return stack
1.21 crook 4744: @item
1.44 crook 4745: A return stack -- for holding the return addresses of colon
1.32 anton 4746: definitions and other (non-FP) data.
1.21 crook 4747:
1.29 crook 4748: @cindex locals stack
1.21 crook 4749: @item
1.44 crook 4750: A locals stack -- for holding local variables.
1.21 crook 4751: @end itemize
4752:
1.1 anton 4753: @menu
4754: * Data stack::
4755: * Floating point stack::
4756: * Return stack::
4757: * Locals stack::
4758: * Stack pointer manipulation::
4759: @end menu
4760:
4761: @node Data stack, Floating point stack, Stack Manipulation, Stack Manipulation
4762: @subsection Data stack
4763: @cindex data stack manipulation words
4764: @cindex stack manipulations words, data stack
4765:
1.44 crook 4766:
1.1 anton 4767: doc-drop
4768: doc-nip
4769: doc-dup
4770: doc-over
4771: doc-tuck
4772: doc-swap
1.21 crook 4773: doc-pick
1.1 anton 4774: doc-rot
4775: doc--rot
4776: doc-?dup
4777: doc-roll
4778: doc-2drop
4779: doc-2nip
4780: doc-2dup
4781: doc-2over
4782: doc-2tuck
4783: doc-2swap
4784: doc-2rot
4785:
1.44 crook 4786:
1.1 anton 4787: @node Floating point stack, Return stack, Data stack, Stack Manipulation
4788: @subsection Floating point stack
4789: @cindex floating-point stack manipulation words
4790: @cindex stack manipulation words, floating-point stack
4791:
1.32 anton 4792: Whilst every sane Forth has a separate floating-point stack, it is not
4793: strictly required; an ANS Forth system could theoretically keep
4794: floating-point numbers on the data stack. As an additional difficulty,
4795: you don't know how many cells a floating-point number takes. It is
4796: reportedly possible to write words in a way that they work also for a
4797: unified stack model, but we do not recommend trying it. Instead, just
4798: say that your program has an environmental dependency on a separate
4799: floating-point stack.
4800:
4801: doc-floating-stack
4802:
1.1 anton 4803: doc-fdrop
4804: doc-fnip
4805: doc-fdup
4806: doc-fover
4807: doc-ftuck
4808: doc-fswap
1.21 crook 4809: doc-fpick
1.1 anton 4810: doc-frot
4811:
1.44 crook 4812:
1.1 anton 4813: @node Return stack, Locals stack, Floating point stack, Stack Manipulation
4814: @subsection Return stack
4815: @cindex return stack manipulation words
4816: @cindex stack manipulation words, return stack
4817:
1.32 anton 4818: @cindex return stack and locals
4819: @cindex locals and return stack
4820: A Forth system is allowed to keep local variables on the
4821: return stack. This is reasonable, as local variables usually eliminate
4822: the need to use the return stack explicitly. So, if you want to produce
4823: a standard compliant program and you are using local variables in a
4824: word, forget about return stack manipulations in that word (refer to the
4825: standard document for the exact rules).
4826:
1.1 anton 4827: doc->r
4828: doc-r>
4829: doc-r@
4830: doc-rdrop
4831: doc-2>r
4832: doc-2r>
4833: doc-2r@
4834: doc-2rdrop
4835:
1.44 crook 4836:
1.1 anton 4837: @node Locals stack, Stack pointer manipulation, Return stack, Stack Manipulation
4838: @subsection Locals stack
4839:
1.78 anton 4840: Gforth uses an extra locals stack. It is described, along with the
4841: reasons for its existence, in @ref{Locals implementation}.
1.21 crook 4842:
1.1 anton 4843: @node Stack pointer manipulation, , Locals stack, Stack Manipulation
4844: @subsection Stack pointer manipulation
4845: @cindex stack pointer manipulation words
4846:
1.44 crook 4847: @c removed s0 r0 l0 -- they are obsolete aliases for sp0 rp0 lp0
1.21 crook 4848: doc-sp0
1.1 anton 4849: doc-sp@
4850: doc-sp!
1.21 crook 4851: doc-fp0
1.1 anton 4852: doc-fp@
4853: doc-fp!
1.21 crook 4854: doc-rp0
1.1 anton 4855: doc-rp@
4856: doc-rp!
1.21 crook 4857: doc-lp0
1.1 anton 4858: doc-lp@
4859: doc-lp!
4860:
1.44 crook 4861:
1.1 anton 4862: @node Memory, Control Structures, Stack Manipulation, Words
4863: @section Memory
1.26 crook 4864: @cindex memory words
1.1 anton 4865:
1.32 anton 4866: @menu
4867: * Memory model::
4868: * Dictionary allocation::
4869: * Heap Allocation::
4870: * Memory Access::
4871: * Address arithmetic::
4872: * Memory Blocks::
4873: @end menu
4874:
1.67 anton 4875: In addition to the standard Forth memory allocation words, there is also
4876: a @uref{http://www.complang.tuwien.ac.at/forth/garbage-collection.zip,
4877: garbage collector}.
4878:
1.32 anton 4879: @node Memory model, Dictionary allocation, Memory, Memory
4880: @subsection ANS Forth and Gforth memory models
4881:
4882: @c The ANS Forth description is a mess (e.g., is the heap part of
4883: @c the dictionary?), so let's not stick to closely with it.
4884:
1.67 anton 4885: ANS Forth considers a Forth system as consisting of several address
4886: spaces, of which only @dfn{data space} is managed and accessible with
4887: the memory words. Memory not necessarily in data space includes the
4888: stacks, the code (called code space) and the headers (called name
4889: space). In Gforth everything is in data space, but the code for the
4890: primitives is usually read-only.
1.32 anton 4891:
4892: Data space is divided into a number of areas: The (data space portion of
4893: the) dictionary@footnote{Sometimes, the term @dfn{dictionary} is used to
4894: refer to the search data structure embodied in word lists and headers,
4895: because it is used for looking up names, just as you would in a
4896: conventional dictionary.}, the heap, and a number of system-allocated
4897: buffers.
4898:
1.68 anton 4899: @cindex address arithmetic restrictions, ANS vs. Gforth
4900: @cindex contiguous regions, ANS vs. Gforth
1.32 anton 4901: In ANS Forth data space is also divided into contiguous regions. You
4902: can only use address arithmetic within a contiguous region, not between
4903: them. Usually each allocation gives you one contiguous region, but the
1.33 anton 4904: dictionary allocation words have additional rules (@pxref{Dictionary
1.32 anton 4905: allocation}).
4906:
4907: Gforth provides one big address space, and address arithmetic can be
4908: performed between any addresses. However, in the dictionary headers or
4909: code are interleaved with data, so almost the only contiguous data space
4910: regions there are those described by ANS Forth as contiguous; but you
4911: can be sure that the dictionary is allocated towards increasing
4912: addresses even between contiguous regions. The memory order of
4913: allocations in the heap is platform-dependent (and possibly different
4914: from one run to the next).
4915:
1.27 crook 4916:
1.32 anton 4917: @node Dictionary allocation, Heap Allocation, Memory model, Memory
4918: @subsection Dictionary allocation
1.27 crook 4919: @cindex reserving data space
4920: @cindex data space - reserving some
4921:
1.32 anton 4922: Dictionary allocation is a stack-oriented allocation scheme, i.e., if
4923: you want to deallocate X, you also deallocate everything
4924: allocated after X.
4925:
1.68 anton 4926: @cindex contiguous regions in dictionary allocation
1.32 anton 4927: The allocations using the words below are contiguous and grow the region
4928: towards increasing addresses. Other words that allocate dictionary
4929: memory of any kind (i.e., defining words including @code{:noname}) end
4930: the contiguous region and start a new one.
4931:
4932: In ANS Forth only @code{create}d words are guaranteed to produce an
4933: address that is the start of the following contiguous region. In
4934: particular, the cell allocated by @code{variable} is not guaranteed to
4935: be contiguous with following @code{allot}ed memory.
4936:
4937: You can deallocate memory by using @code{allot} with a negative argument
4938: (with some restrictions, see @code{allot}). For larger deallocations use
4939: @code{marker}.
1.27 crook 4940:
1.29 crook 4941:
1.27 crook 4942: doc-here
4943: doc-unused
4944: doc-allot
4945: doc-c,
1.29 crook 4946: doc-f,
1.27 crook 4947: doc-,
4948: doc-2,
4949:
1.32 anton 4950: Memory accesses have to be aligned (@pxref{Address arithmetic}). So of
4951: course you should allocate memory in an aligned way, too. I.e., before
4952: allocating allocating a cell, @code{here} must be cell-aligned, etc.
4953: The words below align @code{here} if it is not already. Basically it is
4954: only already aligned for a type, if the last allocation was a multiple
4955: of the size of this type and if @code{here} was aligned for this type
4956: before.
4957:
4958: After freshly @code{create}ing a word, @code{here} is @code{align}ed in
4959: ANS Forth (@code{maxalign}ed in Gforth).
4960:
4961: doc-align
4962: doc-falign
4963: doc-sfalign
4964: doc-dfalign
4965: doc-maxalign
4966: doc-cfalign
4967:
4968:
4969: @node Heap Allocation, Memory Access, Dictionary allocation, Memory
4970: @subsection Heap allocation
4971: @cindex heap allocation
4972: @cindex dynamic allocation of memory
4973: @cindex memory-allocation word set
4974:
1.68 anton 4975: @cindex contiguous regions and heap allocation
1.32 anton 4976: Heap allocation supports deallocation of allocated memory in any
4977: order. Dictionary allocation is not affected by it (i.e., it does not
4978: end a contiguous region). In Gforth, these words are implemented using
4979: the standard C library calls malloc(), free() and resize().
4980:
1.68 anton 4981: The memory region produced by one invocation of @code{allocate} or
4982: @code{resize} is internally contiguous. There is no contiguity between
4983: such a region and any other region (including others allocated from the
4984: heap).
4985:
1.32 anton 4986: doc-allocate
4987: doc-free
4988: doc-resize
4989:
1.27 crook 4990:
1.32 anton 4991: @node Memory Access, Address arithmetic, Heap Allocation, Memory
1.1 anton 4992: @subsection Memory Access
4993: @cindex memory access words
4994:
4995: doc-@
4996: doc-!
4997: doc-+!
4998: doc-c@
4999: doc-c!
5000: doc-2@
5001: doc-2!
5002: doc-f@
5003: doc-f!
5004: doc-sf@
5005: doc-sf!
5006: doc-df@
5007: doc-df!
1.144 anton 5008: doc-sw@
5009: doc-uw@
5010: doc-w!
5011: doc-sl@
5012: doc-ul@
5013: doc-l!
1.68 anton 5014:
1.32 anton 5015: @node Address arithmetic, Memory Blocks, Memory Access, Memory
5016: @subsection Address arithmetic
1.1 anton 5017: @cindex address arithmetic words
5018:
1.67 anton 5019: Address arithmetic is the foundation on which you can build data
5020: structures like arrays, records (@pxref{Structures}) and objects
5021: (@pxref{Object-oriented Forth}).
1.32 anton 5022:
1.68 anton 5023: @cindex address unit
5024: @cindex au (address unit)
1.1 anton 5025: ANS Forth does not specify the sizes of the data types. Instead, it
5026: offers a number of words for computing sizes and doing address
1.29 crook 5027: arithmetic. Address arithmetic is performed in terms of address units
5028: (aus); on most systems the address unit is one byte. Note that a
5029: character may have more than one au, so @code{chars} is no noop (on
1.68 anton 5030: platforms where it is a noop, it compiles to nothing).
1.1 anton 5031:
1.67 anton 5032: The basic address arithmetic words are @code{+} and @code{-}. E.g., if
5033: you have the address of a cell, perform @code{1 cells +}, and you will
5034: have the address of the next cell.
5035:
1.68 anton 5036: @cindex contiguous regions and address arithmetic
1.67 anton 5037: In ANS Forth you can perform address arithmetic only within a contiguous
5038: region, i.e., if you have an address into one region, you can only add
5039: and subtract such that the result is still within the region; you can
5040: only subtract or compare addresses from within the same contiguous
5041: region. Reasons: several contiguous regions can be arranged in memory
5042: in any way; on segmented systems addresses may have unusual
5043: representations, such that address arithmetic only works within a
5044: region. Gforth provides a few more guarantees (linear address space,
5045: dictionary grows upwards), but in general I have found it easy to stay
5046: within contiguous regions (exception: computing and comparing to the
5047: address just beyond the end of an array).
5048:
1.1 anton 5049: @cindex alignment of addresses for types
5050: ANS Forth also defines words for aligning addresses for specific
5051: types. Many computers require that accesses to specific data types
5052: must only occur at specific addresses; e.g., that cells may only be
5053: accessed at addresses divisible by 4. Even if a machine allows unaligned
5054: accesses, it can usually perform aligned accesses faster.
5055:
5056: For the performance-conscious: alignment operations are usually only
5057: necessary during the definition of a data structure, not during the
5058: (more frequent) accesses to it.
5059:
5060: ANS Forth defines no words for character-aligning addresses. This is not
5061: an oversight, but reflects the fact that addresses that are not
5062: char-aligned have no use in the standard and therefore will not be
5063: created.
5064:
5065: @cindex @code{CREATE} and alignment
1.29 crook 5066: ANS Forth guarantees that addresses returned by @code{CREATE}d words
1.1 anton 5067: are cell-aligned; in addition, Gforth guarantees that these addresses
5068: are aligned for all purposes.
5069:
1.26 crook 5070: Note that the ANS Forth word @code{char} has nothing to do with address
5071: arithmetic.
1.1 anton 5072:
1.44 crook 5073:
1.1 anton 5074: doc-chars
5075: doc-char+
5076: doc-cells
5077: doc-cell+
5078: doc-cell
5079: doc-aligned
5080: doc-floats
5081: doc-float+
5082: doc-float
5083: doc-faligned
5084: doc-sfloats
5085: doc-sfloat+
5086: doc-sfaligned
5087: doc-dfloats
5088: doc-dfloat+
5089: doc-dfaligned
5090: doc-maxaligned
5091: doc-cfaligned
5092: doc-address-unit-bits
1.145 anton 5093: doc-/w
5094: doc-/l
1.44 crook 5095:
1.32 anton 5096: @node Memory Blocks, , Address arithmetic, Memory
1.1 anton 5097: @subsection Memory Blocks
5098: @cindex memory block words
1.27 crook 5099: @cindex character strings - moving and copying
5100:
1.49 anton 5101: Memory blocks often represent character strings; For ways of storing
5102: character strings in memory see @ref{String Formats}. For other
5103: string-processing words see @ref{Displaying characters and strings}.
1.1 anton 5104:
1.67 anton 5105: A few of these words work on address unit blocks. In that case, you
5106: usually have to insert @code{CHARS} before the word when working on
5107: character strings. Most words work on character blocks, and expect a
5108: char-aligned address.
5109:
5110: When copying characters between overlapping memory regions, use
5111: @code{chars move} or choose carefully between @code{cmove} and
5112: @code{cmove>}.
1.44 crook 5113:
1.1 anton 5114: doc-move
5115: doc-erase
5116: doc-cmove
5117: doc-cmove>
5118: doc-fill
5119: doc-blank
1.21 crook 5120: doc-compare
1.111 anton 5121: doc-str=
5122: doc-str<
5123: doc-string-prefix?
1.21 crook 5124: doc-search
1.27 crook 5125: doc--trailing
5126: doc-/string
1.82 anton 5127: doc-bounds
1.141 anton 5128: doc-pad
1.111 anton 5129:
1.27 crook 5130: @comment TODO examples
5131:
1.1 anton 5132:
1.26 crook 5133: @node Control Structures, Defining Words, Memory, Words
1.1 anton 5134: @section Control Structures
5135: @cindex control structures
5136:
1.33 anton 5137: Control structures in Forth cannot be used interpretively, only in a
5138: colon definition@footnote{To be precise, they have no interpretation
5139: semantics (@pxref{Interpretation and Compilation Semantics}).}. We do
5140: not like this limitation, but have not seen a satisfying way around it
5141: yet, although many schemes have been proposed.
1.1 anton 5142:
5143: @menu
1.33 anton 5144: * Selection:: IF ... ELSE ... ENDIF
5145: * Simple Loops:: BEGIN ...
1.29 crook 5146: * Counted Loops:: DO
1.67 anton 5147: * Arbitrary control structures::
5148: * Calls and returns::
1.1 anton 5149: * Exception Handling::
5150: @end menu
5151:
5152: @node Selection, Simple Loops, Control Structures, Control Structures
5153: @subsection Selection
5154: @cindex selection control structures
5155: @cindex control structures for selection
5156:
5157: @cindex @code{IF} control structure
5158: @example
1.29 crook 5159: @i{flag}
1.1 anton 5160: IF
1.29 crook 5161: @i{code}
1.1 anton 5162: ENDIF
5163: @end example
1.21 crook 5164: @noindent
1.33 anton 5165:
1.44 crook 5166: If @i{flag} is non-zero (as far as @code{IF} etc. are concerned, a cell
5167: with any bit set represents truth) @i{code} is executed.
1.33 anton 5168:
1.1 anton 5169: @example
1.29 crook 5170: @i{flag}
1.1 anton 5171: IF
1.29 crook 5172: @i{code1}
1.1 anton 5173: ELSE
1.29 crook 5174: @i{code2}
1.1 anton 5175: ENDIF
5176: @end example
5177:
1.44 crook 5178: If @var{flag} is true, @i{code1} is executed, otherwise @i{code2} is
5179: executed.
1.33 anton 5180:
1.1 anton 5181: You can use @code{THEN} instead of @code{ENDIF}. Indeed, @code{THEN} is
5182: standard, and @code{ENDIF} is not, although it is quite popular. We
5183: recommend using @code{ENDIF}, because it is less confusing for people
5184: who also know other languages (and is not prone to reinforcing negative
5185: prejudices against Forth in these people). Adding @code{ENDIF} to a
5186: system that only supplies @code{THEN} is simple:
5187: @example
1.82 anton 5188: : ENDIF POSTPONE then ; immediate
1.1 anton 5189: @end example
5190:
5191: [According to @cite{Webster's New Encyclopedic Dictionary}, @dfn{then
5192: (adv.)} has the following meanings:
5193: @quotation
5194: ... 2b: following next after in order ... 3d: as a necessary consequence
5195: (if you were there, then you saw them).
5196: @end quotation
5197: Forth's @code{THEN} has the meaning 2b, whereas @code{THEN} in Pascal
5198: and many other programming languages has the meaning 3d.]
5199:
1.21 crook 5200: Gforth also provides the words @code{?DUP-IF} and @code{?DUP-0=-IF}, so
1.1 anton 5201: you can avoid using @code{?dup}. Using these alternatives is also more
1.26 crook 5202: efficient than using @code{?dup}. Definitions in ANS Forth
1.1 anton 5203: for @code{ENDIF}, @code{?DUP-IF} and @code{?DUP-0=-IF} are provided in
5204: @file{compat/control.fs}.
5205:
5206: @cindex @code{CASE} control structure
5207: @example
1.29 crook 5208: @i{n}
1.1 anton 5209: CASE
1.29 crook 5210: @i{n1} OF @i{code1} ENDOF
5211: @i{n2} OF @i{code2} ENDOF
1.1 anton 5212: @dots{}
1.68 anton 5213: ( n ) @i{default-code} ( n )
1.131 anton 5214: ENDCASE ( )
1.1 anton 5215: @end example
5216:
1.131 anton 5217: Executes the first @i{codei}, where the @i{ni} is equal to @i{n}. If
5218: no @i{ni} matches, the optional @i{default-code} is executed. The
5219: optional default case can be added by simply writing the code after
5220: the last @code{ENDOF}. It may use @i{n}, which is on top of the stack,
5221: but must not consume it. The value @i{n} is consumed by this
5222: construction (either by a OF that matches, or by the ENDCASE, if no OF
5223: matches).
1.1 anton 5224:
1.69 anton 5225: @progstyle
1.131 anton 5226: To keep the code understandable, you should ensure that you change the
5227: stack in the same way (wrt. number and types of stack items consumed
5228: and pushed) on all paths through a selection construct.
1.69 anton 5229:
1.1 anton 5230: @node Simple Loops, Counted Loops, Selection, Control Structures
5231: @subsection Simple Loops
5232: @cindex simple loops
5233: @cindex loops without count
5234:
5235: @cindex @code{WHILE} loop
5236: @example
5237: BEGIN
1.29 crook 5238: @i{code1}
5239: @i{flag}
1.1 anton 5240: WHILE
1.29 crook 5241: @i{code2}
1.1 anton 5242: REPEAT
5243: @end example
5244:
1.29 crook 5245: @i{code1} is executed and @i{flag} is computed. If it is true,
5246: @i{code2} is executed and the loop is restarted; If @i{flag} is
1.1 anton 5247: false, execution continues after the @code{REPEAT}.
5248:
5249: @cindex @code{UNTIL} loop
5250: @example
5251: BEGIN
1.29 crook 5252: @i{code}
5253: @i{flag}
1.1 anton 5254: UNTIL
5255: @end example
5256:
1.29 crook 5257: @i{code} is executed. The loop is restarted if @code{flag} is false.
1.1 anton 5258:
1.69 anton 5259: @progstyle
5260: To keep the code understandable, a complete iteration of the loop should
5261: not change the number and types of the items on the stacks.
5262:
1.1 anton 5263: @cindex endless loop
5264: @cindex loops, endless
5265: @example
5266: BEGIN
1.29 crook 5267: @i{code}
1.1 anton 5268: AGAIN
5269: @end example
5270:
5271: This is an endless loop.
5272:
5273: @node Counted Loops, Arbitrary control structures, Simple Loops, Control Structures
5274: @subsection Counted Loops
5275: @cindex counted loops
5276: @cindex loops, counted
5277: @cindex @code{DO} loops
5278:
5279: The basic counted loop is:
5280: @example
1.29 crook 5281: @i{limit} @i{start}
1.1 anton 5282: ?DO
1.29 crook 5283: @i{body}
1.1 anton 5284: LOOP
5285: @end example
5286:
1.29 crook 5287: This performs one iteration for every integer, starting from @i{start}
5288: and up to, but excluding @i{limit}. The counter, or @i{index}, can be
1.21 crook 5289: accessed with @code{i}. For example, the loop:
1.1 anton 5290: @example
5291: 10 0 ?DO
5292: i .
5293: LOOP
5294: @end example
1.21 crook 5295: @noindent
5296: prints @code{0 1 2 3 4 5 6 7 8 9}
5297:
1.1 anton 5298: The index of the innermost loop can be accessed with @code{i}, the index
5299: of the next loop with @code{j}, and the index of the third loop with
5300: @code{k}.
5301:
1.44 crook 5302:
1.1 anton 5303: doc-i
5304: doc-j
5305: doc-k
5306:
1.44 crook 5307:
1.1 anton 5308: The loop control data are kept on the return stack, so there are some
1.21 crook 5309: restrictions on mixing return stack accesses and counted loop words. In
5310: particuler, if you put values on the return stack outside the loop, you
5311: cannot read them inside the loop@footnote{well, not in a way that is
5312: portable.}. If you put values on the return stack within a loop, you
5313: have to remove them before the end of the loop and before accessing the
5314: index of the loop.
1.1 anton 5315:
5316: There are several variations on the counted loop:
5317:
1.21 crook 5318: @itemize @bullet
5319: @item
5320: @code{LEAVE} leaves the innermost counted loop immediately; execution
5321: continues after the associated @code{LOOP} or @code{NEXT}. For example:
5322:
5323: @example
5324: 10 0 ?DO i DUP . 3 = IF LEAVE THEN LOOP
5325: @end example
5326: prints @code{0 1 2 3}
5327:
1.1 anton 5328:
1.21 crook 5329: @item
5330: @code{UNLOOP} prepares for an abnormal loop exit, e.g., via
5331: @code{EXIT}. @code{UNLOOP} removes the loop control parameters from the
5332: return stack so @code{EXIT} can get to its return address. For example:
5333:
5334: @example
5335: : demo 10 0 ?DO i DUP . 3 = IF UNLOOP EXIT THEN LOOP ." Done" ;
5336: @end example
5337: prints @code{0 1 2 3}
5338:
5339:
5340: @item
1.29 crook 5341: If @i{start} is greater than @i{limit}, a @code{?DO} loop is entered
1.1 anton 5342: (and @code{LOOP} iterates until they become equal by wrap-around
5343: arithmetic). This behaviour is usually not what you want. Therefore,
5344: Gforth offers @code{+DO} and @code{U+DO} (as replacements for
1.29 crook 5345: @code{?DO}), which do not enter the loop if @i{start} is greater than
5346: @i{limit}; @code{+DO} is for signed loop parameters, @code{U+DO} for
1.1 anton 5347: unsigned loop parameters.
5348:
1.21 crook 5349: @item
5350: @code{?DO} can be replaced by @code{DO}. @code{DO} always enters
5351: the loop, independent of the loop parameters. Do not use @code{DO}, even
5352: if you know that the loop is entered in any case. Such knowledge tends
5353: to become invalid during maintenance of a program, and then the
5354: @code{DO} will make trouble.
5355:
5356: @item
1.29 crook 5357: @code{LOOP} can be replaced with @code{@i{n} +LOOP}; this updates the
5358: index by @i{n} instead of by 1. The loop is terminated when the border
5359: between @i{limit-1} and @i{limit} is crossed. E.g.:
1.1 anton 5360:
1.21 crook 5361: @example
5362: 4 0 +DO i . 2 +LOOP
5363: @end example
5364: @noindent
5365: prints @code{0 2}
5366:
5367: @example
5368: 4 1 +DO i . 2 +LOOP
5369: @end example
5370: @noindent
5371: prints @code{1 3}
1.1 anton 5372:
1.68 anton 5373: @item
1.1 anton 5374: @cindex negative increment for counted loops
5375: @cindex counted loops with negative increment
1.29 crook 5376: The behaviour of @code{@i{n} +LOOP} is peculiar when @i{n} is negative:
1.1 anton 5377:
1.21 crook 5378: @example
5379: -1 0 ?DO i . -1 +LOOP
5380: @end example
5381: @noindent
5382: prints @code{0 -1}
1.1 anton 5383:
1.21 crook 5384: @example
5385: 0 0 ?DO i . -1 +LOOP
5386: @end example
5387: prints nothing.
1.1 anton 5388:
1.29 crook 5389: Therefore we recommend avoiding @code{@i{n} +LOOP} with negative
5390: @i{n}. One alternative is @code{@i{u} -LOOP}, which reduces the
5391: index by @i{u} each iteration. The loop is terminated when the border
5392: between @i{limit+1} and @i{limit} is crossed. Gforth also provides
1.1 anton 5393: @code{-DO} and @code{U-DO} for down-counting loops. E.g.:
5394:
1.21 crook 5395: @example
5396: -2 0 -DO i . 1 -LOOP
5397: @end example
5398: @noindent
5399: prints @code{0 -1}
1.1 anton 5400:
1.21 crook 5401: @example
5402: -1 0 -DO i . 1 -LOOP
5403: @end example
5404: @noindent
5405: prints @code{0}
5406:
5407: @example
5408: 0 0 -DO i . 1 -LOOP
5409: @end example
5410: @noindent
5411: prints nothing.
1.1 anton 5412:
1.21 crook 5413: @end itemize
1.1 anton 5414:
5415: Unfortunately, @code{+DO}, @code{U+DO}, @code{-DO}, @code{U-DO} and
1.26 crook 5416: @code{-LOOP} are not defined in ANS Forth. However, an implementation
5417: for these words that uses only standard words is provided in
5418: @file{compat/loops.fs}.
1.1 anton 5419:
5420:
5421: @cindex @code{FOR} loops
1.26 crook 5422: Another counted loop is:
1.1 anton 5423: @example
1.29 crook 5424: @i{n}
1.1 anton 5425: FOR
1.29 crook 5426: @i{body}
1.1 anton 5427: NEXT
5428: @end example
5429: This is the preferred loop of native code compiler writers who are too
1.26 crook 5430: lazy to optimize @code{?DO} loops properly. This loop structure is not
1.29 crook 5431: defined in ANS Forth. In Gforth, this loop iterates @i{n+1} times;
5432: @code{i} produces values starting with @i{n} and ending with 0. Other
1.26 crook 5433: Forth systems may behave differently, even if they support @code{FOR}
5434: loops. To avoid problems, don't use @code{FOR} loops.
1.1 anton 5435:
5436: @node Arbitrary control structures, Calls and returns, Counted Loops, Control Structures
5437: @subsection Arbitrary control structures
5438: @cindex control structures, user-defined
5439:
5440: @cindex control-flow stack
5441: ANS Forth permits and supports using control structures in a non-nested
5442: way. Information about incomplete control structures is stored on the
5443: control-flow stack. This stack may be implemented on the Forth data
5444: stack, and this is what we have done in Gforth.
5445:
5446: @cindex @code{orig}, control-flow stack item
5447: @cindex @code{dest}, control-flow stack item
5448: An @i{orig} entry represents an unresolved forward branch, a @i{dest}
5449: entry represents a backward branch target. A few words are the basis for
5450: building any control structure possible (except control structures that
5451: need storage, like calls, coroutines, and backtracking).
5452:
1.44 crook 5453:
1.1 anton 5454: doc-if
5455: doc-ahead
5456: doc-then
5457: doc-begin
5458: doc-until
5459: doc-again
5460: doc-cs-pick
5461: doc-cs-roll
5462:
1.44 crook 5463:
1.21 crook 5464: The Standard words @code{CS-PICK} and @code{CS-ROLL} allow you to
5465: manipulate the control-flow stack in a portable way. Without them, you
5466: would need to know how many stack items are occupied by a control-flow
5467: entry (many systems use one cell. In Gforth they currently take three,
5468: but this may change in the future).
5469:
1.1 anton 5470: Some standard control structure words are built from these words:
5471:
1.44 crook 5472:
1.1 anton 5473: doc-else
5474: doc-while
5475: doc-repeat
5476:
1.44 crook 5477:
5478: @noindent
1.1 anton 5479: Gforth adds some more control-structure words:
5480:
1.44 crook 5481:
1.1 anton 5482: doc-endif
5483: doc-?dup-if
5484: doc-?dup-0=-if
5485:
1.44 crook 5486:
5487: @noindent
1.1 anton 5488: Counted loop words constitute a separate group of words:
5489:
1.44 crook 5490:
1.1 anton 5491: doc-?do
5492: doc-+do
5493: doc-u+do
5494: doc--do
5495: doc-u-do
5496: doc-do
5497: doc-for
5498: doc-loop
5499: doc-+loop
5500: doc--loop
5501: doc-next
5502: doc-leave
5503: doc-?leave
5504: doc-unloop
5505: doc-done
5506:
1.44 crook 5507:
1.21 crook 5508: The standard does not allow using @code{CS-PICK} and @code{CS-ROLL} on
5509: @i{do-sys}. Gforth allows it, but it's your job to ensure that for
1.1 anton 5510: every @code{?DO} etc. there is exactly one @code{UNLOOP} on any path
5511: through the definition (@code{LOOP} etc. compile an @code{UNLOOP} on the
5512: fall-through path). Also, you have to ensure that all @code{LEAVE}s are
5513: resolved (by using one of the loop-ending words or @code{DONE}).
5514:
1.44 crook 5515: @noindent
1.26 crook 5516: Another group of control structure words are:
1.1 anton 5517:
1.44 crook 5518:
1.1 anton 5519: doc-case
5520: doc-endcase
5521: doc-of
5522: doc-endof
5523:
1.44 crook 5524:
1.21 crook 5525: @i{case-sys} and @i{of-sys} cannot be processed using @code{CS-PICK} and
5526: @code{CS-ROLL}.
1.1 anton 5527:
5528: @subsubsection Programming Style
1.47 crook 5529: @cindex control structures programming style
5530: @cindex programming style, arbitrary control structures
1.1 anton 5531:
5532: In order to ensure readability we recommend that you do not create
5533: arbitrary control structures directly, but define new control structure
5534: words for the control structure you want and use these words in your
1.26 crook 5535: program. For example, instead of writing:
1.1 anton 5536:
5537: @example
1.26 crook 5538: BEGIN
1.1 anton 5539: ...
1.26 crook 5540: IF [ 1 CS-ROLL ]
1.1 anton 5541: ...
1.26 crook 5542: AGAIN THEN
1.1 anton 5543: @end example
5544:
1.21 crook 5545: @noindent
1.1 anton 5546: we recommend defining control structure words, e.g.,
5547:
5548: @example
1.26 crook 5549: : WHILE ( DEST -- ORIG DEST )
5550: POSTPONE IF
5551: 1 CS-ROLL ; immediate
5552:
5553: : REPEAT ( orig dest -- )
5554: POSTPONE AGAIN
5555: POSTPONE THEN ; immediate
1.1 anton 5556: @end example
5557:
1.21 crook 5558: @noindent
1.1 anton 5559: and then using these to create the control structure:
5560:
5561: @example
1.26 crook 5562: BEGIN
1.1 anton 5563: ...
1.26 crook 5564: WHILE
1.1 anton 5565: ...
1.26 crook 5566: REPEAT
1.1 anton 5567: @end example
5568:
5569: That's much easier to read, isn't it? Of course, @code{REPEAT} and
5570: @code{WHILE} are predefined, so in this example it would not be
5571: necessary to define them.
5572:
5573: @node Calls and returns, Exception Handling, Arbitrary control structures, Control Structures
5574: @subsection Calls and returns
5575: @cindex calling a definition
5576: @cindex returning from a definition
5577:
1.3 anton 5578: @cindex recursive definitions
5579: A definition can be called simply be writing the name of the definition
1.26 crook 5580: to be called. Normally a definition is invisible during its own
1.3 anton 5581: definition. If you want to write a directly recursive definition, you
1.26 crook 5582: can use @code{recursive} to make the current definition visible, or
5583: @code{recurse} to call the current definition directly.
1.3 anton 5584:
1.44 crook 5585:
1.3 anton 5586: doc-recursive
5587: doc-recurse
5588:
1.44 crook 5589:
1.21 crook 5590: @comment TODO add example of the two recursion methods
1.12 anton 5591: @quotation
5592: @progstyle
5593: I prefer using @code{recursive} to @code{recurse}, because calling the
5594: definition by name is more descriptive (if the name is well-chosen) than
5595: the somewhat cryptic @code{recurse}. E.g., in a quicksort
5596: implementation, it is much better to read (and think) ``now sort the
5597: partitions'' than to read ``now do a recursive call''.
5598: @end quotation
1.3 anton 5599:
1.29 crook 5600: For mutual recursion, use @code{Defer}red words, like this:
1.3 anton 5601:
5602: @example
1.28 crook 5603: Defer foo
1.3 anton 5604:
5605: : bar ( ... -- ... )
5606: ... foo ... ;
5607:
5608: :noname ( ... -- ... )
5609: ... bar ... ;
5610: IS foo
5611: @end example
5612:
1.170 pazsan 5613: Deferred words are discussed in more detail in @ref{Deferred Words}.
1.33 anton 5614:
1.26 crook 5615: The current definition returns control to the calling definition when
1.33 anton 5616: the end of the definition is reached or @code{EXIT} is encountered.
1.1 anton 5617:
5618: doc-exit
5619: doc-;s
5620:
1.44 crook 5621:
1.1 anton 5622: @node Exception Handling, , Calls and returns, Control Structures
5623: @subsection Exception Handling
1.26 crook 5624: @cindex exceptions
1.1 anton 5625:
1.68 anton 5626: @c quit is a very bad idea for error handling,
5627: @c because it does not translate into a THROW
5628: @c it also does not belong into this chapter
5629:
5630: If a word detects an error condition that it cannot handle, it can
5631: @code{throw} an exception. In the simplest case, this will terminate
5632: your program, and report an appropriate error.
1.21 crook 5633:
1.68 anton 5634: doc-throw
1.1 anton 5635:
1.69 anton 5636: @code{Throw} consumes a cell-sized error number on the stack. There are
5637: some predefined error numbers in ANS Forth (see @file{errors.fs}). In
5638: Gforth (and most other systems) you can use the iors produced by various
5639: words as error numbers (e.g., a typical use of @code{allocate} is
5640: @code{allocate throw}). Gforth also provides the word @code{exception}
5641: to define your own error numbers (with decent error reporting); an ANS
5642: Forth version of this word (but without the error messages) is available
5643: in @code{compat/except.fs}. And finally, you can use your own error
1.68 anton 5644: numbers (anything outside the range -4095..0), but won't get nice error
5645: messages, only numbers. For example, try:
5646:
5647: @example
1.69 anton 5648: -10 throw \ ANS defined
5649: -267 throw \ system defined
5650: s" my error" exception throw \ user defined
5651: 7 throw \ arbitrary number
1.68 anton 5652: @end example
5653:
5654: doc---exception-exception
1.1 anton 5655:
1.69 anton 5656: A common idiom to @code{THROW} a specific error if a flag is true is
5657: this:
5658:
5659: @example
5660: @code{( flag ) 0<> @i{errno} and throw}
5661: @end example
5662:
5663: Your program can provide exception handlers to catch exceptions. An
5664: exception handler can be used to correct the problem, or to clean up
5665: some data structures and just throw the exception to the next exception
5666: handler. Note that @code{throw} jumps to the dynamically innermost
5667: exception handler. The system's exception handler is outermost, and just
5668: prints an error and restarts command-line interpretation (or, in batch
5669: mode (i.e., while processing the shell command line), leaves Gforth).
1.1 anton 5670:
1.68 anton 5671: The ANS Forth way to catch exceptions is @code{catch}:
1.1 anton 5672:
1.68 anton 5673: doc-catch
1.160 anton 5674: doc-nothrow
1.68 anton 5675:
5676: The most common use of exception handlers is to clean up the state when
5677: an error happens. E.g.,
1.1 anton 5678:
1.26 crook 5679: @example
1.68 anton 5680: base @ >r hex \ actually the hex should be inside foo, or we h
5681: ['] foo catch ( nerror|0 )
5682: r> base !
1.69 anton 5683: ( nerror|0 ) throw \ pass it on
1.26 crook 5684: @end example
1.1 anton 5685:
1.69 anton 5686: A use of @code{catch} for handling the error @code{myerror} might look
5687: like this:
1.44 crook 5688:
1.68 anton 5689: @example
1.69 anton 5690: ['] foo catch
5691: CASE
1.160 anton 5692: myerror OF ... ( do something about it ) nothrow ENDOF
1.69 anton 5693: dup throw \ default: pass other errors on, do nothing on non-errors
5694: ENDCASE
1.68 anton 5695: @end example
1.44 crook 5696:
1.68 anton 5697: Having to wrap the code into a separate word is often cumbersome,
5698: therefore Gforth provides an alternative syntax:
1.1 anton 5699:
5700: @example
1.69 anton 5701: TRY
1.68 anton 5702: @i{code1}
1.172 anton 5703: IFERROR
5704: @i{code2}
5705: THEN
5706: @i{code3}
1.69 anton 5707: ENDTRY
1.1 anton 5708: @end example
5709:
1.172 anton 5710: This performs @i{code1}. If @i{code1} completes normally, execution
5711: continues with @i{code3}. If @i{code1} or there is an exception
5712: before @code{endtry}, the stacks are reset to the state during
5713: @code{try}, the throw value is pushed on the data stack, and execution
5714: constinues at @i{code2}, and finally falls through the @i{code3}.
1.26 crook 5715:
1.68 anton 5716: doc-try
5717: doc-endtry
1.172 anton 5718: doc-iferror
5719:
5720: If you don't need @i{code2}, you can write @code{restore} instead of
5721: @code{iferror then}:
5722:
5723: @example
5724: TRY
5725: @i{code1}
5726: RESTORE
5727: @i{code3}
5728: ENDTRY
5729: @end example
1.26 crook 5730:
1.172 anton 5731: @cindex unwind-protect
1.69 anton 5732: The cleanup example from above in this syntax:
1.26 crook 5733:
1.68 anton 5734: @example
1.172 anton 5735: base @ @{ oldbase @}
5736: TRY
1.68 anton 5737: hex foo \ now the hex is placed correctly
1.69 anton 5738: 0 \ value for throw
1.172 anton 5739: RESTORE
5740: oldbase base !
5741: ENDTRY
5742: throw
1.1 anton 5743: @end example
5744:
1.172 anton 5745: An additional advantage of this variant is that an exception between
5746: @code{restore} and @code{endtry} (e.g., from the user pressing
5747: @kbd{Ctrl-C}) restarts the execution of the code after @code{restore},
5748: so the base will be restored under all circumstances.
5749:
5750: However, you have to ensure that this code does not cause an exception
5751: itself, otherwise the @code{iferror}/@code{restore} code will loop.
5752: Moreover, you should also make sure that the stack contents needed by
5753: the @code{iferror}/@code{restore} code exist everywhere between
5754: @code{try} and @code{endtry}; in our example this is achived by
5755: putting the data in a local before the @code{try} (you cannot use the
5756: return stack because the exception frame (@i{sys1}) is in the way
5757: there).
5758:
5759: This kind of usage corresponds to Lisp's @code{unwind-protect}.
5760:
5761: @cindex @code{recover} (old Gforth versions)
5762: If you do not want this exception-restarting behaviour, you achieve
5763: this as follows:
5764:
5765: @example
5766: TRY
5767: @i{code1}
5768: ENDTRY-IFERROR
5769: @i{code2}
5770: THEN
5771: @end example
5772:
5773: If there is an exception in @i{code1}, then @i{code2} is executed,
5774: otherwise execution continues behind the @code{then} (or in a possible
5775: @code{else} branch). This corresponds to the construct
5776:
5777: @example
5778: TRY
5779: @i{code1}
5780: RECOVER
5781: @i{code2}
5782: ENDTRY
5783: @end example
5784:
5785: in Gforth before version 0.7. So you can directly replace
5786: @code{recover}-using code; however, we recommend that you check if it
5787: would not be better to use one of the other @code{try} variants while
5788: you are at it.
5789:
1.173 ! anton 5790: To ease the transition, Gforth provides two compatibility files:
! 5791: @file{endtry-iferror.fs} provides the @code{try ... endtry-iferror
! 5792: ... then} syntax (but not @code{iferror} or @code{restore}) for old
! 5793: systems; @file{recover-endtry.fs} provides the @code{try ... recover
! 5794: ... endtry} syntax on new systems, so you can use that file as a
! 5795: stopgap to run old programs. Both files work on any system (they just
! 5796: do nothing if the system already has the syntax it implements), so you
! 5797: can unconditionally @code{require} one of these files, even if you use
! 5798: a mix old and new systems.
! 5799:
1.172 anton 5800: doc-restore
5801: doc-endtry-iferror
5802:
5803: Here's the error handling example:
1.1 anton 5804:
1.68 anton 5805: @example
1.69 anton 5806: TRY
1.68 anton 5807: foo
1.172 anton 5808: ENDTRY-IFERROR
1.69 anton 5809: CASE
1.160 anton 5810: myerror OF ... ( do something about it ) nothrow ENDOF
1.69 anton 5811: throw \ pass other errors on
5812: ENDCASE
1.172 anton 5813: THEN
1.68 anton 5814: @end example
1.1 anton 5815:
1.69 anton 5816: @progstyle
5817: As usual, you should ensure that the stack depth is statically known at
5818: the end: either after the @code{throw} for passing on errors, or after
5819: the @code{ENDTRY} (or, if you use @code{catch}, after the end of the
5820: selection construct for handling the error).
5821:
1.68 anton 5822: There are two alternatives to @code{throw}: @code{Abort"} is conditional
5823: and you can provide an error message. @code{Abort} just produces an
5824: ``Aborted'' error.
1.1 anton 5825:
1.68 anton 5826: The problem with these words is that exception handlers cannot
5827: differentiate between different @code{abort"}s; they just look like
5828: @code{-2 throw} to them (the error message cannot be accessed by
5829: standard programs). Similar @code{abort} looks like @code{-1 throw} to
5830: exception handlers.
1.44 crook 5831:
1.68 anton 5832: doc-abort"
1.26 crook 5833: doc-abort
1.29 crook 5834:
5835:
1.44 crook 5836:
1.29 crook 5837: @c -------------------------------------------------------------
1.47 crook 5838: @node Defining Words, Interpretation and Compilation Semantics, Control Structures, Words
1.29 crook 5839: @section Defining Words
5840: @cindex defining words
5841:
1.47 crook 5842: Defining words are used to extend Forth by creating new entries in the dictionary.
5843:
1.29 crook 5844: @menu
1.67 anton 5845: * CREATE::
1.44 crook 5846: * Variables:: Variables and user variables
1.67 anton 5847: * Constants::
1.44 crook 5848: * Values:: Initialised variables
1.67 anton 5849: * Colon Definitions::
1.44 crook 5850: * Anonymous Definitions:: Definitions without names
1.69 anton 5851: * Supplying names:: Passing definition names as strings
1.67 anton 5852: * User-defined Defining Words::
1.170 pazsan 5853: * Deferred Words:: Allow forward references
1.67 anton 5854: * Aliases::
1.29 crook 5855: @end menu
5856:
1.44 crook 5857: @node CREATE, Variables, Defining Words, Defining Words
5858: @subsection @code{CREATE}
1.29 crook 5859: @cindex simple defining words
5860: @cindex defining words, simple
5861:
5862: Defining words are used to create new entries in the dictionary. The
5863: simplest defining word is @code{CREATE}. @code{CREATE} is used like
5864: this:
5865:
5866: @example
5867: CREATE new-word1
5868: @end example
5869:
1.69 anton 5870: @code{CREATE} is a parsing word, i.e., it takes an argument from the
5871: input stream (@code{new-word1} in our example). It generates a
5872: dictionary entry for @code{new-word1}. When @code{new-word1} is
5873: executed, all that it does is leave an address on the stack. The address
5874: represents the value of the data space pointer (@code{HERE}) at the time
5875: that @code{new-word1} was defined. Therefore, @code{CREATE} is a way of
5876: associating a name with the address of a region of memory.
1.29 crook 5877:
1.34 anton 5878: doc-create
5879:
1.69 anton 5880: Note that in ANS Forth guarantees only for @code{create} that its body
5881: is in dictionary data space (i.e., where @code{here}, @code{allot}
5882: etc. work, @pxref{Dictionary allocation}). Also, in ANS Forth only
5883: @code{create}d words can be modified with @code{does>}
5884: (@pxref{User-defined Defining Words}). And in ANS Forth @code{>body}
5885: can only be applied to @code{create}d words.
5886:
1.29 crook 5887: By extending this example to reserve some memory in data space, we end
1.69 anton 5888: up with something like a @i{variable}. Here are two different ways to do
5889: it:
1.29 crook 5890:
5891: @example
5892: CREATE new-word2 1 cells allot \ reserve 1 cell - initial value undefined
5893: CREATE new-word3 4 , \ reserve 1 cell and initialise it (to 4)
5894: @end example
5895:
5896: The variable can be examined and modified using @code{@@} (``fetch'') and
5897: @code{!} (``store'') like this:
5898:
5899: @example
5900: new-word2 @@ . \ get address, fetch from it and display
5901: 1234 new-word2 ! \ new value, get address, store to it
5902: @end example
5903:
1.44 crook 5904: @cindex arrays
5905: A similar mechanism can be used to create arrays. For example, an
5906: 80-character text input buffer:
1.29 crook 5907:
5908: @example
1.44 crook 5909: CREATE text-buf 80 chars allot
5910:
1.168 anton 5911: text-buf 0 chars + c@@ \ the 1st character (offset 0)
5912: text-buf 3 chars + c@@ \ the 4th character (offset 3)
1.44 crook 5913: @end example
1.29 crook 5914:
1.44 crook 5915: You can build arbitrarily complex data structures by allocating
1.49 anton 5916: appropriate areas of memory. For further discussions of this, and to
1.66 anton 5917: learn about some Gforth tools that make it easier,
1.49 anton 5918: @xref{Structures}.
1.44 crook 5919:
5920:
5921: @node Variables, Constants, CREATE, Defining Words
5922: @subsection Variables
5923: @cindex variables
5924:
5925: The previous section showed how a sequence of commands could be used to
5926: generate a variable. As a final refinement, the whole code sequence can
5927: be wrapped up in a defining word (pre-empting the subject of the next
5928: section), making it easier to create new variables:
5929:
5930: @example
5931: : myvariableX ( "name" -- a-addr ) CREATE 1 cells allot ;
5932: : myvariable0 ( "name" -- a-addr ) CREATE 0 , ;
5933:
5934: myvariableX foo \ variable foo starts off with an unknown value
5935: myvariable0 joe \ whilst joe is initialised to 0
1.29 crook 5936:
5937: 45 3 * foo ! \ set foo to 135
5938: 1234 joe ! \ set joe to 1234
5939: 3 joe +! \ increment joe by 3.. to 1237
5940: @end example
5941:
5942: Not surprisingly, there is no need to define @code{myvariable}, since
1.44 crook 5943: Forth already has a definition @code{Variable}. ANS Forth does not
1.69 anton 5944: guarantee that a @code{Variable} is initialised when it is created
5945: (i.e., it may behave like @code{myvariableX}). In contrast, Gforth's
5946: @code{Variable} initialises the variable to 0 (i.e., it behaves exactly
5947: like @code{myvariable0}). Forth also provides @code{2Variable} and
1.47 crook 5948: @code{fvariable} for double and floating-point variables, respectively
1.69 anton 5949: -- they are initialised to 0. and 0e in Gforth. If you use a @code{Variable} to
1.47 crook 5950: store a boolean, you can use @code{on} and @code{off} to toggle its
5951: state.
1.29 crook 5952:
1.34 anton 5953: doc-variable
5954: doc-2variable
5955: doc-fvariable
5956:
1.29 crook 5957: @cindex user variables
5958: @cindex user space
5959: The defining word @code{User} behaves in the same way as @code{Variable}.
5960: The difference is that it reserves space in @i{user (data) space} rather
5961: than normal data space. In a Forth system that has a multi-tasker, each
5962: task has its own set of user variables.
5963:
1.34 anton 5964: doc-user
1.67 anton 5965: @c doc-udp
5966: @c doc-uallot
1.34 anton 5967:
1.29 crook 5968: @comment TODO is that stuff about user variables strictly correct? Is it
5969: @comment just terminal tasks that have user variables?
5970: @comment should document tasker.fs (with some examples) elsewhere
5971: @comment in this manual, then expand on user space and user variables.
5972:
1.44 crook 5973: @node Constants, Values, Variables, Defining Words
5974: @subsection Constants
5975: @cindex constants
5976:
5977: @code{Constant} allows you to declare a fixed value and refer to it by
5978: name. For example:
1.29 crook 5979:
5980: @example
5981: 12 Constant INCHES-PER-FOOT
5982: 3E+08 fconstant SPEED-O-LIGHT
5983: @end example
5984:
5985: A @code{Variable} can be both read and written, so its run-time
5986: behaviour is to supply an address through which its current value can be
5987: manipulated. In contrast, the value of a @code{Constant} cannot be
5988: changed once it has been declared@footnote{Well, often it can be -- but
5989: not in a Standard, portable way. It's safer to use a @code{Value} (read
5990: on).} so it's not necessary to supply the address -- it is more
5991: efficient to return the value of the constant directly. That's exactly
5992: what happens; the run-time effect of a constant is to put its value on
1.49 anton 5993: the top of the stack (You can find one
5994: way of implementing @code{Constant} in @ref{User-defined Defining Words}).
1.29 crook 5995:
1.69 anton 5996: Forth also provides @code{2Constant} and @code{fconstant} for defining
1.29 crook 5997: double and floating-point constants, respectively.
5998:
1.34 anton 5999: doc-constant
6000: doc-2constant
6001: doc-fconstant
6002:
6003: @c that's too deep, and it's not necessarily true for all ANS Forths. - anton
1.44 crook 6004: @c nac-> How could that not be true in an ANS Forth? You can't define a
6005: @c constant, use it and then delete the definition of the constant..
1.69 anton 6006:
6007: @c anton->An ANS Forth system can compile a constant to a literal; On
6008: @c decompilation you would see only the number, just as if it had been used
6009: @c in the first place. The word will stay, of course, but it will only be
6010: @c used by the text interpreter (no run-time duties, except when it is
6011: @c POSTPONEd or somesuch).
6012:
6013: @c nac:
1.44 crook 6014: @c I agree that it's rather deep, but IMO it is an important difference
6015: @c relative to other programming languages.. often it's annoying: it
6016: @c certainly changes my programming style relative to C.
6017:
1.69 anton 6018: @c anton: In what way?
6019:
1.29 crook 6020: Constants in Forth behave differently from their equivalents in other
6021: programming languages. In other languages, a constant (such as an EQU in
6022: assembler or a #define in C) only exists at compile-time; in the
6023: executable program the constant has been translated into an absolute
6024: number and, unless you are using a symbolic debugger, it's impossible to
6025: know what abstract thing that number represents. In Forth a constant has
1.44 crook 6026: an entry in the header space and remains there after the code that uses
6027: it has been defined. In fact, it must remain in the dictionary since it
6028: has run-time duties to perform. For example:
1.29 crook 6029:
6030: @example
6031: 12 Constant INCHES-PER-FOOT
6032: : FEET-TO-INCHES ( n1 -- n2 ) INCHES-PER-FOOT * ;
6033: @end example
6034:
6035: @cindex in-lining of constants
6036: When @code{FEET-TO-INCHES} is executed, it will in turn execute the xt
6037: associated with the constant @code{INCHES-PER-FOOT}. If you use
6038: @code{see} to decompile the definition of @code{FEET-TO-INCHES}, you can
6039: see that it makes a call to @code{INCHES-PER-FOOT}. Some Forth compilers
6040: attempt to optimise constants by in-lining them where they are used. You
6041: can force Gforth to in-line a constant like this:
6042:
6043: @example
6044: : FEET-TO-INCHES ( n1 -- n2 ) [ INCHES-PER-FOOT ] LITERAL * ;
6045: @end example
6046:
6047: If you use @code{see} to decompile @i{this} version of
6048: @code{FEET-TO-INCHES}, you can see that @code{INCHES-PER-FOOT} is no
1.49 anton 6049: longer present. To understand how this works, read
6050: @ref{Interpret/Compile states}, and @ref{Literals}.
1.29 crook 6051:
6052: In-lining constants in this way might improve execution time
6053: fractionally, and can ensure that a constant is now only referenced at
6054: compile-time. However, the definition of the constant still remains in
6055: the dictionary. Some Forth compilers provide a mechanism for controlling
6056: a second dictionary for holding transient words such that this second
6057: dictionary can be deleted later in order to recover memory
6058: space. However, there is no standard way of doing this.
6059:
6060:
1.44 crook 6061: @node Values, Colon Definitions, Constants, Defining Words
6062: @subsection Values
6063: @cindex values
1.34 anton 6064:
1.69 anton 6065: A @code{Value} behaves like a @code{Constant}, but it can be changed.
6066: @code{TO} is a parsing word that changes a @code{Values}. In Gforth
6067: (not in ANS Forth) you can access (and change) a @code{value} also with
6068: @code{>body}.
6069:
6070: Here are some
6071: examples:
1.29 crook 6072:
6073: @example
1.69 anton 6074: 12 Value APPLES \ Define APPLES with an initial value of 12
6075: 34 TO APPLES \ Change the value of APPLES. TO is a parsing word
6076: 1 ' APPLES >body +! \ Increment APPLES. Non-standard usage.
6077: APPLES \ puts 35 on the top of the stack.
1.29 crook 6078: @end example
6079:
1.44 crook 6080: doc-value
6081: doc-to
1.29 crook 6082:
1.35 anton 6083:
1.69 anton 6084:
1.44 crook 6085: @node Colon Definitions, Anonymous Definitions, Values, Defining Words
6086: @subsection Colon Definitions
6087: @cindex colon definitions
1.35 anton 6088:
6089: @example
1.44 crook 6090: : name ( ... -- ... )
6091: word1 word2 word3 ;
1.29 crook 6092: @end example
6093:
1.44 crook 6094: @noindent
6095: Creates a word called @code{name} that, upon execution, executes
6096: @code{word1 word2 word3}. @code{name} is a @dfn{(colon) definition}.
1.29 crook 6097:
1.49 anton 6098: The explanation above is somewhat superficial. For simple examples of
6099: colon definitions see @ref{Your first definition}. For an in-depth
1.66 anton 6100: discussion of some of the issues involved, @xref{Interpretation and
1.49 anton 6101: Compilation Semantics}.
1.29 crook 6102:
1.44 crook 6103: doc-:
6104: doc-;
1.1 anton 6105:
1.34 anton 6106:
1.69 anton 6107: @node Anonymous Definitions, Supplying names, Colon Definitions, Defining Words
1.44 crook 6108: @subsection Anonymous Definitions
6109: @cindex colon definitions
6110: @cindex defining words without name
1.34 anton 6111:
1.44 crook 6112: Sometimes you want to define an @dfn{anonymous word}; a word without a
6113: name. You can do this with:
1.1 anton 6114:
1.44 crook 6115: doc-:noname
1.1 anton 6116:
1.44 crook 6117: This leaves the execution token for the word on the stack after the
6118: closing @code{;}. Here's an example in which a deferred word is
6119: initialised with an @code{xt} from an anonymous colon definition:
1.1 anton 6120:
1.29 crook 6121: @example
1.44 crook 6122: Defer deferred
6123: :noname ( ... -- ... )
6124: ... ;
6125: IS deferred
1.29 crook 6126: @end example
1.26 crook 6127:
1.44 crook 6128: @noindent
6129: Gforth provides an alternative way of doing this, using two separate
6130: words:
1.27 crook 6131:
1.44 crook 6132: doc-noname
6133: @cindex execution token of last defined word
1.116 anton 6134: doc-latestxt
1.1 anton 6135:
1.44 crook 6136: @noindent
6137: The previous example can be rewritten using @code{noname} and
1.116 anton 6138: @code{latestxt}:
1.1 anton 6139:
1.26 crook 6140: @example
1.44 crook 6141: Defer deferred
6142: noname : ( ... -- ... )
6143: ... ;
1.116 anton 6144: latestxt IS deferred
1.26 crook 6145: @end example
1.1 anton 6146:
1.29 crook 6147: @noindent
1.44 crook 6148: @code{noname} works with any defining word, not just @code{:}.
6149:
1.116 anton 6150: @code{latestxt} also works when the last word was not defined as
1.71 anton 6151: @code{noname}. It does not work for combined words, though. It also has
6152: the useful property that is is valid as soon as the header for a
6153: definition has been built. Thus:
1.44 crook 6154:
6155: @example
1.116 anton 6156: latestxt . : foo [ latestxt . ] ; ' foo .
1.44 crook 6157: @end example
1.1 anton 6158:
1.44 crook 6159: @noindent
6160: prints 3 numbers; the last two are the same.
1.26 crook 6161:
1.69 anton 6162: @node Supplying names, User-defined Defining Words, Anonymous Definitions, Defining Words
6163: @subsection Supplying the name of a defined word
6164: @cindex names for defined words
6165: @cindex defining words, name given in a string
6166:
6167: By default, a defining word takes the name for the defined word from the
6168: input stream. Sometimes you want to supply the name from a string. You
6169: can do this with:
6170:
6171: doc-nextname
6172:
6173: For example:
6174:
6175: @example
6176: s" foo" nextname create
6177: @end example
6178:
6179: @noindent
6180: is equivalent to:
6181:
6182: @example
6183: create foo
6184: @end example
6185:
6186: @noindent
6187: @code{nextname} works with any defining word.
6188:
1.1 anton 6189:
1.170 pazsan 6190: @node User-defined Defining Words, Deferred Words, Supplying names, Defining Words
1.26 crook 6191: @subsection User-defined Defining Words
6192: @cindex user-defined defining words
6193: @cindex defining words, user-defined
1.1 anton 6194:
1.29 crook 6195: You can create a new defining word by wrapping defining-time code around
6196: an existing defining word and putting the sequence in a colon
1.69 anton 6197: definition.
6198:
6199: @c anton: This example is very complex and leads in a quite different
6200: @c direction from the CREATE-DOES> stuff that follows. It should probably
6201: @c be done elsewhere, or as a subsubsection of this subsection (or as a
6202: @c subsection of Defining Words)
6203:
6204: For example, suppose that you have a word @code{stats} that
1.29 crook 6205: gathers statistics about colon definitions given the @i{xt} of the
6206: definition, and you want every colon definition in your application to
6207: make a call to @code{stats}. You can define and use a new version of
6208: @code{:} like this:
6209:
6210: @example
6211: : stats ( xt -- ) DUP ." (Gathering statistics for " . ." )"
6212: ... ; \ other code
6213:
1.116 anton 6214: : my: : latestxt postpone literal ['] stats compile, ;
1.29 crook 6215:
6216: my: foo + - ;
6217: @end example
6218:
6219: When @code{foo} is defined using @code{my:} these steps occur:
6220:
6221: @itemize @bullet
6222: @item
6223: @code{my:} is executed.
6224: @item
6225: The @code{:} within the definition (the one between @code{my:} and
1.116 anton 6226: @code{latestxt}) is executed, and does just what it always does; it parses
1.29 crook 6227: the input stream for a name, builds a dictionary header for the name
6228: @code{foo} and switches @code{state} from interpret to compile.
6229: @item
1.116 anton 6230: The word @code{latestxt} is executed. It puts the @i{xt} for the word that is
1.29 crook 6231: being defined -- @code{foo} -- onto the stack.
6232: @item
6233: The code that was produced by @code{postpone literal} is executed; this
6234: causes the value on the stack to be compiled as a literal in the code
6235: area of @code{foo}.
6236: @item
6237: The code @code{['] stats} compiles a literal into the definition of
6238: @code{my:}. When @code{compile,} is executed, that literal -- the
6239: execution token for @code{stats} -- is layed down in the code area of
6240: @code{foo} , following the literal@footnote{Strictly speaking, the
6241: mechanism that @code{compile,} uses to convert an @i{xt} into something
6242: in the code area is implementation-dependent. A threaded implementation
6243: might spit out the execution token directly whilst another
6244: implementation might spit out a native code sequence.}.
6245: @item
6246: At this point, the execution of @code{my:} is complete, and control
6247: returns to the text interpreter. The text interpreter is in compile
6248: state, so subsequent text @code{+ -} is compiled into the definition of
6249: @code{foo} and the @code{;} terminates the definition as always.
6250: @end itemize
6251:
6252: You can use @code{see} to decompile a word that was defined using
6253: @code{my:} and see how it is different from a normal @code{:}
6254: definition. For example:
6255:
6256: @example
6257: : bar + - ; \ like foo but using : rather than my:
6258: see bar
6259: : bar
6260: + - ;
6261: see foo
6262: : foo
6263: 107645672 stats + - ;
6264:
1.140 anton 6265: \ use ' foo . to show that 107645672 is the xt for foo
1.29 crook 6266: @end example
6267:
6268: You can use techniques like this to make new defining words in terms of
6269: @i{any} existing defining word.
1.1 anton 6270:
6271:
1.29 crook 6272: @cindex defining defining words
1.26 crook 6273: @cindex @code{CREATE} ... @code{DOES>}
6274: If you want the words defined with your defining words to behave
6275: differently from words defined with standard defining words, you can
6276: write your defining word like this:
1.1 anton 6277:
6278: @example
1.26 crook 6279: : def-word ( "name" -- )
1.29 crook 6280: CREATE @i{code1}
1.26 crook 6281: DOES> ( ... -- ... )
1.29 crook 6282: @i{code2} ;
1.26 crook 6283:
6284: def-word name
1.1 anton 6285: @end example
6286:
1.29 crook 6287: @cindex child words
6288: This fragment defines a @dfn{defining word} @code{def-word} and then
6289: executes it. When @code{def-word} executes, it @code{CREATE}s a new
6290: word, @code{name}, and executes the code @i{code1}. The code @i{code2}
6291: is not executed at this time. The word @code{name} is sometimes called a
6292: @dfn{child} of @code{def-word}.
6293:
6294: When you execute @code{name}, the address of the body of @code{name} is
6295: put on the data stack and @i{code2} is executed (the address of the body
6296: of @code{name} is the address @code{HERE} returns immediately after the
1.69 anton 6297: @code{CREATE}, i.e., the address a @code{create}d word returns by
6298: default).
6299:
6300: @c anton:
6301: @c www.dictionary.com says:
6302: @c at·a·vism: 1.The reappearance of a characteristic in an organism after
6303: @c several generations of absence, usually caused by the chance
6304: @c recombination of genes. 2.An individual or a part that exhibits
6305: @c atavism. Also called throwback. 3.The return of a trait or recurrence
6306: @c of previous behavior after a period of absence.
6307: @c
6308: @c Doesn't seem to fit.
1.29 crook 6309:
1.69 anton 6310: @c @cindex atavism in child words
1.33 anton 6311: You can use @code{def-word} to define a set of child words that behave
1.69 anton 6312: similarly; they all have a common run-time behaviour determined by
6313: @i{code2}. Typically, the @i{code1} sequence builds a data area in the
6314: body of the child word. The structure of the data is common to all
6315: children of @code{def-word}, but the data values are specific -- and
6316: private -- to each child word. When a child word is executed, the
6317: address of its private data area is passed as a parameter on TOS to be
6318: used and manipulated@footnote{It is legitimate both to read and write to
6319: this data area.} by @i{code2}.
1.29 crook 6320:
6321: The two fragments of code that make up the defining words act (are
6322: executed) at two completely separate times:
1.1 anton 6323:
1.29 crook 6324: @itemize @bullet
6325: @item
6326: At @i{define time}, the defining word executes @i{code1} to generate a
6327: child word
6328: @item
6329: At @i{child execution time}, when a child word is invoked, @i{code2}
6330: is executed, using parameters (data) that are private and specific to
6331: the child word.
6332: @end itemize
6333:
1.44 crook 6334: Another way of understanding the behaviour of @code{def-word} and
6335: @code{name} is to say that, if you make the following definitions:
1.33 anton 6336: @example
6337: : def-word1 ( "name" -- )
6338: CREATE @i{code1} ;
6339:
6340: : action1 ( ... -- ... )
6341: @i{code2} ;
6342:
6343: def-word1 name1
6344: @end example
6345:
1.44 crook 6346: @noindent
6347: Then using @code{name1 action1} is equivalent to using @code{name}.
1.1 anton 6348:
1.29 crook 6349: The classic example is that you can define @code{CONSTANT} in this way:
1.26 crook 6350:
1.1 anton 6351: @example
1.29 crook 6352: : CONSTANT ( w "name" -- )
6353: CREATE ,
1.26 crook 6354: DOES> ( -- w )
6355: @@ ;
1.1 anton 6356: @end example
6357:
1.29 crook 6358: @comment There is a beautiful description of how this works and what
6359: @comment it does in the Forthwrite 100th edition.. as well as an elegant
6360: @comment commentary on the Counting Fruits problem.
6361:
6362: When you create a constant with @code{5 CONSTANT five}, a set of
6363: define-time actions take place; first a new word @code{five} is created,
6364: then the value 5 is laid down in the body of @code{five} with
1.44 crook 6365: @code{,}. When @code{five} is executed, the address of the body is put on
1.29 crook 6366: the stack, and @code{@@} retrieves the value 5. The word @code{five} has
6367: no code of its own; it simply contains a data field and a pointer to the
6368: code that follows @code{DOES>} in its defining word. That makes words
6369: created in this way very compact.
6370:
6371: The final example in this section is intended to remind you that space
6372: reserved in @code{CREATE}d words is @i{data} space and therefore can be
6373: both read and written by a Standard program@footnote{Exercise: use this
6374: example as a starting point for your own implementation of @code{Value}
6375: and @code{TO} -- if you get stuck, investigate the behaviour of @code{'} and
6376: @code{[']}.}:
6377:
6378: @example
6379: : foo ( "name" -- )
6380: CREATE -1 ,
6381: DOES> ( -- )
1.33 anton 6382: @@ . ;
1.29 crook 6383:
6384: foo first-word
6385: foo second-word
6386:
6387: 123 ' first-word >BODY !
6388: @end example
6389:
6390: If @code{first-word} had been a @code{CREATE}d word, we could simply
6391: have executed it to get the address of its data field. However, since it
6392: was defined to have @code{DOES>} actions, its execution semantics are to
6393: perform those @code{DOES>} actions. To get the address of its data field
6394: it's necessary to use @code{'} to get its xt, then @code{>BODY} to
6395: translate the xt into the address of the data field. When you execute
6396: @code{first-word}, it will display @code{123}. When you execute
6397: @code{second-word} it will display @code{-1}.
1.26 crook 6398:
6399: @cindex stack effect of @code{DOES>}-parts
6400: @cindex @code{DOES>}-parts, stack effect
1.29 crook 6401: In the examples above the stack comment after the @code{DOES>} specifies
1.26 crook 6402: the stack effect of the defined words, not the stack effect of the
6403: following code (the following code expects the address of the body on
6404: the top of stack, which is not reflected in the stack comment). This is
6405: the convention that I use and recommend (it clashes a bit with using
6406: locals declarations for stack effect specification, though).
1.1 anton 6407:
1.53 anton 6408: @menu
6409: * CREATE..DOES> applications::
6410: * CREATE..DOES> details::
1.63 anton 6411: * Advanced does> usage example::
1.155 anton 6412: * Const-does>::
1.53 anton 6413: @end menu
6414:
6415: @node CREATE..DOES> applications, CREATE..DOES> details, User-defined Defining Words, User-defined Defining Words
1.26 crook 6416: @subsubsection Applications of @code{CREATE..DOES>}
6417: @cindex @code{CREATE} ... @code{DOES>}, applications
1.1 anton 6418:
1.26 crook 6419: You may wonder how to use this feature. Here are some usage patterns:
1.1 anton 6420:
1.26 crook 6421: @cindex factoring similar colon definitions
6422: When you see a sequence of code occurring several times, and you can
6423: identify a meaning, you will factor it out as a colon definition. When
6424: you see similar colon definitions, you can factor them using
6425: @code{CREATE..DOES>}. E.g., an assembler usually defines several words
6426: that look very similar:
1.1 anton 6427: @example
1.26 crook 6428: : ori, ( reg-target reg-source n -- )
6429: 0 asm-reg-reg-imm ;
6430: : andi, ( reg-target reg-source n -- )
6431: 1 asm-reg-reg-imm ;
1.1 anton 6432: @end example
6433:
1.26 crook 6434: @noindent
6435: This could be factored with:
6436: @example
6437: : reg-reg-imm ( op-code -- )
6438: CREATE ,
6439: DOES> ( reg-target reg-source n -- )
6440: @@ asm-reg-reg-imm ;
6441:
6442: 0 reg-reg-imm ori,
6443: 1 reg-reg-imm andi,
6444: @end example
1.1 anton 6445:
1.26 crook 6446: @cindex currying
6447: Another view of @code{CREATE..DOES>} is to consider it as a crude way to
6448: supply a part of the parameters for a word (known as @dfn{currying} in
6449: the functional language community). E.g., @code{+} needs two
6450: parameters. Creating versions of @code{+} with one parameter fixed can
6451: be done like this:
1.82 anton 6452:
1.1 anton 6453: @example
1.82 anton 6454: : curry+ ( n1 "name" -- )
1.26 crook 6455: CREATE ,
6456: DOES> ( n2 -- n1+n2 )
6457: @@ + ;
6458:
6459: 3 curry+ 3+
6460: -2 curry+ 2-
1.1 anton 6461: @end example
6462:
1.91 anton 6463:
1.63 anton 6464: @node CREATE..DOES> details, Advanced does> usage example, CREATE..DOES> applications, User-defined Defining Words
1.26 crook 6465: @subsubsection The gory details of @code{CREATE..DOES>}
6466: @cindex @code{CREATE} ... @code{DOES>}, details
1.1 anton 6467:
1.26 crook 6468: doc-does>
1.1 anton 6469:
1.26 crook 6470: @cindex @code{DOES>} in a separate definition
6471: This means that you need not use @code{CREATE} and @code{DOES>} in the
6472: same definition; you can put the @code{DOES>}-part in a separate
1.29 crook 6473: definition. This allows us to, e.g., select among different @code{DOES>}-parts:
1.26 crook 6474: @example
6475: : does1
6476: DOES> ( ... -- ... )
1.44 crook 6477: ... ;
6478:
6479: : does2
6480: DOES> ( ... -- ... )
6481: ... ;
6482:
6483: : def-word ( ... -- ... )
6484: create ...
6485: IF
6486: does1
6487: ELSE
6488: does2
6489: ENDIF ;
6490: @end example
6491:
6492: In this example, the selection of whether to use @code{does1} or
1.69 anton 6493: @code{does2} is made at definition-time; at the time that the child word is
1.44 crook 6494: @code{CREATE}d.
6495:
6496: @cindex @code{DOES>} in interpretation state
6497: In a standard program you can apply a @code{DOES>}-part only if the last
6498: word was defined with @code{CREATE}. In Gforth, the @code{DOES>}-part
6499: will override the behaviour of the last word defined in any case. In a
6500: standard program, you can use @code{DOES>} only in a colon
6501: definition. In Gforth, you can also use it in interpretation state, in a
6502: kind of one-shot mode; for example:
6503: @example
6504: CREATE name ( ... -- ... )
6505: @i{initialization}
6506: DOES>
6507: @i{code} ;
6508: @end example
6509:
6510: @noindent
6511: is equivalent to the standard:
6512: @example
6513: :noname
6514: DOES>
6515: @i{code} ;
6516: CREATE name EXECUTE ( ... -- ... )
6517: @i{initialization}
6518: @end example
6519:
1.53 anton 6520: doc->body
6521:
1.152 pazsan 6522: @node Advanced does> usage example, Const-does>, CREATE..DOES> details, User-defined Defining Words
1.63 anton 6523: @subsubsection Advanced does> usage example
6524:
6525: The MIPS disassembler (@file{arch/mips/disasm.fs}) contains many words
6526: for disassembling instructions, that follow a very repetetive scheme:
6527:
6528: @example
6529: :noname @var{disasm-operands} s" @var{inst-name}" type ;
6530: @var{entry-num} cells @var{table} + !
6531: @end example
6532:
6533: Of course, this inspires the idea to factor out the commonalities to
6534: allow a definition like
6535:
6536: @example
6537: @var{disasm-operands} @var{entry-num} @var{table} define-inst @var{inst-name}
6538: @end example
6539:
6540: The parameters @var{disasm-operands} and @var{table} are usually
1.69 anton 6541: correlated. Moreover, before I wrote the disassembler, there already
6542: existed code that defines instructions like this:
1.63 anton 6543:
6544: @example
6545: @var{entry-num} @var{inst-format} @var{inst-name}
6546: @end example
6547:
6548: This code comes from the assembler and resides in
6549: @file{arch/mips/insts.fs}.
6550:
6551: So I had to define the @var{inst-format} words that performed the scheme
6552: above when executed. At first I chose to use run-time code-generation:
6553:
6554: @example
6555: : @var{inst-format} ( entry-num "name" -- ; compiled code: addr w -- )
6556: :noname Postpone @var{disasm-operands}
6557: name Postpone sliteral Postpone type Postpone ;
6558: swap cells @var{table} + ! ;
6559: @end example
6560:
6561: Note that this supplies the other two parameters of the scheme above.
1.44 crook 6562:
1.63 anton 6563: An alternative would have been to write this using
6564: @code{create}/@code{does>}:
6565:
6566: @example
6567: : @var{inst-format} ( entry-num "name" -- )
6568: here name string, ( entry-num c-addr ) \ parse and save "name"
6569: noname create , ( entry-num )
1.116 anton 6570: latestxt swap cells @var{table} + !
1.63 anton 6571: does> ( addr w -- )
6572: \ disassemble instruction w at addr
6573: @@ >r
6574: @var{disasm-operands}
6575: r> count type ;
6576: @end example
6577:
6578: Somehow the first solution is simpler, mainly because it's simpler to
6579: shift a string from definition-time to use-time with @code{sliteral}
6580: than with @code{string,} and friends.
6581:
6582: I wrote a lot of words following this scheme and soon thought about
6583: factoring out the commonalities among them. Note that this uses a
6584: two-level defining word, i.e., a word that defines ordinary defining
6585: words.
6586:
6587: This time a solution involving @code{postpone} and friends seemed more
6588: difficult (try it as an exercise), so I decided to use a
6589: @code{create}/@code{does>} word; since I was already at it, I also used
6590: @code{create}/@code{does>} for the lower level (try using
6591: @code{postpone} etc. as an exercise), resulting in the following
6592: definition:
6593:
6594: @example
6595: : define-format ( disasm-xt table-xt -- )
6596: \ define an instruction format that uses disasm-xt for
6597: \ disassembling and enters the defined instructions into table
6598: \ table-xt
6599: create 2,
6600: does> ( u "inst" -- )
6601: \ defines an anonymous word for disassembling instruction inst,
6602: \ and enters it as u-th entry into table-xt
6603: 2@@ swap here name string, ( u table-xt disasm-xt c-addr ) \ remember string
6604: noname create 2, \ define anonymous word
1.116 anton 6605: execute latestxt swap ! \ enter xt of defined word into table-xt
1.63 anton 6606: does> ( addr w -- )
6607: \ disassemble instruction w at addr
6608: 2@@ >r ( addr w disasm-xt R: c-addr )
6609: execute ( R: c-addr ) \ disassemble operands
6610: r> count type ; \ print name
6611: @end example
6612:
6613: Note that the tables here (in contrast to above) do the @code{cells +}
6614: by themselves (that's why you have to pass an xt). This word is used in
6615: the following way:
6616:
6617: @example
6618: ' @var{disasm-operands} ' @var{table} define-format @var{inst-format}
6619: @end example
6620:
1.71 anton 6621: As shown above, the defined instruction format is then used like this:
6622:
6623: @example
6624: @var{entry-num} @var{inst-format} @var{inst-name}
6625: @end example
6626:
1.63 anton 6627: In terms of currying, this kind of two-level defining word provides the
6628: parameters in three stages: first @var{disasm-operands} and @var{table},
6629: then @var{entry-num} and @var{inst-name}, finally @code{addr w}, i.e.,
6630: the instruction to be disassembled.
6631:
6632: Of course this did not quite fit all the instruction format names used
6633: in @file{insts.fs}, so I had to define a few wrappers that conditioned
6634: the parameters into the right form.
6635:
6636: If you have trouble following this section, don't worry. First, this is
6637: involved and takes time (and probably some playing around) to
6638: understand; second, this is the first two-level
6639: @code{create}/@code{does>} word I have written in seventeen years of
6640: Forth; and if I did not have @file{insts.fs} to start with, I may well
6641: have elected to use just a one-level defining word (with some repeating
6642: of parameters when using the defining word). So it is not necessary to
6643: understand this, but it may improve your understanding of Forth.
1.44 crook 6644:
6645:
1.152 pazsan 6646: @node Const-does>, , Advanced does> usage example, User-defined Defining Words
1.91 anton 6647: @subsubsection @code{Const-does>}
6648:
6649: A frequent use of @code{create}...@code{does>} is for transferring some
6650: values from definition-time to run-time. Gforth supports this use with
6651:
6652: doc-const-does>
6653:
6654: A typical use of this word is:
6655:
6656: @example
6657: : curry+ ( n1 "name" -- )
6658: 1 0 CONST-DOES> ( n2 -- n1+n2 )
6659: + ;
6660:
6661: 3 curry+ 3+
6662: @end example
6663:
6664: Here the @code{1 0} means that 1 cell and 0 floats are transferred from
6665: definition to run-time.
6666:
6667: The advantages of using @code{const-does>} are:
6668:
6669: @itemize
6670:
6671: @item
6672: You don't have to deal with storing and retrieving the values, i.e.,
6673: your program becomes more writable and readable.
6674:
6675: @item
6676: When using @code{does>}, you have to introduce a @code{@@} that cannot
6677: be optimized away (because you could change the data using
6678: @code{>body}...@code{!}); @code{const-does>} avoids this problem.
6679:
6680: @end itemize
6681:
6682: An ANS Forth implementation of @code{const-does>} is available in
6683: @file{compat/const-does.fs}.
6684:
6685:
1.170 pazsan 6686: @node Deferred Words, Aliases, User-defined Defining Words, Defining Words
6687: @subsection Deferred Words
1.44 crook 6688: @cindex deferred words
6689:
6690: The defining word @code{Defer} allows you to define a word by name
6691: without defining its behaviour; the definition of its behaviour is
6692: deferred. Here are two situation where this can be useful:
6693:
6694: @itemize @bullet
6695: @item
6696: Where you want to allow the behaviour of a word to be altered later, and
6697: for all precompiled references to the word to change when its behaviour
6698: is changed.
6699: @item
6700: For mutual recursion; @xref{Calls and returns}.
6701: @end itemize
6702:
6703: In the following example, @code{foo} always invokes the version of
6704: @code{greet} that prints ``@code{Good morning}'' whilst @code{bar}
6705: always invokes the version that prints ``@code{Hello}''. There is no way
6706: of getting @code{foo} to use the later version without re-ordering the
6707: source code and recompiling it.
6708:
6709: @example
6710: : greet ." Good morning" ;
6711: : foo ... greet ... ;
6712: : greet ." Hello" ;
6713: : bar ... greet ... ;
6714: @end example
6715:
6716: This problem can be solved by defining @code{greet} as a @code{Defer}red
6717: word. The behaviour of a @code{Defer}red word can be defined and
6718: redefined at any time by using @code{IS} to associate the xt of a
6719: previously-defined word with it. The previous example becomes:
6720:
6721: @example
1.69 anton 6722: Defer greet ( -- )
1.44 crook 6723: : foo ... greet ... ;
6724: : bar ... greet ... ;
1.69 anton 6725: : greet1 ( -- ) ." Good morning" ;
6726: : greet2 ( -- ) ." Hello" ;
1.132 anton 6727: ' greet2 IS greet \ make greet behave like greet2
1.44 crook 6728: @end example
6729:
1.69 anton 6730: @progstyle
6731: You should write a stack comment for every deferred word, and put only
6732: XTs into deferred words that conform to this stack effect. Otherwise
6733: it's too difficult to use the deferred word.
6734:
1.44 crook 6735: A deferred word can be used to improve the statistics-gathering example
6736: from @ref{User-defined Defining Words}; rather than edit the
6737: application's source code to change every @code{:} to a @code{my:}, do
6738: this:
6739:
6740: @example
6741: : real: : ; \ retain access to the original
6742: defer : \ redefine as a deferred word
1.132 anton 6743: ' my: IS : \ use special version of :
1.44 crook 6744: \
6745: \ load application here
6746: \
1.132 anton 6747: ' real: IS : \ go back to the original
1.44 crook 6748: @end example
6749:
6750:
1.132 anton 6751: One thing to note is that @code{IS} has special compilation semantics,
6752: such that it parses the name at compile time (like @code{TO}):
1.44 crook 6753:
6754: @example
6755: : set-greet ( xt -- )
1.132 anton 6756: IS greet ;
1.44 crook 6757:
6758: ' greet1 set-greet
6759: @end example
6760:
1.132 anton 6761: In situations where @code{IS} does not fit, use @code{defer!} instead.
6762:
1.69 anton 6763: A deferred word can only inherit execution semantics from the xt
6764: (because that is all that an xt can represent -- for more discussion of
6765: this @pxref{Tokens for Words}); by default it will have default
6766: interpretation and compilation semantics deriving from this execution
6767: semantics. However, you can change the interpretation and compilation
6768: semantics of the deferred word in the usual ways:
1.44 crook 6769:
6770: @example
1.132 anton 6771: : bar .... ; immediate
1.44 crook 6772: Defer fred immediate
6773: Defer jim
6774:
1.132 anton 6775: ' bar IS jim \ jim has default semantics
6776: ' bar IS fred \ fred is immediate
1.44 crook 6777: @end example
6778:
6779: doc-defer
1.132 anton 6780: doc-defer!
1.44 crook 6781: doc-is
1.132 anton 6782: doc-defer@
6783: doc-action-of
1.44 crook 6784: @comment TODO document these: what's defers [is]
6785: doc-defers
6786:
6787: @c Use @code{words-deferred} to see a list of deferred words.
6788:
1.132 anton 6789: Definitions of these words (except @code{defers}) in ANS Forth are
6790: provided in @file{compat/defer.fs}.
1.44 crook 6791:
6792:
1.170 pazsan 6793: @node Aliases, , Deferred Words, Defining Words
1.44 crook 6794: @subsection Aliases
6795: @cindex aliases
1.1 anton 6796:
1.44 crook 6797: The defining word @code{Alias} allows you to define a word by name that
6798: has the same behaviour as some other word. Here are two situation where
6799: this can be useful:
1.1 anton 6800:
1.44 crook 6801: @itemize @bullet
6802: @item
6803: When you want access to a word's definition from a different word list
6804: (for an example of this, see the definition of the @code{Root} word list
6805: in the Gforth source).
6806: @item
6807: When you want to create a synonym; a definition that can be known by
6808: either of two names (for example, @code{THEN} and @code{ENDIF} are
6809: aliases).
6810: @end itemize
1.1 anton 6811:
1.69 anton 6812: Like deferred words, an alias has default compilation and interpretation
6813: semantics at the beginning (not the modifications of the other word),
6814: but you can change them in the usual ways (@code{immediate},
6815: @code{compile-only}). For example:
1.1 anton 6816:
6817: @example
1.44 crook 6818: : foo ... ; immediate
6819:
6820: ' foo Alias bar \ bar is not an immediate word
6821: ' foo Alias fooby immediate \ fooby is an immediate word
1.1 anton 6822: @end example
6823:
1.44 crook 6824: Words that are aliases have the same xt, different headers in the
6825: dictionary, and consequently different name tokens (@pxref{Tokens for
6826: Words}) and possibly different immediate flags. An alias can only have
6827: default or immediate compilation semantics; you can define aliases for
6828: combined words with @code{interpret/compile:} -- see @ref{Combined words}.
1.1 anton 6829:
1.44 crook 6830: doc-alias
1.1 anton 6831:
6832:
1.47 crook 6833: @node Interpretation and Compilation Semantics, Tokens for Words, Defining Words, Words
6834: @section Interpretation and Compilation Semantics
1.26 crook 6835: @cindex semantics, interpretation and compilation
1.1 anton 6836:
1.71 anton 6837: @c !! state and ' are used without explanation
6838: @c example for immediate/compile-only? or is the tutorial enough
6839:
1.26 crook 6840: @cindex interpretation semantics
1.71 anton 6841: The @dfn{interpretation semantics} of a (named) word are what the text
1.26 crook 6842: interpreter does when it encounters the word in interpret state. It also
6843: appears in some other contexts, e.g., the execution token returned by
1.71 anton 6844: @code{' @i{word}} identifies the interpretation semantics of @i{word}
6845: (in other words, @code{' @i{word} execute} is equivalent to
1.29 crook 6846: interpret-state text interpretation of @code{@i{word}}).
1.1 anton 6847:
1.26 crook 6848: @cindex compilation semantics
1.71 anton 6849: The @dfn{compilation semantics} of a (named) word are what the text
6850: interpreter does when it encounters the word in compile state. It also
6851: appears in other contexts, e.g, @code{POSTPONE @i{word}}
6852: compiles@footnote{In standard terminology, ``appends to the current
6853: definition''.} the compilation semantics of @i{word}.
1.1 anton 6854:
1.26 crook 6855: @cindex execution semantics
6856: The standard also talks about @dfn{execution semantics}. They are used
6857: only for defining the interpretation and compilation semantics of many
6858: words. By default, the interpretation semantics of a word are to
6859: @code{execute} its execution semantics, and the compilation semantics of
6860: a word are to @code{compile,} its execution semantics.@footnote{In
6861: standard terminology: The default interpretation semantics are its
6862: execution semantics; the default compilation semantics are to append its
6863: execution semantics to the execution semantics of the current
6864: definition.}
6865:
1.71 anton 6866: Unnamed words (@pxref{Anonymous Definitions}) cannot be encountered by
6867: the text interpreter, ticked, or @code{postpone}d, so they have no
6868: interpretation or compilation semantics. Their behaviour is represented
6869: by their XT (@pxref{Tokens for Words}), and we call it execution
6870: semantics, too.
6871:
1.26 crook 6872: @comment TODO expand, make it co-operate with new sections on text interpreter.
6873:
6874: @cindex immediate words
6875: @cindex compile-only words
6876: You can change the semantics of the most-recently defined word:
6877:
1.44 crook 6878:
1.26 crook 6879: doc-immediate
6880: doc-compile-only
6881: doc-restrict
6882:
1.82 anton 6883: By convention, words with non-default compilation semantics (e.g.,
6884: immediate words) often have names surrounded with brackets (e.g.,
6885: @code{[']}, @pxref{Execution token}).
1.44 crook 6886:
1.26 crook 6887: Note that ticking (@code{'}) a compile-only word gives an error
6888: (``Interpreting a compile-only word'').
1.1 anton 6889:
1.47 crook 6890: @menu
1.67 anton 6891: * Combined words::
1.47 crook 6892: @end menu
1.44 crook 6893:
1.71 anton 6894:
1.48 anton 6895: @node Combined words, , Interpretation and Compilation Semantics, Interpretation and Compilation Semantics
1.44 crook 6896: @subsection Combined Words
6897: @cindex combined words
6898:
6899: Gforth allows you to define @dfn{combined words} -- words that have an
6900: arbitrary combination of interpretation and compilation semantics.
6901:
1.26 crook 6902: doc-interpret/compile:
1.1 anton 6903:
1.26 crook 6904: This feature was introduced for implementing @code{TO} and @code{S"}. I
6905: recommend that you do not define such words, as cute as they may be:
6906: they make it hard to get at both parts of the word in some contexts.
6907: E.g., assume you want to get an execution token for the compilation
6908: part. Instead, define two words, one that embodies the interpretation
6909: part, and one that embodies the compilation part. Once you have done
6910: that, you can define a combined word with @code{interpret/compile:} for
6911: the convenience of your users.
1.1 anton 6912:
1.26 crook 6913: You might try to use this feature to provide an optimizing
6914: implementation of the default compilation semantics of a word. For
6915: example, by defining:
1.1 anton 6916: @example
1.26 crook 6917: :noname
6918: foo bar ;
6919: :noname
6920: POSTPONE foo POSTPONE bar ;
1.29 crook 6921: interpret/compile: opti-foobar
1.1 anton 6922: @end example
1.26 crook 6923:
1.23 crook 6924: @noindent
1.26 crook 6925: as an optimizing version of:
6926:
1.1 anton 6927: @example
1.26 crook 6928: : foobar
6929: foo bar ;
1.1 anton 6930: @end example
6931:
1.26 crook 6932: Unfortunately, this does not work correctly with @code{[compile]},
6933: because @code{[compile]} assumes that the compilation semantics of all
6934: @code{interpret/compile:} words are non-default. I.e., @code{[compile]
1.29 crook 6935: opti-foobar} would compile compilation semantics, whereas
6936: @code{[compile] foobar} would compile interpretation semantics.
1.1 anton 6937:
1.26 crook 6938: @cindex state-smart words (are a bad idea)
1.82 anton 6939: @anchor{state-smartness}
1.29 crook 6940: Some people try to use @dfn{state-smart} words to emulate the feature provided
1.26 crook 6941: by @code{interpret/compile:} (words are state-smart if they check
6942: @code{STATE} during execution). E.g., they would try to code
6943: @code{foobar} like this:
1.1 anton 6944:
1.26 crook 6945: @example
6946: : foobar
6947: STATE @@
6948: IF ( compilation state )
6949: POSTPONE foo POSTPONE bar
6950: ELSE
6951: foo bar
6952: ENDIF ; immediate
6953: @end example
1.1 anton 6954:
1.26 crook 6955: Although this works if @code{foobar} is only processed by the text
6956: interpreter, it does not work in other contexts (like @code{'} or
6957: @code{POSTPONE}). E.g., @code{' foobar} will produce an execution token
6958: for a state-smart word, not for the interpretation semantics of the
6959: original @code{foobar}; when you execute this execution token (directly
6960: with @code{EXECUTE} or indirectly through @code{COMPILE,}) in compile
6961: state, the result will not be what you expected (i.e., it will not
6962: perform @code{foo bar}). State-smart words are a bad idea. Simply don't
6963: write them@footnote{For a more detailed discussion of this topic, see
1.66 anton 6964: M. Anton Ertl,
6965: @cite{@uref{http://www.complang.tuwien.ac.at/papers/ertl98.ps.gz,@code{State}-smartness---Why
6966: it is Evil and How to Exorcise it}}, EuroForth '98.}!
1.1 anton 6967:
1.26 crook 6968: @cindex defining words with arbitrary semantics combinations
6969: It is also possible to write defining words that define words with
6970: arbitrary combinations of interpretation and compilation semantics. In
6971: general, they look like this:
1.1 anton 6972:
1.26 crook 6973: @example
6974: : def-word
6975: create-interpret/compile
1.29 crook 6976: @i{code1}
1.26 crook 6977: interpretation>
1.29 crook 6978: @i{code2}
1.26 crook 6979: <interpretation
6980: compilation>
1.29 crook 6981: @i{code3}
1.26 crook 6982: <compilation ;
6983: @end example
1.1 anton 6984:
1.29 crook 6985: For a @i{word} defined with @code{def-word}, the interpretation
6986: semantics are to push the address of the body of @i{word} and perform
6987: @i{code2}, and the compilation semantics are to push the address of
6988: the body of @i{word} and perform @i{code3}. E.g., @code{constant}
1.26 crook 6989: can also be defined like this (except that the defined constants don't
6990: behave correctly when @code{[compile]}d):
1.1 anton 6991:
1.26 crook 6992: @example
6993: : constant ( n "name" -- )
6994: create-interpret/compile
6995: ,
6996: interpretation> ( -- n )
6997: @@
6998: <interpretation
6999: compilation> ( compilation. -- ; run-time. -- n )
7000: @@ postpone literal
7001: <compilation ;
7002: @end example
1.1 anton 7003:
1.44 crook 7004:
1.26 crook 7005: doc-create-interpret/compile
7006: doc-interpretation>
7007: doc-<interpretation
7008: doc-compilation>
7009: doc-<compilation
1.1 anton 7010:
1.44 crook 7011:
1.29 crook 7012: Words defined with @code{interpret/compile:} and
1.26 crook 7013: @code{create-interpret/compile} have an extended header structure that
7014: differs from other words; however, unless you try to access them with
7015: plain address arithmetic, you should not notice this. Words for
7016: accessing the header structure usually know how to deal with this; e.g.,
1.29 crook 7017: @code{'} @i{word} @code{>body} also gives you the body of a word created
7018: with @code{create-interpret/compile}.
1.1 anton 7019:
1.44 crook 7020:
1.47 crook 7021: @c -------------------------------------------------------------
1.81 anton 7022: @node Tokens for Words, Compiling words, Interpretation and Compilation Semantics, Words
1.47 crook 7023: @section Tokens for Words
7024: @cindex tokens for words
7025:
7026: This section describes the creation and use of tokens that represent
7027: words.
7028:
1.71 anton 7029: @menu
7030: * Execution token:: represents execution/interpretation semantics
7031: * Compilation token:: represents compilation semantics
7032: * Name token:: represents named words
7033: @end menu
1.47 crook 7034:
1.71 anton 7035: @node Execution token, Compilation token, Tokens for Words, Tokens for Words
7036: @subsection Execution token
1.47 crook 7037:
7038: @cindex xt
7039: @cindex execution token
1.71 anton 7040: An @dfn{execution token} (@i{XT}) represents some behaviour of a word.
7041: You can use @code{execute} to invoke this behaviour.
1.47 crook 7042:
1.71 anton 7043: @cindex tick (')
7044: You can use @code{'} to get an execution token that represents the
7045: interpretation semantics of a named word:
1.47 crook 7046:
7047: @example
1.97 anton 7048: 5 ' . ( n xt )
7049: execute ( ) \ execute the xt (i.e., ".")
1.71 anton 7050: @end example
1.47 crook 7051:
1.71 anton 7052: doc-'
7053:
7054: @code{'} parses at run-time; there is also a word @code{[']} that parses
7055: when it is compiled, and compiles the resulting XT:
7056:
7057: @example
7058: : foo ['] . execute ;
7059: 5 foo
7060: : bar ' execute ; \ by contrast,
7061: 5 bar . \ ' parses "." when bar executes
7062: @end example
7063:
7064: doc-[']
7065:
7066: If you want the execution token of @i{word}, write @code{['] @i{word}}
7067: in compiled code and @code{' @i{word}} in interpreted code. Gforth's
7068: @code{'} and @code{[']} behave somewhat unusually by complaining about
7069: compile-only words (because these words have no interpretation
7070: semantics). You might get what you want by using @code{COMP' @i{word}
7071: DROP} or @code{[COMP'] @i{word} DROP} (for details @pxref{Compilation
7072: token}).
7073:
1.116 anton 7074: Another way to get an XT is @code{:noname} or @code{latestxt}
1.71 anton 7075: (@pxref{Anonymous Definitions}). For anonymous words this gives an xt
7076: for the only behaviour the word has (the execution semantics). For
1.116 anton 7077: named words, @code{latestxt} produces an XT for the same behaviour it
1.71 anton 7078: would produce if the word was defined anonymously.
7079:
7080: @example
7081: :noname ." hello" ;
7082: execute
1.47 crook 7083: @end example
7084:
1.71 anton 7085: An XT occupies one cell and can be manipulated like any other cell.
7086:
1.47 crook 7087: @cindex code field address
7088: @cindex CFA
1.71 anton 7089: In ANS Forth the XT is just an abstract data type (i.e., defined by the
7090: operations that produce or consume it). For old hands: In Gforth, the
7091: XT is implemented as a code field address (CFA).
7092:
7093: doc-execute
7094: doc-perform
7095:
7096: @node Compilation token, Name token, Execution token, Tokens for Words
7097: @subsection Compilation token
1.47 crook 7098:
7099: @cindex compilation token
1.71 anton 7100: @cindex CT (compilation token)
7101: Gforth represents the compilation semantics of a named word by a
1.47 crook 7102: @dfn{compilation token} consisting of two cells: @i{w xt}. The top cell
7103: @i{xt} is an execution token. The compilation semantics represented by
7104: the compilation token can be performed with @code{execute}, which
7105: consumes the whole compilation token, with an additional stack effect
7106: determined by the represented compilation semantics.
7107:
7108: At present, the @i{w} part of a compilation token is an execution token,
7109: and the @i{xt} part represents either @code{execute} or
7110: @code{compile,}@footnote{Depending upon the compilation semantics of the
7111: word. If the word has default compilation semantics, the @i{xt} will
7112: represent @code{compile,}. Otherwise (e.g., for immediate words), the
7113: @i{xt} will represent @code{execute}.}. However, don't rely on that
7114: knowledge, unless necessary; future versions of Gforth may introduce
7115: unusual compilation tokens (e.g., a compilation token that represents
7116: the compilation semantics of a literal).
7117:
1.71 anton 7118: You can perform the compilation semantics represented by the compilation
7119: token with @code{execute}. You can compile the compilation semantics
7120: with @code{postpone,}. I.e., @code{COMP' @i{word} postpone,} is
7121: equivalent to @code{postpone @i{word}}.
7122:
7123: doc-[comp']
7124: doc-comp'
7125: doc-postpone,
7126:
7127: @node Name token, , Compilation token, Tokens for Words
7128: @subsection Name token
1.47 crook 7129:
7130: @cindex name token
1.116 anton 7131: Gforth represents named words by the @dfn{name token}, (@i{nt}). Name
7132: token is an abstract data type that occurs as argument or result of the
7133: words below.
7134:
7135: @c !! put this elswhere?
1.47 crook 7136: @cindex name field address
7137: @cindex NFA
1.116 anton 7138: The closest thing to the nt in older Forth systems is the name field
7139: address (NFA), but there are significant differences: in older Forth
7140: systems each word had a unique NFA, LFA, CFA and PFA (in this order, or
7141: LFA, NFA, CFA, PFA) and there were words for getting from one to the
7142: next. In contrast, in Gforth 0@dots{}n nts correspond to one xt; there
7143: is a link field in the structure identified by the name token, but
7144: searching usually uses a hash table external to these structures; the
7145: name in Gforth has a cell-wide count-and-flags field, and the nt is not
7146: implemented as the address of that count field.
1.47 crook 7147:
7148: doc-find-name
1.116 anton 7149: doc-latest
7150: doc->name
1.47 crook 7151: doc-name>int
7152: doc-name?int
7153: doc-name>comp
7154: doc-name>string
1.109 anton 7155: doc-id.
7156: doc-.name
7157: doc-.id
1.47 crook 7158:
1.81 anton 7159: @c ----------------------------------------------------------
7160: @node Compiling words, The Text Interpreter, Tokens for Words, Words
7161: @section Compiling words
7162: @cindex compiling words
7163: @cindex macros
7164:
7165: In contrast to most other languages, Forth has no strict boundary
1.82 anton 7166: between compilation and run-time. E.g., you can run arbitrary code
7167: between defining words (or for computing data used by defining words
7168: like @code{constant}). Moreover, @code{Immediate} (@pxref{Interpretation
7169: and Compilation Semantics} and @code{[}...@code{]} (see below) allow
7170: running arbitrary code while compiling a colon definition (exception:
7171: you must not allot dictionary space).
7172:
7173: @menu
7174: * Literals:: Compiling data values
7175: * Macros:: Compiling words
7176: @end menu
7177:
7178: @node Literals, Macros, Compiling words, Compiling words
7179: @subsection Literals
7180: @cindex Literals
7181:
7182: The simplest and most frequent example is to compute a literal during
7183: compilation. E.g., the following definition prints an array of strings,
7184: one string per line:
7185:
7186: @example
7187: : .strings ( addr u -- ) \ gforth
7188: 2* cells bounds U+DO
7189: cr i 2@@ type
7190: 2 cells +LOOP ;
7191: @end example
1.81 anton 7192:
1.82 anton 7193: With a simple-minded compiler like Gforth's, this computes @code{2
7194: cells} on every loop iteration. You can compute this value once and for
7195: all at compile time and compile it into the definition like this:
7196:
7197: @example
7198: : .strings ( addr u -- ) \ gforth
7199: 2* cells bounds U+DO
7200: cr i 2@@ type
7201: [ 2 cells ] literal +LOOP ;
7202: @end example
7203:
7204: @code{[} switches the text interpreter to interpret state (you will get
7205: an @code{ok} prompt if you type this example interactively and insert a
7206: newline between @code{[} and @code{]}), so it performs the
7207: interpretation semantics of @code{2 cells}; this computes a number.
7208: @code{]} switches the text interpreter back into compile state. It then
7209: performs @code{Literal}'s compilation semantics, which are to compile
7210: this number into the current word. You can decompile the word with
7211: @code{see .strings} to see the effect on the compiled code.
1.81 anton 7212:
1.82 anton 7213: You can also optimize the @code{2* cells} into @code{[ 2 cells ] literal
7214: *} in this way.
1.81 anton 7215:
1.82 anton 7216: doc-[
7217: doc-]
1.81 anton 7218: doc-literal
7219: doc-]L
1.82 anton 7220:
7221: There are also words for compiling other data types than single cells as
7222: literals:
7223:
1.81 anton 7224: doc-2literal
7225: doc-fliteral
1.82 anton 7226: doc-sliteral
7227:
7228: @cindex colon-sys, passing data across @code{:}
7229: @cindex @code{:}, passing data across
7230: You might be tempted to pass data from outside a colon definition to the
7231: inside on the data stack. This does not work, because @code{:} puhes a
7232: colon-sys, making stuff below unaccessible. E.g., this does not work:
7233:
7234: @example
7235: 5 : foo literal ; \ error: "unstructured"
7236: @end example
7237:
7238: Instead, you have to pass the value in some other way, e.g., through a
7239: variable:
7240:
7241: @example
7242: variable temp
7243: 5 temp !
7244: : foo [ temp @@ ] literal ;
7245: @end example
7246:
7247:
7248: @node Macros, , Literals, Compiling words
7249: @subsection Macros
7250: @cindex Macros
7251: @cindex compiling compilation semantics
7252:
7253: @code{Literal} and friends compile data values into the current
7254: definition. You can also write words that compile other words into the
7255: current definition. E.g.,
7256:
7257: @example
7258: : compile-+ ( -- ) \ compiled code: ( n1 n2 -- n )
7259: POSTPONE + ;
7260:
7261: : foo ( n1 n2 -- n )
7262: [ compile-+ ] ;
7263: 1 2 foo .
7264: @end example
7265:
7266: This is equivalent to @code{: foo + ;} (@code{see foo} to check this).
7267: What happens in this example? @code{Postpone} compiles the compilation
7268: semantics of @code{+} into @code{compile-+}; later the text interpreter
7269: executes @code{compile-+} and thus the compilation semantics of +, which
7270: compile (the execution semantics of) @code{+} into
7271: @code{foo}.@footnote{A recent RFI answer requires that compiling words
7272: should only be executed in compile state, so this example is not
7273: guaranteed to work on all standard systems, but on any decent system it
7274: will work.}
7275:
7276: doc-postpone
7277: doc-[compile]
7278:
7279: Compiling words like @code{compile-+} are usually immediate (or similar)
7280: so you do not have to switch to interpret state to execute them;
7281: mopifying the last example accordingly produces:
7282:
7283: @example
7284: : [compile-+] ( compilation: --; interpretation: -- )
7285: \ compiled code: ( n1 n2 -- n )
7286: POSTPONE + ; immediate
7287:
7288: : foo ( n1 n2 -- n )
7289: [compile-+] ;
7290: 1 2 foo .
7291: @end example
7292:
7293: Immediate compiling words are similar to macros in other languages (in
7294: particular, Lisp). The important differences to macros in, e.g., C are:
7295:
7296: @itemize @bullet
7297:
7298: @item
7299: You use the same language for defining and processing macros, not a
7300: separate preprocessing language and processor.
7301:
7302: @item
7303: Consequently, the full power of Forth is available in macro definitions.
7304: E.g., you can perform arbitrarily complex computations, or generate
7305: different code conditionally or in a loop (e.g., @pxref{Advanced macros
7306: Tutorial}). This power is very useful when writing a parser generators
7307: or other code-generating software.
7308:
7309: @item
7310: Macros defined using @code{postpone} etc. deal with the language at a
7311: higher level than strings; name binding happens at macro definition
7312: time, so you can avoid the pitfalls of name collisions that can happen
7313: in C macros. Of course, Forth is a liberal language and also allows to
7314: shoot yourself in the foot with text-interpreted macros like
7315:
7316: @example
7317: : [compile-+] s" +" evaluate ; immediate
7318: @end example
7319:
7320: Apart from binding the name at macro use time, using @code{evaluate}
7321: also makes your definition @code{state}-smart (@pxref{state-smartness}).
7322: @end itemize
7323:
7324: You may want the macro to compile a number into a word. The word to do
7325: it is @code{literal}, but you have to @code{postpone} it, so its
7326: compilation semantics take effect when the macro is executed, not when
7327: it is compiled:
7328:
7329: @example
7330: : [compile-5] ( -- ) \ compiled code: ( -- n )
7331: 5 POSTPONE literal ; immediate
7332:
7333: : foo [compile-5] ;
7334: foo .
7335: @end example
7336:
7337: You may want to pass parameters to a macro, that the macro should
7338: compile into the current definition. If the parameter is a number, then
7339: you can use @code{postpone literal} (similar for other values).
7340:
7341: If you want to pass a word that is to be compiled, the usual way is to
7342: pass an execution token and @code{compile,} it:
7343:
7344: @example
7345: : twice1 ( xt -- ) \ compiled code: ... -- ...
7346: dup compile, compile, ;
7347:
7348: : 2+ ( n1 -- n2 )
7349: [ ' 1+ twice1 ] ;
7350: @end example
7351:
7352: doc-compile,
7353:
7354: An alternative available in Gforth, that allows you to pass compile-only
7355: words as parameters is to use the compilation token (@pxref{Compilation
7356: token}). The same example in this technique:
7357:
7358: @example
7359: : twice ( ... ct -- ... ) \ compiled code: ... -- ...
7360: 2dup 2>r execute 2r> execute ;
7361:
7362: : 2+ ( n1 -- n2 )
7363: [ comp' 1+ twice ] ;
7364: @end example
7365:
7366: In the example above @code{2>r} and @code{2r>} ensure that @code{twice}
7367: works even if the executed compilation semantics has an effect on the
7368: data stack.
7369:
7370: You can also define complete definitions with these words; this provides
7371: an alternative to using @code{does>} (@pxref{User-defined Defining
7372: Words}). E.g., instead of
7373:
7374: @example
7375: : curry+ ( n1 "name" -- )
7376: CREATE ,
7377: DOES> ( n2 -- n1+n2 )
7378: @@ + ;
7379: @end example
7380:
7381: you could define
7382:
7383: @example
7384: : curry+ ( n1 "name" -- )
7385: \ name execution: ( n2 -- n1+n2 )
7386: >r : r> POSTPONE literal POSTPONE + POSTPONE ; ;
1.81 anton 7387:
1.82 anton 7388: -3 curry+ 3-
7389: see 3-
7390: @end example
1.81 anton 7391:
1.82 anton 7392: The sequence @code{>r : r>} is necessary, because @code{:} puts a
7393: colon-sys on the data stack that makes everything below it unaccessible.
1.81 anton 7394:
1.82 anton 7395: This way of writing defining words is sometimes more, sometimes less
7396: convenient than using @code{does>} (@pxref{Advanced does> usage
7397: example}). One advantage of this method is that it can be optimized
7398: better, because the compiler knows that the value compiled with
7399: @code{literal} is fixed, whereas the data associated with a
7400: @code{create}d word can be changed.
1.47 crook 7401:
1.26 crook 7402: @c ----------------------------------------------------------
1.111 anton 7403: @node The Text Interpreter, The Input Stream, Compiling words, Words
1.26 crook 7404: @section The Text Interpreter
7405: @cindex interpreter - outer
7406: @cindex text interpreter
7407: @cindex outer interpreter
1.1 anton 7408:
1.34 anton 7409: @c Should we really describe all these ugly details? IMO the text
7410: @c interpreter should be much cleaner, but that may not be possible within
7411: @c ANS Forth. - anton
1.44 crook 7412: @c nac-> I wanted to explain how it works to show how you can exploit
7413: @c it in your own programs. When I was writing a cross-compiler, figuring out
7414: @c some of these gory details was very helpful to me. None of the textbooks
7415: @c I've seen cover it, and the most modern Forth textbook -- Forth Inc's,
7416: @c seems to positively avoid going into too much detail for some of
7417: @c the internals.
1.34 anton 7418:
1.71 anton 7419: @c anton: ok. I wonder, though, if this is the right place; for some stuff
7420: @c it is; for the ugly details, I would prefer another place. I wonder
7421: @c whether we should have a chapter before "Words" that describes some
7422: @c basic concepts referred to in words, and a chapter after "Words" that
7423: @c describes implementation details.
7424:
1.29 crook 7425: The text interpreter@footnote{This is an expanded version of the
7426: material in @ref{Introducing the Text Interpreter}.} is an endless loop
1.34 anton 7427: that processes input from the current input device. It is also called
7428: the outer interpreter, in contrast to the inner interpreter
7429: (@pxref{Engine}) which executes the compiled Forth code on interpretive
7430: implementations.
1.27 crook 7431:
1.29 crook 7432: @cindex interpret state
7433: @cindex compile state
7434: The text interpreter operates in one of two states: @dfn{interpret
7435: state} and @dfn{compile state}. The current state is defined by the
1.71 anton 7436: aptly-named variable @code{state}.
1.29 crook 7437:
7438: This section starts by describing how the text interpreter behaves when
7439: it is in interpret state, processing input from the user input device --
7440: the keyboard. This is the mode that a Forth system is in after it starts
7441: up.
7442:
7443: @cindex input buffer
7444: @cindex terminal input buffer
7445: The text interpreter works from an area of memory called the @dfn{input
7446: buffer}@footnote{When the text interpreter is processing input from the
7447: keyboard, this area of memory is called the @dfn{terminal input buffer}
7448: (TIB) and is addressed by the (obsolescent) words @code{TIB} and
7449: @code{#TIB}.}, which stores your keyboard input when you press the
1.30 anton 7450: @key{RET} key. Starting at the beginning of the input buffer, it skips
1.29 crook 7451: leading spaces (called @dfn{delimiters}) then parses a string (a
7452: sequence of non-space characters) until it reaches either a space
7453: character or the end of the buffer. Having parsed a string, it makes two
7454: attempts to process it:
1.27 crook 7455:
1.29 crook 7456: @cindex dictionary
1.27 crook 7457: @itemize @bullet
7458: @item
1.29 crook 7459: It looks for the string in a @dfn{dictionary} of definitions. If the
7460: string is found, the string names a @dfn{definition} (also known as a
7461: @dfn{word}) and the dictionary search returns information that allows
7462: the text interpreter to perform the word's @dfn{interpretation
7463: semantics}. In most cases, this simply means that the word will be
7464: executed.
1.27 crook 7465: @item
7466: If the string is not found in the dictionary, the text interpreter
1.29 crook 7467: attempts to treat it as a number, using the rules described in
7468: @ref{Number Conversion}. If the string represents a legal number in the
7469: current radix, the number is pushed onto a parameter stack (the data
7470: stack for integers, the floating-point stack for floating-point
7471: numbers).
7472: @end itemize
7473:
7474: If both attempts fail, or if the word is found in the dictionary but has
7475: no interpretation semantics@footnote{This happens if the word was
7476: defined as @code{COMPILE-ONLY}.} the text interpreter discards the
7477: remainder of the input buffer, issues an error message and waits for
7478: more input. If one of the attempts succeeds, the text interpreter
7479: repeats the parsing process until the whole of the input buffer has been
7480: processed, at which point it prints the status message ``@code{ ok}''
7481: and waits for more input.
7482:
1.71 anton 7483: @c anton: this should be in the input stream subsection (or below it)
7484:
1.29 crook 7485: @cindex parse area
7486: The text interpreter keeps track of its position in the input buffer by
7487: updating a variable called @code{>IN} (pronounced ``to-in''). The value
7488: of @code{>IN} starts out as 0, indicating an offset of 0 from the start
7489: of the input buffer. The region from offset @code{>IN @@} to the end of
7490: the input buffer is called the @dfn{parse area}@footnote{In other words,
7491: the text interpreter processes the contents of the input buffer by
7492: parsing strings from the parse area until the parse area is empty.}.
7493: This example shows how @code{>IN} changes as the text interpreter parses
7494: the input buffer:
7495:
7496: @example
7497: : remaining >IN @@ SOURCE 2 PICK - -ROT + SWAP
7498: CR ." ->" TYPE ." <-" ; IMMEDIATE
7499:
7500: 1 2 3 remaining + remaining .
7501:
7502: : foo 1 2 3 remaining SWAP remaining ;
7503: @end example
7504:
7505: @noindent
7506: The result is:
7507:
7508: @example
7509: ->+ remaining .<-
7510: ->.<-5 ok
7511:
7512: ->SWAP remaining ;-<
7513: ->;<- ok
7514: @end example
7515:
7516: @cindex parsing words
7517: The value of @code{>IN} can also be modified by a word in the input
7518: buffer that is executed by the text interpreter. This means that a word
7519: can ``trick'' the text interpreter into either skipping a section of the
7520: input buffer@footnote{This is how parsing words work.} or into parsing a
7521: section twice. For example:
1.27 crook 7522:
1.29 crook 7523: @example
1.71 anton 7524: : lat ." <<foo>>" ;
7525: : flat ." <<bar>>" >IN DUP @@ 3 - SWAP ! ;
1.29 crook 7526: @end example
7527:
7528: @noindent
7529: When @code{flat} is executed, this output is produced@footnote{Exercise
7530: for the reader: what would happen if the @code{3} were replaced with
7531: @code{4}?}:
7532:
7533: @example
1.71 anton 7534: <<bar>><<foo>>
1.29 crook 7535: @end example
7536:
1.71 anton 7537: This technique can be used to work around some of the interoperability
7538: problems of parsing words. Of course, it's better to avoid parsing
7539: words where possible.
7540:
1.29 crook 7541: @noindent
7542: Two important notes about the behaviour of the text interpreter:
1.27 crook 7543:
7544: @itemize @bullet
7545: @item
7546: It processes each input string to completion before parsing additional
1.29 crook 7547: characters from the input buffer.
7548: @item
7549: It treats the input buffer as a read-only region (and so must your code).
7550: @end itemize
7551:
7552: @noindent
7553: When the text interpreter is in compile state, its behaviour changes in
7554: these ways:
7555:
7556: @itemize @bullet
7557: @item
7558: If a parsed string is found in the dictionary, the text interpreter will
7559: perform the word's @dfn{compilation semantics}. In most cases, this
7560: simply means that the execution semantics of the word will be appended
7561: to the current definition.
1.27 crook 7562: @item
1.29 crook 7563: When a number is encountered, it is compiled into the current definition
7564: (as a literal) rather than being pushed onto a parameter stack.
7565: @item
7566: If an error occurs, @code{state} is modified to put the text interpreter
7567: back into interpret state.
7568: @item
7569: Each time a line is entered from the keyboard, Gforth prints
7570: ``@code{ compiled}'' rather than `` @code{ok}''.
7571: @end itemize
7572:
7573: @cindex text interpreter - input sources
7574: When the text interpreter is using an input device other than the
7575: keyboard, its behaviour changes in these ways:
7576:
7577: @itemize @bullet
7578: @item
7579: When the parse area is empty, the text interpreter attempts to refill
7580: the input buffer from the input source. When the input source is
1.71 anton 7581: exhausted, the input source is set back to the previous input source.
1.29 crook 7582: @item
7583: It doesn't print out ``@code{ ok}'' or ``@code{ compiled}'' messages each
7584: time the parse area is emptied.
7585: @item
7586: If an error occurs, the input source is set back to the user input
7587: device.
1.27 crook 7588: @end itemize
1.21 crook 7589:
1.49 anton 7590: You can read about this in more detail in @ref{Input Sources}.
1.44 crook 7591:
1.26 crook 7592: doc->in
1.27 crook 7593: doc-source
7594:
1.26 crook 7595: doc-tib
7596: doc-#tib
1.1 anton 7597:
1.44 crook 7598:
1.26 crook 7599: @menu
1.67 anton 7600: * Input Sources::
7601: * Number Conversion::
7602: * Interpret/Compile states::
7603: * Interpreter Directives::
1.26 crook 7604: @end menu
1.1 anton 7605:
1.29 crook 7606: @node Input Sources, Number Conversion, The Text Interpreter, The Text Interpreter
7607: @subsection Input Sources
7608: @cindex input sources
7609: @cindex text interpreter - input sources
7610:
1.44 crook 7611: By default, the text interpreter processes input from the user input
1.29 crook 7612: device (the keyboard) when Forth starts up. The text interpreter can
7613: process input from any of these sources:
7614:
7615: @itemize @bullet
7616: @item
7617: The user input device -- the keyboard.
7618: @item
7619: A file, using the words described in @ref{Forth source files}.
7620: @item
7621: A block, using the words described in @ref{Blocks}.
7622: @item
7623: A text string, using @code{evaluate}.
7624: @end itemize
7625:
7626: A program can identify the current input device from the values of
7627: @code{source-id} and @code{blk}.
7628:
1.44 crook 7629:
1.29 crook 7630: doc-source-id
7631: doc-blk
7632:
7633: doc-save-input
7634: doc-restore-input
7635:
7636: doc-evaluate
1.111 anton 7637: doc-query
1.1 anton 7638:
1.29 crook 7639:
1.44 crook 7640:
1.29 crook 7641: @node Number Conversion, Interpret/Compile states, Input Sources, The Text Interpreter
1.26 crook 7642: @subsection Number Conversion
7643: @cindex number conversion
7644: @cindex double-cell numbers, input format
7645: @cindex input format for double-cell numbers
7646: @cindex single-cell numbers, input format
7647: @cindex input format for single-cell numbers
7648: @cindex floating-point numbers, input format
7649: @cindex input format for floating-point numbers
1.1 anton 7650:
1.29 crook 7651: This section describes the rules that the text interpreter uses when it
7652: tries to convert a string into a number.
1.1 anton 7653:
1.26 crook 7654: Let <digit> represent any character that is a legal digit in the current
1.29 crook 7655: number base@footnote{For example, 0-9 when the number base is decimal or
7656: 0-9, A-F when the number base is hexadecimal.}.
1.1 anton 7657:
1.26 crook 7658: Let <decimal digit> represent any character in the range 0-9.
1.1 anton 7659:
1.29 crook 7660: Let @{@i{a b}@} represent the @i{optional} presence of any of the characters
7661: in the braces (@i{a} or @i{b} or neither).
1.1 anton 7662:
1.26 crook 7663: Let * represent any number of instances of the previous character
7664: (including none).
1.1 anton 7665:
1.26 crook 7666: Let any other character represent itself.
1.1 anton 7667:
1.29 crook 7668: @noindent
1.26 crook 7669: Now, the conversion rules are:
1.21 crook 7670:
1.26 crook 7671: @itemize @bullet
7672: @item
7673: A string of the form <digit><digit>* is treated as a single-precision
1.29 crook 7674: (cell-sized) positive integer. Examples are 0 123 6784532 32343212343456 42
1.26 crook 7675: @item
7676: A string of the form -<digit><digit>* is treated as a single-precision
1.29 crook 7677: (cell-sized) negative integer, and is represented using 2's-complement
1.26 crook 7678: arithmetic. Examples are -45 -5681 -0
7679: @item
7680: A string of the form <digit><digit>*.<digit>* is treated as a double-precision
1.29 crook 7681: (double-cell-sized) positive integer. Examples are 3465. 3.465 34.65
7682: (all three of these represent the same number).
1.26 crook 7683: @item
7684: A string of the form -<digit><digit>*.<digit>* is treated as a
1.29 crook 7685: double-precision (double-cell-sized) negative integer, and is
1.26 crook 7686: represented using 2's-complement arithmetic. Examples are -3465. -3.465
1.29 crook 7687: -34.65 (all three of these represent the same number).
1.26 crook 7688: @item
1.29 crook 7689: A string of the form @{+ -@}<decimal digit>@{.@}<decimal digit>*@{e
7690: E@}@{+ -@}<decimal digit><decimal digit>* is treated as a floating-point
1.35 anton 7691: number. Examples are 1e 1e0 1.e 1.e0 +1e+0 (which all represent the same
1.29 crook 7692: number) +12.E-4
1.26 crook 7693: @end itemize
1.1 anton 7694:
1.26 crook 7695: By default, the number base used for integer number conversion is given
1.35 anton 7696: by the contents of the variable @code{base}. Note that a lot of
7697: confusion can result from unexpected values of @code{base}. If you
7698: change @code{base} anywhere, make sure to save the old value and restore
7699: it afterwards. In general I recommend keeping @code{base} decimal, and
7700: using the prefixes described below for the popular non-decimal bases.
1.1 anton 7701:
1.29 crook 7702: doc-dpl
1.26 crook 7703: doc-base
7704: doc-hex
7705: doc-decimal
1.1 anton 7706:
1.26 crook 7707: @cindex '-prefix for character strings
7708: @cindex &-prefix for decimal numbers
1.133 anton 7709: @cindex #-prefix for decimal numbers
1.26 crook 7710: @cindex %-prefix for binary numbers
7711: @cindex $-prefix for hexadecimal numbers
1.133 anton 7712: @cindex 0x-prefix for hexadecimal numbers
1.35 anton 7713: Gforth allows you to override the value of @code{base} by using a
1.29 crook 7714: prefix@footnote{Some Forth implementations provide a similar scheme by
7715: implementing @code{$} etc. as parsing words that process the subsequent
7716: number in the input stream and push it onto the stack. For example, see
7717: @cite{Number Conversion and Literals}, by Wil Baden; Forth Dimensions
7718: 20(3) pages 26--27. In such implementations, unlike in Gforth, a space
7719: is required between the prefix and the number.} before the first digit
1.133 anton 7720: of an (integer) number. The following prefixes are supported:
1.1 anton 7721:
1.26 crook 7722: @itemize @bullet
7723: @item
1.35 anton 7724: @code{&} -- decimal
1.26 crook 7725: @item
1.133 anton 7726: @code{#} -- decimal
7727: @item
1.35 anton 7728: @code{%} -- binary
1.26 crook 7729: @item
1.35 anton 7730: @code{$} -- hexadecimal
1.26 crook 7731: @item
1.133 anton 7732: @code{0x} -- hexadecimal, if base<33.
7733: @item
7734: @code{'} -- numeric value (e.g., ASCII code) of next character; an
7735: optional @code{'} may be present after the character.
1.26 crook 7736: @end itemize
1.1 anton 7737:
1.26 crook 7738: Here are some examples, with the equivalent decimal number shown after
7739: in braces:
1.1 anton 7740:
1.26 crook 7741: -$41 (-65), %1001101 (205), %1001.0001 (145 - a double-precision number),
1.133 anton 7742: 'A (65),
7743: -'a' (-97),
1.26 crook 7744: &905 (905), $abc (2478), $ABC (2478).
1.1 anton 7745:
1.26 crook 7746: @cindex number conversion - traps for the unwary
1.29 crook 7747: @noindent
1.26 crook 7748: Number conversion has a number of traps for the unwary:
1.1 anton 7749:
1.26 crook 7750: @itemize @bullet
7751: @item
7752: You cannot determine the current number base using the code sequence
1.35 anton 7753: @code{base @@ .} -- the number base is always 10 in the current number
7754: base. Instead, use something like @code{base @@ dec.}
1.26 crook 7755: @item
7756: If the number base is set to a value greater than 14 (for example,
7757: hexadecimal), the number 123E4 is ambiguous; the conversion rules allow
7758: it to be intepreted as either a single-precision integer or a
7759: floating-point number (Gforth treats it as an integer). The ambiguity
7760: can be resolved by explicitly stating the sign of the mantissa and/or
7761: exponent: 123E+4 or +123E4 -- if the number base is decimal, no
7762: ambiguity arises; either representation will be treated as a
7763: floating-point number.
7764: @item
1.29 crook 7765: There is a word @code{bin} but it does @i{not} set the number base!
1.26 crook 7766: It is used to specify file types.
7767: @item
1.72 anton 7768: ANS Forth requires the @code{.} of a double-precision number to be the
7769: final character in the string. Gforth allows the @code{.} to be
7770: anywhere after the first digit.
1.26 crook 7771: @item
7772: The number conversion process does not check for overflow.
7773: @item
1.72 anton 7774: In an ANS Forth program @code{base} is required to be decimal when
7775: converting floating-point numbers. In Gforth, number conversion to
7776: floating-point numbers always uses base &10, irrespective of the value
7777: of @code{base}.
1.26 crook 7778: @end itemize
1.1 anton 7779:
1.49 anton 7780: You can read numbers into your programs with the words described in
7781: @ref{Input}.
1.1 anton 7782:
1.82 anton 7783: @node Interpret/Compile states, Interpreter Directives, Number Conversion, The Text Interpreter
1.26 crook 7784: @subsection Interpret/Compile states
7785: @cindex Interpret/Compile states
1.1 anton 7786:
1.29 crook 7787: A standard program is not permitted to change @code{state}
7788: explicitly. However, it can change @code{state} implicitly, using the
7789: words @code{[} and @code{]}. When @code{[} is executed it switches
7790: @code{state} to interpret state, and therefore the text interpreter
7791: starts interpreting. When @code{]} is executed it switches @code{state}
7792: to compile state and therefore the text interpreter starts
1.44 crook 7793: compiling. The most common usage for these words is for switching into
7794: interpret state and back from within a colon definition; this technique
1.49 anton 7795: can be used to compile a literal (for an example, @pxref{Literals}) or
7796: for conditional compilation (for an example, @pxref{Interpreter
7797: Directives}).
1.44 crook 7798:
1.35 anton 7799:
7800: @c This is a bad example: It's non-standard, and it's not necessary.
7801: @c However, I can't think of a good example for switching into compile
7802: @c state when there is no current word (@code{state}-smart words are not a
7803: @c good reason). So maybe we should use an example for switching into
7804: @c interpret @code{state} in a colon def. - anton
1.44 crook 7805: @c nac-> I agree. I started out by putting in the example, then realised
7806: @c that it was non-ANS, so wrote more words around it. I hope this
7807: @c re-written version is acceptable to you. I do want to keep the example
7808: @c as it is helpful for showing what is and what is not portable, particularly
7809: @c where it outlaws a style in common use.
7810:
1.72 anton 7811: @c anton: it's more important to show what's portable. After we have done
1.83 anton 7812: @c that, we can also show what's not. In any case, I have written a
7813: @c section Compiling Words which also deals with [ ].
1.35 anton 7814:
1.95 anton 7815: @c !! The following example does not work in Gforth 0.5.9 or later.
1.29 crook 7816:
1.95 anton 7817: @c @code{[} and @code{]} also give you the ability to switch into compile
7818: @c state and back, but we cannot think of any useful Standard application
7819: @c for this ability. Pre-ANS Forth textbooks have examples like this:
7820:
7821: @c @example
7822: @c : AA ." this is A" ;
7823: @c : BB ." this is B" ;
7824: @c : CC ." this is C" ;
7825:
7826: @c create table ] aa bb cc [
7827:
7828: @c : go ( n -- ) \ n is offset into table.. 0 for 1st entry
7829: @c cells table + @@ execute ;
7830: @c @end example
7831:
7832: @c This example builds a jump table; @code{0 go} will display ``@code{this
7833: @c is A}''. Using @code{[} and @code{]} in this example is equivalent to
7834: @c defining @code{table} like this:
7835:
7836: @c @example
7837: @c create table ' aa COMPILE, ' bb COMPILE, ' cc COMPILE,
7838: @c @end example
7839:
7840: @c The problem with this code is that the definition of @code{table} is not
7841: @c portable -- it @i{compile}s execution tokens into code space. Whilst it
7842: @c @i{may} work on systems where code space and data space co-incide, the
7843: @c Standard only allows data space to be assigned for a @code{CREATE}d
7844: @c word. In addition, the Standard only allows @code{@@} to access data
7845: @c space, whilst this example is using it to access code space. The only
7846: @c portable, Standard way to build this table is to build it in data space,
7847: @c like this:
7848:
7849: @c @example
7850: @c create table ' aa , ' bb , ' cc ,
7851: @c @end example
1.29 crook 7852:
1.95 anton 7853: @c doc-state
1.44 crook 7854:
1.29 crook 7855:
1.82 anton 7856: @node Interpreter Directives, , Interpret/Compile states, The Text Interpreter
1.26 crook 7857: @subsection Interpreter Directives
7858: @cindex interpreter directives
1.72 anton 7859: @cindex conditional compilation
1.1 anton 7860:
1.29 crook 7861: These words are usually used in interpret state; typically to control
7862: which parts of a source file are processed by the text
1.26 crook 7863: interpreter. There are only a few ANS Forth Standard words, but Gforth
7864: supplements these with a rich set of immediate control structure words
7865: to compensate for the fact that the non-immediate versions can only be
1.29 crook 7866: used in compile state (@pxref{Control Structures}). Typical usages:
7867:
7868: @example
1.72 anton 7869: FALSE Constant HAVE-ASSEMBLER
1.29 crook 7870: .
7871: .
1.72 anton 7872: HAVE-ASSEMBLER [IF]
1.29 crook 7873: : ASSEMBLER-FEATURE
7874: ...
7875: ;
7876: [ENDIF]
7877: .
7878: .
7879: : SEE
7880: ... \ general-purpose SEE code
1.72 anton 7881: [ HAVE-ASSEMBLER [IF] ]
1.29 crook 7882: ... \ assembler-specific SEE code
7883: [ [ENDIF] ]
7884: ;
7885: @end example
1.1 anton 7886:
1.44 crook 7887:
1.26 crook 7888: doc-[IF]
7889: doc-[ELSE]
7890: doc-[THEN]
7891: doc-[ENDIF]
1.1 anton 7892:
1.26 crook 7893: doc-[IFDEF]
7894: doc-[IFUNDEF]
1.1 anton 7895:
1.26 crook 7896: doc-[?DO]
7897: doc-[DO]
7898: doc-[FOR]
7899: doc-[LOOP]
7900: doc-[+LOOP]
7901: doc-[NEXT]
1.1 anton 7902:
1.26 crook 7903: doc-[BEGIN]
7904: doc-[UNTIL]
7905: doc-[AGAIN]
7906: doc-[WHILE]
7907: doc-[REPEAT]
1.1 anton 7908:
1.27 crook 7909:
1.26 crook 7910: @c -------------------------------------------------------------
1.111 anton 7911: @node The Input Stream, Word Lists, The Text Interpreter, Words
7912: @section The Input Stream
7913: @cindex input stream
7914:
7915: @c !! integrate this better with the "Text Interpreter" section
7916: The text interpreter reads from the input stream, which can come from
7917: several sources (@pxref{Input Sources}). Some words, in particular
7918: defining words, but also words like @code{'}, read parameters from the
7919: input stream instead of from the stack.
7920:
7921: Such words are called parsing words, because they parse the input
7922: stream. Parsing words are hard to use in other words, because it is
7923: hard to pass program-generated parameters through the input stream.
7924: They also usually have an unintuitive combination of interpretation and
7925: compilation semantics when implemented naively, leading to various
7926: approaches that try to produce a more intuitive behaviour
7927: (@pxref{Combined words}).
7928:
7929: It should be obvious by now that parsing words are a bad idea. If you
7930: want to implement a parsing word for convenience, also provide a factor
7931: of the word that does not parse, but takes the parameters on the stack.
7932: To implement the parsing word on top if it, you can use the following
7933: words:
7934:
7935: @c anton: these belong in the input stream section
7936: doc-parse
1.138 anton 7937: doc-parse-name
1.111 anton 7938: doc-parse-word
7939: doc-name
7940: doc-word
7941: doc-\"-parse
7942: doc-refill
7943:
7944: Conversely, if you have the bad luck (or lack of foresight) to have to
7945: deal with parsing words without having such factors, how do you pass a
7946: string that is not in the input stream to it?
7947:
7948: doc-execute-parsing
7949:
1.146 anton 7950: A definition of this word in ANS Forth is provided in
7951: @file{compat/execute-parsing.fs}.
7952:
1.111 anton 7953: If you want to run a parsing word on a file, the following word should
7954: help:
7955:
7956: doc-execute-parsing-file
7957:
7958: @c -------------------------------------------------------------
7959: @node Word Lists, Environmental Queries, The Input Stream, Words
1.26 crook 7960: @section Word Lists
7961: @cindex word lists
1.32 anton 7962: @cindex header space
1.1 anton 7963:
1.36 anton 7964: A wordlist is a list of named words; you can add new words and look up
7965: words by name (and you can remove words in a restricted way with
7966: markers). Every named (and @code{reveal}ed) word is in one wordlist.
7967:
7968: @cindex search order stack
7969: The text interpreter searches the wordlists present in the search order
7970: (a stack of wordlists), from the top to the bottom. Within each
7971: wordlist, the search starts conceptually at the newest word; i.e., if
7972: two words in a wordlist have the same name, the newer word is found.
1.1 anton 7973:
1.26 crook 7974: @cindex compilation word list
1.36 anton 7975: New words are added to the @dfn{compilation wordlist} (aka current
7976: wordlist).
1.1 anton 7977:
1.36 anton 7978: @cindex wid
7979: A word list is identified by a cell-sized word list identifier (@i{wid})
7980: in much the same way as a file is identified by a file handle. The
7981: numerical value of the wid has no (portable) meaning, and might change
7982: from session to session.
1.1 anton 7983:
1.29 crook 7984: The ANS Forth ``Search order'' word set is intended to provide a set of
7985: low-level tools that allow various different schemes to be
1.74 anton 7986: implemented. Gforth also provides @code{vocabulary}, a traditional Forth
1.26 crook 7987: word. @file{compat/vocabulary.fs} provides an implementation in ANS
1.45 crook 7988: Forth.
1.1 anton 7989:
1.27 crook 7990: @comment TODO: locals section refers to here, saying that every word list (aka
7991: @comment vocabulary) has its own methods for searching etc. Need to document that.
1.78 anton 7992: @c anton: but better in a separate subsection on wordlist internals
1.1 anton 7993:
1.45 crook 7994: @comment TODO: document markers, reveal, tables, mappedwordlist
7995:
7996: @comment the gforthman- prefix is used to pick out the true definition of a
1.27 crook 7997: @comment word from the source files, rather than some alias.
1.44 crook 7998:
1.26 crook 7999: doc-forth-wordlist
8000: doc-definitions
8001: doc-get-current
8002: doc-set-current
8003: doc-get-order
1.45 crook 8004: doc---gforthman-set-order
1.26 crook 8005: doc-wordlist
1.30 anton 8006: doc-table
1.79 anton 8007: doc->order
1.36 anton 8008: doc-previous
1.26 crook 8009: doc-also
1.45 crook 8010: doc---gforthman-forth
1.26 crook 8011: doc-only
1.45 crook 8012: doc---gforthman-order
1.15 anton 8013:
1.26 crook 8014: doc-find
8015: doc-search-wordlist
1.15 anton 8016:
1.26 crook 8017: doc-words
8018: doc-vlist
1.44 crook 8019: @c doc-words-deferred
1.1 anton 8020:
1.74 anton 8021: @c doc-mappedwordlist @c map-structure undefined, implemantation-specific
1.26 crook 8022: doc-root
8023: doc-vocabulary
8024: doc-seal
8025: doc-vocs
8026: doc-current
8027: doc-context
1.1 anton 8028:
1.44 crook 8029:
1.26 crook 8030: @menu
1.75 anton 8031: * Vocabularies::
1.67 anton 8032: * Why use word lists?::
1.75 anton 8033: * Word list example::
1.26 crook 8034: @end menu
8035:
1.75 anton 8036: @node Vocabularies, Why use word lists?, Word Lists, Word Lists
8037: @subsection Vocabularies
8038: @cindex Vocabularies, detailed explanation
8039:
8040: Here is an example of creating and using a new wordlist using ANS
8041: Forth words:
8042:
8043: @example
8044: wordlist constant my-new-words-wordlist
8045: : my-new-words get-order nip my-new-words-wordlist swap set-order ;
8046:
8047: \ add it to the search order
8048: also my-new-words
8049:
8050: \ alternatively, add it to the search order and make it
8051: \ the compilation word list
8052: also my-new-words definitions
8053: \ type "order" to see the problem
8054: @end example
8055:
8056: The problem with this example is that @code{order} has no way to
8057: associate the name @code{my-new-words} with the wid of the word list (in
8058: Gforth, @code{order} and @code{vocs} will display @code{???} for a wid
8059: that has no associated name). There is no Standard way of associating a
8060: name with a wid.
8061:
8062: In Gforth, this example can be re-coded using @code{vocabulary}, which
8063: associates a name with a wid:
8064:
8065: @example
8066: vocabulary my-new-words
8067:
8068: \ add it to the search order
8069: also my-new-words
8070:
8071: \ alternatively, add it to the search order and make it
8072: \ the compilation word list
8073: my-new-words definitions
8074: \ type "order" to see that the problem is solved
8075: @end example
8076:
8077:
8078: @node Why use word lists?, Word list example, Vocabularies, Word Lists
1.26 crook 8079: @subsection Why use word lists?
8080: @cindex word lists - why use them?
8081:
1.74 anton 8082: Here are some reasons why people use wordlists:
1.26 crook 8083:
8084: @itemize @bullet
1.74 anton 8085:
8086: @c anton: Gforth's hashing implementation makes the search speed
8087: @c independent from the number of words. But it is linear with the number
8088: @c of wordlists that have to be searched, so in effect using more wordlists
8089: @c actually slows down compilation.
8090:
8091: @c @item
8092: @c To improve compilation speed by reducing the number of header space
8093: @c entries that must be searched. This is achieved by creating a new
8094: @c word list that contains all of the definitions that are used in the
8095: @c definition of a Forth system but which would not usually be used by
8096: @c programs running on that system. That word list would be on the search
8097: @c list when the Forth system was compiled but would be removed from the
8098: @c search list for normal operation. This can be a useful technique for
8099: @c low-performance systems (for example, 8-bit processors in embedded
8100: @c systems) but is unlikely to be necessary in high-performance desktop
8101: @c systems.
8102:
1.26 crook 8103: @item
8104: To prevent a set of words from being used outside the context in which
8105: they are valid. Two classic examples of this are an integrated editor
8106: (all of the edit commands are defined in a separate word list; the
8107: search order is set to the editor word list when the editor is invoked;
8108: the old search order is restored when the editor is terminated) and an
8109: integrated assembler (the op-codes for the machine are defined in a
8110: separate word list which is used when a @code{CODE} word is defined).
1.74 anton 8111:
8112: @item
8113: To organize the words of an application or library into a user-visible
8114: set (in @code{forth-wordlist} or some other common wordlist) and a set
8115: of helper words used just for the implementation (hidden in a separate
1.75 anton 8116: wordlist). This keeps @code{words}' output smaller, separates
8117: implementation and interface, and reduces the chance of name conflicts
8118: within the common wordlist.
1.74 anton 8119:
1.26 crook 8120: @item
8121: To prevent a name-space clash between multiple definitions with the same
8122: name. For example, when building a cross-compiler you might have a word
8123: @code{IF} that generates conditional code for your target system. By
8124: placing this definition in a different word list you can control whether
8125: the host system's @code{IF} or the target system's @code{IF} get used in
8126: any particular context by controlling the order of the word lists on the
8127: search order stack.
1.74 anton 8128:
1.26 crook 8129: @end itemize
1.1 anton 8130:
1.74 anton 8131: The downsides of using wordlists are:
8132:
8133: @itemize
8134:
8135: @item
8136: Debugging becomes more cumbersome.
8137:
8138: @item
8139: Name conflicts worked around with wordlists are still there, and you
8140: have to arrange the search order carefully to get the desired results;
8141: if you forget to do that, you get hard-to-find errors (as in any case
8142: where you read the code differently from the compiler; @code{see} can
1.75 anton 8143: help seeing which of several possible words the name resolves to in such
8144: cases). @code{See} displays just the name of the words, not what
8145: wordlist they belong to, so it might be misleading. Using unique names
8146: is a better approach to avoid name conflicts.
1.74 anton 8147:
8148: @item
8149: You have to explicitly undo any changes to the search order. In many
8150: cases it would be more convenient if this happened implicitly. Gforth
8151: currently does not provide such a feature, but it may do so in the
8152: future.
8153: @end itemize
8154:
8155:
1.75 anton 8156: @node Word list example, , Why use word lists?, Word Lists
8157: @subsection Word list example
8158: @cindex word lists - example
1.1 anton 8159:
1.74 anton 8160: The following example is from the
8161: @uref{http://www.complang.tuwien.ac.at/forth/garbage-collection.zip,
8162: garbage collector} and uses wordlists to separate public words from
8163: helper words:
8164:
8165: @example
8166: get-current ( wid )
8167: vocabulary garbage-collector also garbage-collector definitions
8168: ... \ define helper words
8169: ( wid ) set-current \ restore original (i.e., public) compilation wordlist
8170: ... \ define the public (i.e., API) words
8171: \ they can refer to the helper words
8172: previous \ restore original search order (helper words become invisible)
8173: @end example
8174:
1.26 crook 8175: @c -------------------------------------------------------------
8176: @node Environmental Queries, Files, Word Lists, Words
8177: @section Environmental Queries
8178: @cindex environmental queries
1.21 crook 8179:
1.26 crook 8180: ANS Forth introduced the idea of ``environmental queries'' as a way
8181: for a program running on a system to determine certain characteristics of the system.
8182: The Standard specifies a number of strings that might be recognised by a system.
1.21 crook 8183:
1.32 anton 8184: The Standard requires that the header space used for environmental queries
8185: be distinct from the header space used for definitions.
1.21 crook 8186:
1.26 crook 8187: Typically, environmental queries are supported by creating a set of
1.29 crook 8188: definitions in a word list that is @i{only} used during environmental
1.26 crook 8189: queries; that is what Gforth does. There is no Standard way of adding
8190: definitions to the set of recognised environmental queries, but any
8191: implementation that supports the loading of optional word sets must have
8192: some mechanism for doing this (after loading the word set, the
8193: associated environmental query string must return @code{true}). In
8194: Gforth, the word list used to honour environmental queries can be
8195: manipulated just like any other word list.
1.21 crook 8196:
1.44 crook 8197:
1.26 crook 8198: doc-environment?
8199: doc-environment-wordlist
1.21 crook 8200:
1.26 crook 8201: doc-gforth
8202: doc-os-class
1.21 crook 8203:
1.44 crook 8204:
1.26 crook 8205: Note that, whilst the documentation for (e.g.) @code{gforth} shows it
8206: returning two items on the stack, querying it using @code{environment?}
8207: will return an additional item; the @code{true} flag that shows that the
8208: string was recognised.
1.21 crook 8209:
1.26 crook 8210: @comment TODO Document the standard strings or note where they are documented herein
1.21 crook 8211:
1.26 crook 8212: Here are some examples of using environmental queries:
1.21 crook 8213:
1.26 crook 8214: @example
8215: s" address-unit-bits" environment? 0=
8216: [IF]
8217: cr .( environmental attribute address-units-bits unknown... ) cr
1.75 anton 8218: [ELSE]
8219: drop \ ensure balanced stack effect
1.26 crook 8220: [THEN]
1.21 crook 8221:
1.75 anton 8222: \ this might occur in the prelude of a standard program that uses THROW
8223: s" exception" environment? [IF]
8224: 0= [IF]
8225: : throw abort" exception thrown" ;
8226: [THEN]
8227: [ELSE] \ we don't know, so make sure
8228: : throw abort" exception thrown" ;
8229: [THEN]
1.21 crook 8230:
1.26 crook 8231: s" gforth" environment? [IF] .( Gforth version ) TYPE
8232: [ELSE] .( Not Gforth..) [THEN]
1.75 anton 8233:
8234: \ a program using v*
8235: s" gforth" environment? [IF]
8236: s" 0.5.0" compare 0< [IF] \ v* is a primitive since 0.5.0
8237: : v* ( f_addr1 nstride1 f_addr2 nstride2 ucount -- r )
8238: >r swap 2swap swap 0e r> 0 ?DO
8239: dup f@ over + 2swap dup f@ f* f+ over + 2swap
8240: LOOP
8241: 2drop 2drop ;
8242: [THEN]
8243: [ELSE] \
8244: : v* ( f_addr1 nstride1 f_addr2 nstride2 ucount -- r )
8245: ...
8246: [THEN]
1.26 crook 8247: @end example
1.21 crook 8248:
1.26 crook 8249: Here is an example of adding a definition to the environment word list:
1.21 crook 8250:
1.26 crook 8251: @example
8252: get-current environment-wordlist set-current
8253: true constant block
8254: true constant block-ext
8255: set-current
8256: @end example
1.21 crook 8257:
1.26 crook 8258: You can see what definitions are in the environment word list like this:
1.21 crook 8259:
1.26 crook 8260: @example
1.79 anton 8261: environment-wordlist >order words previous
1.26 crook 8262: @end example
1.21 crook 8263:
8264:
1.26 crook 8265: @c -------------------------------------------------------------
8266: @node Files, Blocks, Environmental Queries, Words
8267: @section Files
1.28 crook 8268: @cindex files
8269: @cindex I/O - file-handling
1.21 crook 8270:
1.26 crook 8271: Gforth provides facilities for accessing files that are stored in the
8272: host operating system's file-system. Files that are processed by Gforth
8273: can be divided into two categories:
1.21 crook 8274:
1.23 crook 8275: @itemize @bullet
8276: @item
1.29 crook 8277: Files that are processed by the Text Interpreter (@dfn{Forth source files}).
1.23 crook 8278: @item
1.29 crook 8279: Files that are processed by some other program (@dfn{general files}).
1.26 crook 8280: @end itemize
8281:
8282: @menu
1.48 anton 8283: * Forth source files::
8284: * General files::
1.167 anton 8285: * Redirection::
1.48 anton 8286: * Search Paths::
1.26 crook 8287: @end menu
8288:
8289: @c -------------------------------------------------------------
8290: @node Forth source files, General files, Files, Files
8291: @subsection Forth source files
8292: @cindex including files
8293: @cindex Forth source files
1.21 crook 8294:
1.26 crook 8295: The simplest way to interpret the contents of a file is to use one of
8296: these two formats:
1.21 crook 8297:
1.26 crook 8298: @example
8299: include mysource.fs
8300: s" mysource.fs" included
8301: @end example
1.21 crook 8302:
1.75 anton 8303: You usually want to include a file only if it is not included already
1.26 crook 8304: (by, say, another source file). In that case, you can use one of these
1.45 crook 8305: three formats:
1.21 crook 8306:
1.26 crook 8307: @example
8308: require mysource.fs
8309: needs mysource.fs
8310: s" mysource.fs" required
8311: @end example
1.21 crook 8312:
1.26 crook 8313: @cindex stack effect of included files
8314: @cindex including files, stack effect
1.45 crook 8315: It is good practice to write your source files such that interpreting them
8316: does not change the stack. Source files designed in this way can be used with
1.26 crook 8317: @code{required} and friends without complications. For example:
1.21 crook 8318:
1.26 crook 8319: @example
1.75 anton 8320: 1024 require foo.fs drop
1.26 crook 8321: @end example
1.21 crook 8322:
1.75 anton 8323: Here you want to pass the argument 1024 (e.g., a buffer size) to
8324: @file{foo.fs}. Interpreting @file{foo.fs} has the stack effect ( n -- n
8325: ), which allows its use with @code{require}. Of course with such
8326: parameters to required files, you have to ensure that the first
8327: @code{require} fits for all uses (i.e., @code{require} it early in the
8328: master load file).
1.44 crook 8329:
1.26 crook 8330: doc-include-file
8331: doc-included
1.28 crook 8332: doc-included?
1.26 crook 8333: doc-include
8334: doc-required
8335: doc-require
8336: doc-needs
1.75 anton 8337: @c doc-init-included-files @c internal
8338: doc-sourcefilename
8339: doc-sourceline#
1.44 crook 8340:
1.26 crook 8341: A definition in ANS Forth for @code{required} is provided in
8342: @file{compat/required.fs}.
1.21 crook 8343:
1.26 crook 8344: @c -------------------------------------------------------------
1.167 anton 8345: @node General files, Redirection, Forth source files, Files
1.26 crook 8346: @subsection General files
8347: @cindex general files
8348: @cindex file-handling
1.21 crook 8349:
1.75 anton 8350: Files are opened/created by name and type. The following file access
8351: methods (FAMs) are recognised:
1.44 crook 8352:
1.75 anton 8353: @cindex fam (file access method)
1.26 crook 8354: doc-r/o
8355: doc-r/w
8356: doc-w/o
8357: doc-bin
1.1 anton 8358:
1.44 crook 8359:
1.26 crook 8360: When a file is opened/created, it returns a file identifier,
1.29 crook 8361: @i{wfileid} that is used for all other file commands. All file
8362: commands also return a status value, @i{wior}, that is 0 for a
1.26 crook 8363: successful operation and an implementation-defined non-zero value in the
8364: case of an error.
1.21 crook 8365:
1.44 crook 8366:
1.26 crook 8367: doc-open-file
8368: doc-create-file
1.21 crook 8369:
1.26 crook 8370: doc-close-file
8371: doc-delete-file
8372: doc-rename-file
8373: doc-read-file
8374: doc-read-line
1.154 anton 8375: doc-key-file
8376: doc-key?-file
1.26 crook 8377: doc-write-file
8378: doc-write-line
8379: doc-emit-file
8380: doc-flush-file
1.21 crook 8381:
1.26 crook 8382: doc-file-status
8383: doc-file-position
8384: doc-reposition-file
8385: doc-file-size
8386: doc-resize-file
1.21 crook 8387:
1.93 anton 8388: doc-slurp-file
8389: doc-slurp-fid
1.112 anton 8390: doc-stdin
8391: doc-stdout
8392: doc-stderr
1.44 crook 8393:
1.26 crook 8394: @c ---------------------------------------------------------
1.167 anton 8395: @node Redirection, Search Paths, General files, Files
8396: @subsection Redirection
8397: @cindex Redirection
8398: @cindex Input Redirection
8399: @cindex Output Redirection
8400:
8401: You can redirect the output of @code{type} and @code{emit} and all the
8402: words that use them (all output words that don't have an explicit
8403: target file) to an arbitrary file with the @code{>outfile
8404: ... outfile<} construct, used like this:
8405:
8406: @example
8407: : print-some-warning ( n -- )
8408: stderr >outfile cr ." warning# " . outfile< ;
8409: @end example
8410:
8411: After the @code{outfile<}, the original output direction is restored;
8412: this construct is nestable and safe against exceptions. Similarly,
8413: there is a construct @code{>infile ... infile<} for redirecting the
8414: input of @code{key} and its users (any input word that does not take a
8415: file explicitly).
8416:
8417: If you do not want to redirect the input or output to a file, you can
8418: also make use of the fact that @code{key}, @code{emit} and @code{type}
8419: are deferred words (@pxref{Deferred Words}). However, in that case
8420: you have to worry about the restoration and the protection against
8421: exceptions yourself; also, note that for redirecting the output in
8422: this way, you have to redirect both @code{emit} and @code{type}.
8423:
8424: doc->outfile
8425: doc-outfile<
8426: doc->infile
8427: doc-infile<
8428:
8429: @c ---------------------------------------------------------
8430: @node Search Paths, , Redirection, Files
1.26 crook 8431: @subsection Search Paths
8432: @cindex path for @code{included}
8433: @cindex file search path
8434: @cindex @code{include} search path
8435: @cindex search path for files
1.21 crook 8436:
1.26 crook 8437: If you specify an absolute filename (i.e., a filename starting with
8438: @file{/} or @file{~}, or with @file{:} in the second position (as in
8439: @samp{C:...})) for @code{included} and friends, that file is included
8440: just as you would expect.
1.21 crook 8441:
1.75 anton 8442: If the filename starts with @file{./}, this refers to the directory that
8443: the present file was @code{included} from. This allows files to include
8444: other files relative to their own position (irrespective of the current
8445: working directory or the absolute position). This feature is essential
8446: for libraries consisting of several files, where a file may include
8447: other files from the library. It corresponds to @code{#include "..."}
8448: in C. If the current input source is not a file, @file{.} refers to the
8449: directory of the innermost file being included, or, if there is no file
8450: being included, to the current working directory.
8451:
8452: For relative filenames (not starting with @file{./}), Gforth uses a
8453: search path similar to Forth's search order (@pxref{Word Lists}). It
8454: tries to find the given filename in the directories present in the path,
8455: and includes the first one it finds. There are separate search paths for
8456: Forth source files and general files. If the search path contains the
8457: directory @file{.}, this refers to the directory of the current file, or
8458: the working directory, as if the file had been specified with @file{./}.
1.21 crook 8459:
1.26 crook 8460: Use @file{~+} to refer to the current working directory (as in the
8461: @code{bash}).
1.1 anton 8462:
1.75 anton 8463: @c anton: fold the following subsubsections into this subsection?
1.1 anton 8464:
1.48 anton 8465: @menu
1.75 anton 8466: * Source Search Paths::
1.48 anton 8467: * General Search Paths::
8468: @end menu
8469:
1.26 crook 8470: @c ---------------------------------------------------------
1.75 anton 8471: @node Source Search Paths, General Search Paths, Search Paths, Search Paths
8472: @subsubsection Source Search Paths
8473: @cindex search path control, source files
1.5 anton 8474:
1.26 crook 8475: The search path is initialized when you start Gforth (@pxref{Invoking
1.75 anton 8476: Gforth}). You can display it and change it using @code{fpath} in
8477: combination with the general path handling words.
1.5 anton 8478:
1.75 anton 8479: doc-fpath
8480: @c the functionality of the following words is easily available through
8481: @c fpath and the general path words. The may go away.
8482: @c doc-.fpath
8483: @c doc-fpath+
8484: @c doc-fpath=
8485: @c doc-open-fpath-file
1.44 crook 8486:
8487: @noindent
1.26 crook 8488: Here is an example of using @code{fpath} and @code{require}:
1.5 anton 8489:
1.26 crook 8490: @example
1.75 anton 8491: fpath path= /usr/lib/forth/|./
1.26 crook 8492: require timer.fs
8493: @end example
1.5 anton 8494:
1.75 anton 8495:
1.26 crook 8496: @c ---------------------------------------------------------
1.75 anton 8497: @node General Search Paths, , Source Search Paths, Search Paths
1.26 crook 8498: @subsubsection General Search Paths
1.75 anton 8499: @cindex search path control, source files
1.5 anton 8500:
1.26 crook 8501: Your application may need to search files in several directories, like
8502: @code{included} does. To facilitate this, Gforth allows you to define
8503: and use your own search paths, by providing generic equivalents of the
8504: Forth search path words:
1.5 anton 8505:
1.75 anton 8506: doc-open-path-file
8507: doc-path-allot
8508: doc-clear-path
8509: doc-also-path
1.26 crook 8510: doc-.path
8511: doc-path+
8512: doc-path=
1.5 anton 8513:
1.75 anton 8514: @c anton: better define a word for it, say "path-allot ( ucount -- path-addr )
1.44 crook 8515:
1.75 anton 8516: Here's an example of creating an empty search path:
8517: @c
1.26 crook 8518: @example
1.75 anton 8519: create mypath 500 path-allot \ maximum length 500 chars (is checked)
1.26 crook 8520: @end example
1.5 anton 8521:
1.26 crook 8522: @c -------------------------------------------------------------
8523: @node Blocks, Other I/O, Files, Words
8524: @section Blocks
1.28 crook 8525: @cindex I/O - blocks
8526: @cindex blocks
8527:
8528: When you run Gforth on a modern desk-top computer, it runs under the
8529: control of an operating system which provides certain services. One of
8530: these services is @var{file services}, which allows Forth source code
8531: and data to be stored in files and read into Gforth (@pxref{Files}).
8532:
8533: Traditionally, Forth has been an important programming language on
8534: systems where it has interfaced directly to the underlying hardware with
8535: no intervening operating system. Forth provides a mechanism, called
1.29 crook 8536: @dfn{blocks}, for accessing mass storage on such systems.
1.28 crook 8537:
8538: A block is a 1024-byte data area, which can be used to hold data or
8539: Forth source code. No structure is imposed on the contents of the
8540: block. A block is identified by its number; blocks are numbered
8541: contiguously from 1 to an implementation-defined maximum.
8542:
8543: A typical system that used blocks but no operating system might use a
8544: single floppy-disk drive for mass storage, with the disks formatted to
8545: provide 256-byte sectors. Blocks would be implemented by assigning the
8546: first four sectors of the disk to block 1, the second four sectors to
8547: block 2 and so on, up to the limit of the capacity of the disk. The disk
8548: would not contain any file system information, just the set of blocks.
8549:
1.29 crook 8550: @cindex blocks file
1.28 crook 8551: On systems that do provide file services, blocks are typically
1.29 crook 8552: implemented by storing a sequence of blocks within a single @dfn{blocks
1.28 crook 8553: file}. The size of the blocks file will be an exact multiple of 1024
8554: bytes, corresponding to the number of blocks it contains. This is the
8555: mechanism that Gforth uses.
8556:
1.29 crook 8557: @cindex @file{blocks.fb}
1.75 anton 8558: Only one blocks file can be open at a time. If you use block words without
1.28 crook 8559: having specified a blocks file, Gforth defaults to the blocks file
8560: @file{blocks.fb}. Gforth uses the Forth search path when attempting to
1.75 anton 8561: locate a blocks file (@pxref{Source Search Paths}).
1.28 crook 8562:
1.29 crook 8563: @cindex block buffers
1.28 crook 8564: When you read and write blocks under program control, Gforth uses a
1.29 crook 8565: number of @dfn{block buffers} as intermediate storage. These buffers are
1.28 crook 8566: not used when you use @code{load} to interpret the contents of a block.
8567:
1.75 anton 8568: The behaviour of the block buffers is analagous to that of a cache.
8569: Each block buffer has three states:
1.28 crook 8570:
8571: @itemize @bullet
8572: @item
8573: Unassigned
8574: @item
8575: Assigned-clean
8576: @item
8577: Assigned-dirty
8578: @end itemize
8579:
1.29 crook 8580: Initially, all block buffers are @i{unassigned}. In order to access a
1.28 crook 8581: block, the block (specified by its block number) must be assigned to a
8582: block buffer.
8583:
8584: The assignment of a block to a block buffer is performed by @code{block}
8585: or @code{buffer}. Use @code{block} when you wish to modify the existing
8586: contents of a block. Use @code{buffer} when you don't care about the
8587: existing contents of the block@footnote{The ANS Forth definition of
1.35 anton 8588: @code{buffer} is intended not to cause disk I/O; if the data associated
1.28 crook 8589: with the particular block is already stored in a block buffer due to an
8590: earlier @code{block} command, @code{buffer} will return that block
8591: buffer and the existing contents of the block will be
8592: available. Otherwise, @code{buffer} will simply assign a new, empty
1.29 crook 8593: block buffer for the block.}.
1.28 crook 8594:
1.47 crook 8595: Once a block has been assigned to a block buffer using @code{block} or
1.75 anton 8596: @code{buffer}, that block buffer becomes the @i{current block
8597: buffer}. Data may only be manipulated (read or written) within the
8598: current block buffer.
1.47 crook 8599:
8600: When the contents of the current block buffer has been modified it is
1.48 anton 8601: necessary, @emph{before calling @code{block} or @code{buffer} again}, to
1.75 anton 8602: either abandon the changes (by doing nothing) or mark the block as
8603: changed (assigned-dirty), using @code{update}. Using @code{update} does
8604: not change the blocks file; it simply changes a block buffer's state to
8605: @i{assigned-dirty}. The block will be written implicitly when it's
8606: buffer is needed for another block, or explicitly by @code{flush} or
8607: @code{save-buffers}.
8608:
8609: word @code{Flush} writes all @i{assigned-dirty} blocks back to the
8610: blocks file on disk. Leaving Gforth with @code{bye} also performs a
8611: @code{flush}.
1.28 crook 8612:
1.29 crook 8613: In Gforth, @code{block} and @code{buffer} use a @i{direct-mapped}
1.28 crook 8614: algorithm to assign a block buffer to a block. That means that any
8615: particular block can only be assigned to one specific block buffer,
1.29 crook 8616: called (for the particular operation) the @i{victim buffer}. If the
1.47 crook 8617: victim buffer is @i{unassigned} or @i{assigned-clean} it is allocated to
8618: the new block immediately. If it is @i{assigned-dirty} its current
8619: contents are written back to the blocks file on disk before it is
1.28 crook 8620: allocated to the new block.
8621:
8622: Although no structure is imposed on the contents of a block, it is
8623: traditional to display the contents as 16 lines each of 64 characters. A
8624: block provides a single, continuous stream of input (for example, it
8625: acts as a single parse area) -- there are no end-of-line characters
8626: within a block, and no end-of-file character at the end of a
8627: block. There are two consequences of this:
1.26 crook 8628:
1.28 crook 8629: @itemize @bullet
8630: @item
8631: The last character of one line wraps straight into the first character
8632: of the following line
8633: @item
8634: The word @code{\} -- comment to end of line -- requires special
8635: treatment; in the context of a block it causes all characters until the
8636: end of the current 64-character ``line'' to be ignored.
8637: @end itemize
8638:
8639: In Gforth, when you use @code{block} with a non-existent block number,
1.45 crook 8640: the current blocks file will be extended to the appropriate size and the
1.28 crook 8641: block buffer will be initialised with spaces.
8642:
1.47 crook 8643: Gforth includes a simple block editor (type @code{use blocked.fb 0 list}
8644: for details) but doesn't encourage the use of blocks; the mechanism is
8645: only provided for backward compatibility -- ANS Forth requires blocks to
8646: be available when files are.
1.28 crook 8647:
8648: Common techniques that are used when working with blocks include:
8649:
8650: @itemize @bullet
8651: @item
8652: A screen editor that allows you to edit blocks without leaving the Forth
8653: environment.
8654: @item
8655: Shadow screens; where every code block has an associated block
8656: containing comments (for example: code in odd block numbers, comments in
8657: even block numbers). Typically, the block editor provides a convenient
8658: mechanism to toggle between code and comments.
8659: @item
8660: Load blocks; a single block (typically block 1) contains a number of
8661: @code{thru} commands which @code{load} the whole of the application.
8662: @end itemize
1.26 crook 8663:
1.29 crook 8664: See Frank Sergeant's Pygmy Forth to see just how well blocks can be
8665: integrated into a Forth programming environment.
1.26 crook 8666:
8667: @comment TODO what about errors on open-blocks?
1.44 crook 8668:
1.26 crook 8669: doc-open-blocks
8670: doc-use
1.75 anton 8671: doc-block-offset
1.26 crook 8672: doc-get-block-fid
8673: doc-block-position
1.28 crook 8674:
1.75 anton 8675: doc-list
1.28 crook 8676: doc-scr
8677:
1.45 crook 8678: doc---gforthman-block
1.28 crook 8679: doc-buffer
8680:
1.75 anton 8681: doc-empty-buffers
8682: doc-empty-buffer
1.26 crook 8683: doc-update
1.28 crook 8684: doc-updated?
1.26 crook 8685: doc-save-buffers
1.75 anton 8686: doc-save-buffer
1.26 crook 8687: doc-flush
1.28 crook 8688:
1.26 crook 8689: doc-load
8690: doc-thru
8691: doc-+load
8692: doc-+thru
1.45 crook 8693: doc---gforthman--->
1.26 crook 8694: doc-block-included
8695:
1.44 crook 8696:
1.26 crook 8697: @c -------------------------------------------------------------
1.126 pazsan 8698: @node Other I/O, OS command line arguments, Blocks, Words
1.26 crook 8699: @section Other I/O
1.28 crook 8700: @cindex I/O - keyboard and display
1.26 crook 8701:
8702: @menu
8703: * Simple numeric output:: Predefined formats
8704: * Formatted numeric output:: Formatted (pictured) output
8705: * String Formats:: How Forth stores strings in memory
1.67 anton 8706: * Displaying characters and strings:: Other stuff
1.26 crook 8707: * Input:: Input
1.112 anton 8708: * Pipes:: How to create your own pipes
1.149 pazsan 8709: * Xchars and Unicode:: Non-ASCII characters
1.26 crook 8710: @end menu
8711:
8712: @node Simple numeric output, Formatted numeric output, Other I/O, Other I/O
8713: @subsection Simple numeric output
1.28 crook 8714: @cindex numeric output - simple/free-format
1.5 anton 8715:
1.26 crook 8716: The simplest output functions are those that display numbers from the
8717: data or floating-point stacks. Floating-point output is always displayed
8718: using base 10. Numbers displayed from the data stack use the value stored
8719: in @code{base}.
1.5 anton 8720:
1.44 crook 8721:
1.26 crook 8722: doc-.
8723: doc-dec.
8724: doc-hex.
8725: doc-u.
8726: doc-.r
8727: doc-u.r
8728: doc-d.
8729: doc-ud.
8730: doc-d.r
8731: doc-ud.r
8732: doc-f.
8733: doc-fe.
8734: doc-fs.
1.111 anton 8735: doc-f.rdp
1.44 crook 8736:
1.26 crook 8737: Examples of printing the number 1234.5678E23 in the different floating-point output
8738: formats are shown below:
1.5 anton 8739:
8740: @example
1.26 crook 8741: f. 123456779999999000000000000.
8742: fe. 123.456779999999E24
8743: fs. 1.23456779999999E26
1.5 anton 8744: @end example
8745:
8746:
1.26 crook 8747: @node Formatted numeric output, String Formats, Simple numeric output, Other I/O
8748: @subsection Formatted numeric output
1.28 crook 8749: @cindex formatted numeric output
1.26 crook 8750: @cindex pictured numeric output
1.28 crook 8751: @cindex numeric output - formatted
1.26 crook 8752:
1.29 crook 8753: Forth traditionally uses a technique called @dfn{pictured numeric
1.26 crook 8754: output} for formatted printing of integers. In this technique, digits
8755: are extracted from the number (using the current output radix defined by
8756: @code{base}), converted to ASCII codes and appended to a string that is
8757: built in a scratch-pad area of memory (@pxref{core-idef,
8758: Implementation-defined options, Implementation-defined
8759: options}). Arbitrary characters can be appended to the string during the
8760: extraction process. The completed string is specified by an address
8761: and length and can be manipulated (@code{TYPE}ed, copied, modified)
8762: under program control.
1.5 anton 8763:
1.75 anton 8764: All of the integer output words described in the previous section
8765: (@pxref{Simple numeric output}) are implemented in Gforth using pictured
8766: numeric output.
1.5 anton 8767:
1.47 crook 8768: Three important things to remember about pictured numeric output:
1.5 anton 8769:
1.26 crook 8770: @itemize @bullet
8771: @item
1.28 crook 8772: It always operates on double-precision numbers; to display a
1.49 anton 8773: single-precision number, convert it first (for ways of doing this
8774: @pxref{Double precision}).
1.26 crook 8775: @item
1.28 crook 8776: It always treats the double-precision number as though it were
8777: unsigned. The examples below show ways of printing signed numbers.
1.26 crook 8778: @item
8779: The string is built up from right to left; least significant digit first.
8780: @end itemize
1.5 anton 8781:
1.44 crook 8782:
1.26 crook 8783: doc-<#
1.47 crook 8784: doc-<<#
1.26 crook 8785: doc-#
8786: doc-#s
8787: doc-hold
8788: doc-sign
8789: doc-#>
1.47 crook 8790: doc-#>>
1.5 anton 8791:
1.26 crook 8792: doc-represent
1.111 anton 8793: doc-f>str-rdp
8794: doc-f>buf-rdp
1.5 anton 8795:
1.44 crook 8796:
8797: @noindent
1.26 crook 8798: Here are some examples of using pictured numeric output:
1.5 anton 8799:
8800: @example
1.26 crook 8801: : my-u. ( u -- )
8802: \ Simplest use of pns.. behaves like Standard u.
8803: 0 \ convert to unsigned double
1.75 anton 8804: <<# \ start conversion
1.26 crook 8805: #s \ convert all digits
8806: #> \ complete conversion
1.75 anton 8807: TYPE SPACE \ display, with trailing space
8808: #>> ; \ release hold area
1.5 anton 8809:
1.26 crook 8810: : cents-only ( u -- )
8811: 0 \ convert to unsigned double
1.75 anton 8812: <<# \ start conversion
1.26 crook 8813: # # \ convert two least-significant digits
8814: #> \ complete conversion, discard other digits
1.75 anton 8815: TYPE SPACE \ display, with trailing space
8816: #>> ; \ release hold area
1.5 anton 8817:
1.26 crook 8818: : dollars-and-cents ( u -- )
8819: 0 \ convert to unsigned double
1.75 anton 8820: <<# \ start conversion
1.26 crook 8821: # # \ convert two least-significant digits
8822: [char] . hold \ insert decimal point
8823: #s \ convert remaining digits
8824: [char] $ hold \ append currency symbol
8825: #> \ complete conversion
1.75 anton 8826: TYPE SPACE \ display, with trailing space
8827: #>> ; \ release hold area
1.5 anton 8828:
1.26 crook 8829: : my-. ( n -- )
8830: \ handling negatives.. behaves like Standard .
8831: s>d \ convert to signed double
8832: swap over dabs \ leave sign byte followed by unsigned double
1.75 anton 8833: <<# \ start conversion
1.26 crook 8834: #s \ convert all digits
8835: rot sign \ get at sign byte, append "-" if needed
8836: #> \ complete conversion
1.75 anton 8837: TYPE SPACE \ display, with trailing space
8838: #>> ; \ release hold area
1.5 anton 8839:
1.26 crook 8840: : account. ( n -- )
1.75 anton 8841: \ accountants don't like minus signs, they use parentheses
1.26 crook 8842: \ for negative numbers
8843: s>d \ convert to signed double
8844: swap over dabs \ leave sign byte followed by unsigned double
1.75 anton 8845: <<# \ start conversion
1.26 crook 8846: 2 pick \ get copy of sign byte
8847: 0< IF [char] ) hold THEN \ right-most character of output
8848: #s \ convert all digits
8849: rot \ get at sign byte
8850: 0< IF [char] ( hold THEN
8851: #> \ complete conversion
1.75 anton 8852: TYPE SPACE \ display, with trailing space
8853: #>> ; \ release hold area
8854:
1.5 anton 8855: @end example
8856:
1.26 crook 8857: Here are some examples of using these words:
1.5 anton 8858:
8859: @example
1.26 crook 8860: 1 my-u. 1
8861: hex -1 my-u. decimal FFFFFFFF
8862: 1 cents-only 01
8863: 1234 cents-only 34
8864: 2 dollars-and-cents $0.02
8865: 1234 dollars-and-cents $12.34
8866: 123 my-. 123
8867: -123 my. -123
8868: 123 account. 123
8869: -456 account. (456)
1.5 anton 8870: @end example
8871:
8872:
1.26 crook 8873: @node String Formats, Displaying characters and strings, Formatted numeric output, Other I/O
8874: @subsection String Formats
1.27 crook 8875: @cindex strings - see character strings
8876: @cindex character strings - formats
1.28 crook 8877: @cindex I/O - see character strings
1.75 anton 8878: @cindex counted strings
8879:
8880: @c anton: this does not really belong here; maybe the memory section,
8881: @c or the principles chapter
1.26 crook 8882:
1.27 crook 8883: Forth commonly uses two different methods for representing character
8884: strings:
1.26 crook 8885:
8886: @itemize @bullet
8887: @item
8888: @cindex address of counted string
1.45 crook 8889: @cindex counted string
1.29 crook 8890: As a @dfn{counted string}, represented by a @i{c-addr}. The char
8891: addressed by @i{c-addr} contains a character-count, @i{n}, of the
8892: string and the string occupies the subsequent @i{n} char addresses in
1.26 crook 8893: memory.
8894: @item
1.29 crook 8895: As cell pair on the stack; @i{c-addr u}, where @i{u} is the length
8896: of the string in characters, and @i{c-addr} is the address of the
1.26 crook 8897: first byte of the string.
8898: @end itemize
8899:
8900: ANS Forth encourages the use of the second format when representing
1.75 anton 8901: strings.
1.26 crook 8902:
1.44 crook 8903:
1.26 crook 8904: doc-count
8905:
1.44 crook 8906:
1.49 anton 8907: For words that move, copy and search for strings see @ref{Memory
8908: Blocks}. For words that display characters and strings see
8909: @ref{Displaying characters and strings}.
1.26 crook 8910:
8911: @node Displaying characters and strings, Input, String Formats, Other I/O
8912: @subsection Displaying characters and strings
1.27 crook 8913: @cindex characters - compiling and displaying
8914: @cindex character strings - compiling and displaying
1.26 crook 8915:
8916: This section starts with a glossary of Forth words and ends with a set
8917: of examples.
8918:
1.44 crook 8919:
1.26 crook 8920: doc-bl
8921: doc-space
8922: doc-spaces
8923: doc-emit
8924: doc-toupper
8925: doc-."
8926: doc-.(
1.98 anton 8927: doc-.\"
1.26 crook 8928: doc-type
1.44 crook 8929: doc-typewhite
1.26 crook 8930: doc-cr
1.27 crook 8931: @cindex cursor control
1.26 crook 8932: doc-at-xy
8933: doc-page
8934: doc-s"
1.98 anton 8935: doc-s\"
1.26 crook 8936: doc-c"
8937: doc-char
8938: doc-[char]
8939:
1.44 crook 8940:
8941: @noindent
1.26 crook 8942: As an example, consider the following text, stored in a file @file{test.fs}:
1.5 anton 8943:
8944: @example
1.26 crook 8945: .( text-1)
8946: : my-word
8947: ." text-2" cr
8948: .( text-3)
8949: ;
8950:
8951: ." text-4"
8952:
8953: : my-char
8954: [char] ALPHABET emit
8955: char emit
8956: ;
1.5 anton 8957: @end example
8958:
1.26 crook 8959: When you load this code into Gforth, the following output is generated:
1.5 anton 8960:
1.26 crook 8961: @example
1.30 anton 8962: @kbd{include test.fs @key{RET}} text-1text-3text-4 ok
1.26 crook 8963: @end example
1.5 anton 8964:
1.26 crook 8965: @itemize @bullet
8966: @item
8967: Messages @code{text-1} and @code{text-3} are displayed because @code{.(}
8968: is an immediate word; it behaves in the same way whether it is used inside
8969: or outside a colon definition.
8970: @item
8971: Message @code{text-4} is displayed because of Gforth's added interpretation
8972: semantics for @code{."}.
8973: @item
1.29 crook 8974: Message @code{text-2} is @i{not} displayed, because the text interpreter
1.26 crook 8975: performs the compilation semantics for @code{."} within the definition of
8976: @code{my-word}.
8977: @end itemize
1.5 anton 8978:
1.26 crook 8979: Here are some examples of executing @code{my-word} and @code{my-char}:
1.5 anton 8980:
1.26 crook 8981: @example
1.30 anton 8982: @kbd{my-word @key{RET}} text-2
1.26 crook 8983: ok
1.30 anton 8984: @kbd{my-char fred @key{RET}} Af ok
8985: @kbd{my-char jim @key{RET}} Aj ok
1.26 crook 8986: @end example
1.5 anton 8987:
8988: @itemize @bullet
8989: @item
1.26 crook 8990: Message @code{text-2} is displayed because of the run-time behaviour of
8991: @code{."}.
8992: @item
8993: @code{[char]} compiles the ``A'' from ``ALPHABET'' and puts its display code
8994: on the stack at run-time. @code{emit} always displays the character
8995: when @code{my-char} is executed.
8996: @item
8997: @code{char} parses a string at run-time and the second @code{emit} displays
8998: the first character of the string.
1.5 anton 8999: @item
1.26 crook 9000: If you type @code{see my-char} you can see that @code{[char]} discarded
9001: the text ``LPHABET'' and only compiled the display code for ``A'' into the
9002: definition of @code{my-char}.
1.5 anton 9003: @end itemize
9004:
9005:
9006:
1.112 anton 9007: @node Input, Pipes, Displaying characters and strings, Other I/O
1.26 crook 9008: @subsection Input
9009: @cindex input
1.28 crook 9010: @cindex I/O - see input
9011: @cindex parsing a string
1.5 anton 9012:
1.49 anton 9013: For ways of storing character strings in memory see @ref{String Formats}.
1.5 anton 9014:
1.27 crook 9015: @comment TODO examples for >number >float accept key key? pad parse word refill
1.29 crook 9016: @comment then index them
1.27 crook 9017:
1.44 crook 9018:
1.27 crook 9019: doc-key
9020: doc-key?
1.45 crook 9021: doc-ekey
1.141 anton 9022: doc-ekey>char
1.45 crook 9023: doc-ekey?
1.141 anton 9024:
9025: Gforth recognizes various keys available on ANSI terminals (in MS-DOS
9026: you need the ANSI.SYS driver to get that behaviour). These are the
9027: keyboard events produced by various common keys:
9028:
9029: doc-k-left
9030: doc-k-right
9031: doc-k-up
9032: doc-k-down
9033: doc-k-home
9034: doc-k-end
9035: doc-k-prior
9036: doc-k-next
9037: doc-k-insert
9038: doc-k-delete
9039:
9040: The function keys (aka keypad keys) are:
9041:
9042: doc-k1
9043: doc-k2
9044: doc-k3
9045: doc-k4
9046: doc-k5
9047: doc-k6
9048: doc-k7
9049: doc-k8
9050: doc-k9
9051: doc-k10
9052: doc-k11
9053: doc-k12
9054:
9055: Note that K11 and K12 are not as widely available. The shifted
9056: function keys are also not very widely available:
9057:
9058: doc-s-k1
9059: doc-s-k2
9060: doc-s-k3
9061: doc-s-k4
9062: doc-s-k5
9063: doc-s-k6
9064: doc-s-k7
9065: doc-s-k8
9066: doc-s-k9
9067: doc-s-k10
9068: doc-s-k11
9069: doc-s-k12
9070:
9071: Words for inputting one line from the keyboard:
9072:
9073: doc-accept
9074: doc-edit-line
9075:
9076: Conversion words:
9077:
1.143 anton 9078: doc-s>number?
9079: doc-s>unumber?
1.26 crook 9080: doc->number
9081: doc->float
1.143 anton 9082:
1.141 anton 9083:
1.27 crook 9084: @comment obsolescent words..
1.141 anton 9085: Obsolescent input and conversion words:
9086:
1.27 crook 9087: doc-convert
1.26 crook 9088: doc-expect
1.27 crook 9089: doc-span
1.5 anton 9090:
9091:
1.149 pazsan 9092: @node Pipes, Xchars and Unicode, Input, Other I/O
1.112 anton 9093: @subsection Pipes
9094: @cindex pipes, creating your own
9095:
9096: In addition to using Gforth in pipes created by other processes
9097: (@pxref{Gforth in pipes}), you can create your own pipe with
9098: @code{open-pipe}, and read from or write to it.
9099:
9100: doc-open-pipe
9101: doc-close-pipe
9102:
9103: If you write to a pipe, Gforth can throw a @code{broken-pipe-error}; if
9104: you don't catch this exception, Gforth will catch it and exit, usually
9105: silently (@pxref{Gforth in pipes}). Since you probably do not want
9106: this, you should wrap a @code{catch} or @code{try} block around the code
9107: from @code{open-pipe} to @code{close-pipe}, so you can deal with the
9108: problem yourself, and then return to regular processing.
9109:
9110: doc-broken-pipe-error
9111:
1.155 anton 9112: @node Xchars and Unicode, , Pipes, Other I/O
9113: @subsection Xchars and Unicode
1.149 pazsan 9114:
9115: This chapter needs completion
1.112 anton 9116:
1.121 anton 9117: @node OS command line arguments, Locals, Other I/O, Words
9118: @section OS command line arguments
9119: @cindex OS command line arguments
9120: @cindex command line arguments, OS
9121: @cindex arguments, OS command line
9122:
9123: The usual way to pass arguments to Gforth programs on the command line
9124: is via the @option{-e} option, e.g.
9125:
9126: @example
9127: gforth -e "123 456" foo.fs -e bye
9128: @end example
9129:
9130: However, you may want to interpret the command-line arguments directly.
9131: In that case, you can access the (image-specific) command-line arguments
1.123 anton 9132: through @code{next-arg}:
1.121 anton 9133:
1.123 anton 9134: doc-next-arg
1.121 anton 9135:
1.123 anton 9136: Here's an example program @file{echo.fs} for @code{next-arg}:
1.121 anton 9137:
9138: @example
9139: : echo ( -- )
1.122 anton 9140: begin
1.123 anton 9141: next-arg 2dup 0 0 d<> while
9142: type space
9143: repeat
9144: 2drop ;
1.121 anton 9145:
9146: echo cr bye
9147: @end example
9148:
9149: This can be invoked with
9150:
9151: @example
9152: gforth echo.fs hello world
9153: @end example
1.123 anton 9154:
9155: and it will print
9156:
9157: @example
9158: hello world
9159: @end example
9160:
9161: The next lower level of dealing with the OS command line are the
9162: following words:
9163:
9164: doc-arg
9165: doc-shift-args
9166:
9167: Finally, at the lowest level Gforth provides the following words:
9168:
9169: doc-argc
9170: doc-argv
1.121 anton 9171:
1.78 anton 9172: @c -------------------------------------------------------------
1.126 pazsan 9173: @node Locals, Structures, OS command line arguments, Words
1.78 anton 9174: @section Locals
9175: @cindex locals
9176:
9177: Local variables can make Forth programming more enjoyable and Forth
9178: programs easier to read. Unfortunately, the locals of ANS Forth are
9179: laden with restrictions. Therefore, we provide not only the ANS Forth
9180: locals wordset, but also our own, more powerful locals wordset (we
9181: implemented the ANS Forth locals wordset through our locals wordset).
1.44 crook 9182:
1.78 anton 9183: The ideas in this section have also been published in M. Anton Ertl,
9184: @cite{@uref{http://www.complang.tuwien.ac.at/papers/ertl94l.ps.gz,
9185: Automatic Scoping of Local Variables}}, EuroForth '94.
1.12 anton 9186:
9187: @menu
1.78 anton 9188: * Gforth locals::
9189: * ANS Forth locals::
1.5 anton 9190: @end menu
9191:
1.78 anton 9192: @node Gforth locals, ANS Forth locals, Locals, Locals
9193: @subsection Gforth locals
9194: @cindex Gforth locals
9195: @cindex locals, Gforth style
1.5 anton 9196:
1.78 anton 9197: Locals can be defined with
1.44 crook 9198:
1.78 anton 9199: @example
9200: @{ local1 local2 ... -- comment @}
9201: @end example
9202: or
9203: @example
9204: @{ local1 local2 ... @}
9205: @end example
1.5 anton 9206:
1.78 anton 9207: E.g.,
9208: @example
9209: : max @{ n1 n2 -- n3 @}
9210: n1 n2 > if
9211: n1
9212: else
9213: n2
9214: endif ;
9215: @end example
1.44 crook 9216:
1.78 anton 9217: The similarity of locals definitions with stack comments is intended. A
9218: locals definition often replaces the stack comment of a word. The order
9219: of the locals corresponds to the order in a stack comment and everything
9220: after the @code{--} is really a comment.
1.77 anton 9221:
1.78 anton 9222: This similarity has one disadvantage: It is too easy to confuse locals
9223: declarations with stack comments, causing bugs and making them hard to
9224: find. However, this problem can be avoided by appropriate coding
9225: conventions: Do not use both notations in the same program. If you do,
9226: they should be distinguished using additional means, e.g. by position.
1.77 anton 9227:
1.78 anton 9228: @cindex types of locals
9229: @cindex locals types
9230: The name of the local may be preceded by a type specifier, e.g.,
9231: @code{F:} for a floating point value:
1.5 anton 9232:
1.78 anton 9233: @example
9234: : CX* @{ F: Ar F: Ai F: Br F: Bi -- Cr Ci @}
9235: \ complex multiplication
9236: Ar Br f* Ai Bi f* f-
9237: Ar Bi f* Ai Br f* f+ ;
9238: @end example
1.44 crook 9239:
1.78 anton 9240: @cindex flavours of locals
9241: @cindex locals flavours
9242: @cindex value-flavoured locals
9243: @cindex variable-flavoured locals
9244: Gforth currently supports cells (@code{W:}, @code{W^}), doubles
9245: (@code{D:}, @code{D^}), floats (@code{F:}, @code{F^}) and characters
9246: (@code{C:}, @code{C^}) in two flavours: a value-flavoured local (defined
9247: with @code{W:}, @code{D:} etc.) produces its value and can be changed
9248: with @code{TO}. A variable-flavoured local (defined with @code{W^} etc.)
9249: produces its address (which becomes invalid when the variable's scope is
9250: left). E.g., the standard word @code{emit} can be defined in terms of
9251: @code{type} like this:
1.5 anton 9252:
1.78 anton 9253: @example
9254: : emit @{ C^ char* -- @}
9255: char* 1 type ;
9256: @end example
1.5 anton 9257:
1.78 anton 9258: @cindex default type of locals
9259: @cindex locals, default type
9260: A local without type specifier is a @code{W:} local. Both flavours of
9261: locals are initialized with values from the data or FP stack.
1.44 crook 9262:
1.78 anton 9263: Currently there is no way to define locals with user-defined data
9264: structures, but we are working on it.
1.5 anton 9265:
1.78 anton 9266: Gforth allows defining locals everywhere in a colon definition. This
9267: poses the following questions:
1.5 anton 9268:
1.78 anton 9269: @menu
9270: * Where are locals visible by name?::
9271: * How long do locals live?::
9272: * Locals programming style::
9273: * Locals implementation::
9274: @end menu
1.44 crook 9275:
1.78 anton 9276: @node Where are locals visible by name?, How long do locals live?, Gforth locals, Gforth locals
9277: @subsubsection Where are locals visible by name?
9278: @cindex locals visibility
9279: @cindex visibility of locals
9280: @cindex scope of locals
1.5 anton 9281:
1.78 anton 9282: Basically, the answer is that locals are visible where you would expect
9283: it in block-structured languages, and sometimes a little longer. If you
9284: want to restrict the scope of a local, enclose its definition in
9285: @code{SCOPE}...@code{ENDSCOPE}.
1.5 anton 9286:
9287:
1.78 anton 9288: doc-scope
9289: doc-endscope
1.5 anton 9290:
9291:
1.78 anton 9292: These words behave like control structure words, so you can use them
9293: with @code{CS-PICK} and @code{CS-ROLL} to restrict the scope in
9294: arbitrary ways.
1.77 anton 9295:
1.78 anton 9296: If you want a more exact answer to the visibility question, here's the
9297: basic principle: A local is visible in all places that can only be
9298: reached through the definition of the local@footnote{In compiler
9299: construction terminology, all places dominated by the definition of the
9300: local.}. In other words, it is not visible in places that can be reached
9301: without going through the definition of the local. E.g., locals defined
9302: in @code{IF}...@code{ENDIF} are visible until the @code{ENDIF}, locals
9303: defined in @code{BEGIN}...@code{UNTIL} are visible after the
9304: @code{UNTIL} (until, e.g., a subsequent @code{ENDSCOPE}).
1.77 anton 9305:
1.78 anton 9306: The reasoning behind this solution is: We want to have the locals
9307: visible as long as it is meaningful. The user can always make the
9308: visibility shorter by using explicit scoping. In a place that can
9309: only be reached through the definition of a local, the meaning of a
9310: local name is clear. In other places it is not: How is the local
9311: initialized at the control flow path that does not contain the
9312: definition? Which local is meant, if the same name is defined twice in
9313: two independent control flow paths?
1.77 anton 9314:
1.78 anton 9315: This should be enough detail for nearly all users, so you can skip the
9316: rest of this section. If you really must know all the gory details and
9317: options, read on.
1.77 anton 9318:
1.78 anton 9319: In order to implement this rule, the compiler has to know which places
9320: are unreachable. It knows this automatically after @code{AHEAD},
9321: @code{AGAIN}, @code{EXIT} and @code{LEAVE}; in other cases (e.g., after
9322: most @code{THROW}s), you can use the word @code{UNREACHABLE} to tell the
9323: compiler that the control flow never reaches that place. If
9324: @code{UNREACHABLE} is not used where it could, the only consequence is
9325: that the visibility of some locals is more limited than the rule above
9326: says. If @code{UNREACHABLE} is used where it should not (i.e., if you
9327: lie to the compiler), buggy code will be produced.
1.77 anton 9328:
1.5 anton 9329:
1.78 anton 9330: doc-unreachable
1.5 anton 9331:
1.23 crook 9332:
1.78 anton 9333: Another problem with this rule is that at @code{BEGIN}, the compiler
9334: does not know which locals will be visible on the incoming
9335: back-edge. All problems discussed in the following are due to this
9336: ignorance of the compiler (we discuss the problems using @code{BEGIN}
9337: loops as examples; the discussion also applies to @code{?DO} and other
9338: loops). Perhaps the most insidious example is:
1.26 crook 9339: @example
1.78 anton 9340: AHEAD
9341: BEGIN
9342: x
9343: [ 1 CS-ROLL ] THEN
9344: @{ x @}
9345: ...
9346: UNTIL
1.26 crook 9347: @end example
1.23 crook 9348:
1.78 anton 9349: This should be legal according to the visibility rule. The use of
9350: @code{x} can only be reached through the definition; but that appears
9351: textually below the use.
9352:
9353: From this example it is clear that the visibility rules cannot be fully
9354: implemented without major headaches. Our implementation treats common
9355: cases as advertised and the exceptions are treated in a safe way: The
9356: compiler makes a reasonable guess about the locals visible after a
9357: @code{BEGIN}; if it is too pessimistic, the
9358: user will get a spurious error about the local not being defined; if the
9359: compiler is too optimistic, it will notice this later and issue a
9360: warning. In the case above the compiler would complain about @code{x}
9361: being undefined at its use. You can see from the obscure examples in
9362: this section that it takes quite unusual control structures to get the
9363: compiler into trouble, and even then it will often do fine.
1.23 crook 9364:
1.78 anton 9365: If the @code{BEGIN} is reachable from above, the most optimistic guess
9366: is that all locals visible before the @code{BEGIN} will also be
9367: visible after the @code{BEGIN}. This guess is valid for all loops that
9368: are entered only through the @code{BEGIN}, in particular, for normal
9369: @code{BEGIN}...@code{WHILE}...@code{REPEAT} and
9370: @code{BEGIN}...@code{UNTIL} loops and it is implemented in our
9371: compiler. When the branch to the @code{BEGIN} is finally generated by
9372: @code{AGAIN} or @code{UNTIL}, the compiler checks the guess and
9373: warns the user if it was too optimistic:
1.26 crook 9374: @example
1.78 anton 9375: IF
9376: @{ x @}
9377: BEGIN
9378: \ x ?
9379: [ 1 cs-roll ] THEN
9380: ...
9381: UNTIL
1.26 crook 9382: @end example
1.23 crook 9383:
1.78 anton 9384: Here, @code{x} lives only until the @code{BEGIN}, but the compiler
9385: optimistically assumes that it lives until the @code{THEN}. It notices
9386: this difference when it compiles the @code{UNTIL} and issues a
9387: warning. The user can avoid the warning, and make sure that @code{x}
9388: is not used in the wrong area by using explicit scoping:
9389: @example
9390: IF
9391: SCOPE
9392: @{ x @}
9393: ENDSCOPE
9394: BEGIN
9395: [ 1 cs-roll ] THEN
9396: ...
9397: UNTIL
9398: @end example
1.23 crook 9399:
1.78 anton 9400: Since the guess is optimistic, there will be no spurious error messages
9401: about undefined locals.
1.44 crook 9402:
1.78 anton 9403: If the @code{BEGIN} is not reachable from above (e.g., after
9404: @code{AHEAD} or @code{EXIT}), the compiler cannot even make an
9405: optimistic guess, as the locals visible after the @code{BEGIN} may be
9406: defined later. Therefore, the compiler assumes that no locals are
9407: visible after the @code{BEGIN}. However, the user can use
9408: @code{ASSUME-LIVE} to make the compiler assume that the same locals are
9409: visible at the BEGIN as at the point where the top control-flow stack
9410: item was created.
1.23 crook 9411:
1.44 crook 9412:
1.78 anton 9413: doc-assume-live
1.26 crook 9414:
1.23 crook 9415:
1.78 anton 9416: @noindent
9417: E.g.,
9418: @example
9419: @{ x @}
9420: AHEAD
9421: ASSUME-LIVE
9422: BEGIN
9423: x
9424: [ 1 CS-ROLL ] THEN
9425: ...
9426: UNTIL
9427: @end example
1.44 crook 9428:
1.78 anton 9429: Other cases where the locals are defined before the @code{BEGIN} can be
9430: handled by inserting an appropriate @code{CS-ROLL} before the
9431: @code{ASSUME-LIVE} (and changing the control-flow stack manipulation
9432: behind the @code{ASSUME-LIVE}).
1.23 crook 9433:
1.78 anton 9434: Cases where locals are defined after the @code{BEGIN} (but should be
9435: visible immediately after the @code{BEGIN}) can only be handled by
9436: rearranging the loop. E.g., the ``most insidious'' example above can be
9437: arranged into:
9438: @example
9439: BEGIN
9440: @{ x @}
9441: ... 0=
9442: WHILE
9443: x
9444: REPEAT
9445: @end example
1.44 crook 9446:
1.78 anton 9447: @node How long do locals live?, Locals programming style, Where are locals visible by name?, Gforth locals
9448: @subsubsection How long do locals live?
9449: @cindex locals lifetime
9450: @cindex lifetime of locals
1.23 crook 9451:
1.78 anton 9452: The right answer for the lifetime question would be: A local lives at
9453: least as long as it can be accessed. For a value-flavoured local this
9454: means: until the end of its visibility. However, a variable-flavoured
9455: local could be accessed through its address far beyond its visibility
9456: scope. Ultimately, this would mean that such locals would have to be
9457: garbage collected. Since this entails un-Forth-like implementation
9458: complexities, I adopted the same cowardly solution as some other
9459: languages (e.g., C): The local lives only as long as it is visible;
9460: afterwards its address is invalid (and programs that access it
9461: afterwards are erroneous).
1.23 crook 9462:
1.78 anton 9463: @node Locals programming style, Locals implementation, How long do locals live?, Gforth locals
9464: @subsubsection Locals programming style
9465: @cindex locals programming style
9466: @cindex programming style, locals
1.23 crook 9467:
1.78 anton 9468: The freedom to define locals anywhere has the potential to change
9469: programming styles dramatically. In particular, the need to use the
9470: return stack for intermediate storage vanishes. Moreover, all stack
9471: manipulations (except @code{PICK}s and @code{ROLL}s with run-time
9472: determined arguments) can be eliminated: If the stack items are in the
9473: wrong order, just write a locals definition for all of them; then
9474: write the items in the order you want.
1.23 crook 9475:
1.78 anton 9476: This seems a little far-fetched and eliminating stack manipulations is
9477: unlikely to become a conscious programming objective. Still, the number
9478: of stack manipulations will be reduced dramatically if local variables
9479: are used liberally (e.g., compare @code{max} (@pxref{Gforth locals}) with
9480: a traditional implementation of @code{max}).
1.23 crook 9481:
1.78 anton 9482: This shows one potential benefit of locals: making Forth programs more
9483: readable. Of course, this benefit will only be realized if the
9484: programmers continue to honour the principle of factoring instead of
9485: using the added latitude to make the words longer.
1.23 crook 9486:
1.78 anton 9487: @cindex single-assignment style for locals
9488: Using @code{TO} can and should be avoided. Without @code{TO},
9489: every value-flavoured local has only a single assignment and many
9490: advantages of functional languages apply to Forth. I.e., programs are
9491: easier to analyse, to optimize and to read: It is clear from the
9492: definition what the local stands for, it does not turn into something
9493: different later.
1.23 crook 9494:
1.78 anton 9495: E.g., a definition using @code{TO} might look like this:
9496: @example
9497: : strcmp @{ addr1 u1 addr2 u2 -- n @}
9498: u1 u2 min 0
9499: ?do
9500: addr1 c@@ addr2 c@@ -
9501: ?dup-if
9502: unloop exit
9503: then
9504: addr1 char+ TO addr1
9505: addr2 char+ TO addr2
9506: loop
9507: u1 u2 - ;
1.26 crook 9508: @end example
1.78 anton 9509: Here, @code{TO} is used to update @code{addr1} and @code{addr2} at
9510: every loop iteration. @code{strcmp} is a typical example of the
9511: readability problems of using @code{TO}. When you start reading
9512: @code{strcmp}, you think that @code{addr1} refers to the start of the
9513: string. Only near the end of the loop you realize that it is something
9514: else.
1.23 crook 9515:
1.78 anton 9516: This can be avoided by defining two locals at the start of the loop that
9517: are initialized with the right value for the current iteration.
9518: @example
9519: : strcmp @{ addr1 u1 addr2 u2 -- n @}
9520: addr1 addr2
9521: u1 u2 min 0
9522: ?do @{ s1 s2 @}
9523: s1 c@@ s2 c@@ -
9524: ?dup-if
9525: unloop exit
9526: then
9527: s1 char+ s2 char+
9528: loop
9529: 2drop
9530: u1 u2 - ;
9531: @end example
9532: Here it is clear from the start that @code{s1} has a different value
9533: in every loop iteration.
1.23 crook 9534:
1.78 anton 9535: @node Locals implementation, , Locals programming style, Gforth locals
9536: @subsubsection Locals implementation
9537: @cindex locals implementation
9538: @cindex implementation of locals
1.23 crook 9539:
1.78 anton 9540: @cindex locals stack
9541: Gforth uses an extra locals stack. The most compelling reason for
9542: this is that the return stack is not float-aligned; using an extra stack
9543: also eliminates the problems and restrictions of using the return stack
9544: as locals stack. Like the other stacks, the locals stack grows toward
9545: lower addresses. A few primitives allow an efficient implementation:
9546:
9547:
9548: doc-@local#
9549: doc-f@local#
9550: doc-laddr#
9551: doc-lp+!#
9552: doc-lp!
9553: doc->l
9554: doc-f>l
9555:
9556:
9557: In addition to these primitives, some specializations of these
9558: primitives for commonly occurring inline arguments are provided for
9559: efficiency reasons, e.g., @code{@@local0} as specialization of
9560: @code{@@local#} for the inline argument 0. The following compiling words
9561: compile the right specialized version, or the general version, as
9562: appropriate:
1.23 crook 9563:
1.5 anton 9564:
1.107 dvdkhlng 9565: @c doc-compile-@local
9566: @c doc-compile-f@local
1.78 anton 9567: doc-compile-lp+!
1.5 anton 9568:
9569:
1.78 anton 9570: Combinations of conditional branches and @code{lp+!#} like
9571: @code{?branch-lp+!#} (the locals pointer is only changed if the branch
9572: is taken) are provided for efficiency and correctness in loops.
1.5 anton 9573:
1.78 anton 9574: A special area in the dictionary space is reserved for keeping the
9575: local variable names. @code{@{} switches the dictionary pointer to this
9576: area and @code{@}} switches it back and generates the locals
9577: initializing code. @code{W:} etc.@ are normal defining words. This
9578: special area is cleared at the start of every colon definition.
1.5 anton 9579:
1.78 anton 9580: @cindex word list for defining locals
9581: A special feature of Gforth's dictionary is used to implement the
9582: definition of locals without type specifiers: every word list (aka
9583: vocabulary) has its own methods for searching
9584: etc. (@pxref{Word Lists}). For the present purpose we defined a word list
9585: with a special search method: When it is searched for a word, it
9586: actually creates that word using @code{W:}. @code{@{} changes the search
9587: order to first search the word list containing @code{@}}, @code{W:} etc.,
9588: and then the word list for defining locals without type specifiers.
1.5 anton 9589:
1.78 anton 9590: The lifetime rules support a stack discipline within a colon
9591: definition: The lifetime of a local is either nested with other locals
9592: lifetimes or it does not overlap them.
1.23 crook 9593:
1.78 anton 9594: At @code{BEGIN}, @code{IF}, and @code{AHEAD} no code for locals stack
9595: pointer manipulation is generated. Between control structure words
9596: locals definitions can push locals onto the locals stack. @code{AGAIN}
9597: is the simplest of the other three control flow words. It has to
9598: restore the locals stack depth of the corresponding @code{BEGIN}
9599: before branching. The code looks like this:
9600: @format
9601: @code{lp+!#} current-locals-size @minus{} dest-locals-size
9602: @code{branch} <begin>
9603: @end format
1.26 crook 9604:
1.78 anton 9605: @code{UNTIL} is a little more complicated: If it branches back, it
9606: must adjust the stack just like @code{AGAIN}. But if it falls through,
9607: the locals stack must not be changed. The compiler generates the
9608: following code:
9609: @format
9610: @code{?branch-lp+!#} <begin> current-locals-size @minus{} dest-locals-size
9611: @end format
9612: The locals stack pointer is only adjusted if the branch is taken.
1.44 crook 9613:
1.78 anton 9614: @code{THEN} can produce somewhat inefficient code:
9615: @format
9616: @code{lp+!#} current-locals-size @minus{} orig-locals-size
9617: <orig target>:
9618: @code{lp+!#} orig-locals-size @minus{} new-locals-size
9619: @end format
9620: The second @code{lp+!#} adjusts the locals stack pointer from the
9621: level at the @i{orig} point to the level after the @code{THEN}. The
9622: first @code{lp+!#} adjusts the locals stack pointer from the current
9623: level to the level at the orig point, so the complete effect is an
9624: adjustment from the current level to the right level after the
9625: @code{THEN}.
1.26 crook 9626:
1.78 anton 9627: @cindex locals information on the control-flow stack
9628: @cindex control-flow stack items, locals information
9629: In a conventional Forth implementation a dest control-flow stack entry
9630: is just the target address and an orig entry is just the address to be
9631: patched. Our locals implementation adds a word list to every orig or dest
9632: item. It is the list of locals visible (or assumed visible) at the point
9633: described by the entry. Our implementation also adds a tag to identify
9634: the kind of entry, in particular to differentiate between live and dead
9635: (reachable and unreachable) orig entries.
1.26 crook 9636:
1.78 anton 9637: A few unusual operations have to be performed on locals word lists:
1.44 crook 9638:
1.5 anton 9639:
1.78 anton 9640: doc-common-list
9641: doc-sub-list?
9642: doc-list-size
1.52 anton 9643:
9644:
1.78 anton 9645: Several features of our locals word list implementation make these
9646: operations easy to implement: The locals word lists are organised as
9647: linked lists; the tails of these lists are shared, if the lists
9648: contain some of the same locals; and the address of a name is greater
9649: than the address of the names behind it in the list.
1.5 anton 9650:
1.78 anton 9651: Another important implementation detail is the variable
9652: @code{dead-code}. It is used by @code{BEGIN} and @code{THEN} to
9653: determine if they can be reached directly or only through the branch
9654: that they resolve. @code{dead-code} is set by @code{UNREACHABLE},
9655: @code{AHEAD}, @code{EXIT} etc., and cleared at the start of a colon
9656: definition, by @code{BEGIN} and usually by @code{THEN}.
1.5 anton 9657:
1.78 anton 9658: Counted loops are similar to other loops in most respects, but
9659: @code{LEAVE} requires special attention: It performs basically the same
9660: service as @code{AHEAD}, but it does not create a control-flow stack
9661: entry. Therefore the information has to be stored elsewhere;
9662: traditionally, the information was stored in the target fields of the
9663: branches created by the @code{LEAVE}s, by organizing these fields into a
9664: linked list. Unfortunately, this clever trick does not provide enough
9665: space for storing our extended control flow information. Therefore, we
9666: introduce another stack, the leave stack. It contains the control-flow
9667: stack entries for all unresolved @code{LEAVE}s.
1.44 crook 9668:
1.78 anton 9669: Local names are kept until the end of the colon definition, even if
9670: they are no longer visible in any control-flow path. In a few cases
9671: this may lead to increased space needs for the locals name area, but
9672: usually less than reclaiming this space would cost in code size.
1.5 anton 9673:
1.44 crook 9674:
1.78 anton 9675: @node ANS Forth locals, , Gforth locals, Locals
9676: @subsection ANS Forth locals
9677: @cindex locals, ANS Forth style
1.5 anton 9678:
1.78 anton 9679: The ANS Forth locals wordset does not define a syntax for locals, but
9680: words that make it possible to define various syntaxes. One of the
9681: possible syntaxes is a subset of the syntax we used in the Gforth locals
9682: wordset, i.e.:
1.29 crook 9683:
9684: @example
1.78 anton 9685: @{ local1 local2 ... -- comment @}
9686: @end example
9687: @noindent
9688: or
9689: @example
9690: @{ local1 local2 ... @}
1.29 crook 9691: @end example
9692:
1.78 anton 9693: The order of the locals corresponds to the order in a stack comment. The
9694: restrictions are:
1.5 anton 9695:
1.78 anton 9696: @itemize @bullet
9697: @item
9698: Locals can only be cell-sized values (no type specifiers are allowed).
9699: @item
9700: Locals can be defined only outside control structures.
9701: @item
9702: Locals can interfere with explicit usage of the return stack. For the
9703: exact (and long) rules, see the standard. If you don't use return stack
9704: accessing words in a definition using locals, you will be all right. The
9705: purpose of this rule is to make locals implementation on the return
9706: stack easier.
9707: @item
9708: The whole definition must be in one line.
9709: @end itemize
1.5 anton 9710:
1.78 anton 9711: Locals defined in ANS Forth behave like @code{VALUE}s
9712: (@pxref{Values}). I.e., they are initialized from the stack. Using their
9713: name produces their value. Their value can be changed using @code{TO}.
1.77 anton 9714:
1.78 anton 9715: Since the syntax above is supported by Gforth directly, you need not do
9716: anything to use it. If you want to port a program using this syntax to
9717: another ANS Forth system, use @file{compat/anslocal.fs} to implement the
9718: syntax on the other system.
1.5 anton 9719:
1.78 anton 9720: Note that a syntax shown in the standard, section A.13 looks
9721: similar, but is quite different in having the order of locals
9722: reversed. Beware!
1.5 anton 9723:
1.78 anton 9724: The ANS Forth locals wordset itself consists of one word:
1.5 anton 9725:
1.78 anton 9726: doc-(local)
1.5 anton 9727:
1.78 anton 9728: The ANS Forth locals extension wordset defines a syntax using
9729: @code{locals|}, but it is so awful that we strongly recommend not to use
9730: it. We have implemented this syntax to make porting to Gforth easy, but
9731: do not document it here. The problem with this syntax is that the locals
9732: are defined in an order reversed with respect to the standard stack
9733: comment notation, making programs harder to read, and easier to misread
9734: and miswrite. The only merit of this syntax is that it is easy to
9735: implement using the ANS Forth locals wordset.
1.53 anton 9736:
9737:
1.78 anton 9738: @c ----------------------------------------------------------
9739: @node Structures, Object-oriented Forth, Locals, Words
9740: @section Structures
9741: @cindex structures
9742: @cindex records
1.53 anton 9743:
1.78 anton 9744: This section presents the structure package that comes with Gforth. A
9745: version of the package implemented in ANS Forth is available in
9746: @file{compat/struct.fs}. This package was inspired by a posting on
9747: comp.lang.forth in 1989 (unfortunately I don't remember, by whom;
9748: possibly John Hayes). A version of this section has been published in
9749: M. Anton Ertl,
9750: @uref{http://www.complang.tuwien.ac.at/forth/objects/structs.html, Yet
9751: Another Forth Structures Package}, Forth Dimensions 19(3), pages
9752: 13--16. Marcel Hendrix provided helpful comments.
1.53 anton 9753:
1.78 anton 9754: @menu
9755: * Why explicit structure support?::
9756: * Structure Usage::
9757: * Structure Naming Convention::
9758: * Structure Implementation::
9759: * Structure Glossary::
9760: @end menu
1.55 anton 9761:
1.78 anton 9762: @node Why explicit structure support?, Structure Usage, Structures, Structures
9763: @subsection Why explicit structure support?
1.53 anton 9764:
1.78 anton 9765: @cindex address arithmetic for structures
9766: @cindex structures using address arithmetic
9767: If we want to use a structure containing several fields, we could simply
9768: reserve memory for it, and access the fields using address arithmetic
9769: (@pxref{Address arithmetic}). As an example, consider a structure with
9770: the following fields
1.57 anton 9771:
1.78 anton 9772: @table @code
9773: @item a
9774: is a float
9775: @item b
9776: is a cell
9777: @item c
9778: is a float
9779: @end table
1.57 anton 9780:
1.78 anton 9781: Given the (float-aligned) base address of the structure we get the
9782: address of the field
1.52 anton 9783:
1.78 anton 9784: @table @code
9785: @item a
9786: without doing anything further.
9787: @item b
9788: with @code{float+}
9789: @item c
9790: with @code{float+ cell+ faligned}
9791: @end table
1.52 anton 9792:
1.78 anton 9793: It is easy to see that this can become quite tiring.
1.52 anton 9794:
1.78 anton 9795: Moreover, it is not very readable, because seeing a
9796: @code{cell+} tells us neither which kind of structure is
9797: accessed nor what field is accessed; we have to somehow infer the kind
9798: of structure, and then look up in the documentation, which field of
9799: that structure corresponds to that offset.
1.53 anton 9800:
1.78 anton 9801: Finally, this kind of address arithmetic also causes maintenance
9802: troubles: If you add or delete a field somewhere in the middle of the
9803: structure, you have to find and change all computations for the fields
9804: afterwards.
1.52 anton 9805:
1.78 anton 9806: So, instead of using @code{cell+} and friends directly, how
9807: about storing the offsets in constants:
1.52 anton 9808:
1.78 anton 9809: @example
9810: 0 constant a-offset
9811: 0 float+ constant b-offset
9812: 0 float+ cell+ faligned c-offset
9813: @end example
1.64 pazsan 9814:
1.78 anton 9815: Now we can get the address of field @code{x} with @code{x-offset
9816: +}. This is much better in all respects. Of course, you still
9817: have to change all later offset definitions if you add a field. You can
9818: fix this by declaring the offsets in the following way:
1.57 anton 9819:
1.78 anton 9820: @example
9821: 0 constant a-offset
9822: a-offset float+ constant b-offset
9823: b-offset cell+ faligned constant c-offset
9824: @end example
1.57 anton 9825:
1.78 anton 9826: Since we always use the offsets with @code{+}, we could use a defining
9827: word @code{cfield} that includes the @code{+} in the action of the
9828: defined word:
1.64 pazsan 9829:
1.78 anton 9830: @example
9831: : cfield ( n "name" -- )
9832: create ,
9833: does> ( name execution: addr1 -- addr2 )
9834: @@ + ;
1.64 pazsan 9835:
1.78 anton 9836: 0 cfield a
9837: 0 a float+ cfield b
9838: 0 b cell+ faligned cfield c
9839: @end example
1.64 pazsan 9840:
1.78 anton 9841: Instead of @code{x-offset +}, we now simply write @code{x}.
1.64 pazsan 9842:
1.78 anton 9843: The structure field words now can be used quite nicely. However,
9844: their definition is still a bit cumbersome: We have to repeat the
9845: name, the information about size and alignment is distributed before
9846: and after the field definitions etc. The structure package presented
9847: here addresses these problems.
1.64 pazsan 9848:
1.78 anton 9849: @node Structure Usage, Structure Naming Convention, Why explicit structure support?, Structures
9850: @subsection Structure Usage
9851: @cindex structure usage
1.57 anton 9852:
1.78 anton 9853: @cindex @code{field} usage
9854: @cindex @code{struct} usage
9855: @cindex @code{end-struct} usage
9856: You can define a structure for a (data-less) linked list with:
1.57 anton 9857: @example
1.78 anton 9858: struct
9859: cell% field list-next
9860: end-struct list%
1.57 anton 9861: @end example
9862:
1.78 anton 9863: With the address of the list node on the stack, you can compute the
9864: address of the field that contains the address of the next node with
9865: @code{list-next}. E.g., you can determine the length of a list
9866: with:
1.57 anton 9867:
9868: @example
1.78 anton 9869: : list-length ( list -- n )
9870: \ "list" is a pointer to the first element of a linked list
9871: \ "n" is the length of the list
9872: 0 BEGIN ( list1 n1 )
9873: over
9874: WHILE ( list1 n1 )
9875: 1+ swap list-next @@ swap
9876: REPEAT
9877: nip ;
1.57 anton 9878: @end example
9879:
1.78 anton 9880: You can reserve memory for a list node in the dictionary with
9881: @code{list% %allot}, which leaves the address of the list node on the
9882: stack. For the equivalent allocation on the heap you can use @code{list%
9883: %alloc} (or, for an @code{allocate}-like stack effect (i.e., with ior),
9884: use @code{list% %allocate}). You can get the the size of a list
9885: node with @code{list% %size} and its alignment with @code{list%
9886: %alignment}.
9887:
9888: Note that in ANS Forth the body of a @code{create}d word is
9889: @code{aligned} but not necessarily @code{faligned};
9890: therefore, if you do a:
1.57 anton 9891:
9892: @example
1.78 anton 9893: create @emph{name} foo% %allot drop
1.57 anton 9894: @end example
9895:
1.78 anton 9896: @noindent
9897: then the memory alloted for @code{foo%} is guaranteed to start at the
9898: body of @code{@emph{name}} only if @code{foo%} contains only character,
9899: cell and double fields. Therefore, if your structure contains floats,
9900: better use
1.57 anton 9901:
9902: @example
1.78 anton 9903: foo% %allot constant @emph{name}
1.57 anton 9904: @end example
9905:
1.78 anton 9906: @cindex structures containing structures
9907: You can include a structure @code{foo%} as a field of
9908: another structure, like this:
1.65 anton 9909: @example
1.78 anton 9910: struct
9911: ...
9912: foo% field ...
9913: ...
9914: end-struct ...
1.65 anton 9915: @end example
1.52 anton 9916:
1.78 anton 9917: @cindex structure extension
9918: @cindex extended records
9919: Instead of starting with an empty structure, you can extend an
9920: existing structure. E.g., a plain linked list without data, as defined
9921: above, is hardly useful; You can extend it to a linked list of integers,
9922: like this:@footnote{This feature is also known as @emph{extended
9923: records}. It is the main innovation in the Oberon language; in other
9924: words, adding this feature to Modula-2 led Wirth to create a new
9925: language, write a new compiler etc. Adding this feature to Forth just
9926: required a few lines of code.}
1.52 anton 9927:
1.78 anton 9928: @example
9929: list%
9930: cell% field intlist-int
9931: end-struct intlist%
9932: @end example
1.55 anton 9933:
1.78 anton 9934: @code{intlist%} is a structure with two fields:
9935: @code{list-next} and @code{intlist-int}.
1.55 anton 9936:
1.78 anton 9937: @cindex structures containing arrays
9938: You can specify an array type containing @emph{n} elements of
9939: type @code{foo%} like this:
1.55 anton 9940:
9941: @example
1.78 anton 9942: foo% @emph{n} *
1.56 anton 9943: @end example
1.55 anton 9944:
1.78 anton 9945: You can use this array type in any place where you can use a normal
9946: type, e.g., when defining a @code{field}, or with
9947: @code{%allot}.
9948:
9949: @cindex first field optimization
9950: The first field is at the base address of a structure and the word for
9951: this field (e.g., @code{list-next}) actually does not change the address
9952: on the stack. You may be tempted to leave it away in the interest of
9953: run-time and space efficiency. This is not necessary, because the
9954: structure package optimizes this case: If you compile a first-field
9955: words, no code is generated. So, in the interest of readability and
9956: maintainability you should include the word for the field when accessing
9957: the field.
1.52 anton 9958:
9959:
1.78 anton 9960: @node Structure Naming Convention, Structure Implementation, Structure Usage, Structures
9961: @subsection Structure Naming Convention
9962: @cindex structure naming convention
1.52 anton 9963:
1.78 anton 9964: The field names that come to (my) mind are often quite generic, and,
9965: if used, would cause frequent name clashes. E.g., many structures
9966: probably contain a @code{counter} field. The structure names
9967: that come to (my) mind are often also the logical choice for the names
9968: of words that create such a structure.
1.52 anton 9969:
1.78 anton 9970: Therefore, I have adopted the following naming conventions:
1.52 anton 9971:
1.78 anton 9972: @itemize @bullet
9973: @cindex field naming convention
9974: @item
9975: The names of fields are of the form
9976: @code{@emph{struct}-@emph{field}}, where
9977: @code{@emph{struct}} is the basic name of the structure, and
9978: @code{@emph{field}} is the basic name of the field. You can
9979: think of field words as converting the (address of the)
9980: structure into the (address of the) field.
1.52 anton 9981:
1.78 anton 9982: @cindex structure naming convention
9983: @item
9984: The names of structures are of the form
9985: @code{@emph{struct}%}, where
9986: @code{@emph{struct}} is the basic name of the structure.
9987: @end itemize
1.52 anton 9988:
1.78 anton 9989: This naming convention does not work that well for fields of extended
9990: structures; e.g., the integer list structure has a field
9991: @code{intlist-int}, but has @code{list-next}, not
9992: @code{intlist-next}.
1.53 anton 9993:
1.78 anton 9994: @node Structure Implementation, Structure Glossary, Structure Naming Convention, Structures
9995: @subsection Structure Implementation
9996: @cindex structure implementation
9997: @cindex implementation of structures
1.52 anton 9998:
1.78 anton 9999: The central idea in the implementation is to pass the data about the
10000: structure being built on the stack, not in some global
10001: variable. Everything else falls into place naturally once this design
10002: decision is made.
1.53 anton 10003:
1.78 anton 10004: The type description on the stack is of the form @emph{align
10005: size}. Keeping the size on the top-of-stack makes dealing with arrays
10006: very simple.
1.53 anton 10007:
1.78 anton 10008: @code{field} is a defining word that uses @code{Create}
10009: and @code{DOES>}. The body of the field contains the offset
10010: of the field, and the normal @code{DOES>} action is simply:
1.53 anton 10011:
10012: @example
1.78 anton 10013: @@ +
1.53 anton 10014: @end example
10015:
1.78 anton 10016: @noindent
10017: i.e., add the offset to the address, giving the stack effect
10018: @i{addr1 -- addr2} for a field.
10019:
10020: @cindex first field optimization, implementation
10021: This simple structure is slightly complicated by the optimization
10022: for fields with offset 0, which requires a different
10023: @code{DOES>}-part (because we cannot rely on there being
10024: something on the stack if such a field is invoked during
10025: compilation). Therefore, we put the different @code{DOES>}-parts
10026: in separate words, and decide which one to invoke based on the
10027: offset. For a zero offset, the field is basically a noop; it is
10028: immediate, and therefore no code is generated when it is compiled.
1.53 anton 10029:
1.78 anton 10030: @node Structure Glossary, , Structure Implementation, Structures
10031: @subsection Structure Glossary
10032: @cindex structure glossary
1.53 anton 10033:
1.5 anton 10034:
1.78 anton 10035: doc-%align
10036: doc-%alignment
10037: doc-%alloc
10038: doc-%allocate
10039: doc-%allot
10040: doc-cell%
10041: doc-char%
10042: doc-dfloat%
10043: doc-double%
10044: doc-end-struct
10045: doc-field
10046: doc-float%
10047: doc-naligned
10048: doc-sfloat%
10049: doc-%size
10050: doc-struct
1.54 anton 10051:
10052:
1.26 crook 10053: @c -------------------------------------------------------------
1.78 anton 10054: @node Object-oriented Forth, Programming Tools, Structures, Words
10055: @section Object-oriented Forth
10056:
10057: Gforth comes with three packages for object-oriented programming:
10058: @file{objects.fs}, @file{oof.fs}, and @file{mini-oof.fs}; none of them
10059: is preloaded, so you have to @code{include} them before use. The most
10060: important differences between these packages (and others) are discussed
10061: in @ref{Comparison with other object models}. All packages are written
10062: in ANS Forth and can be used with any other ANS Forth.
1.5 anton 10063:
1.78 anton 10064: @menu
10065: * Why object-oriented programming?::
10066: * Object-Oriented Terminology::
10067: * Objects::
10068: * OOF::
10069: * Mini-OOF::
10070: * Comparison with other object models::
10071: @end menu
1.5 anton 10072:
1.78 anton 10073: @c ----------------------------------------------------------------
10074: @node Why object-oriented programming?, Object-Oriented Terminology, Object-oriented Forth, Object-oriented Forth
10075: @subsection Why object-oriented programming?
10076: @cindex object-oriented programming motivation
10077: @cindex motivation for object-oriented programming
1.44 crook 10078:
1.78 anton 10079: Often we have to deal with several data structures (@emph{objects}),
10080: that have to be treated similarly in some respects, but differently in
10081: others. Graphical objects are the textbook example: circles, triangles,
10082: dinosaurs, icons, and others, and we may want to add more during program
10083: development. We want to apply some operations to any graphical object,
10084: e.g., @code{draw} for displaying it on the screen. However, @code{draw}
10085: has to do something different for every kind of object.
10086: @comment TODO add some other operations eg perimeter, area
10087: @comment and tie in to concrete examples later..
1.5 anton 10088:
1.78 anton 10089: We could implement @code{draw} as a big @code{CASE}
10090: control structure that executes the appropriate code depending on the
10091: kind of object to be drawn. This would be not be very elegant, and,
10092: moreover, we would have to change @code{draw} every time we add
10093: a new kind of graphical object (say, a spaceship).
1.44 crook 10094:
1.78 anton 10095: What we would rather do is: When defining spaceships, we would tell
10096: the system: ``Here's how you @code{draw} a spaceship; you figure
10097: out the rest''.
1.5 anton 10098:
1.78 anton 10099: This is the problem that all systems solve that (rightfully) call
10100: themselves object-oriented; the object-oriented packages presented here
10101: solve this problem (and not much else).
10102: @comment TODO ?list properties of oo systems.. oo vs o-based?
1.44 crook 10103:
1.78 anton 10104: @c ------------------------------------------------------------------------
10105: @node Object-Oriented Terminology, Objects, Why object-oriented programming?, Object-oriented Forth
10106: @subsection Object-Oriented Terminology
10107: @cindex object-oriented terminology
10108: @cindex terminology for object-oriented programming
1.5 anton 10109:
1.78 anton 10110: This section is mainly for reference, so you don't have to understand
10111: all of it right away. The terminology is mainly Smalltalk-inspired. In
10112: short:
1.44 crook 10113:
1.78 anton 10114: @table @emph
10115: @cindex class
10116: @item class
10117: a data structure definition with some extras.
1.5 anton 10118:
1.78 anton 10119: @cindex object
10120: @item object
10121: an instance of the data structure described by the class definition.
1.5 anton 10122:
1.78 anton 10123: @cindex instance variables
10124: @item instance variables
10125: fields of the data structure.
1.5 anton 10126:
1.78 anton 10127: @cindex selector
10128: @cindex method selector
10129: @cindex virtual function
10130: @item selector
10131: (or @emph{method selector}) a word (e.g.,
10132: @code{draw}) that performs an operation on a variety of data
10133: structures (classes). A selector describes @emph{what} operation to
10134: perform. In C++ terminology: a (pure) virtual function.
1.5 anton 10135:
1.78 anton 10136: @cindex method
10137: @item method
10138: the concrete definition that performs the operation
10139: described by the selector for a specific class. A method specifies
10140: @emph{how} the operation is performed for a specific class.
1.5 anton 10141:
1.78 anton 10142: @cindex selector invocation
10143: @cindex message send
10144: @cindex invoking a selector
10145: @item selector invocation
10146: a call of a selector. One argument of the call (the TOS (top-of-stack))
10147: is used for determining which method is used. In Smalltalk terminology:
10148: a message (consisting of the selector and the other arguments) is sent
10149: to the object.
1.5 anton 10150:
1.78 anton 10151: @cindex receiving object
10152: @item receiving object
10153: the object used for determining the method executed by a selector
10154: invocation. In the @file{objects.fs} model, it is the object that is on
10155: the TOS when the selector is invoked. (@emph{Receiving} comes from
10156: the Smalltalk @emph{message} terminology.)
1.5 anton 10157:
1.78 anton 10158: @cindex child class
10159: @cindex parent class
10160: @cindex inheritance
10161: @item child class
10162: a class that has (@emph{inherits}) all properties (instance variables,
10163: selectors, methods) from a @emph{parent class}. In Smalltalk
10164: terminology: The subclass inherits from the superclass. In C++
10165: terminology: The derived class inherits from the base class.
1.5 anton 10166:
1.78 anton 10167: @end table
1.5 anton 10168:
1.78 anton 10169: @c If you wonder about the message sending terminology, it comes from
10170: @c a time when each object had it's own task and objects communicated via
10171: @c message passing; eventually the Smalltalk developers realized that
10172: @c they can do most things through simple (indirect) calls. They kept the
10173: @c terminology.
1.5 anton 10174:
1.78 anton 10175: @c --------------------------------------------------------------
10176: @node Objects, OOF, Object-Oriented Terminology, Object-oriented Forth
10177: @subsection The @file{objects.fs} model
10178: @cindex objects
10179: @cindex object-oriented programming
1.26 crook 10180:
1.78 anton 10181: @cindex @file{objects.fs}
10182: @cindex @file{oof.fs}
1.26 crook 10183:
1.78 anton 10184: This section describes the @file{objects.fs} package. This material also
10185: has been published in M. Anton Ertl,
10186: @cite{@uref{http://www.complang.tuwien.ac.at/forth/objects/objects.html,
10187: Yet Another Forth Objects Package}}, Forth Dimensions 19(2), pages
10188: 37--43.
10189: @c McKewan's and Zsoter's packages
1.26 crook 10190:
1.78 anton 10191: This section assumes that you have read @ref{Structures}.
1.5 anton 10192:
1.78 anton 10193: The techniques on which this model is based have been used to implement
10194: the parser generator, Gray, and have also been used in Gforth for
10195: implementing the various flavours of word lists (hashed or not,
10196: case-sensitive or not, special-purpose word lists for locals etc.).
1.5 anton 10197:
10198:
1.26 crook 10199: @menu
1.78 anton 10200: * Properties of the Objects model::
10201: * Basic Objects Usage::
10202: * The Objects base class::
10203: * Creating objects::
10204: * Object-Oriented Programming Style::
10205: * Class Binding::
10206: * Method conveniences::
10207: * Classes and Scoping::
10208: * Dividing classes::
10209: * Object Interfaces::
10210: * Objects Implementation::
10211: * Objects Glossary::
1.26 crook 10212: @end menu
1.5 anton 10213:
1.78 anton 10214: Marcel Hendrix provided helpful comments on this section.
1.5 anton 10215:
1.78 anton 10216: @node Properties of the Objects model, Basic Objects Usage, Objects, Objects
10217: @subsubsection Properties of the @file{objects.fs} model
10218: @cindex @file{objects.fs} properties
1.5 anton 10219:
1.78 anton 10220: @itemize @bullet
10221: @item
10222: It is straightforward to pass objects on the stack. Passing
10223: selectors on the stack is a little less convenient, but possible.
1.44 crook 10224:
1.78 anton 10225: @item
10226: Objects are just data structures in memory, and are referenced by their
10227: address. You can create words for objects with normal defining words
10228: like @code{constant}. Likewise, there is no difference between instance
10229: variables that contain objects and those that contain other data.
1.5 anton 10230:
1.78 anton 10231: @item
10232: Late binding is efficient and easy to use.
1.44 crook 10233:
1.78 anton 10234: @item
10235: It avoids parsing, and thus avoids problems with state-smartness
10236: and reduced extensibility; for convenience there are a few parsing
10237: words, but they have non-parsing counterparts. There are also a few
10238: defining words that parse. This is hard to avoid, because all standard
10239: defining words parse (except @code{:noname}); however, such
10240: words are not as bad as many other parsing words, because they are not
10241: state-smart.
1.5 anton 10242:
1.78 anton 10243: @item
10244: It does not try to incorporate everything. It does a few things and does
10245: them well (IMO). In particular, this model was not designed to support
10246: information hiding (although it has features that may help); you can use
10247: a separate package for achieving this.
1.5 anton 10248:
1.78 anton 10249: @item
10250: It is layered; you don't have to learn and use all features to use this
10251: model. Only a few features are necessary (@pxref{Basic Objects Usage},
10252: @pxref{The Objects base class}, @pxref{Creating objects}.), the others
10253: are optional and independent of each other.
1.5 anton 10254:
1.78 anton 10255: @item
10256: An implementation in ANS Forth is available.
1.5 anton 10257:
1.78 anton 10258: @end itemize
1.5 anton 10259:
1.44 crook 10260:
1.78 anton 10261: @node Basic Objects Usage, The Objects base class, Properties of the Objects model, Objects
10262: @subsubsection Basic @file{objects.fs} Usage
10263: @cindex basic objects usage
10264: @cindex objects, basic usage
1.5 anton 10265:
1.78 anton 10266: You can define a class for graphical objects like this:
1.44 crook 10267:
1.78 anton 10268: @cindex @code{class} usage
10269: @cindex @code{end-class} usage
10270: @cindex @code{selector} usage
1.5 anton 10271: @example
1.78 anton 10272: object class \ "object" is the parent class
10273: selector draw ( x y graphical -- )
10274: end-class graphical
10275: @end example
10276:
10277: This code defines a class @code{graphical} with an
10278: operation @code{draw}. We can perform the operation
10279: @code{draw} on any @code{graphical} object, e.g.:
10280:
10281: @example
10282: 100 100 t-rex draw
1.26 crook 10283: @end example
1.5 anton 10284:
1.78 anton 10285: @noindent
10286: where @code{t-rex} is a word (say, a constant) that produces a
10287: graphical object.
10288:
10289: @comment TODO add a 2nd operation eg perimeter.. and use for
10290: @comment a concrete example
1.5 anton 10291:
1.78 anton 10292: @cindex abstract class
10293: How do we create a graphical object? With the present definitions,
10294: we cannot create a useful graphical object. The class
10295: @code{graphical} describes graphical objects in general, but not
10296: any concrete graphical object type (C++ users would call it an
10297: @emph{abstract class}); e.g., there is no method for the selector
10298: @code{draw} in the class @code{graphical}.
1.5 anton 10299:
1.78 anton 10300: For concrete graphical objects, we define child classes of the
10301: class @code{graphical}, e.g.:
1.5 anton 10302:
1.78 anton 10303: @cindex @code{overrides} usage
10304: @cindex @code{field} usage in class definition
1.26 crook 10305: @example
1.78 anton 10306: graphical class \ "graphical" is the parent class
10307: cell% field circle-radius
1.5 anton 10308:
1.78 anton 10309: :noname ( x y circle -- )
10310: circle-radius @@ draw-circle ;
10311: overrides draw
1.5 anton 10312:
1.78 anton 10313: :noname ( n-radius circle -- )
10314: circle-radius ! ;
10315: overrides construct
1.5 anton 10316:
1.78 anton 10317: end-class circle
10318: @end example
1.44 crook 10319:
1.78 anton 10320: Here we define a class @code{circle} as a child of @code{graphical},
10321: with field @code{circle-radius} (which behaves just like a field
10322: (@pxref{Structures}); it defines (using @code{overrides}) new methods
10323: for the selectors @code{draw} and @code{construct} (@code{construct} is
10324: defined in @code{object}, the parent class of @code{graphical}).
1.5 anton 10325:
1.78 anton 10326: Now we can create a circle on the heap (i.e.,
10327: @code{allocate}d memory) with:
1.44 crook 10328:
1.78 anton 10329: @cindex @code{heap-new} usage
1.5 anton 10330: @example
1.78 anton 10331: 50 circle heap-new constant my-circle
1.5 anton 10332: @end example
10333:
1.78 anton 10334: @noindent
10335: @code{heap-new} invokes @code{construct}, thus
10336: initializing the field @code{circle-radius} with 50. We can draw
10337: this new circle at (100,100) with:
1.5 anton 10338:
10339: @example
1.78 anton 10340: 100 100 my-circle draw
1.5 anton 10341: @end example
10342:
1.78 anton 10343: @cindex selector invocation, restrictions
10344: @cindex class definition, restrictions
10345: Note: You can only invoke a selector if the object on the TOS
10346: (the receiving object) belongs to the class where the selector was
10347: defined or one of its descendents; e.g., you can invoke
10348: @code{draw} only for objects belonging to @code{graphical}
10349: or its descendents (e.g., @code{circle}). Immediately before
10350: @code{end-class}, the search order has to be the same as
10351: immediately after @code{class}.
10352:
10353: @node The Objects base class, Creating objects, Basic Objects Usage, Objects
10354: @subsubsection The @file{object.fs} base class
10355: @cindex @code{object} class
10356:
10357: When you define a class, you have to specify a parent class. So how do
10358: you start defining classes? There is one class available from the start:
10359: @code{object}. It is ancestor for all classes and so is the
10360: only class that has no parent. It has two selectors: @code{construct}
10361: and @code{print}.
10362:
10363: @node Creating objects, Object-Oriented Programming Style, The Objects base class, Objects
10364: @subsubsection Creating objects
10365: @cindex creating objects
10366: @cindex object creation
10367: @cindex object allocation options
10368:
10369: @cindex @code{heap-new} discussion
10370: @cindex @code{dict-new} discussion
10371: @cindex @code{construct} discussion
10372: You can create and initialize an object of a class on the heap with
10373: @code{heap-new} ( ... class -- object ) and in the dictionary
10374: (allocation with @code{allot}) with @code{dict-new} (
10375: ... class -- object ). Both words invoke @code{construct}, which
10376: consumes the stack items indicated by "..." above.
10377:
10378: @cindex @code{init-object} discussion
10379: @cindex @code{class-inst-size} discussion
10380: If you want to allocate memory for an object yourself, you can get its
10381: alignment and size with @code{class-inst-size 2@@} ( class --
10382: align size ). Once you have memory for an object, you can initialize
10383: it with @code{init-object} ( ... class object -- );
10384: @code{construct} does only a part of the necessary work.
10385:
10386: @node Object-Oriented Programming Style, Class Binding, Creating objects, Objects
10387: @subsubsection Object-Oriented Programming Style
10388: @cindex object-oriented programming style
10389: @cindex programming style, object-oriented
1.5 anton 10390:
1.78 anton 10391: This section is not exhaustive.
1.5 anton 10392:
1.78 anton 10393: @cindex stack effects of selectors
10394: @cindex selectors and stack effects
10395: In general, it is a good idea to ensure that all methods for the
10396: same selector have the same stack effect: when you invoke a selector,
10397: you often have no idea which method will be invoked, so, unless all
10398: methods have the same stack effect, you will not know the stack effect
10399: of the selector invocation.
1.5 anton 10400:
1.78 anton 10401: One exception to this rule is methods for the selector
10402: @code{construct}. We know which method is invoked, because we
10403: specify the class to be constructed at the same place. Actually, I
10404: defined @code{construct} as a selector only to give the users a
10405: convenient way to specify initialization. The way it is used, a
10406: mechanism different from selector invocation would be more natural
10407: (but probably would take more code and more space to explain).
1.5 anton 10408:
1.78 anton 10409: @node Class Binding, Method conveniences, Object-Oriented Programming Style, Objects
10410: @subsubsection Class Binding
10411: @cindex class binding
10412: @cindex early binding
1.5 anton 10413:
1.78 anton 10414: @cindex late binding
10415: Normal selector invocations determine the method at run-time depending
10416: on the class of the receiving object. This run-time selection is called
10417: @i{late binding}.
1.5 anton 10418:
1.78 anton 10419: Sometimes it's preferable to invoke a different method. For example,
10420: you might want to use the simple method for @code{print}ing
10421: @code{object}s instead of the possibly long-winded @code{print} method
10422: of the receiver class. You can achieve this by replacing the invocation
10423: of @code{print} with:
1.5 anton 10424:
1.78 anton 10425: @cindex @code{[bind]} usage
1.5 anton 10426: @example
1.78 anton 10427: [bind] object print
1.5 anton 10428: @end example
10429:
1.78 anton 10430: @noindent
10431: in compiled code or:
10432:
10433: @cindex @code{bind} usage
1.5 anton 10434: @example
1.78 anton 10435: bind object print
1.5 anton 10436: @end example
10437:
1.78 anton 10438: @cindex class binding, alternative to
10439: @noindent
10440: in interpreted code. Alternatively, you can define the method with a
10441: name (e.g., @code{print-object}), and then invoke it through the
10442: name. Class binding is just a (often more convenient) way to achieve
10443: the same effect; it avoids name clutter and allows you to invoke
10444: methods directly without naming them first.
1.5 anton 10445:
1.78 anton 10446: @cindex superclass binding
10447: @cindex parent class binding
10448: A frequent use of class binding is this: When we define a method
10449: for a selector, we often want the method to do what the selector does
10450: in the parent class, and a little more. There is a special word for
10451: this purpose: @code{[parent]}; @code{[parent]
10452: @emph{selector}} is equivalent to @code{[bind] @emph{parent
10453: selector}}, where @code{@emph{parent}} is the parent
10454: class of the current class. E.g., a method definition might look like:
1.44 crook 10455:
1.78 anton 10456: @cindex @code{[parent]} usage
10457: @example
10458: :noname
10459: dup [parent] foo \ do parent's foo on the receiving object
10460: ... \ do some more
10461: ; overrides foo
10462: @end example
1.6 pazsan 10463:
1.78 anton 10464: @cindex class binding as optimization
10465: In @cite{Object-oriented programming in ANS Forth} (Forth Dimensions,
10466: March 1997), Andrew McKewan presents class binding as an optimization
10467: technique. I recommend not using it for this purpose unless you are in
10468: an emergency. Late binding is pretty fast with this model anyway, so the
10469: benefit of using class binding is small; the cost of using class binding
10470: where it is not appropriate is reduced maintainability.
1.44 crook 10471:
1.78 anton 10472: While we are at programming style questions: You should bind
10473: selectors only to ancestor classes of the receiving object. E.g., say,
10474: you know that the receiving object is of class @code{foo} or its
10475: descendents; then you should bind only to @code{foo} and its
10476: ancestors.
1.12 anton 10477:
1.78 anton 10478: @node Method conveniences, Classes and Scoping, Class Binding, Objects
10479: @subsubsection Method conveniences
10480: @cindex method conveniences
1.44 crook 10481:
1.78 anton 10482: In a method you usually access the receiving object pretty often. If
10483: you define the method as a plain colon definition (e.g., with
10484: @code{:noname}), you may have to do a lot of stack
10485: gymnastics. To avoid this, you can define the method with @code{m:
10486: ... ;m}. E.g., you could define the method for
10487: @code{draw}ing a @code{circle} with
1.6 pazsan 10488:
1.78 anton 10489: @cindex @code{this} usage
10490: @cindex @code{m:} usage
10491: @cindex @code{;m} usage
10492: @example
10493: m: ( x y circle -- )
10494: ( x y ) this circle-radius @@ draw-circle ;m
10495: @end example
1.6 pazsan 10496:
1.78 anton 10497: @cindex @code{exit} in @code{m: ... ;m}
10498: @cindex @code{exitm} discussion
10499: @cindex @code{catch} in @code{m: ... ;m}
10500: When this method is executed, the receiver object is removed from the
10501: stack; you can access it with @code{this} (admittedly, in this
10502: example the use of @code{m: ... ;m} offers no advantage). Note
10503: that I specify the stack effect for the whole method (i.e. including
10504: the receiver object), not just for the code between @code{m:}
10505: and @code{;m}. You cannot use @code{exit} in
10506: @code{m:...;m}; instead, use
10507: @code{exitm}.@footnote{Moreover, for any word that calls
10508: @code{catch} and was defined before loading
10509: @code{objects.fs}, you have to redefine it like I redefined
10510: @code{catch}: @code{: catch this >r catch r> to-this ;}}
1.12 anton 10511:
1.78 anton 10512: @cindex @code{inst-var} usage
10513: You will frequently use sequences of the form @code{this
10514: @emph{field}} (in the example above: @code{this
10515: circle-radius}). If you use the field only in this way, you can
10516: define it with @code{inst-var} and eliminate the
10517: @code{this} before the field name. E.g., the @code{circle}
10518: class above could also be defined with:
1.6 pazsan 10519:
1.78 anton 10520: @example
10521: graphical class
10522: cell% inst-var radius
1.6 pazsan 10523:
1.78 anton 10524: m: ( x y circle -- )
10525: radius @@ draw-circle ;m
10526: overrides draw
1.6 pazsan 10527:
1.78 anton 10528: m: ( n-radius circle -- )
10529: radius ! ;m
10530: overrides construct
1.6 pazsan 10531:
1.78 anton 10532: end-class circle
10533: @end example
1.6 pazsan 10534:
1.78 anton 10535: @code{radius} can only be used in @code{circle} and its
10536: descendent classes and inside @code{m:...;m}.
1.6 pazsan 10537:
1.78 anton 10538: @cindex @code{inst-value} usage
10539: You can also define fields with @code{inst-value}, which is
10540: to @code{inst-var} what @code{value} is to
10541: @code{variable}. You can change the value of such a field with
10542: @code{[to-inst]}. E.g., we could also define the class
10543: @code{circle} like this:
1.44 crook 10544:
1.78 anton 10545: @example
10546: graphical class
10547: inst-value radius
1.6 pazsan 10548:
1.78 anton 10549: m: ( x y circle -- )
10550: radius draw-circle ;m
10551: overrides draw
1.44 crook 10552:
1.78 anton 10553: m: ( n-radius circle -- )
10554: [to-inst] radius ;m
10555: overrides construct
1.6 pazsan 10556:
1.78 anton 10557: end-class circle
10558: @end example
1.6 pazsan 10559:
1.78 anton 10560: @c !! :m is easy to confuse with m:. Another name would be better.
1.6 pazsan 10561:
1.78 anton 10562: @c Finally, you can define named methods with @code{:m}. One use of this
10563: @c feature is the definition of words that occur only in one class and are
10564: @c not intended to be overridden, but which still need method context
10565: @c (e.g., for accessing @code{inst-var}s). Another use is for methods that
10566: @c would be bound frequently, if defined anonymously.
1.6 pazsan 10567:
10568:
1.78 anton 10569: @node Classes and Scoping, Dividing classes, Method conveniences, Objects
10570: @subsubsection Classes and Scoping
10571: @cindex classes and scoping
10572: @cindex scoping and classes
1.6 pazsan 10573:
1.78 anton 10574: Inheritance is frequent, unlike structure extension. This exacerbates
10575: the problem with the field name convention (@pxref{Structure Naming
10576: Convention}): One always has to remember in which class the field was
10577: originally defined; changing a part of the class structure would require
10578: changes for renaming in otherwise unaffected code.
1.6 pazsan 10579:
1.78 anton 10580: @cindex @code{inst-var} visibility
10581: @cindex @code{inst-value} visibility
10582: To solve this problem, I added a scoping mechanism (which was not in my
10583: original charter): A field defined with @code{inst-var} (or
10584: @code{inst-value}) is visible only in the class where it is defined and in
10585: the descendent classes of this class. Using such fields only makes
10586: sense in @code{m:}-defined methods in these classes anyway.
1.6 pazsan 10587:
1.78 anton 10588: This scoping mechanism allows us to use the unadorned field name,
10589: because name clashes with unrelated words become much less likely.
1.6 pazsan 10590:
1.78 anton 10591: @cindex @code{protected} discussion
10592: @cindex @code{private} discussion
10593: Once we have this mechanism, we can also use it for controlling the
10594: visibility of other words: All words defined after
10595: @code{protected} are visible only in the current class and its
10596: descendents. @code{public} restores the compilation
10597: (i.e. @code{current}) word list that was in effect before. If you
10598: have several @code{protected}s without an intervening
10599: @code{public} or @code{set-current}, @code{public}
10600: will restore the compilation word list in effect before the first of
10601: these @code{protected}s.
1.6 pazsan 10602:
1.78 anton 10603: @node Dividing classes, Object Interfaces, Classes and Scoping, Objects
10604: @subsubsection Dividing classes
10605: @cindex Dividing classes
10606: @cindex @code{methods}...@code{end-methods}
1.6 pazsan 10607:
1.78 anton 10608: You may want to do the definition of methods separate from the
10609: definition of the class, its selectors, fields, and instance variables,
10610: i.e., separate the implementation from the definition. You can do this
10611: in the following way:
1.6 pazsan 10612:
1.78 anton 10613: @example
10614: graphical class
10615: inst-value radius
10616: end-class circle
1.6 pazsan 10617:
1.78 anton 10618: ... \ do some other stuff
1.6 pazsan 10619:
1.78 anton 10620: circle methods \ now we are ready
1.44 crook 10621:
1.78 anton 10622: m: ( x y circle -- )
10623: radius draw-circle ;m
10624: overrides draw
1.6 pazsan 10625:
1.78 anton 10626: m: ( n-radius circle -- )
10627: [to-inst] radius ;m
10628: overrides construct
1.44 crook 10629:
1.78 anton 10630: end-methods
10631: @end example
1.7 pazsan 10632:
1.78 anton 10633: You can use several @code{methods}...@code{end-methods} sections. The
10634: only things you can do to the class in these sections are: defining
10635: methods, and overriding the class's selectors. You must not define new
10636: selectors or fields.
1.7 pazsan 10637:
1.78 anton 10638: Note that you often have to override a selector before using it. In
10639: particular, you usually have to override @code{construct} with a new
10640: method before you can invoke @code{heap-new} and friends. E.g., you
10641: must not create a circle before the @code{overrides construct} sequence
10642: in the example above.
1.7 pazsan 10643:
1.78 anton 10644: @node Object Interfaces, Objects Implementation, Dividing classes, Objects
10645: @subsubsection Object Interfaces
10646: @cindex object interfaces
10647: @cindex interfaces for objects
1.7 pazsan 10648:
1.78 anton 10649: In this model you can only call selectors defined in the class of the
10650: receiving objects or in one of its ancestors. If you call a selector
10651: with a receiving object that is not in one of these classes, the
10652: result is undefined; if you are lucky, the program crashes
10653: immediately.
1.7 pazsan 10654:
1.78 anton 10655: @cindex selectors common to hardly-related classes
10656: Now consider the case when you want to have a selector (or several)
10657: available in two classes: You would have to add the selector to a
10658: common ancestor class, in the worst case to @code{object}. You
10659: may not want to do this, e.g., because someone else is responsible for
10660: this ancestor class.
1.7 pazsan 10661:
1.78 anton 10662: The solution for this problem is interfaces. An interface is a
10663: collection of selectors. If a class implements an interface, the
10664: selectors become available to the class and its descendents. A class
10665: can implement an unlimited number of interfaces. For the problem
10666: discussed above, we would define an interface for the selector(s), and
10667: both classes would implement the interface.
1.7 pazsan 10668:
1.78 anton 10669: As an example, consider an interface @code{storage} for
10670: writing objects to disk and getting them back, and a class
10671: @code{foo} that implements it. The code would look like this:
1.7 pazsan 10672:
1.78 anton 10673: @cindex @code{interface} usage
10674: @cindex @code{end-interface} usage
10675: @cindex @code{implementation} usage
10676: @example
10677: interface
10678: selector write ( file object -- )
10679: selector read1 ( file object -- )
10680: end-interface storage
1.13 pazsan 10681:
1.78 anton 10682: bar class
10683: storage implementation
1.13 pazsan 10684:
1.78 anton 10685: ... overrides write
10686: ... overrides read1
10687: ...
10688: end-class foo
10689: @end example
1.13 pazsan 10690:
1.78 anton 10691: @noindent
10692: (I would add a word @code{read} @i{( file -- object )} that uses
10693: @code{read1} internally, but that's beyond the point illustrated
10694: here.)
1.13 pazsan 10695:
1.78 anton 10696: Note that you cannot use @code{protected} in an interface; and
10697: of course you cannot define fields.
1.13 pazsan 10698:
1.78 anton 10699: In the Neon model, all selectors are available for all classes;
10700: therefore it does not need interfaces. The price you pay in this model
10701: is slower late binding, and therefore, added complexity to avoid late
10702: binding.
1.13 pazsan 10703:
1.78 anton 10704: @node Objects Implementation, Objects Glossary, Object Interfaces, Objects
10705: @subsubsection @file{objects.fs} Implementation
10706: @cindex @file{objects.fs} implementation
1.13 pazsan 10707:
1.78 anton 10708: @cindex @code{object-map} discussion
10709: An object is a piece of memory, like one of the data structures
10710: described with @code{struct...end-struct}. It has a field
10711: @code{object-map} that points to the method map for the object's
10712: class.
1.13 pazsan 10713:
1.78 anton 10714: @cindex method map
10715: @cindex virtual function table
10716: The @emph{method map}@footnote{This is Self terminology; in C++
10717: terminology: virtual function table.} is an array that contains the
10718: execution tokens (@i{xt}s) of the methods for the object's class. Each
10719: selector contains an offset into a method map.
1.13 pazsan 10720:
1.78 anton 10721: @cindex @code{selector} implementation, class
10722: @code{selector} is a defining word that uses
10723: @code{CREATE} and @code{DOES>}. The body of the
10724: selector contains the offset; the @code{DOES>} action for a
10725: class selector is, basically:
1.8 pazsan 10726:
10727: @example
1.78 anton 10728: ( object addr ) @@ over object-map @@ + @@ execute
1.13 pazsan 10729: @end example
10730:
1.78 anton 10731: Since @code{object-map} is the first field of the object, it
10732: does not generate any code. As you can see, calling a selector has a
10733: small, constant cost.
1.26 crook 10734:
1.78 anton 10735: @cindex @code{current-interface} discussion
10736: @cindex class implementation and representation
10737: A class is basically a @code{struct} combined with a method
10738: map. During the class definition the alignment and size of the class
10739: are passed on the stack, just as with @code{struct}s, so
10740: @code{field} can also be used for defining class
10741: fields. However, passing more items on the stack would be
10742: inconvenient, so @code{class} builds a data structure in memory,
10743: which is accessed through the variable
10744: @code{current-interface}. After its definition is complete, the
10745: class is represented on the stack by a pointer (e.g., as parameter for
10746: a child class definition).
1.26 crook 10747:
1.78 anton 10748: A new class starts off with the alignment and size of its parent,
10749: and a copy of the parent's method map. Defining new fields extends the
10750: size and alignment; likewise, defining new selectors extends the
10751: method map. @code{overrides} just stores a new @i{xt} in the method
10752: map at the offset given by the selector.
1.13 pazsan 10753:
1.78 anton 10754: @cindex class binding, implementation
10755: Class binding just gets the @i{xt} at the offset given by the selector
10756: from the class's method map and @code{compile,}s (in the case of
10757: @code{[bind]}) it.
1.13 pazsan 10758:
1.78 anton 10759: @cindex @code{this} implementation
10760: @cindex @code{catch} and @code{this}
10761: @cindex @code{this} and @code{catch}
10762: I implemented @code{this} as a @code{value}. At the
10763: start of an @code{m:...;m} method the old @code{this} is
10764: stored to the return stack and restored at the end; and the object on
10765: the TOS is stored @code{TO this}. This technique has one
10766: disadvantage: If the user does not leave the method via
10767: @code{;m}, but via @code{throw} or @code{exit},
10768: @code{this} is not restored (and @code{exit} may
10769: crash). To deal with the @code{throw} problem, I have redefined
10770: @code{catch} to save and restore @code{this}; the same
10771: should be done with any word that can catch an exception. As for
10772: @code{exit}, I simply forbid it (as a replacement, there is
10773: @code{exitm}).
1.13 pazsan 10774:
1.78 anton 10775: @cindex @code{inst-var} implementation
10776: @code{inst-var} is just the same as @code{field}, with
10777: a different @code{DOES>} action:
1.13 pazsan 10778: @example
1.78 anton 10779: @@ this +
1.8 pazsan 10780: @end example
1.78 anton 10781: Similar for @code{inst-value}.
1.8 pazsan 10782:
1.78 anton 10783: @cindex class scoping implementation
10784: Each class also has a word list that contains the words defined with
10785: @code{inst-var} and @code{inst-value}, and its protected
10786: words. It also has a pointer to its parent. @code{class} pushes
10787: the word lists of the class and all its ancestors onto the search order stack,
10788: and @code{end-class} drops them.
1.20 pazsan 10789:
1.78 anton 10790: @cindex interface implementation
10791: An interface is like a class without fields, parent and protected
10792: words; i.e., it just has a method map. If a class implements an
10793: interface, its method map contains a pointer to the method map of the
10794: interface. The positive offsets in the map are reserved for class
10795: methods, therefore interface map pointers have negative
10796: offsets. Interfaces have offsets that are unique throughout the
10797: system, unlike class selectors, whose offsets are only unique for the
10798: classes where the selector is available (invokable).
1.20 pazsan 10799:
1.78 anton 10800: This structure means that interface selectors have to perform one
10801: indirection more than class selectors to find their method. Their body
10802: contains the interface map pointer offset in the class method map, and
10803: the method offset in the interface method map. The
10804: @code{does>} action for an interface selector is, basically:
1.20 pazsan 10805:
10806: @example
1.78 anton 10807: ( object selector-body )
10808: 2dup selector-interface @@ ( object selector-body object interface-offset )
10809: swap object-map @@ + @@ ( object selector-body map )
10810: swap selector-offset @@ + @@ execute
1.20 pazsan 10811: @end example
10812:
1.78 anton 10813: where @code{object-map} and @code{selector-offset} are
10814: first fields and generate no code.
1.20 pazsan 10815:
1.78 anton 10816: As a concrete example, consider the following code:
1.20 pazsan 10817:
10818: @example
1.78 anton 10819: interface
10820: selector if1sel1
10821: selector if1sel2
10822: end-interface if1
1.20 pazsan 10823:
1.78 anton 10824: object class
10825: if1 implementation
10826: selector cl1sel1
10827: cell% inst-var cl1iv1
1.20 pazsan 10828:
1.78 anton 10829: ' m1 overrides construct
10830: ' m2 overrides if1sel1
10831: ' m3 overrides if1sel2
10832: ' m4 overrides cl1sel2
10833: end-class cl1
1.20 pazsan 10834:
1.78 anton 10835: create obj1 object dict-new drop
10836: create obj2 cl1 dict-new drop
10837: @end example
1.20 pazsan 10838:
1.78 anton 10839: The data structure created by this code (including the data structure
10840: for @code{object}) is shown in the
10841: @uref{objects-implementation.eps,figure}, assuming a cell size of 4.
10842: @comment TODO add this diagram..
1.20 pazsan 10843:
1.78 anton 10844: @node Objects Glossary, , Objects Implementation, Objects
10845: @subsubsection @file{objects.fs} Glossary
10846: @cindex @file{objects.fs} Glossary
1.20 pazsan 10847:
10848:
1.78 anton 10849: doc---objects-bind
10850: doc---objects-<bind>
10851: doc---objects-bind'
10852: doc---objects-[bind]
10853: doc---objects-class
10854: doc---objects-class->map
10855: doc---objects-class-inst-size
10856: doc---objects-class-override!
1.79 anton 10857: doc---objects-class-previous
10858: doc---objects-class>order
1.78 anton 10859: doc---objects-construct
10860: doc---objects-current'
10861: doc---objects-[current]
10862: doc---objects-current-interface
10863: doc---objects-dict-new
10864: doc---objects-end-class
10865: doc---objects-end-class-noname
10866: doc---objects-end-interface
10867: doc---objects-end-interface-noname
10868: doc---objects-end-methods
10869: doc---objects-exitm
10870: doc---objects-heap-new
10871: doc---objects-implementation
10872: doc---objects-init-object
10873: doc---objects-inst-value
10874: doc---objects-inst-var
10875: doc---objects-interface
10876: doc---objects-m:
10877: doc---objects-:m
10878: doc---objects-;m
10879: doc---objects-method
10880: doc---objects-methods
10881: doc---objects-object
10882: doc---objects-overrides
10883: doc---objects-[parent]
10884: doc---objects-print
10885: doc---objects-protected
10886: doc---objects-public
10887: doc---objects-selector
10888: doc---objects-this
10889: doc---objects-<to-inst>
10890: doc---objects-[to-inst]
10891: doc---objects-to-this
10892: doc---objects-xt-new
1.20 pazsan 10893:
10894:
1.78 anton 10895: @c -------------------------------------------------------------
10896: @node OOF, Mini-OOF, Objects, Object-oriented Forth
10897: @subsection The @file{oof.fs} model
10898: @cindex oof
10899: @cindex object-oriented programming
1.20 pazsan 10900:
1.78 anton 10901: @cindex @file{objects.fs}
10902: @cindex @file{oof.fs}
1.20 pazsan 10903:
1.78 anton 10904: This section describes the @file{oof.fs} package.
1.20 pazsan 10905:
1.78 anton 10906: The package described in this section has been used in bigFORTH since 1991, and
10907: used for two large applications: a chromatographic system used to
10908: create new medicaments, and a graphic user interface library (MINOS).
1.20 pazsan 10909:
1.78 anton 10910: You can find a description (in German) of @file{oof.fs} in @cite{Object
10911: oriented bigFORTH} by Bernd Paysan, published in @cite{Vierte Dimension}
10912: 10(2), 1994.
1.20 pazsan 10913:
1.78 anton 10914: @menu
10915: * Properties of the OOF model::
10916: * Basic OOF Usage::
10917: * The OOF base class::
10918: * Class Declaration::
10919: * Class Implementation::
10920: @end menu
1.20 pazsan 10921:
1.78 anton 10922: @node Properties of the OOF model, Basic OOF Usage, OOF, OOF
10923: @subsubsection Properties of the @file{oof.fs} model
10924: @cindex @file{oof.fs} properties
1.20 pazsan 10925:
1.78 anton 10926: @itemize @bullet
10927: @item
10928: This model combines object oriented programming with information
10929: hiding. It helps you writing large application, where scoping is
10930: necessary, because it provides class-oriented scoping.
1.20 pazsan 10931:
1.78 anton 10932: @item
10933: Named objects, object pointers, and object arrays can be created,
10934: selector invocation uses the ``object selector'' syntax. Selector invocation
10935: to objects and/or selectors on the stack is a bit less convenient, but
10936: possible.
1.44 crook 10937:
1.78 anton 10938: @item
10939: Selector invocation and instance variable usage of the active object is
10940: straightforward, since both make use of the active object.
1.44 crook 10941:
1.78 anton 10942: @item
10943: Late binding is efficient and easy to use.
1.20 pazsan 10944:
1.78 anton 10945: @item
10946: State-smart objects parse selectors. However, extensibility is provided
10947: using a (parsing) selector @code{postpone} and a selector @code{'}.
1.20 pazsan 10948:
1.78 anton 10949: @item
10950: An implementation in ANS Forth is available.
1.20 pazsan 10951:
1.78 anton 10952: @end itemize
1.23 crook 10953:
10954:
1.78 anton 10955: @node Basic OOF Usage, The OOF base class, Properties of the OOF model, OOF
10956: @subsubsection Basic @file{oof.fs} Usage
10957: @cindex @file{oof.fs} usage
1.23 crook 10958:
1.78 anton 10959: This section uses the same example as for @code{objects} (@pxref{Basic Objects Usage}).
1.23 crook 10960:
1.78 anton 10961: You can define a class for graphical objects like this:
1.23 crook 10962:
1.78 anton 10963: @cindex @code{class} usage
10964: @cindex @code{class;} usage
10965: @cindex @code{method} usage
10966: @example
10967: object class graphical \ "object" is the parent class
1.139 pazsan 10968: method draw ( x y -- )
1.78 anton 10969: class;
10970: @end example
1.23 crook 10971:
1.78 anton 10972: This code defines a class @code{graphical} with an
10973: operation @code{draw}. We can perform the operation
10974: @code{draw} on any @code{graphical} object, e.g.:
1.23 crook 10975:
1.78 anton 10976: @example
10977: 100 100 t-rex draw
10978: @end example
1.23 crook 10979:
1.78 anton 10980: @noindent
10981: where @code{t-rex} is an object or object pointer, created with e.g.
10982: @code{graphical : t-rex}.
1.23 crook 10983:
1.78 anton 10984: @cindex abstract class
10985: How do we create a graphical object? With the present definitions,
10986: we cannot create a useful graphical object. The class
10987: @code{graphical} describes graphical objects in general, but not
10988: any concrete graphical object type (C++ users would call it an
10989: @emph{abstract class}); e.g., there is no method for the selector
10990: @code{draw} in the class @code{graphical}.
1.23 crook 10991:
1.78 anton 10992: For concrete graphical objects, we define child classes of the
10993: class @code{graphical}, e.g.:
1.23 crook 10994:
1.78 anton 10995: @example
10996: graphical class circle \ "graphical" is the parent class
10997: cell var circle-radius
10998: how:
10999: : draw ( x y -- )
11000: circle-radius @@ draw-circle ;
1.23 crook 11001:
1.139 pazsan 11002: : init ( n-radius -- )
1.78 anton 11003: circle-radius ! ;
11004: class;
11005: @end example
1.1 anton 11006:
1.78 anton 11007: Here we define a class @code{circle} as a child of @code{graphical},
11008: with a field @code{circle-radius}; it defines new methods for the
11009: selectors @code{draw} and @code{init} (@code{init} is defined in
11010: @code{object}, the parent class of @code{graphical}).
1.1 anton 11011:
1.78 anton 11012: Now we can create a circle in the dictionary with:
1.1 anton 11013:
1.78 anton 11014: @example
11015: 50 circle : my-circle
11016: @end example
1.21 crook 11017:
1.78 anton 11018: @noindent
11019: @code{:} invokes @code{init}, thus initializing the field
11020: @code{circle-radius} with 50. We can draw this new circle at (100,100)
11021: with:
1.1 anton 11022:
1.78 anton 11023: @example
11024: 100 100 my-circle draw
11025: @end example
1.1 anton 11026:
1.78 anton 11027: @cindex selector invocation, restrictions
11028: @cindex class definition, restrictions
11029: Note: You can only invoke a selector if the receiving object belongs to
11030: the class where the selector was defined or one of its descendents;
11031: e.g., you can invoke @code{draw} only for objects belonging to
11032: @code{graphical} or its descendents (e.g., @code{circle}). The scoping
11033: mechanism will check if you try to invoke a selector that is not
11034: defined in this class hierarchy, so you'll get an error at compilation
11035: time.
1.1 anton 11036:
11037:
1.78 anton 11038: @node The OOF base class, Class Declaration, Basic OOF Usage, OOF
11039: @subsubsection The @file{oof.fs} base class
11040: @cindex @file{oof.fs} base class
1.1 anton 11041:
1.78 anton 11042: When you define a class, you have to specify a parent class. So how do
11043: you start defining classes? There is one class available from the start:
11044: @code{object}. You have to use it as ancestor for all classes. It is the
11045: only class that has no parent. Classes are also objects, except that
11046: they don't have instance variables; class manipulation such as
11047: inheritance or changing definitions of a class is handled through
11048: selectors of the class @code{object}.
1.1 anton 11049:
1.78 anton 11050: @code{object} provides a number of selectors:
1.1 anton 11051:
1.78 anton 11052: @itemize @bullet
11053: @item
11054: @code{class} for subclassing, @code{definitions} to add definitions
11055: later on, and @code{class?} to get type informations (is the class a
11056: subclass of the class passed on the stack?).
1.1 anton 11057:
1.78 anton 11058: doc---object-class
11059: doc---object-definitions
11060: doc---object-class?
1.1 anton 11061:
11062:
1.26 crook 11063: @item
1.78 anton 11064: @code{init} and @code{dispose} as constructor and destructor of the
11065: object. @code{init} is invocated after the object's memory is allocated,
11066: while @code{dispose} also handles deallocation. Thus if you redefine
11067: @code{dispose}, you have to call the parent's dispose with @code{super
11068: dispose}, too.
11069:
11070: doc---object-init
11071: doc---object-dispose
11072:
1.1 anton 11073:
1.26 crook 11074: @item
1.78 anton 11075: @code{new}, @code{new[]}, @code{:}, @code{ptr}, @code{asptr}, and
11076: @code{[]} to create named and unnamed objects and object arrays or
11077: object pointers.
11078:
11079: doc---object-new
11080: doc---object-new[]
11081: doc---object-:
11082: doc---object-ptr
11083: doc---object-asptr
11084: doc---object-[]
11085:
1.1 anton 11086:
1.26 crook 11087: @item
1.78 anton 11088: @code{::} and @code{super} for explicit scoping. You should use explicit
11089: scoping only for super classes or classes with the same set of instance
11090: variables. Explicitly-scoped selectors use early binding.
1.21 crook 11091:
1.78 anton 11092: doc---object-::
11093: doc---object-super
1.21 crook 11094:
11095:
1.26 crook 11096: @item
1.78 anton 11097: @code{self} to get the address of the object
1.21 crook 11098:
1.78 anton 11099: doc---object-self
1.21 crook 11100:
11101:
1.78 anton 11102: @item
11103: @code{bind}, @code{bound}, @code{link}, and @code{is} to assign object
11104: pointers and instance defers.
1.21 crook 11105:
1.78 anton 11106: doc---object-bind
11107: doc---object-bound
11108: doc---object-link
11109: doc---object-is
1.21 crook 11110:
11111:
1.78 anton 11112: @item
11113: @code{'} to obtain selector tokens, @code{send} to invocate selectors
11114: form the stack, and @code{postpone} to generate selector invocation code.
1.21 crook 11115:
1.78 anton 11116: doc---object-'
11117: doc---object-postpone
1.21 crook 11118:
11119:
1.78 anton 11120: @item
11121: @code{with} and @code{endwith} to select the active object from the
11122: stack, and enable its scope. Using @code{with} and @code{endwith}
11123: also allows you to create code using selector @code{postpone} without being
11124: trapped by the state-smart objects.
1.21 crook 11125:
1.78 anton 11126: doc---object-with
11127: doc---object-endwith
1.21 crook 11128:
11129:
1.78 anton 11130: @end itemize
1.21 crook 11131:
1.78 anton 11132: @node Class Declaration, Class Implementation, The OOF base class, OOF
11133: @subsubsection Class Declaration
11134: @cindex class declaration
1.21 crook 11135:
1.78 anton 11136: @itemize @bullet
11137: @item
11138: Instance variables
1.21 crook 11139:
1.78 anton 11140: doc---oof-var
1.21 crook 11141:
11142:
1.78 anton 11143: @item
11144: Object pointers
1.21 crook 11145:
1.78 anton 11146: doc---oof-ptr
11147: doc---oof-asptr
1.21 crook 11148:
11149:
1.78 anton 11150: @item
11151: Instance defers
1.21 crook 11152:
1.78 anton 11153: doc---oof-defer
1.21 crook 11154:
11155:
1.78 anton 11156: @item
11157: Method selectors
1.21 crook 11158:
1.78 anton 11159: doc---oof-early
11160: doc---oof-method
1.21 crook 11161:
11162:
1.78 anton 11163: @item
11164: Class-wide variables
1.21 crook 11165:
1.78 anton 11166: doc---oof-static
1.21 crook 11167:
11168:
1.78 anton 11169: @item
11170: End declaration
1.1 anton 11171:
1.78 anton 11172: doc---oof-how:
11173: doc---oof-class;
1.21 crook 11174:
11175:
1.78 anton 11176: @end itemize
1.21 crook 11177:
1.78 anton 11178: @c -------------------------------------------------------------
11179: @node Class Implementation, , Class Declaration, OOF
11180: @subsubsection Class Implementation
11181: @cindex class implementation
1.21 crook 11182:
1.78 anton 11183: @c -------------------------------------------------------------
11184: @node Mini-OOF, Comparison with other object models, OOF, Object-oriented Forth
11185: @subsection The @file{mini-oof.fs} model
11186: @cindex mini-oof
1.21 crook 11187:
1.78 anton 11188: Gforth's third object oriented Forth package is a 12-liner. It uses a
1.79 anton 11189: mixture of the @file{objects.fs} and the @file{oof.fs} syntax,
1.78 anton 11190: and reduces to the bare minimum of features. This is based on a posting
11191: of Bernd Paysan in comp.lang.forth.
1.21 crook 11192:
1.78 anton 11193: @menu
11194: * Basic Mini-OOF Usage::
11195: * Mini-OOF Example::
11196: * Mini-OOF Implementation::
11197: @end menu
1.21 crook 11198:
1.78 anton 11199: @c -------------------------------------------------------------
11200: @node Basic Mini-OOF Usage, Mini-OOF Example, Mini-OOF, Mini-OOF
11201: @subsubsection Basic @file{mini-oof.fs} Usage
11202: @cindex mini-oof usage
1.21 crook 11203:
1.78 anton 11204: There is a base class (@code{class}, which allocates one cell for the
11205: object pointer) plus seven other words: to define a method, a variable,
11206: a class; to end a class, to resolve binding, to allocate an object and
11207: to compile a class method.
11208: @comment TODO better description of the last one
1.26 crook 11209:
1.21 crook 11210:
1.78 anton 11211: doc-object
11212: doc-method
11213: doc-var
11214: doc-class
11215: doc-end-class
11216: doc-defines
11217: doc-new
11218: doc-::
1.21 crook 11219:
11220:
11221:
1.78 anton 11222: @c -------------------------------------------------------------
11223: @node Mini-OOF Example, Mini-OOF Implementation, Basic Mini-OOF Usage, Mini-OOF
11224: @subsubsection Mini-OOF Example
11225: @cindex mini-oof example
1.1 anton 11226:
1.78 anton 11227: A short example shows how to use this package. This example, in slightly
11228: extended form, is supplied as @file{moof-exm.fs}
11229: @comment TODO could flesh this out with some comments from the Forthwrite article
1.20 pazsan 11230:
1.26 crook 11231: @example
1.78 anton 11232: object class
11233: method init
11234: method draw
11235: end-class graphical
1.26 crook 11236: @end example
1.20 pazsan 11237:
1.78 anton 11238: This code defines a class @code{graphical} with an
11239: operation @code{draw}. We can perform the operation
11240: @code{draw} on any @code{graphical} object, e.g.:
1.20 pazsan 11241:
1.26 crook 11242: @example
1.78 anton 11243: 100 100 t-rex draw
1.26 crook 11244: @end example
1.12 anton 11245:
1.78 anton 11246: where @code{t-rex} is an object or object pointer, created with e.g.
11247: @code{graphical new Constant t-rex}.
1.12 anton 11248:
1.78 anton 11249: For concrete graphical objects, we define child classes of the
11250: class @code{graphical}, e.g.:
1.12 anton 11251:
1.26 crook 11252: @example
11253: graphical class
1.78 anton 11254: cell var circle-radius
11255: end-class circle \ "graphical" is the parent class
1.12 anton 11256:
1.78 anton 11257: :noname ( x y -- )
11258: circle-radius @@ draw-circle ; circle defines draw
11259: :noname ( r -- )
11260: circle-radius ! ; circle defines init
11261: @end example
1.12 anton 11262:
1.78 anton 11263: There is no implicit init method, so we have to define one. The creation
11264: code of the object now has to call init explicitely.
1.21 crook 11265:
1.78 anton 11266: @example
11267: circle new Constant my-circle
11268: 50 my-circle init
1.12 anton 11269: @end example
11270:
1.78 anton 11271: It is also possible to add a function to create named objects with
11272: automatic call of @code{init}, given that all objects have @code{init}
11273: on the same place:
1.38 anton 11274:
1.78 anton 11275: @example
11276: : new: ( .. o "name" -- )
11277: new dup Constant init ;
11278: 80 circle new: large-circle
11279: @end example
1.12 anton 11280:
1.78 anton 11281: We can draw this new circle at (100,100) with:
1.12 anton 11282:
1.78 anton 11283: @example
11284: 100 100 my-circle draw
11285: @end example
1.12 anton 11286:
1.78 anton 11287: @node Mini-OOF Implementation, , Mini-OOF Example, Mini-OOF
11288: @subsubsection @file{mini-oof.fs} Implementation
1.12 anton 11289:
1.78 anton 11290: Object-oriented systems with late binding typically use a
11291: ``vtable''-approach: the first variable in each object is a pointer to a
11292: table, which contains the methods as function pointers. The vtable
11293: may also contain other information.
1.12 anton 11294:
1.79 anton 11295: So first, let's declare selectors:
1.37 anton 11296:
11297: @example
1.79 anton 11298: : method ( m v "name" -- m' v ) Create over , swap cell+ swap
1.78 anton 11299: DOES> ( ... o -- ... ) @@ over @@ + @@ execute ;
11300: @end example
1.37 anton 11301:
1.79 anton 11302: During selector declaration, the number of selectors and instance
11303: variables is on the stack (in address units). @code{method} creates one
11304: selector and increments the selector number. To execute a selector, it
1.78 anton 11305: takes the object, fetches the vtable pointer, adds the offset, and
1.79 anton 11306: executes the method @i{xt} stored there. Each selector takes the object
11307: it is invoked with as top of stack parameter; it passes the parameters
11308: (including the object) unchanged to the appropriate method which should
1.78 anton 11309: consume that object.
1.37 anton 11310:
1.78 anton 11311: Now, we also have to declare instance variables
1.37 anton 11312:
1.78 anton 11313: @example
1.79 anton 11314: : var ( m v size "name" -- m v' ) Create over , +
1.78 anton 11315: DOES> ( o -- addr ) @@ + ;
1.37 anton 11316: @end example
11317:
1.78 anton 11318: As before, a word is created with the current offset. Instance
11319: variables can have different sizes (cells, floats, doubles, chars), so
11320: all we do is take the size and add it to the offset. If your machine
11321: has alignment restrictions, put the proper @code{aligned} or
11322: @code{faligned} before the variable, to adjust the variable
11323: offset. That's why it is on the top of stack.
1.37 anton 11324:
1.78 anton 11325: We need a starting point (the base object) and some syntactic sugar:
1.37 anton 11326:
1.78 anton 11327: @example
11328: Create object 1 cells , 2 cells ,
1.79 anton 11329: : class ( class -- class selectors vars ) dup 2@@ ;
1.78 anton 11330: @end example
1.12 anton 11331:
1.78 anton 11332: For inheritance, the vtable of the parent object has to be
11333: copied when a new, derived class is declared. This gives all the
11334: methods of the parent class, which can be overridden, though.
1.12 anton 11335:
1.78 anton 11336: @example
1.79 anton 11337: : end-class ( class selectors vars "name" -- )
1.78 anton 11338: Create here >r , dup , 2 cells ?DO ['] noop , 1 cells +LOOP
11339: cell+ dup cell+ r> rot @@ 2 cells /string move ;
11340: @end example
1.12 anton 11341:
1.78 anton 11342: The first line creates the vtable, initialized with
11343: @code{noop}s. The second line is the inheritance mechanism, it
11344: copies the xts from the parent vtable.
1.12 anton 11345:
1.78 anton 11346: We still have no way to define new methods, let's do that now:
1.12 anton 11347:
1.26 crook 11348: @example
1.79 anton 11349: : defines ( xt class "name" -- ) ' >body @@ + ! ;
1.78 anton 11350: @end example
1.12 anton 11351:
1.78 anton 11352: To allocate a new object, we need a word, too:
1.12 anton 11353:
1.78 anton 11354: @example
11355: : new ( class -- o ) here over @@ allot swap over ! ;
1.12 anton 11356: @end example
11357:
1.78 anton 11358: Sometimes derived classes want to access the method of the
11359: parent object. There are two ways to achieve this with Mini-OOF:
11360: first, you could use named words, and second, you could look up the
11361: vtable of the parent object.
1.12 anton 11362:
1.78 anton 11363: @example
11364: : :: ( class "name" -- ) ' >body @@ + @@ compile, ;
11365: @end example
1.12 anton 11366:
11367:
1.78 anton 11368: Nothing can be more confusing than a good example, so here is
11369: one. First let's declare a text object (called
11370: @code{button}), that stores text and position:
1.12 anton 11371:
1.78 anton 11372: @example
11373: object class
11374: cell var text
11375: cell var len
11376: cell var x
11377: cell var y
11378: method init
11379: method draw
11380: end-class button
11381: @end example
1.12 anton 11382:
1.78 anton 11383: @noindent
11384: Now, implement the two methods, @code{draw} and @code{init}:
1.21 crook 11385:
1.26 crook 11386: @example
1.78 anton 11387: :noname ( o -- )
11388: >r r@@ x @@ r@@ y @@ at-xy r@@ text @@ r> len @@ type ;
11389: button defines draw
11390: :noname ( addr u o -- )
11391: >r 0 r@@ x ! 0 r@@ y ! r@@ len ! r> text ! ;
11392: button defines init
1.26 crook 11393: @end example
1.12 anton 11394:
1.78 anton 11395: @noindent
11396: To demonstrate inheritance, we define a class @code{bold-button}, with no
1.79 anton 11397: new data and no new selectors:
1.78 anton 11398:
11399: @example
11400: button class
11401: end-class bold-button
1.12 anton 11402:
1.78 anton 11403: : bold 27 emit ." [1m" ;
11404: : normal 27 emit ." [0m" ;
11405: @end example
1.1 anton 11406:
1.78 anton 11407: @noindent
11408: The class @code{bold-button} has a different draw method to
11409: @code{button}, but the new method is defined in terms of the draw method
11410: for @code{button}:
1.20 pazsan 11411:
1.78 anton 11412: @example
11413: :noname bold [ button :: draw ] normal ; bold-button defines draw
11414: @end example
1.21 crook 11415:
1.78 anton 11416: @noindent
1.79 anton 11417: Finally, create two objects and apply selectors:
1.21 crook 11418:
1.26 crook 11419: @example
1.78 anton 11420: button new Constant foo
11421: s" thin foo" foo init
11422: page
11423: foo draw
11424: bold-button new Constant bar
11425: s" fat bar" bar init
11426: 1 bar y !
11427: bar draw
1.26 crook 11428: @end example
1.21 crook 11429:
11430:
1.78 anton 11431: @node Comparison with other object models, , Mini-OOF, Object-oriented Forth
11432: @subsection Comparison with other object models
11433: @cindex comparison of object models
11434: @cindex object models, comparison
11435:
11436: Many object-oriented Forth extensions have been proposed (@cite{A survey
11437: of object-oriented Forths} (SIGPLAN Notices, April 1996) by Bradford
11438: J. Rodriguez and W. F. S. Poehlman lists 17). This section discusses the
11439: relation of the object models described here to two well-known and two
11440: closely-related (by the use of method maps) models. Andras Zsoter
11441: helped us with this section.
11442:
11443: @cindex Neon model
11444: The most popular model currently seems to be the Neon model (see
11445: @cite{Object-oriented programming in ANS Forth} (Forth Dimensions, March
11446: 1997) by Andrew McKewan) but this model has a number of limitations
11447: @footnote{A longer version of this critique can be
11448: found in @cite{On Standardizing Object-Oriented Forth Extensions} (Forth
11449: Dimensions, May 1997) by Anton Ertl.}:
11450:
11451: @itemize @bullet
11452: @item
11453: It uses a @code{@emph{selector object}} syntax, which makes it unnatural
11454: to pass objects on the stack.
1.21 crook 11455:
1.78 anton 11456: @item
11457: It requires that the selector parses the input stream (at
1.79 anton 11458: compile time); this leads to reduced extensibility and to bugs that are
1.78 anton 11459: hard to find.
1.21 crook 11460:
1.78 anton 11461: @item
1.79 anton 11462: It allows using every selector on every object; this eliminates the
11463: need for interfaces, but makes it harder to create efficient
11464: implementations.
1.78 anton 11465: @end itemize
1.21 crook 11466:
1.78 anton 11467: @cindex Pountain's object-oriented model
11468: Another well-known publication is @cite{Object-Oriented Forth} (Academic
11469: Press, London, 1987) by Dick Pountain. However, it is not really about
11470: object-oriented programming, because it hardly deals with late
11471: binding. Instead, it focuses on features like information hiding and
11472: overloading that are characteristic of modular languages like Ada (83).
1.26 crook 11473:
1.78 anton 11474: @cindex Zsoter's object-oriented model
1.79 anton 11475: In @uref{http://www.forth.org/oopf.html, Does late binding have to be
11476: slow?} (Forth Dimensions 18(1) 1996, pages 31-35) Andras Zsoter
11477: describes a model that makes heavy use of an active object (like
11478: @code{this} in @file{objects.fs}): The active object is not only used
11479: for accessing all fields, but also specifies the receiving object of
11480: every selector invocation; you have to change the active object
11481: explicitly with @code{@{ ... @}}, whereas in @file{objects.fs} it
11482: changes more or less implicitly at @code{m: ... ;m}. Such a change at
11483: the method entry point is unnecessary with Zsoter's model, because the
11484: receiving object is the active object already. On the other hand, the
11485: explicit change is absolutely necessary in that model, because otherwise
11486: no one could ever change the active object. An ANS Forth implementation
11487: of this model is available through
11488: @uref{http://www.forth.org/oopf.html}.
1.21 crook 11489:
1.78 anton 11490: @cindex @file{oof.fs}, differences to other models
11491: The @file{oof.fs} model combines information hiding and overloading
11492: resolution (by keeping names in various word lists) with object-oriented
11493: programming. It sets the active object implicitly on method entry, but
11494: also allows explicit changing (with @code{>o...o>} or with
11495: @code{with...endwith}). It uses parsing and state-smart objects and
11496: classes for resolving overloading and for early binding: the object or
11497: class parses the selector and determines the method from this. If the
11498: selector is not parsed by an object or class, it performs a call to the
11499: selector for the active object (late binding), like Zsoter's model.
11500: Fields are always accessed through the active object. The big
11501: disadvantage of this model is the parsing and the state-smartness, which
11502: reduces extensibility and increases the opportunities for subtle bugs;
11503: essentially, you are only safe if you never tick or @code{postpone} an
11504: object or class (Bernd disagrees, but I (Anton) am not convinced).
1.21 crook 11505:
1.78 anton 11506: @cindex @file{mini-oof.fs}, differences to other models
11507: The @file{mini-oof.fs} model is quite similar to a very stripped-down
11508: version of the @file{objects.fs} model, but syntactically it is a
11509: mixture of the @file{objects.fs} and @file{oof.fs} models.
1.21 crook 11510:
11511:
1.78 anton 11512: @c -------------------------------------------------------------
1.150 anton 11513: @node Programming Tools, C Interface, Object-oriented Forth, Words
1.78 anton 11514: @section Programming Tools
11515: @cindex programming tools
1.21 crook 11516:
1.78 anton 11517: @c !! move this and assembler down below OO stuff.
1.21 crook 11518:
1.78 anton 11519: @menu
1.150 anton 11520: * Examining:: Data and Code.
11521: * Forgetting words:: Usually before reloading.
1.78 anton 11522: * Debugging:: Simple and quick.
11523: * Assertions:: Making your programs self-checking.
11524: * Singlestep Debugger:: Executing your program word by word.
11525: @end menu
1.21 crook 11526:
1.78 anton 11527: @node Examining, Forgetting words, Programming Tools, Programming Tools
11528: @subsection Examining data and code
11529: @cindex examining data and code
11530: @cindex data examination
11531: @cindex code examination
1.44 crook 11532:
1.78 anton 11533: The following words inspect the stack non-destructively:
1.21 crook 11534:
1.78 anton 11535: doc-.s
11536: doc-f.s
1.158 anton 11537: doc-maxdepth-.s
1.44 crook 11538:
1.78 anton 11539: There is a word @code{.r} but it does @i{not} display the return stack!
11540: It is used for formatted numeric output (@pxref{Simple numeric output}).
1.21 crook 11541:
1.78 anton 11542: doc-depth
11543: doc-fdepth
11544: doc-clearstack
1.124 anton 11545: doc-clearstacks
1.21 crook 11546:
1.78 anton 11547: The following words inspect memory.
1.21 crook 11548:
1.78 anton 11549: doc-?
11550: doc-dump
1.21 crook 11551:
1.78 anton 11552: And finally, @code{see} allows to inspect code:
1.21 crook 11553:
1.78 anton 11554: doc-see
11555: doc-xt-see
1.111 anton 11556: doc-simple-see
11557: doc-simple-see-range
1.21 crook 11558:
1.78 anton 11559: @node Forgetting words, Debugging, Examining, Programming Tools
11560: @subsection Forgetting words
11561: @cindex words, forgetting
11562: @cindex forgeting words
1.21 crook 11563:
1.78 anton 11564: @c anton: other, maybe better places for this subsection: Defining Words;
11565: @c Dictionary allocation. At least a reference should be there.
1.21 crook 11566:
1.78 anton 11567: Forth allows you to forget words (and everything that was alloted in the
11568: dictonary after them) in a LIFO manner.
1.21 crook 11569:
1.78 anton 11570: doc-marker
1.21 crook 11571:
1.78 anton 11572: The most common use of this feature is during progam development: when
11573: you change a source file, forget all the words it defined and load it
11574: again (since you also forget everything defined after the source file
11575: was loaded, you have to reload that, too). Note that effects like
11576: storing to variables and destroyed system words are not undone when you
11577: forget words. With a system like Gforth, that is fast enough at
11578: starting up and compiling, I find it more convenient to exit and restart
11579: Gforth, as this gives me a clean slate.
1.21 crook 11580:
1.78 anton 11581: Here's an example of using @code{marker} at the start of a source file
11582: that you are debugging; it ensures that you only ever have one copy of
11583: the file's definitions compiled at any time:
1.21 crook 11584:
1.78 anton 11585: @example
11586: [IFDEF] my-code
11587: my-code
11588: [ENDIF]
1.26 crook 11589:
1.78 anton 11590: marker my-code
11591: init-included-files
1.21 crook 11592:
1.78 anton 11593: \ .. definitions start here
11594: \ .
11595: \ .
11596: \ end
11597: @end example
1.21 crook 11598:
1.26 crook 11599:
1.78 anton 11600: @node Debugging, Assertions, Forgetting words, Programming Tools
11601: @subsection Debugging
11602: @cindex debugging
1.21 crook 11603:
1.78 anton 11604: Languages with a slow edit/compile/link/test development loop tend to
11605: require sophisticated tracing/stepping debuggers to facilate debugging.
1.21 crook 11606:
1.78 anton 11607: A much better (faster) way in fast-compiling languages is to add
11608: printing code at well-selected places, let the program run, look at
11609: the output, see where things went wrong, add more printing code, etc.,
11610: until the bug is found.
1.21 crook 11611:
1.78 anton 11612: The simple debugging aids provided in @file{debugs.fs}
11613: are meant to support this style of debugging.
1.21 crook 11614:
1.78 anton 11615: The word @code{~~} prints debugging information (by default the source
11616: location and the stack contents). It is easy to insert. If you use Emacs
11617: it is also easy to remove (@kbd{C-x ~} in the Emacs Forth mode to
11618: query-replace them with nothing). The deferred words
1.101 anton 11619: @code{printdebugdata} and @code{.debugline} control the output of
1.78 anton 11620: @code{~~}. The default source location output format works well with
11621: Emacs' compilation mode, so you can step through the program at the
11622: source level using @kbd{C-x `} (the advantage over a stepping debugger
11623: is that you can step in any direction and you know where the crash has
11624: happened or where the strange data has occurred).
1.21 crook 11625:
1.78 anton 11626: doc-~~
11627: doc-printdebugdata
1.101 anton 11628: doc-.debugline
1.21 crook 11629:
1.106 anton 11630: @cindex filenames in @code{~~} output
11631: @code{~~} (and assertions) will usually print the wrong file name if a
11632: marker is executed in the same file after their occurance. They will
11633: print @samp{*somewhere*} as file name if a marker is executed in the
11634: same file before their occurance.
11635:
11636:
1.78 anton 11637: @node Assertions, Singlestep Debugger, Debugging, Programming Tools
11638: @subsection Assertions
11639: @cindex assertions
1.21 crook 11640:
1.78 anton 11641: It is a good idea to make your programs self-checking, especially if you
11642: make an assumption that may become invalid during maintenance (for
11643: example, that a certain field of a data structure is never zero). Gforth
11644: supports @dfn{assertions} for this purpose. They are used like this:
1.21 crook 11645:
11646: @example
1.78 anton 11647: assert( @i{flag} )
1.26 crook 11648: @end example
11649:
1.78 anton 11650: The code between @code{assert(} and @code{)} should compute a flag, that
11651: should be true if everything is alright and false otherwise. It should
11652: not change anything else on the stack. The overall stack effect of the
11653: assertion is @code{( -- )}. E.g.
1.21 crook 11654:
1.26 crook 11655: @example
1.78 anton 11656: assert( 1 1 + 2 = ) \ what we learn in school
11657: assert( dup 0<> ) \ assert that the top of stack is not zero
11658: assert( false ) \ this code should not be reached
1.21 crook 11659: @end example
11660:
1.78 anton 11661: The need for assertions is different at different times. During
11662: debugging, we want more checking, in production we sometimes care more
11663: for speed. Therefore, assertions can be turned off, i.e., the assertion
11664: becomes a comment. Depending on the importance of an assertion and the
11665: time it takes to check it, you may want to turn off some assertions and
11666: keep others turned on. Gforth provides several levels of assertions for
11667: this purpose:
11668:
11669:
11670: doc-assert0(
11671: doc-assert1(
11672: doc-assert2(
11673: doc-assert3(
11674: doc-assert(
11675: doc-)
1.21 crook 11676:
11677:
1.78 anton 11678: The variable @code{assert-level} specifies the highest assertions that
11679: are turned on. I.e., at the default @code{assert-level} of one,
11680: @code{assert0(} and @code{assert1(} assertions perform checking, while
11681: @code{assert2(} and @code{assert3(} assertions are treated as comments.
1.26 crook 11682:
1.78 anton 11683: The value of @code{assert-level} is evaluated at compile-time, not at
11684: run-time. Therefore you cannot turn assertions on or off at run-time;
11685: you have to set the @code{assert-level} appropriately before compiling a
11686: piece of code. You can compile different pieces of code at different
11687: @code{assert-level}s (e.g., a trusted library at level 1 and
11688: newly-written code at level 3).
1.26 crook 11689:
11690:
1.78 anton 11691: doc-assert-level
1.26 crook 11692:
11693:
1.78 anton 11694: If an assertion fails, a message compatible with Emacs' compilation mode
11695: is produced and the execution is aborted (currently with @code{ABORT"}.
11696: If there is interest, we will introduce a special throw code. But if you
11697: intend to @code{catch} a specific condition, using @code{throw} is
11698: probably more appropriate than an assertion).
1.106 anton 11699:
11700: @cindex filenames in assertion output
11701: Assertions (and @code{~~}) will usually print the wrong file name if a
11702: marker is executed in the same file after their occurance. They will
11703: print @samp{*somewhere*} as file name if a marker is executed in the
11704: same file before their occurance.
1.44 crook 11705:
1.78 anton 11706: Definitions in ANS Forth for these assertion words are provided
11707: in @file{compat/assert.fs}.
1.26 crook 11708:
1.44 crook 11709:
1.78 anton 11710: @node Singlestep Debugger, , Assertions, Programming Tools
11711: @subsection Singlestep Debugger
11712: @cindex singlestep Debugger
11713: @cindex debugging Singlestep
1.44 crook 11714:
1.159 anton 11715: The singlestep debugger works only with the engine @code{gforth-ditc}.
1.112 anton 11716:
1.78 anton 11717: When you create a new word there's often the need to check whether it
11718: behaves correctly or not. You can do this by typing @code{dbg
11719: badword}. A debug session might look like this:
1.26 crook 11720:
1.78 anton 11721: @example
11722: : badword 0 DO i . LOOP ; ok
11723: 2 dbg badword
11724: : badword
11725: Scanning code...
1.44 crook 11726:
1.78 anton 11727: Nesting debugger ready!
1.44 crook 11728:
1.78 anton 11729: 400D4738 8049BC4 0 -> [ 2 ] 00002 00000
11730: 400D4740 8049F68 DO -> [ 0 ]
11731: 400D4744 804A0C8 i -> [ 1 ] 00000
11732: 400D4748 400C5E60 . -> 0 [ 0 ]
11733: 400D474C 8049D0C LOOP -> [ 0 ]
11734: 400D4744 804A0C8 i -> [ 1 ] 00001
11735: 400D4748 400C5E60 . -> 1 [ 0 ]
11736: 400D474C 8049D0C LOOP -> [ 0 ]
11737: 400D4758 804B384 ; -> ok
11738: @end example
1.21 crook 11739:
1.78 anton 11740: Each line displayed is one step. You always have to hit return to
11741: execute the next word that is displayed. If you don't want to execute
11742: the next word in a whole, you have to type @kbd{n} for @code{nest}. Here is
11743: an overview what keys are available:
1.44 crook 11744:
1.78 anton 11745: @table @i
1.44 crook 11746:
1.78 anton 11747: @item @key{RET}
11748: Next; Execute the next word.
1.21 crook 11749:
1.78 anton 11750: @item n
11751: Nest; Single step through next word.
1.44 crook 11752:
1.78 anton 11753: @item u
11754: Unnest; Stop debugging and execute rest of word. If we got to this word
11755: with nest, continue debugging with the calling word.
1.44 crook 11756:
1.78 anton 11757: @item d
11758: Done; Stop debugging and execute rest.
1.21 crook 11759:
1.78 anton 11760: @item s
11761: Stop; Abort immediately.
1.44 crook 11762:
1.78 anton 11763: @end table
1.44 crook 11764:
1.78 anton 11765: Debugging large application with this mechanism is very difficult, because
11766: you have to nest very deeply into the program before the interesting part
11767: begins. This takes a lot of time.
1.26 crook 11768:
1.78 anton 11769: To do it more directly put a @code{BREAK:} command into your source code.
11770: When program execution reaches @code{BREAK:} the single step debugger is
11771: invoked and you have all the features described above.
1.44 crook 11772:
1.78 anton 11773: If you have more than one part to debug it is useful to know where the
11774: program has stopped at the moment. You can do this by the
11775: @code{BREAK" string"} command. This behaves like @code{BREAK:} except that
11776: string is typed out when the ``breakpoint'' is reached.
1.44 crook 11777:
1.26 crook 11778:
1.78 anton 11779: doc-dbg
11780: doc-break:
11781: doc-break"
1.44 crook 11782:
1.150 anton 11783: @c ------------------------------------------------------------
11784: @node C Interface, Assembler and Code Words, Programming Tools, Words
11785: @section C Interface
11786: @cindex C interface
11787: @cindex foreign language interface
11788: @cindex interface to C functions
11789:
11790: Note that the C interface is not yet complete; a better way of
11791: declaring C functions is planned, as well as a way of declaring
11792: structs, unions, and their fields.
11793:
11794: @menu
11795: * Calling C Functions::
11796: * Declaring C Functions::
11797: * Callbacks::
1.155 anton 11798: * Low-Level C Interface Words::
1.150 anton 11799: @end menu
11800:
1.151 pazsan 11801: @node Calling C Functions, Declaring C Functions, C Interface, C Interface
1.150 anton 11802: @subsection Calling C functions
1.155 anton 11803: @cindex C functions, calls to
11804: @cindex calling C functions
1.150 anton 11805:
1.151 pazsan 11806: Once a C function is declared (see @pxref{Declaring C Functions}), you
1.150 anton 11807: can call it as follows: You push the arguments on the stack(s), and
11808: then call the word for the C function. The arguments have to be
11809: pushed in the same order as the arguments appear in the C
11810: documentation (i.e., the first argument is deepest on the stack).
11811: Integer and pointer arguments have to be pushed on the data stack,
11812: floating-point arguments on the FP stack; these arguments are consumed
1.155 anton 11813: by the called C function.
1.150 anton 11814:
1.155 anton 11815: On returning from the C function, the return value, if any, resides on
11816: the appropriate stack: an integer return value is pushed on the data
11817: stack, an FP return value on the FP stack, and a void return value
11818: results in not pushing anything. Note that most C functions have a
11819: return value, even if that is often not used in C; in Forth, you have
11820: to @code{drop} this return value explicitly if you do not use it.
1.150 anton 11821:
11822: By default, an integer argument or return value corresponds to a
11823: single cell, and a floating-point argument or return value corresponds
11824: to a Forth float value; the C interface performs the appropriate
11825: conversions where necessary, on a best-effort basis (in some cases,
11826: there may be some loss).
11827:
11828: As an example, consider the POSIX function @code{lseek()}:
11829:
11830: @example
11831: off_t lseek(int fd, off_t offset, int whence);
11832: @end example
11833:
11834: This function takes three integer arguments, and returns an integer
11835: argument, so a Forth call for setting the current file offset to the
11836: start of the file could look like this:
11837:
11838: @example
11839: fd @@ 0 SEEK_SET lseek -1 = if
11840: ... \ error handling
11841: then
11842: @end example
11843:
11844: You might be worried that an @code{off_t} does not fit into a cell, so
11845: you could not pass larger offsets to lseek, and might get only a part
1.155 anton 11846: of the return values. In that case, in your declaration of the
11847: function (@pxref{Declaring C Functions}) you should declare it to use
11848: double-cells for the off_t argument and return value, and maybe give
11849: the resulting Forth word a different name, like @code{dlseek}; the
11850: result could be called like this:
1.150 anton 11851:
11852: @example
11853: fd @@ 0. SEEK_SET dlseek -1. d= if
11854: ... \ error handling
11855: then
11856: @end example
11857:
11858: Passing and returning structs or unions is currently not supported by
11859: our interface@footnote{If you know the calling convention of your C
11860: compiler, you usually can call such functions in some way, but that
11861: way is usually not portable between platforms, and sometimes not even
11862: between C compilers.}.
11863:
11864: Calling functions with a variable number of arguments (e.g.,
11865: @code{printf()}) is currently only supported by having you declare one
11866: function-calling word for each argument pattern, and calling the
11867: appropriate word for the desired pattern.
11868:
1.155 anton 11869:
1.151 pazsan 11870: @node Declaring C Functions, Callbacks, Calling C Functions, C Interface
1.150 anton 11871: @subsection Declaring C Functions
1.155 anton 11872: @cindex C functions, declarations
11873: @cindex declaring C functions
1.150 anton 11874:
11875: Before you can call @code{lseek} or @code{dlseek}, you have to declare
11876: it. You have to look up in your system what the concrete type for the
11877: abstract type @code{off_t} is; let's assume it is @code{long}. Then
11878: the declarations for these words are:
11879:
11880: @example
11881: library libc libc.so.6
11882: libc lseek int long int (long) lseek ( fd noffset whence -- noffset2 )
11883: libc dlseek int dlong int (dlong) lseek ( fd doffset whence -- doffset2 )
11884: @end example
11885:
11886: The first line defines a Forth word @code{libc} for accessing the C
11887: functions in the shared library @file{libc.so.6} (the name of the
11888: shared library depends on the library and the OS; this example is the
11889: standard C library (containing most of the standard C and Unix
11890: functions) for GNU/Linux systems since about 1998).
11891:
11892: The next two lines define two Forth words for the same C function
11893: @code{lseek()}; the middle line defines @code{lseek ( n1 n2 n3 -- n
1.155 anton 11894: )}, and the last line defines @code{dlseek ( n1 d2 n3 -- d )}.
1.150 anton 11895:
11896: As you can see, the declarations are relatively platform-dependent
11897: (e.g., on one platform @code{off_t} may be a @code{long}, whereas on
11898: another platform it may be a @code{long long}; actually, in this case
11899: you can have this difference even on the same platform), while the
11900: resulting function-calling words are platform-independent, and calls
11901: to them are portable.
11902:
11903: At some point in the future this interface will be superseded by a
11904: more convenient one with fewer portability issues. But the resulting
1.155 anton 11905: words for calling the C function will still have the same interface,
1.156 anton 11906: so you will not need to change the calls.
1.155 anton 11907:
11908: Anyway, here are the words for the current interface:
1.150 anton 11909:
1.155 anton 11910: doc-library
11911: doc-int
11912: doc-dint
11913: doc-uint
11914: doc-udint
11915: doc-long
11916: doc-dlong
11917: doc-ulong
11918: doc-udlong
11919: doc-longlong
11920: doc-dlonglong
11921: doc-ulonglong
11922: doc-udlonglong
1.156 anton 11923: doc-ptr
1.155 anton 11924: doc-cfloat
11925: doc-cdouble
11926: doc-clongdouble
11927: doc-(int)
11928: doc-(dint)
11929: doc-(uint)
11930: doc-(udint)
11931: doc-(long)
11932: doc-(dlong)
11933: doc-(ulong)
11934: doc-(udlong)
11935: doc-(longlong)
11936: doc-(dlonglong)
11937: doc-(ulonglong)
11938: doc-(udlonglong)
1.156 anton 11939: doc-(ptr)
1.155 anton 11940: doc-(cfloat)
11941: doc-(cdouble)
11942: doc-(clongdouble)
1.150 anton 11943:
1.155 anton 11944:
11945: @node Callbacks, Low-Level C Interface Words, Declaring C Functions, C Interface
1.150 anton 11946: @subsection Callbacks
1.155 anton 11947: @cindex Callback functions written in Forth
11948: @cindex C function pointers to Forth words
11949:
11950: In some cases you have to pass a function pointer to a C function,
11951: i.e., the library wants to call back to your application (and the
11952: pointed-to function is called a callback function). You can pass the
11953: address of an existing C function (that you get with @code{lib-sym},
11954: @pxref{Low-Level C Interface Words}), but if there is no appropriate C
11955: function, you probably want to define the function as a Forth word.
11956:
11957: !!!
11958: @c I don't understand the existing callback interface from the example - anton
11959:
11960: doc-callback
11961: doc-callback;
11962: doc-fptr
1.150 anton 11963:
1.165 anton 11964:
11965: @c > > Und dann gibt's noch die fptr-Deklaration, die einem
11966: @c > > C-Funktionspointer entspricht (Deklaration gleich wie bei
11967: @c > > Library-Funktionen, nur ohne den C-Namen, Aufruf mit der
11968: @c > > C-Funktionsadresse auf dem TOS).
11969: @c >
11970: @c > Ja, da bin ich dann ausgestiegen, weil ich aus dem Beispiel nicht
11971: @c > gesehen habe, wozu das gut ist.
11972: @c
11973: @c Irgendwie muss ich den Callback ja testen. Und es soll ja auch
11974: @c vorkommen, dass man von irgendwelchen kranken Interfaces einen
11975: @c Funktionspointer übergeben bekommt, den man dann bei Gelegenheit
11976: @c aufrufen muss. Also kann man den deklarieren, und das damit deklarierte
11977: @c Wort verhält sich dann wie ein EXECUTE für alle C-Funktionen mit
11978: @c demselben Prototyp.
11979:
11980:
1.155 anton 11981: @node Low-Level C Interface Words, , Callbacks, C Interface
11982: @subsection Low-Level C Interface Words
1.44 crook 11983:
1.155 anton 11984: doc-open-lib
11985: doc-lib-sym
1.26 crook 11986:
1.78 anton 11987: @c -------------------------------------------------------------
1.150 anton 11988: @node Assembler and Code Words, Threading Words, C Interface, Words
1.78 anton 11989: @section Assembler and Code Words
11990: @cindex assembler
11991: @cindex code words
1.44 crook 11992:
1.78 anton 11993: @menu
11994: * Code and ;code::
11995: * Common Assembler:: Assembler Syntax
11996: * Common Disassembler::
11997: * 386 Assembler:: Deviations and special cases
11998: * Alpha Assembler:: Deviations and special cases
11999: * MIPS assembler:: Deviations and special cases
1.161 anton 12000: * PowerPC assembler:: Deviations and special cases
1.78 anton 12001: * Other assemblers:: How to write them
12002: @end menu
1.21 crook 12003:
1.78 anton 12004: @node Code and ;code, Common Assembler, Assembler and Code Words, Assembler and Code Words
12005: @subsection @code{Code} and @code{;code}
1.26 crook 12006:
1.78 anton 12007: Gforth provides some words for defining primitives (words written in
12008: machine code), and for defining the machine-code equivalent of
12009: @code{DOES>}-based defining words. However, the machine-independent
12010: nature of Gforth poses a few problems: First of all, Gforth runs on
12011: several architectures, so it can provide no standard assembler. What's
12012: worse is that the register allocation not only depends on the processor,
12013: but also on the @code{gcc} version and options used.
1.44 crook 12014:
1.78 anton 12015: The words that Gforth offers encapsulate some system dependences (e.g.,
12016: the header structure), so a system-independent assembler may be used in
12017: Gforth. If you do not have an assembler, you can compile machine code
12018: directly with @code{,} and @code{c,}@footnote{This isn't portable,
12019: because these words emit stuff in @i{data} space; it works because
12020: Gforth has unified code/data spaces. Assembler isn't likely to be
12021: portable anyway.}.
1.21 crook 12022:
1.44 crook 12023:
1.78 anton 12024: doc-assembler
12025: doc-init-asm
12026: doc-code
12027: doc-end-code
12028: doc-;code
12029: doc-flush-icache
1.44 crook 12030:
1.21 crook 12031:
1.78 anton 12032: If @code{flush-icache} does not work correctly, @code{code} words
12033: etc. will not work (reliably), either.
1.44 crook 12034:
1.78 anton 12035: The typical usage of these @code{code} words can be shown most easily by
12036: analogy to the equivalent high-level defining words:
1.44 crook 12037:
1.78 anton 12038: @example
12039: : foo code foo
12040: <high-level Forth words> <assembler>
12041: ; end-code
12042:
12043: : bar : bar
12044: <high-level Forth words> <high-level Forth words>
12045: CREATE CREATE
12046: <high-level Forth words> <high-level Forth words>
12047: DOES> ;code
12048: <high-level Forth words> <assembler>
12049: ; end-code
12050: @end example
1.21 crook 12051:
1.78 anton 12052: @c anton: the following stuff is also in "Common Assembler", in less detail.
1.44 crook 12053:
1.78 anton 12054: @cindex registers of the inner interpreter
12055: In the assembly code you will want to refer to the inner interpreter's
12056: registers (e.g., the data stack pointer) and you may want to use other
12057: registers for temporary storage. Unfortunately, the register allocation
12058: is installation-dependent.
1.44 crook 12059:
1.78 anton 12060: In particular, @code{ip} (Forth instruction pointer) and @code{rp}
1.100 anton 12061: (return stack pointer) may be in different places in @code{gforth} and
12062: @code{gforth-fast}, or different installations. This means that you
12063: cannot write a @code{NEXT} routine that works reliably on both versions
12064: or different installations; so for doing @code{NEXT}, I recommend
12065: jumping to @code{' noop >code-address}, which contains nothing but a
12066: @code{NEXT}.
1.21 crook 12067:
1.78 anton 12068: For general accesses to the inner interpreter's registers, the easiest
12069: solution is to use explicit register declarations (@pxref{Explicit Reg
12070: Vars, , Variables in Specified Registers, gcc.info, GNU C Manual}) for
12071: all of the inner interpreter's registers: You have to compile Gforth
12072: with @code{-DFORCE_REG} (configure option @code{--enable-force-reg}) and
12073: the appropriate declarations must be present in the @code{machine.h}
12074: file (see @code{mips.h} for an example; you can find a full list of all
12075: declarable register symbols with @code{grep register engine.c}). If you
12076: give explicit registers to all variables that are declared at the
12077: beginning of @code{engine()}, you should be able to use the other
12078: caller-saved registers for temporary storage. Alternatively, you can use
12079: the @code{gcc} option @code{-ffixed-REG} (@pxref{Code Gen Options, ,
12080: Options for Code Generation Conventions, gcc.info, GNU C Manual}) to
12081: reserve a register (however, this restriction on register allocation may
12082: slow Gforth significantly).
1.44 crook 12083:
1.78 anton 12084: If this solution is not viable (e.g., because @code{gcc} does not allow
12085: you to explicitly declare all the registers you need), you have to find
12086: out by looking at the code where the inner interpreter's registers
12087: reside and which registers can be used for temporary storage. You can
12088: get an assembly listing of the engine's code with @code{make engine.s}.
1.44 crook 12089:
1.78 anton 12090: In any case, it is good practice to abstract your assembly code from the
12091: actual register allocation. E.g., if the data stack pointer resides in
12092: register @code{$17}, create an alias for this register called @code{sp},
12093: and use that in your assembly code.
1.21 crook 12094:
1.78 anton 12095: @cindex code words, portable
12096: Another option for implementing normal and defining words efficiently
12097: is to add the desired functionality to the source of Gforth. For normal
12098: words you just have to edit @file{primitives} (@pxref{Automatic
12099: Generation}). Defining words (equivalent to @code{;CODE} words, for fast
12100: defined words) may require changes in @file{engine.c}, @file{kernel.fs},
12101: @file{prims2x.fs}, and possibly @file{cross.fs}.
1.44 crook 12102:
1.78 anton 12103: @node Common Assembler, Common Disassembler, Code and ;code, Assembler and Code Words
12104: @subsection Common Assembler
1.44 crook 12105:
1.78 anton 12106: The assemblers in Gforth generally use a postfix syntax, i.e., the
12107: instruction name follows the operands.
1.21 crook 12108:
1.78 anton 12109: The operands are passed in the usual order (the same that is used in the
12110: manual of the architecture). Since they all are Forth words, they have
12111: to be separated by spaces; you can also use Forth words to compute the
12112: operands.
1.44 crook 12113:
1.78 anton 12114: The instruction names usually end with a @code{,}. This makes it easier
12115: to visually separate instructions if you put several of them on one
12116: line; it also avoids shadowing other Forth words (e.g., @code{and}).
1.21 crook 12117:
1.78 anton 12118: Registers are usually specified by number; e.g., (decimal) @code{11}
12119: specifies registers R11 and F11 on the Alpha architecture (which one,
12120: depends on the instruction). The usual names are also available, e.g.,
12121: @code{s2} for R11 on Alpha.
1.21 crook 12122:
1.78 anton 12123: Control flow is specified similar to normal Forth code (@pxref{Arbitrary
12124: control structures}), with @code{if,}, @code{ahead,}, @code{then,},
12125: @code{begin,}, @code{until,}, @code{again,}, @code{cs-roll},
12126: @code{cs-pick}, @code{else,}, @code{while,}, and @code{repeat,}. The
12127: conditions are specified in a way specific to each assembler.
1.1 anton 12128:
1.78 anton 12129: Note that the register assignments of the Gforth engine can change
12130: between Gforth versions, or even between different compilations of the
12131: same Gforth version (e.g., if you use a different GCC version). So if
12132: you want to refer to Gforth's registers (e.g., the stack pointer or
12133: TOS), I recommend defining your own words for refering to these
12134: registers, and using them later on; then you can easily adapt to a
12135: changed register assignment. The stability of the register assignment
12136: is usually better if you build Gforth with @code{--enable-force-reg}.
1.1 anton 12137:
1.100 anton 12138: The most common use of these registers is to dispatch to the next word
12139: (the @code{next} routine). A portable way to do this is to jump to
12140: @code{' noop >code-address} (of course, this is less efficient than
12141: integrating the @code{next} code and scheduling it well).
1.1 anton 12142:
1.96 anton 12143: Another difference between Gforth version is that the top of stack is
12144: kept in memory in @code{gforth} and, on most platforms, in a register in
12145: @code{gforth-fast}.
12146:
1.78 anton 12147: @node Common Disassembler, 386 Assembler, Common Assembler, Assembler and Code Words
12148: @subsection Common Disassembler
1.127 anton 12149: @cindex disassembler, general
12150: @cindex gdb disassembler
1.1 anton 12151:
1.78 anton 12152: You can disassemble a @code{code} word with @code{see}
12153: (@pxref{Debugging}). You can disassemble a section of memory with
1.1 anton 12154:
1.127 anton 12155: doc-discode
1.44 crook 12156:
1.127 anton 12157: There are two kinds of disassembler for Gforth: The Forth disassembler
12158: (available on some CPUs) and the gdb disassembler (available on
12159: platforms with @command{gdb} and @command{mktemp}). If both are
12160: available, the Forth disassembler is used by default. If you prefer
12161: the gdb disassembler, say
12162:
12163: @example
12164: ' disasm-gdb is discode
12165: @end example
12166:
12167: If neither is available, @code{discode} performs @code{dump}.
12168:
12169: The Forth disassembler generally produces output that can be fed into the
1.78 anton 12170: assembler (i.e., same syntax, etc.). It also includes additional
12171: information in comments. In particular, the address of the instruction
12172: is given in a comment before the instruction.
1.1 anton 12173:
1.127 anton 12174: The gdb disassembler produces output in the same format as the gdb
12175: @code{disassemble} command (@pxref{Machine Code,,Source and machine
12176: code,gdb,Debugging with GDB}), in the default flavour (AT&T syntax for
12177: the 386 and AMD64 architectures).
12178:
1.78 anton 12179: @code{See} may display more or less than the actual code of the word,
12180: because the recognition of the end of the code is unreliable. You can
1.127 anton 12181: use @code{discode} if it did not display enough. It may display more, if
1.78 anton 12182: the code word is not immediately followed by a named word. If you have
1.116 anton 12183: something else there, you can follow the word with @code{align latest ,}
1.78 anton 12184: to ensure that the end is recognized.
1.21 crook 12185:
1.78 anton 12186: @node 386 Assembler, Alpha Assembler, Common Disassembler, Assembler and Code Words
12187: @subsection 386 Assembler
1.44 crook 12188:
1.78 anton 12189: The 386 assembler included in Gforth was written by Bernd Paysan, it's
12190: available under GPL, and originally part of bigFORTH.
1.21 crook 12191:
1.78 anton 12192: The 386 disassembler included in Gforth was written by Andrew McKewan
12193: and is in the public domain.
1.21 crook 12194:
1.91 anton 12195: The disassembler displays code in an Intel-like prefix syntax.
1.21 crook 12196:
1.78 anton 12197: The assembler uses a postfix syntax with reversed parameters.
1.1 anton 12198:
1.78 anton 12199: The assembler includes all instruction of the Athlon, i.e. 486 core
12200: instructions, Pentium and PPro extensions, floating point, MMX, 3Dnow!,
12201: but not ISSE. It's an integrated 16- and 32-bit assembler. Default is 32
12202: bit, you can switch to 16 bit with .86 and back to 32 bit with .386.
1.1 anton 12203:
1.78 anton 12204: There are several prefixes to switch between different operation sizes,
12205: @code{.b} for byte accesses, @code{.w} for word accesses, @code{.d} for
12206: double-word accesses. Addressing modes can be switched with @code{.wa}
12207: for 16 bit addresses, and @code{.da} for 32 bit addresses. You don't
12208: need a prefix for byte register names (@code{AL} et al).
1.1 anton 12209:
1.78 anton 12210: For floating point operations, the prefixes are @code{.fs} (IEEE
12211: single), @code{.fl} (IEEE double), @code{.fx} (extended), @code{.fw}
12212: (word), @code{.fd} (double-word), and @code{.fq} (quad-word).
1.21 crook 12213:
1.78 anton 12214: The MMX opcodes don't have size prefixes, they are spelled out like in
12215: the Intel assembler. Instead of move from and to memory, there are
12216: PLDQ/PLDD and PSTQ/PSTD.
1.21 crook 12217:
1.78 anton 12218: The registers lack the 'e' prefix; even in 32 bit mode, eax is called
12219: ax. Immediate values are indicated by postfixing them with @code{#},
1.91 anton 12220: e.g., @code{3 #}. Here are some examples of addressing modes in various
12221: syntaxes:
1.21 crook 12222:
1.26 crook 12223: @example
1.91 anton 12224: Gforth Intel (NASM) AT&T (gas) Name
12225: .w ax ax %ax register (16 bit)
12226: ax eax %eax register (32 bit)
12227: 3 # offset 3 $3 immediate
12228: 1000 #) byte ptr 1000 1000 displacement
12229: bx ) [ebx] (%ebx) base
12230: 100 di d) 100[edi] 100(%edi) base+displacement
12231: 20 ax *4 i#) 20[eax*4] 20(,%eax,4) (index*scale)+displacement
12232: di ax *4 i) [edi][eax*4] (%edi,%eax,4) base+(index*scale)
12233: 4 bx cx di) 4[ebx][ecx] 4(%ebx,%ecx) base+index+displacement
12234: 12 sp ax *2 di) 12[esp][eax*2] 12(%esp,%eax,2) base+(index*scale)+displacement
12235: @end example
12236:
12237: You can use @code{L)} and @code{LI)} instead of @code{D)} and
12238: @code{DI)} to enforce 32-bit displacement fields (useful for
12239: later patching).
1.21 crook 12240:
1.78 anton 12241: Some example of instructions are:
1.1 anton 12242:
12243: @example
1.78 anton 12244: ax bx mov \ move ebx,eax
12245: 3 # ax mov \ mov eax,3
1.137 pazsan 12246: 100 di d) ax mov \ mov eax,100[edi]
1.78 anton 12247: 4 bx cx di) ax mov \ mov eax,4[ebx][ecx]
12248: .w ax bx mov \ mov bx,ax
1.1 anton 12249: @end example
12250:
1.78 anton 12251: The following forms are supported for binary instructions:
1.1 anton 12252:
12253: @example
1.78 anton 12254: <reg> <reg> <inst>
12255: <n> # <reg> <inst>
12256: <mem> <reg> <inst>
12257: <reg> <mem> <inst>
1.136 pazsan 12258: <n> # <mem> <inst>
1.1 anton 12259: @end example
12260:
1.136 pazsan 12261: The shift/rotate syntax is:
1.1 anton 12262:
1.26 crook 12263: @example
1.78 anton 12264: <reg/mem> 1 # shl \ shortens to shift without immediate
12265: <reg/mem> 4 # shl
12266: <reg/mem> cl shl
1.26 crook 12267: @end example
1.1 anton 12268:
1.78 anton 12269: Precede string instructions (@code{movs} etc.) with @code{.b} to get
12270: the byte version.
1.1 anton 12271:
1.78 anton 12272: The control structure words @code{IF} @code{UNTIL} etc. must be preceded
12273: by one of these conditions: @code{vs vc u< u>= 0= 0<> u<= u> 0< 0>= ps
12274: pc < >= <= >}. (Note that most of these words shadow some Forth words
12275: when @code{assembler} is in front of @code{forth} in the search path,
12276: e.g., in @code{code} words). Currently the control structure words use
12277: one stack item, so you have to use @code{roll} instead of @code{cs-roll}
12278: to shuffle them (you can also use @code{swap} etc.).
1.21 crook 12279:
1.78 anton 12280: Here is an example of a @code{code} word (assumes that the stack pointer
12281: is in esi and the TOS is in ebx):
1.21 crook 12282:
1.26 crook 12283: @example
1.78 anton 12284: code my+ ( n1 n2 -- n )
12285: 4 si D) bx add
12286: 4 # si add
12287: Next
12288: end-code
1.26 crook 12289: @end example
1.21 crook 12290:
1.161 anton 12291:
1.78 anton 12292: @node Alpha Assembler, MIPS assembler, 386 Assembler, Assembler and Code Words
12293: @subsection Alpha Assembler
1.21 crook 12294:
1.78 anton 12295: The Alpha assembler and disassembler were originally written by Bernd
12296: Thallner.
1.26 crook 12297:
1.78 anton 12298: The register names @code{a0}--@code{a5} are not available to avoid
12299: shadowing hex numbers.
1.2 jwilke 12300:
1.78 anton 12301: Immediate forms of arithmetic instructions are distinguished by a
12302: @code{#} just before the @code{,}, e.g., @code{and#,} (note: @code{lda,}
12303: does not count as arithmetic instruction).
1.2 jwilke 12304:
1.78 anton 12305: You have to specify all operands to an instruction, even those that
12306: other assemblers consider optional, e.g., the destination register for
12307: @code{br,}, or the destination register and hint for @code{jmp,}.
1.2 jwilke 12308:
1.78 anton 12309: You can specify conditions for @code{if,} by removing the first @code{b}
12310: and the trailing @code{,} from a branch with a corresponding name; e.g.,
1.2 jwilke 12311:
1.26 crook 12312: @example
1.78 anton 12313: 11 fgt if, \ if F11>0e
12314: ...
12315: endif,
1.26 crook 12316: @end example
1.2 jwilke 12317:
1.78 anton 12318: @code{fbgt,} gives @code{fgt}.
12319:
1.161 anton 12320: @node MIPS assembler, PowerPC assembler, Alpha Assembler, Assembler and Code Words
1.78 anton 12321: @subsection MIPS assembler
1.2 jwilke 12322:
1.78 anton 12323: The MIPS assembler was originally written by Christian Pirker.
1.2 jwilke 12324:
1.78 anton 12325: Currently the assembler and disassembler only cover the MIPS-I
12326: architecture (R3000), and don't support FP instructions.
1.2 jwilke 12327:
1.78 anton 12328: The register names @code{$a0}--@code{$a3} are not available to avoid
12329: shadowing hex numbers.
1.2 jwilke 12330:
1.78 anton 12331: Because there is no way to distinguish registers from immediate values,
12332: you have to explicitly use the immediate forms of instructions, i.e.,
12333: @code{addiu,}, not just @code{addu,} (@command{as} does this
12334: implicitly).
1.2 jwilke 12335:
1.78 anton 12336: If the architecture manual specifies several formats for the instruction
12337: (e.g., for @code{jalr,}), you usually have to use the one with more
12338: arguments (i.e., two for @code{jalr,}). When in doubt, see
12339: @code{arch/mips/testasm.fs} for an example of correct use.
1.2 jwilke 12340:
1.78 anton 12341: Branches and jumps in the MIPS architecture have a delay slot. You have
12342: to fill it yourself (the simplest way is to use @code{nop,}), the
12343: assembler does not do it for you (unlike @command{as}). Even
12344: @code{if,}, @code{ahead,}, @code{until,}, @code{again,}, @code{while,},
12345: @code{else,} and @code{repeat,} need a delay slot. Since @code{begin,}
12346: and @code{then,} just specify branch targets, they are not affected.
1.2 jwilke 12347:
1.78 anton 12348: Note that you must not put branches, jumps, or @code{li,} into the delay
12349: slot: @code{li,} may expand to several instructions, and control flow
12350: instructions may not be put into the branch delay slot in any case.
1.2 jwilke 12351:
1.78 anton 12352: For branches the argument specifying the target is a relative address;
12353: You have to add the address of the delay slot to get the absolute
12354: address.
1.1 anton 12355:
1.78 anton 12356: The MIPS architecture also has load delay slots and restrictions on
12357: using @code{mfhi,} and @code{mflo,}; you have to order the instructions
12358: yourself to satisfy these restrictions, the assembler does not do it for
12359: you.
1.1 anton 12360:
1.78 anton 12361: You can specify the conditions for @code{if,} etc. by taking a
12362: conditional branch and leaving away the @code{b} at the start and the
12363: @code{,} at the end. E.g.,
1.1 anton 12364:
1.26 crook 12365: @example
1.78 anton 12366: 4 5 eq if,
12367: ... \ do something if $4 equals $5
12368: then,
1.26 crook 12369: @end example
1.1 anton 12370:
1.161 anton 12371:
12372: @node PowerPC assembler, Other assemblers, MIPS assembler, Assembler and Code Words
12373: @subsection PowerPC assembler
12374:
1.162 anton 12375: The PowerPC assembler and disassembler were contributed by Michal
1.161 anton 12376: Revucky.
12377:
1.162 anton 12378: This assembler does not follow the convention of ending mnemonic names
12379: with a ``,'', so some mnemonic names shadow regular Forth words (in
12380: particular: @code{and or xor fabs}); so if you want to use the Forth
12381: words, you have to make them visible first, e.g., with @code{also
12382: forth}.
12383:
1.161 anton 12384: Registers are referred to by their number, e.g., @code{9} means the
12385: integer register 9 or the FP register 9 (depending on the
12386: instruction).
12387:
12388: Because there is no way to distinguish registers from immediate values,
12389: you have to explicitly use the immediate forms of instructions, i.e.,
1.162 anton 12390: @code{addi,}, not just @code{add,}.
1.161 anton 12391:
1.162 anton 12392: The assembler and disassembler usually support the most general form
1.161 anton 12393: of an instruction, but usually not the shorter forms (especially for
12394: branches).
12395:
12396:
12397:
12398: @node Other assemblers, , PowerPC assembler, Assembler and Code Words
1.78 anton 12399: @subsection Other assemblers
12400:
12401: If you want to contribute another assembler/disassembler, please contact
1.103 anton 12402: us (@email{anton@@mips.complang.tuwien.ac.at}) to check if we have such
12403: an assembler already. If you are writing them from scratch, please use
12404: a similar syntax style as the one we use (i.e., postfix, commas at the
12405: end of the instruction names, @pxref{Common Assembler}); make the output
12406: of the disassembler be valid input for the assembler, and keep the style
1.78 anton 12407: similar to the style we used.
12408:
12409: Hints on implementation: The most important part is to have a good test
12410: suite that contains all instructions. Once you have that, the rest is
12411: easy. For actual coding you can take a look at
12412: @file{arch/mips/disasm.fs} to get some ideas on how to use data for both
12413: the assembler and disassembler, avoiding redundancy and some potential
12414: bugs. You can also look at that file (and @pxref{Advanced does> usage
12415: example}) to get ideas how to factor a disassembler.
12416:
12417: Start with the disassembler, because it's easier to reuse data from the
12418: disassembler for the assembler than the other way round.
1.1 anton 12419:
1.78 anton 12420: For the assembler, take a look at @file{arch/alpha/asm.fs}, which shows
12421: how simple it can be.
1.1 anton 12422:
1.161 anton 12423:
12424:
12425:
1.78 anton 12426: @c -------------------------------------------------------------
12427: @node Threading Words, Passing Commands to the OS, Assembler and Code Words, Words
12428: @section Threading Words
12429: @cindex threading words
1.1 anton 12430:
1.78 anton 12431: @cindex code address
12432: These words provide access to code addresses and other threading stuff
12433: in Gforth (and, possibly, other interpretive Forths). It more or less
12434: abstracts away the differences between direct and indirect threading
12435: (and, for direct threading, the machine dependences). However, at
12436: present this wordset is still incomplete. It is also pretty low-level;
12437: some day it will hopefully be made unnecessary by an internals wordset
12438: that abstracts implementation details away completely.
1.1 anton 12439:
1.78 anton 12440: The terminology used here stems from indirect threaded Forth systems; in
12441: such a system, the XT of a word is represented by the CFA (code field
12442: address) of a word; the CFA points to a cell that contains the code
12443: address. The code address is the address of some machine code that
12444: performs the run-time action of invoking the word (e.g., the
12445: @code{dovar:} routine pushes the address of the body of the word (a
12446: variable) on the stack
12447: ).
1.1 anton 12448:
1.78 anton 12449: @cindex code address
12450: @cindex code field address
12451: In an indirect threaded Forth, you can get the code address of @i{name}
12452: with @code{' @i{name} @@}; in Gforth you can get it with @code{' @i{name}
12453: >code-address}, independent of the threading method.
1.1 anton 12454:
1.78 anton 12455: doc-threading-method
12456: doc->code-address
12457: doc-code-address!
1.1 anton 12458:
1.78 anton 12459: @cindex @code{does>}-handler
12460: @cindex @code{does>}-code
12461: For a word defined with @code{DOES>}, the code address usually points to
12462: a jump instruction (the @dfn{does-handler}) that jumps to the dodoes
12463: routine (in Gforth on some platforms, it can also point to the dodoes
12464: routine itself). What you are typically interested in, though, is
12465: whether a word is a @code{DOES>}-defined word, and what Forth code it
12466: executes; @code{>does-code} tells you that.
1.1 anton 12467:
1.78 anton 12468: doc->does-code
1.1 anton 12469:
1.78 anton 12470: To create a @code{DOES>}-defined word with the following basic words,
12471: you have to set up a @code{DOES>}-handler with @code{does-handler!};
12472: @code{/does-handler} aus behind you have to place your executable Forth
12473: code. Finally you have to create a word and modify its behaviour with
12474: @code{does-handler!}.
1.1 anton 12475:
1.78 anton 12476: doc-does-code!
12477: doc-does-handler!
12478: doc-/does-handler
1.1 anton 12479:
1.78 anton 12480: The code addresses produced by various defining words are produced by
12481: the following words:
1.1 anton 12482:
1.78 anton 12483: doc-docol:
12484: doc-docon:
12485: doc-dovar:
12486: doc-douser:
12487: doc-dodefer:
12488: doc-dofield:
1.1 anton 12489:
1.99 anton 12490: @cindex definer
12491: The following two words generalize @code{>code-address},
12492: @code{>does-code}, @code{code-address!}, and @code{does-code!}:
12493:
12494: doc->definer
12495: doc-definer!
12496:
1.26 crook 12497: @c -------------------------------------------------------------
1.78 anton 12498: @node Passing Commands to the OS, Keeping track of Time, Threading Words, Words
1.21 crook 12499: @section Passing Commands to the Operating System
12500: @cindex operating system - passing commands
12501: @cindex shell commands
12502:
12503: Gforth allows you to pass an arbitrary string to the host operating
12504: system shell (if such a thing exists) for execution.
12505:
12506: doc-sh
12507: doc-system
12508: doc-$?
1.23 crook 12509: doc-getenv
1.44 crook 12510:
1.26 crook 12511: @c -------------------------------------------------------------
1.47 crook 12512: @node Keeping track of Time, Miscellaneous Words, Passing Commands to the OS, Words
12513: @section Keeping track of Time
12514: @cindex time-related words
12515:
12516: doc-ms
12517: doc-time&date
1.79 anton 12518: doc-utime
12519: doc-cputime
1.47 crook 12520:
12521:
12522: @c -------------------------------------------------------------
12523: @node Miscellaneous Words, , Keeping track of Time, Words
1.21 crook 12524: @section Miscellaneous Words
12525: @cindex miscellaneous words
12526:
1.29 crook 12527: @comment TODO find homes for these
12528:
1.26 crook 12529: These section lists the ANS Forth words that are not documented
1.21 crook 12530: elsewhere in this manual. Ultimately, they all need proper homes.
12531:
1.68 anton 12532: doc-quit
1.44 crook 12533:
1.26 crook 12534: The following ANS Forth words are not currently supported by Gforth
1.27 crook 12535: (@pxref{ANS conformance}):
1.21 crook 12536:
12537: @code{EDITOR}
12538: @code{EMIT?}
12539: @code{FORGET}
12540:
1.24 anton 12541: @c ******************************************************************
12542: @node Error messages, Tools, Words, Top
12543: @chapter Error messages
12544: @cindex error messages
12545: @cindex backtrace
12546:
12547: A typical Gforth error message looks like this:
12548:
12549: @example
1.86 anton 12550: in file included from \evaluated string/:-1
1.24 anton 12551: in file included from ./yyy.fs:1
12552: ./xxx.fs:4: Invalid memory address
1.134 anton 12553: >>>bar<<<
1.79 anton 12554: Backtrace:
1.25 anton 12555: $400E664C @@
12556: $400E6664 foo
1.24 anton 12557: @end example
12558:
12559: The message identifying the error is @code{Invalid memory address}. The
12560: error happened when text-interpreting line 4 of the file
12561: @file{./xxx.fs}. This line is given (it contains @code{bar}), and the
12562: word on the line where the error happened, is pointed out (with
1.134 anton 12563: @code{>>>} and @code{<<<}).
1.24 anton 12564:
12565: The file containing the error was included in line 1 of @file{./yyy.fs},
12566: and @file{yyy.fs} was included from a non-file (in this case, by giving
12567: @file{yyy.fs} as command-line parameter to Gforth).
12568:
12569: At the end of the error message you find a return stack dump that can be
12570: interpreted as a backtrace (possibly empty). On top you find the top of
12571: the return stack when the @code{throw} happened, and at the bottom you
12572: find the return stack entry just above the return stack of the topmost
12573: text interpreter.
12574:
12575: To the right of most return stack entries you see a guess for the word
12576: that pushed that return stack entry as its return address. This gives a
12577: backtrace. In our case we see that @code{bar} called @code{foo}, and
12578: @code{foo} called @code{@@} (and @code{@@} had an @emph{Invalid memory
12579: address} exception).
12580:
12581: Note that the backtrace is not perfect: We don't know which return stack
12582: entries are return addresses (so we may get false positives); and in
12583: some cases (e.g., for @code{abort"}) we cannot determine from the return
12584: address the word that pushed the return address, so for some return
12585: addresses you see no names in the return stack dump.
1.25 anton 12586:
12587: @cindex @code{catch} and backtraces
12588: The return stack dump represents the return stack at the time when a
12589: specific @code{throw} was executed. In programs that make use of
12590: @code{catch}, it is not necessarily clear which @code{throw} should be
12591: used for the return stack dump (e.g., consider one @code{throw} that
12592: indicates an error, which is caught, and during recovery another error
1.160 anton 12593: happens; which @code{throw} should be used for the stack dump?).
12594: Gforth presents the return stack dump for the first @code{throw} after
12595: the last executed (not returned-to) @code{catch} or @code{nothrow};
12596: this works well in the usual case. To get the right backtrace, you
12597: usually want to insert @code{nothrow} or @code{['] false catch drop}
12598: after a @code{catch} if the error is not rethrown.
1.25 anton 12599:
12600: @cindex @code{gforth-fast} and backtraces
12601: @cindex @code{gforth-fast}, difference from @code{gforth}
12602: @cindex backtraces with @code{gforth-fast}
12603: @cindex return stack dump with @code{gforth-fast}
1.79 anton 12604: @code{Gforth} is able to do a return stack dump for throws generated
1.25 anton 12605: from primitives (e.g., invalid memory address, stack empty etc.);
12606: @code{gforth-fast} is only able to do a return stack dump from a
1.96 anton 12607: directly called @code{throw} (including @code{abort} etc.). Given an
1.30 anton 12608: exception caused by a primitive in @code{gforth-fast}, you will
12609: typically see no return stack dump at all; however, if the exception is
12610: caught by @code{catch} (e.g., for restoring some state), and then
12611: @code{throw}n again, the return stack dump will be for the first such
12612: @code{throw}.
1.2 jwilke 12613:
1.5 anton 12614: @c ******************************************************************
1.24 anton 12615: @node Tools, ANS conformance, Error messages, Top
1.1 anton 12616: @chapter Tools
12617:
12618: @menu
12619: * ANS Report:: Report the words used, sorted by wordset.
1.127 anton 12620: * Stack depth changes:: Where does this stack item come from?
1.1 anton 12621: @end menu
12622:
12623: See also @ref{Emacs and Gforth}.
12624:
1.126 pazsan 12625: @node ANS Report, Stack depth changes, Tools, Tools
1.1 anton 12626: @section @file{ans-report.fs}: Report the words used, sorted by wordset
12627: @cindex @file{ans-report.fs}
12628: @cindex report the words used in your program
12629: @cindex words used in your program
12630:
12631: If you want to label a Forth program as ANS Forth Program, you must
12632: document which wordsets the program uses; for extension wordsets, it is
12633: helpful to list the words the program requires from these wordsets
12634: (because Forth systems are allowed to provide only some words of them).
12635:
12636: The @file{ans-report.fs} tool makes it easy for you to determine which
12637: words from which wordset and which non-ANS words your application
12638: uses. You simply have to include @file{ans-report.fs} before loading the
12639: program you want to check. After loading your program, you can get the
12640: report with @code{print-ans-report}. A typical use is to run this as
12641: batch job like this:
12642: @example
12643: gforth ans-report.fs myprog.fs -e "print-ans-report bye"
12644: @end example
12645:
12646: The output looks like this (for @file{compat/control.fs}):
12647: @example
12648: The program uses the following words
12649: from CORE :
12650: : POSTPONE THEN ; immediate ?dup IF 0=
12651: from BLOCK-EXT :
12652: \
12653: from FILE :
12654: (
12655: @end example
12656:
12657: @subsection Caveats
12658:
12659: Note that @file{ans-report.fs} just checks which words are used, not whether
12660: they are used in an ANS Forth conforming way!
12661:
12662: Some words are defined in several wordsets in the
12663: standard. @file{ans-report.fs} reports them for only one of the
12664: wordsets, and not necessarily the one you expect. It depends on usage
12665: which wordset is the right one to specify. E.g., if you only use the
12666: compilation semantics of @code{S"}, it is a Core word; if you also use
12667: its interpretation semantics, it is a File word.
1.124 anton 12668:
12669:
1.127 anton 12670: @node Stack depth changes, , ANS Report, Tools
1.124 anton 12671: @section Stack depth changes during interpretation
12672: @cindex @file{depth-changes.fs}
12673: @cindex depth changes during interpretation
12674: @cindex stack depth changes during interpretation
12675: @cindex items on the stack after interpretation
12676:
12677: Sometimes you notice that, after loading a file, there are items left
12678: on the stack. The tool @file{depth-changes.fs} helps you find out
12679: quickly where in the file these stack items are coming from.
12680:
12681: The simplest way of using @file{depth-changes.fs} is to include it
12682: before the file(s) you want to check, e.g.:
12683:
12684: @example
12685: gforth depth-changes.fs my-file.fs
12686: @end example
12687:
12688: This will compare the stack depths of the data and FP stack at every
12689: empty line (in interpretation state) against these depths at the last
12690: empty line (in interpretation state). If the depths are not equal,
12691: the position in the file and the stack contents are printed with
12692: @code{~~} (@pxref{Debugging}). This indicates that a stack depth
12693: change has occured in the paragraph of non-empty lines before the
12694: indicated line. It is a good idea to leave an empty line at the end
12695: of the file, so the last paragraph is checked, too.
12696:
12697: Checking only at empty lines usually works well, but sometimes you
12698: have big blocks of non-empty lines (e.g., when building a big table),
12699: and you want to know where in this block the stack depth changed. You
12700: can check all interpreted lines with
12701:
12702: @example
12703: gforth depth-changes.fs -e "' all-lines is depth-changes-filter" my-file.fs
12704: @end example
12705:
12706: This checks the stack depth at every end-of-line. So the depth change
12707: occured in the line reported by the @code{~~} (not in the line
12708: before).
12709:
12710: Note that, while this offers better accuracy in indicating where the
12711: stack depth changes, it will often report many intentional stack depth
12712: changes (e.g., when an interpreted computation stretches across
12713: several lines). You can suppress the checking of some lines by
12714: putting backslashes at the end of these lines (not followed by white
12715: space), and using
12716:
12717: @example
12718: gforth depth-changes.fs -e "' most-lines is depth-changes-filter" my-file.fs
12719: @end example
1.1 anton 12720:
12721: @c ******************************************************************
1.65 anton 12722: @node ANS conformance, Standard vs Extensions, Tools, Top
1.1 anton 12723: @chapter ANS conformance
12724: @cindex ANS conformance of Gforth
12725:
12726: To the best of our knowledge, Gforth is an
12727:
12728: ANS Forth System
12729: @itemize @bullet
12730: @item providing the Core Extensions word set
12731: @item providing the Block word set
12732: @item providing the Block Extensions word set
12733: @item providing the Double-Number word set
12734: @item providing the Double-Number Extensions word set
12735: @item providing the Exception word set
12736: @item providing the Exception Extensions word set
12737: @item providing the Facility word set
1.40 anton 12738: @item providing @code{EKEY}, @code{EKEY>CHAR}, @code{EKEY?}, @code{MS} and @code{TIME&DATE} from the Facility Extensions word set
1.1 anton 12739: @item providing the File Access word set
12740: @item providing the File Access Extensions word set
12741: @item providing the Floating-Point word set
12742: @item providing the Floating-Point Extensions word set
12743: @item providing the Locals word set
12744: @item providing the Locals Extensions word set
12745: @item providing the Memory-Allocation word set
12746: @item providing the Memory-Allocation Extensions word set (that one's easy)
12747: @item providing the Programming-Tools word set
12748: @item providing @code{;CODE}, @code{AHEAD}, @code{ASSEMBLER}, @code{BYE}, @code{CODE}, @code{CS-PICK}, @code{CS-ROLL}, @code{STATE}, @code{[ELSE]}, @code{[IF]}, @code{[THEN]} from the Programming-Tools Extensions word set
12749: @item providing the Search-Order word set
12750: @item providing the Search-Order Extensions word set
12751: @item providing the String word set
12752: @item providing the String Extensions word set (another easy one)
12753: @end itemize
12754:
1.118 anton 12755: Gforth has the following environmental restrictions:
12756:
12757: @cindex environmental restrictions
12758: @itemize @bullet
12759: @item
12760: While processing the OS command line, if an exception is not caught,
12761: Gforth exits with a non-zero exit code instyead of performing QUIT.
12762:
12763: @item
12764: When an @code{throw} is performed after a @code{query}, Gforth does not
12765: allways restore the input source specification in effect at the
12766: corresponding catch.
12767:
12768: @end itemize
12769:
12770:
1.1 anton 12771: @cindex system documentation
12772: In addition, ANS Forth systems are required to document certain
12773: implementation choices. This chapter tries to meet these
12774: requirements. In many cases it gives a way to ask the system for the
12775: information instead of providing the information directly, in
12776: particular, if the information depends on the processor, the operating
12777: system or the installation options chosen, or if they are likely to
12778: change during the maintenance of Gforth.
12779:
12780: @comment The framework for the rest has been taken from pfe.
12781:
12782: @menu
12783: * The Core Words::
12784: * The optional Block word set::
12785: * The optional Double Number word set::
12786: * The optional Exception word set::
12787: * The optional Facility word set::
12788: * The optional File-Access word set::
12789: * The optional Floating-Point word set::
12790: * The optional Locals word set::
12791: * The optional Memory-Allocation word set::
12792: * The optional Programming-Tools word set::
12793: * The optional Search-Order word set::
12794: @end menu
12795:
12796:
12797: @c =====================================================================
12798: @node The Core Words, The optional Block word set, ANS conformance, ANS conformance
12799: @comment node-name, next, previous, up
12800: @section The Core Words
12801: @c =====================================================================
12802: @cindex core words, system documentation
12803: @cindex system documentation, core words
12804:
12805: @menu
12806: * core-idef:: Implementation Defined Options
12807: * core-ambcond:: Ambiguous Conditions
12808: * core-other:: Other System Documentation
12809: @end menu
12810:
12811: @c ---------------------------------------------------------------------
12812: @node core-idef, core-ambcond, The Core Words, The Core Words
12813: @subsection Implementation Defined Options
12814: @c ---------------------------------------------------------------------
12815: @cindex core words, implementation-defined options
12816: @cindex implementation-defined options, core words
12817:
12818:
12819: @table @i
12820: @item (Cell) aligned addresses:
12821: @cindex cell-aligned addresses
12822: @cindex aligned addresses
12823: processor-dependent. Gforth's alignment words perform natural alignment
12824: (e.g., an address aligned for a datum of size 8 is divisible by
12825: 8). Unaligned accesses usually result in a @code{-23 THROW}.
12826:
12827: @item @code{EMIT} and non-graphic characters:
12828: @cindex @code{EMIT} and non-graphic characters
12829: @cindex non-graphic characters and @code{EMIT}
12830: The character is output using the C library function (actually, macro)
12831: @code{putc}.
12832:
12833: @item character editing of @code{ACCEPT} and @code{EXPECT}:
12834: @cindex character editing of @code{ACCEPT} and @code{EXPECT}
12835: @cindex editing in @code{ACCEPT} and @code{EXPECT}
12836: @cindex @code{ACCEPT}, editing
12837: @cindex @code{EXPECT}, editing
12838: This is modeled on the GNU readline library (@pxref{Readline
12839: Interaction, , Command Line Editing, readline, The GNU Readline
12840: Library}) with Emacs-like key bindings. @kbd{Tab} deviates a little by
12841: producing a full word completion every time you type it (instead of
1.28 crook 12842: producing the common prefix of all completions). @xref{Command-line editing}.
1.1 anton 12843:
12844: @item character set:
12845: @cindex character set
12846: The character set of your computer and display device. Gforth is
12847: 8-bit-clean (but some other component in your system may make trouble).
12848:
12849: @item Character-aligned address requirements:
12850: @cindex character-aligned address requirements
12851: installation-dependent. Currently a character is represented by a C
12852: @code{unsigned char}; in the future we might switch to @code{wchar_t}
12853: (Comments on that requested).
12854:
12855: @item character-set extensions and matching of names:
12856: @cindex character-set extensions and matching of names
1.26 crook 12857: @cindex case-sensitivity for name lookup
12858: @cindex name lookup, case-sensitivity
12859: @cindex locale and case-sensitivity
1.21 crook 12860: Any character except the ASCII NUL character can be used in a
1.1 anton 12861: name. Matching is case-insensitive (except in @code{TABLE}s). The
1.47 crook 12862: matching is performed using the C library function @code{strncasecmp}, whose
1.1 anton 12863: function is probably influenced by the locale. E.g., the @code{C} locale
12864: does not know about accents and umlauts, so they are matched
12865: case-sensitively in that locale. For portability reasons it is best to
12866: write programs such that they work in the @code{C} locale. Then one can
12867: use libraries written by a Polish programmer (who might use words
12868: containing ISO Latin-2 encoded characters) and by a French programmer
12869: (ISO Latin-1) in the same program (of course, @code{WORDS} will produce
12870: funny results for some of the words (which ones, depends on the font you
12871: are using)). Also, the locale you prefer may not be available in other
12872: operating systems. Hopefully, Unicode will solve these problems one day.
12873:
12874: @item conditions under which control characters match a space delimiter:
12875: @cindex space delimiters
12876: @cindex control characters as delimiters
1.117 anton 12877: If @code{word} is called with the space character as a delimiter, all
1.1 anton 12878: white-space characters (as identified by the C macro @code{isspace()})
1.117 anton 12879: are delimiters. @code{Parse}, on the other hand, treats space like other
1.138 anton 12880: delimiters. @code{Parse-name}, which is used by the outer
1.1 anton 12881: interpreter (aka text interpreter) by default, treats all white-space
12882: characters as delimiters.
12883:
1.26 crook 12884: @item format of the control-flow stack:
12885: @cindex control-flow stack, format
12886: The data stack is used as control-flow stack. The size of a control-flow
1.1 anton 12887: stack item in cells is given by the constant @code{cs-item-size}. At the
12888: time of this writing, an item consists of a (pointer to a) locals list
12889: (third), an address in the code (second), and a tag for identifying the
12890: item (TOS). The following tags are used: @code{defstart},
12891: @code{live-orig}, @code{dead-orig}, @code{dest}, @code{do-dest},
12892: @code{scopestart}.
12893:
12894: @item conversion of digits > 35
12895: @cindex digits > 35
12896: The characters @code{[\]^_'} are the digits with the decimal value
12897: 36@minus{}41. There is no way to input many of the larger digits.
12898:
12899: @item display after input terminates in @code{ACCEPT} and @code{EXPECT}:
12900: @cindex @code{EXPECT}, display after end of input
12901: @cindex @code{ACCEPT}, display after end of input
12902: The cursor is moved to the end of the entered string. If the input is
12903: terminated using the @kbd{Return} key, a space is typed.
12904:
12905: @item exception abort sequence of @code{ABORT"}:
12906: @cindex exception abort sequence of @code{ABORT"}
12907: @cindex @code{ABORT"}, exception abort sequence
12908: The error string is stored into the variable @code{"error} and a
12909: @code{-2 throw} is performed.
12910:
12911: @item input line terminator:
12912: @cindex input line terminator
12913: @cindex line terminator on input
1.26 crook 12914: @cindex newline character on input
1.1 anton 12915: For interactive input, @kbd{C-m} (CR) and @kbd{C-j} (LF) terminate
12916: lines. One of these characters is typically produced when you type the
12917: @kbd{Enter} or @kbd{Return} key.
12918:
12919: @item maximum size of a counted string:
12920: @cindex maximum size of a counted string
12921: @cindex counted string, maximum size
12922: @code{s" /counted-string" environment? drop .}. Currently 255 characters
1.79 anton 12923: on all platforms, but this may change.
1.1 anton 12924:
12925: @item maximum size of a parsed string:
12926: @cindex maximum size of a parsed string
12927: @cindex parsed string, maximum size
12928: Given by the constant @code{/line}. Currently 255 characters.
12929:
12930: @item maximum size of a definition name, in characters:
12931: @cindex maximum size of a definition name, in characters
12932: @cindex name, maximum length
1.113 anton 12933: MAXU/8
1.1 anton 12934:
12935: @item maximum string length for @code{ENVIRONMENT?}, in characters:
12936: @cindex maximum string length for @code{ENVIRONMENT?}, in characters
12937: @cindex @code{ENVIRONMENT?} string length, maximum
1.113 anton 12938: MAXU/8
1.1 anton 12939:
12940: @item method of selecting the user input device:
12941: @cindex user input device, method of selecting
12942: The user input device is the standard input. There is currently no way to
12943: change it from within Gforth. However, the input can typically be
12944: redirected in the command line that starts Gforth.
12945:
12946: @item method of selecting the user output device:
12947: @cindex user output device, method of selecting
12948: @code{EMIT} and @code{TYPE} output to the file-id stored in the value
1.10 anton 12949: @code{outfile-id} (@code{stdout} by default). Gforth uses unbuffered
12950: output when the user output device is a terminal, otherwise the output
12951: is buffered.
1.1 anton 12952:
12953: @item methods of dictionary compilation:
12954: What are we expected to document here?
12955:
12956: @item number of bits in one address unit:
12957: @cindex number of bits in one address unit
12958: @cindex address unit, size in bits
12959: @code{s" address-units-bits" environment? drop .}. 8 in all current
1.79 anton 12960: platforms.
1.1 anton 12961:
12962: @item number representation and arithmetic:
12963: @cindex number representation and arithmetic
1.79 anton 12964: Processor-dependent. Binary two's complement on all current platforms.
1.1 anton 12965:
12966: @item ranges for integer types:
12967: @cindex ranges for integer types
12968: @cindex integer types, ranges
12969: Installation-dependent. Make environmental queries for @code{MAX-N},
12970: @code{MAX-U}, @code{MAX-D} and @code{MAX-UD}. The lower bounds for
12971: unsigned (and positive) types is 0. The lower bound for signed types on
12972: two's complement and one's complement machines machines can be computed
12973: by adding 1 to the upper bound.
12974:
12975: @item read-only data space regions:
12976: @cindex read-only data space regions
12977: @cindex data-space, read-only regions
12978: The whole Forth data space is writable.
12979:
12980: @item size of buffer at @code{WORD}:
12981: @cindex size of buffer at @code{WORD}
12982: @cindex @code{WORD} buffer size
12983: @code{PAD HERE - .}. 104 characters on 32-bit machines. The buffer is
12984: shared with the pictured numeric output string. If overwriting
12985: @code{PAD} is acceptable, it is as large as the remaining dictionary
12986: space, although only as much can be sensibly used as fits in a counted
12987: string.
12988:
12989: @item size of one cell in address units:
12990: @cindex cell size
12991: @code{1 cells .}.
12992:
12993: @item size of one character in address units:
12994: @cindex char size
1.79 anton 12995: @code{1 chars .}. 1 on all current platforms.
1.1 anton 12996:
12997: @item size of the keyboard terminal buffer:
12998: @cindex size of the keyboard terminal buffer
12999: @cindex terminal buffer, size
13000: Varies. You can determine the size at a specific time using @code{lp@@
13001: tib - .}. It is shared with the locals stack and TIBs of files that
13002: include the current file. You can change the amount of space for TIBs
13003: and locals stack at Gforth startup with the command line option
13004: @code{-l}.
13005:
13006: @item size of the pictured numeric output buffer:
13007: @cindex size of the pictured numeric output buffer
13008: @cindex pictured numeric output buffer, size
13009: @code{PAD HERE - .}. 104 characters on 32-bit machines. The buffer is
13010: shared with @code{WORD}.
13011:
13012: @item size of the scratch area returned by @code{PAD}:
13013: @cindex size of the scratch area returned by @code{PAD}
13014: @cindex @code{PAD} size
13015: The remainder of dictionary space. @code{unused pad here - - .}.
13016:
13017: @item system case-sensitivity characteristics:
13018: @cindex case-sensitivity characteristics
1.26 crook 13019: Dictionary searches are case-insensitive (except in
1.1 anton 13020: @code{TABLE}s). However, as explained above under @i{character-set
13021: extensions}, the matching for non-ASCII characters is determined by the
13022: locale you are using. In the default @code{C} locale all non-ASCII
13023: characters are matched case-sensitively.
13024:
13025: @item system prompt:
13026: @cindex system prompt
13027: @cindex prompt
13028: @code{ ok} in interpret state, @code{ compiled} in compile state.
13029:
13030: @item division rounding:
13031: @cindex division rounding
1.166 anton 13032: The ordinary division words @code{/ mod /mod */ */mod} perform floored
13033: division (with the default installation of Gforth). You can check
13034: this with @code{s" floored" environment? drop .}. If you write
13035: programs that need a specific division rounding, best use
13036: @code{fm/mod} or @code{sm/rem} for portability.
1.1 anton 13037:
13038: @item values of @code{STATE} when true:
13039: @cindex @code{STATE} values
13040: -1.
13041:
13042: @item values returned after arithmetic overflow:
13043: On two's complement machines, arithmetic is performed modulo
13044: 2**bits-per-cell for single arithmetic and 4**bits-per-cell for double
1.164 anton 13045: arithmetic (with appropriate mapping for signed types). Division by
13046: zero typically results in a @code{-55 throw} (Floating-point
13047: unidentified fault) or @code{-10 throw} (divide by zero). Integer
1.166 anton 13048: division overflow can result in these throws, or in @code{-11 throw};
13049: in @code{gforth-fast} division overflow and divide by zero may also
13050: result in returning bogus results without producing an exception.
1.1 anton 13051:
13052: @item whether the current definition can be found after @t{DOES>}:
13053: @cindex @t{DOES>}, visibility of current definition
13054: No.
13055:
13056: @end table
13057:
13058: @c ---------------------------------------------------------------------
13059: @node core-ambcond, core-other, core-idef, The Core Words
13060: @subsection Ambiguous conditions
13061: @c ---------------------------------------------------------------------
13062: @cindex core words, ambiguous conditions
13063: @cindex ambiguous conditions, core words
13064:
13065: @table @i
13066:
13067: @item a name is neither a word nor a number:
13068: @cindex name not found
1.26 crook 13069: @cindex undefined word
1.80 anton 13070: @code{-13 throw} (Undefined word).
1.1 anton 13071:
13072: @item a definition name exceeds the maximum length allowed:
1.26 crook 13073: @cindex word name too long
1.1 anton 13074: @code{-19 throw} (Word name too long)
13075:
13076: @item addressing a region not inside the various data spaces of the forth system:
13077: @cindex Invalid memory address
1.32 anton 13078: The stacks, code space and header space are accessible. Machine code space is
1.1 anton 13079: typically readable. Accessing other addresses gives results dependent on
13080: the operating system. On decent systems: @code{-9 throw} (Invalid memory
13081: address).
13082:
13083: @item argument type incompatible with parameter:
1.26 crook 13084: @cindex argument type mismatch
1.1 anton 13085: This is usually not caught. Some words perform checks, e.g., the control
13086: flow words, and issue a @code{ABORT"} or @code{-12 THROW} (Argument type
13087: mismatch).
13088:
13089: @item attempting to obtain the execution token of a word with undefined execution semantics:
13090: @cindex Interpreting a compile-only word, for @code{'} etc.
13091: @cindex execution token of words with undefined execution semantics
13092: @code{-14 throw} (Interpreting a compile-only word). In some cases, you
13093: get an execution token for @code{compile-only-error} (which performs a
13094: @code{-14 throw} when executed).
13095:
13096: @item dividing by zero:
13097: @cindex dividing by zero
13098: @cindex floating point unidentified fault, integer division
1.80 anton 13099: On some platforms, this produces a @code{-10 throw} (Division by
1.24 anton 13100: zero); on other systems, this typically results in a @code{-55 throw}
13101: (Floating-point unidentified fault).
1.1 anton 13102:
13103: @item insufficient data stack or return stack space:
13104: @cindex insufficient data stack or return stack space
13105: @cindex stack overflow
1.26 crook 13106: @cindex address alignment exception, stack overflow
1.1 anton 13107: @cindex Invalid memory address, stack overflow
13108: Depending on the operating system, the installation, and the invocation
13109: of Gforth, this is either checked by the memory management hardware, or
1.24 anton 13110: it is not checked. If it is checked, you typically get a @code{-3 throw}
13111: (Stack overflow), @code{-5 throw} (Return stack overflow), or @code{-9
13112: throw} (Invalid memory address) (depending on the platform and how you
13113: achieved the overflow) as soon as the overflow happens. If it is not
13114: checked, overflows typically result in mysterious illegal memory
13115: accesses, producing @code{-9 throw} (Invalid memory address) or
13116: @code{-23 throw} (Address alignment exception); they might also destroy
13117: the internal data structure of @code{ALLOCATE} and friends, resulting in
13118: various errors in these words.
1.1 anton 13119:
13120: @item insufficient space for loop control parameters:
13121: @cindex insufficient space for loop control parameters
1.80 anton 13122: Like other return stack overflows.
1.1 anton 13123:
13124: @item insufficient space in the dictionary:
13125: @cindex insufficient space in the dictionary
13126: @cindex dictionary overflow
1.12 anton 13127: If you try to allot (either directly with @code{allot}, or indirectly
13128: with @code{,}, @code{create} etc.) more memory than available in the
13129: dictionary, you get a @code{-8 throw} (Dictionary overflow). If you try
13130: to access memory beyond the end of the dictionary, the results are
13131: similar to stack overflows.
1.1 anton 13132:
13133: @item interpreting a word with undefined interpretation semantics:
13134: @cindex interpreting a word with undefined interpretation semantics
13135: @cindex Interpreting a compile-only word
13136: For some words, we have defined interpretation semantics. For the
13137: others: @code{-14 throw} (Interpreting a compile-only word).
13138:
13139: @item modifying the contents of the input buffer or a string literal:
13140: @cindex modifying the contents of the input buffer or a string literal
13141: These are located in writable memory and can be modified.
13142:
13143: @item overflow of the pictured numeric output string:
13144: @cindex overflow of the pictured numeric output string
13145: @cindex pictured numeric output string, overflow
1.24 anton 13146: @code{-17 throw} (Pictured numeric ouput string overflow).
1.1 anton 13147:
13148: @item parsed string overflow:
13149: @cindex parsed string overflow
13150: @code{PARSE} cannot overflow. @code{WORD} does not check for overflow.
13151:
13152: @item producing a result out of range:
13153: @cindex result out of range
13154: On two's complement machines, arithmetic is performed modulo
13155: 2**bits-per-cell for single arithmetic and 4**bits-per-cell for double
1.166 anton 13156: arithmetic (with appropriate mapping for signed types). Division by
13157: zero typically results in a @code{-10 throw} (divide by zero) or
13158: @code{-55 throw} (floating point unidentified fault). Overflow on
13159: division may result in these errors or in @code{-11 throw} (result out
13160: of range). @code{Gforth-fast} may silently produce bogus results on
13161: division overflow or division by zero. @code{Convert} and
1.24 anton 13162: @code{>number} currently overflow silently.
1.1 anton 13163:
13164: @item reading from an empty data or return stack:
13165: @cindex stack empty
13166: @cindex stack underflow
1.24 anton 13167: @cindex return stack underflow
1.1 anton 13168: The data stack is checked by the outer (aka text) interpreter after
13169: every word executed. If it has underflowed, a @code{-4 throw} (Stack
13170: underflow) is performed. Apart from that, stacks may be checked or not,
1.24 anton 13171: depending on operating system, installation, and invocation. If they are
13172: caught by a check, they typically result in @code{-4 throw} (Stack
13173: underflow), @code{-6 throw} (Return stack underflow) or @code{-9 throw}
13174: (Invalid memory address), depending on the platform and which stack
13175: underflows and by how much. Note that even if the system uses checking
13176: (through the MMU), your program may have to underflow by a significant
13177: number of stack items to trigger the reaction (the reason for this is
13178: that the MMU, and therefore the checking, works with a page-size
13179: granularity). If there is no checking, the symptoms resulting from an
13180: underflow are similar to those from an overflow. Unbalanced return
1.80 anton 13181: stack errors can result in a variety of symptoms, including @code{-9 throw}
1.24 anton 13182: (Invalid memory address) and Illegal Instruction (typically @code{-260
13183: throw}).
1.1 anton 13184:
13185: @item unexpected end of the input buffer, resulting in an attempt to use a zero-length string as a name:
13186: @cindex unexpected end of the input buffer
13187: @cindex zero-length string as a name
13188: @cindex Attempt to use zero-length string as a name
13189: @code{Create} and its descendants perform a @code{-16 throw} (Attempt to
13190: use zero-length string as a name). Words like @code{'} probably will not
13191: find what they search. Note that it is possible to create zero-length
13192: names with @code{nextname} (should it not?).
13193:
13194: @item @code{>IN} greater than input buffer:
13195: @cindex @code{>IN} greater than input buffer
13196: The next invocation of a parsing word returns a string with length 0.
13197:
13198: @item @code{RECURSE} appears after @code{DOES>}:
13199: @cindex @code{RECURSE} appears after @code{DOES>}
13200: Compiles a recursive call to the defining word, not to the defined word.
13201:
13202: @item argument input source different than current input source for @code{RESTORE-INPUT}:
13203: @cindex argument input source different than current input source for @code{RESTORE-INPUT}
1.26 crook 13204: @cindex argument type mismatch, @code{RESTORE-INPUT}
1.1 anton 13205: @cindex @code{RESTORE-INPUT}, Argument type mismatch
13206: @code{-12 THROW}. Note that, once an input file is closed (e.g., because
13207: the end of the file was reached), its source-id may be
13208: reused. Therefore, restoring an input source specification referencing a
13209: closed file may lead to unpredictable results instead of a @code{-12
13210: THROW}.
13211:
13212: In the future, Gforth may be able to restore input source specifications
13213: from other than the current input source.
13214:
13215: @item data space containing definitions gets de-allocated:
13216: @cindex data space containing definitions gets de-allocated
13217: Deallocation with @code{allot} is not checked. This typically results in
13218: memory access faults or execution of illegal instructions.
13219:
13220: @item data space read/write with incorrect alignment:
13221: @cindex data space read/write with incorrect alignment
13222: @cindex alignment faults
1.26 crook 13223: @cindex address alignment exception
1.1 anton 13224: Processor-dependent. Typically results in a @code{-23 throw} (Address
1.12 anton 13225: alignment exception). Under Linux-Intel on a 486 or later processor with
1.1 anton 13226: alignment turned on, incorrect alignment results in a @code{-9 throw}
13227: (Invalid memory address). There are reportedly some processors with
1.12 anton 13228: alignment restrictions that do not report violations.
1.1 anton 13229:
13230: @item data space pointer not properly aligned, @code{,}, @code{C,}:
13231: @cindex data space pointer not properly aligned, @code{,}, @code{C,}
13232: Like other alignment errors.
13233:
13234: @item less than u+2 stack items (@code{PICK} and @code{ROLL}):
13235: Like other stack underflows.
13236:
13237: @item loop control parameters not available:
13238: @cindex loop control parameters not available
13239: Not checked. The counted loop words simply assume that the top of return
13240: stack items are loop control parameters and behave accordingly.
13241:
13242: @item most recent definition does not have a name (@code{IMMEDIATE}):
13243: @cindex most recent definition does not have a name (@code{IMMEDIATE})
13244: @cindex last word was headerless
13245: @code{abort" last word was headerless"}.
13246:
13247: @item name not defined by @code{VALUE} used by @code{TO}:
13248: @cindex name not defined by @code{VALUE} used by @code{TO}
13249: @cindex @code{TO} on non-@code{VALUE}s
13250: @cindex Invalid name argument, @code{TO}
13251: @code{-32 throw} (Invalid name argument) (unless name is a local or was
13252: defined by @code{CONSTANT}; in the latter case it just changes the constant).
13253:
13254: @item name not found (@code{'}, @code{POSTPONE}, @code{[']}, @code{[COMPILE]}):
13255: @cindex name not found (@code{'}, @code{POSTPONE}, @code{[']}, @code{[COMPILE]})
1.26 crook 13256: @cindex undefined word, @code{'}, @code{POSTPONE}, @code{[']}, @code{[COMPILE]}
1.1 anton 13257: @code{-13 throw} (Undefined word)
13258:
13259: @item parameters are not of the same type (@code{DO}, @code{?DO}, @code{WITHIN}):
13260: @cindex parameters are not of the same type (@code{DO}, @code{?DO}, @code{WITHIN})
13261: Gforth behaves as if they were of the same type. I.e., you can predict
13262: the behaviour by interpreting all parameters as, e.g., signed.
13263:
13264: @item @code{POSTPONE} or @code{[COMPILE]} applied to @code{TO}:
13265: @cindex @code{POSTPONE} or @code{[COMPILE]} applied to @code{TO}
13266: Assume @code{: X POSTPONE TO ; IMMEDIATE}. @code{X} performs the
13267: compilation semantics of @code{TO}.
13268:
13269: @item String longer than a counted string returned by @code{WORD}:
1.26 crook 13270: @cindex string longer than a counted string returned by @code{WORD}
1.1 anton 13271: @cindex @code{WORD}, string overflow
13272: Not checked. The string will be ok, but the count will, of course,
13273: contain only the least significant bits of the length.
13274:
13275: @item u greater than or equal to the number of bits in a cell (@code{LSHIFT}, @code{RSHIFT}):
13276: @cindex @code{LSHIFT}, large shift counts
13277: @cindex @code{RSHIFT}, large shift counts
13278: Processor-dependent. Typical behaviours are returning 0 and using only
13279: the low bits of the shift count.
13280:
13281: @item word not defined via @code{CREATE}:
13282: @cindex @code{>BODY} of non-@code{CREATE}d words
13283: @code{>BODY} produces the PFA of the word no matter how it was defined.
13284:
13285: @cindex @code{DOES>} of non-@code{CREATE}d words
13286: @code{DOES>} changes the execution semantics of the last defined word no
13287: matter how it was defined. E.g., @code{CONSTANT DOES>} is equivalent to
13288: @code{CREATE , DOES>}.
13289:
13290: @item words improperly used outside @code{<#} and @code{#>}:
13291: Not checked. As usual, you can expect memory faults.
13292:
13293: @end table
13294:
13295:
13296: @c ---------------------------------------------------------------------
13297: @node core-other, , core-ambcond, The Core Words
13298: @subsection Other system documentation
13299: @c ---------------------------------------------------------------------
13300: @cindex other system documentation, core words
13301: @cindex core words, other system documentation
13302:
13303: @table @i
13304: @item nonstandard words using @code{PAD}:
13305: @cindex @code{PAD} use by nonstandard words
13306: None.
13307:
13308: @item operator's terminal facilities available:
13309: @cindex operator's terminal facilities available
1.80 anton 13310: After processing the OS's command line, Gforth goes into interactive mode,
1.1 anton 13311: and you can give commands to Gforth interactively. The actual facilities
13312: available depend on how you invoke Gforth.
13313:
13314: @item program data space available:
13315: @cindex program data space available
13316: @cindex data space available
13317: @code{UNUSED .} gives the remaining dictionary space. The total
13318: dictionary space can be specified with the @code{-m} switch
13319: (@pxref{Invoking Gforth}) when Gforth starts up.
13320:
13321: @item return stack space available:
13322: @cindex return stack space available
13323: You can compute the total return stack space in cells with
13324: @code{s" RETURN-STACK-CELLS" environment? drop .}. You can specify it at
13325: startup time with the @code{-r} switch (@pxref{Invoking Gforth}).
13326:
13327: @item stack space available:
13328: @cindex stack space available
13329: You can compute the total data stack space in cells with
13330: @code{s" STACK-CELLS" environment? drop .}. You can specify it at
13331: startup time with the @code{-d} switch (@pxref{Invoking Gforth}).
13332:
13333: @item system dictionary space required, in address units:
13334: @cindex system dictionary space required, in address units
13335: Type @code{here forthstart - .} after startup. At the time of this
13336: writing, this gives 80080 (bytes) on a 32-bit system.
13337: @end table
13338:
13339:
13340: @c =====================================================================
13341: @node The optional Block word set, The optional Double Number word set, The Core Words, ANS conformance
13342: @section The optional Block word set
13343: @c =====================================================================
13344: @cindex system documentation, block words
13345: @cindex block words, system documentation
13346:
13347: @menu
13348: * block-idef:: Implementation Defined Options
13349: * block-ambcond:: Ambiguous Conditions
13350: * block-other:: Other System Documentation
13351: @end menu
13352:
13353:
13354: @c ---------------------------------------------------------------------
13355: @node block-idef, block-ambcond, The optional Block word set, The optional Block word set
13356: @subsection Implementation Defined Options
13357: @c ---------------------------------------------------------------------
13358: @cindex implementation-defined options, block words
13359: @cindex block words, implementation-defined options
13360:
13361: @table @i
13362: @item the format for display by @code{LIST}:
13363: @cindex @code{LIST} display format
13364: First the screen number is displayed, then 16 lines of 64 characters,
13365: each line preceded by the line number.
13366:
13367: @item the length of a line affected by @code{\}:
13368: @cindex length of a line affected by @code{\}
13369: @cindex @code{\}, line length in blocks
13370: 64 characters.
13371: @end table
13372:
13373:
13374: @c ---------------------------------------------------------------------
13375: @node block-ambcond, block-other, block-idef, The optional Block word set
13376: @subsection Ambiguous conditions
13377: @c ---------------------------------------------------------------------
13378: @cindex block words, ambiguous conditions
13379: @cindex ambiguous conditions, block words
13380:
13381: @table @i
13382: @item correct block read was not possible:
13383: @cindex block read not possible
13384: Typically results in a @code{throw} of some OS-derived value (between
13385: -512 and -2048). If the blocks file was just not long enough, blanks are
13386: supplied for the missing portion.
13387:
13388: @item I/O exception in block transfer:
13389: @cindex I/O exception in block transfer
13390: @cindex block transfer, I/O exception
13391: Typically results in a @code{throw} of some OS-derived value (between
13392: -512 and -2048).
13393:
13394: @item invalid block number:
13395: @cindex invalid block number
13396: @cindex block number invalid
13397: @code{-35 throw} (Invalid block number)
13398:
13399: @item a program directly alters the contents of @code{BLK}:
13400: @cindex @code{BLK}, altering @code{BLK}
13401: The input stream is switched to that other block, at the same
13402: position. If the storing to @code{BLK} happens when interpreting
13403: non-block input, the system will get quite confused when the block ends.
13404:
13405: @item no current block buffer for @code{UPDATE}:
13406: @cindex @code{UPDATE}, no current block buffer
13407: @code{UPDATE} has no effect.
13408:
13409: @end table
13410:
13411: @c ---------------------------------------------------------------------
13412: @node block-other, , block-ambcond, The optional Block word set
13413: @subsection Other system documentation
13414: @c ---------------------------------------------------------------------
13415: @cindex other system documentation, block words
13416: @cindex block words, other system documentation
13417:
13418: @table @i
13419: @item any restrictions a multiprogramming system places on the use of buffer addresses:
13420: No restrictions (yet).
13421:
13422: @item the number of blocks available for source and data:
13423: depends on your disk space.
13424:
13425: @end table
13426:
13427:
13428: @c =====================================================================
13429: @node The optional Double Number word set, The optional Exception word set, The optional Block word set, ANS conformance
13430: @section The optional Double Number word set
13431: @c =====================================================================
13432: @cindex system documentation, double words
13433: @cindex double words, system documentation
13434:
13435: @menu
13436: * double-ambcond:: Ambiguous Conditions
13437: @end menu
13438:
13439:
13440: @c ---------------------------------------------------------------------
13441: @node double-ambcond, , The optional Double Number word set, The optional Double Number word set
13442: @subsection Ambiguous conditions
13443: @c ---------------------------------------------------------------------
13444: @cindex double words, ambiguous conditions
13445: @cindex ambiguous conditions, double words
13446:
13447: @table @i
1.29 crook 13448: @item @i{d} outside of range of @i{n} in @code{D>S}:
13449: @cindex @code{D>S}, @i{d} out of range of @i{n}
13450: The least significant cell of @i{d} is produced.
1.1 anton 13451:
13452: @end table
13453:
13454:
13455: @c =====================================================================
13456: @node The optional Exception word set, The optional Facility word set, The optional Double Number word set, ANS conformance
13457: @section The optional Exception word set
13458: @c =====================================================================
13459: @cindex system documentation, exception words
13460: @cindex exception words, system documentation
13461:
13462: @menu
13463: * exception-idef:: Implementation Defined Options
13464: @end menu
13465:
13466:
13467: @c ---------------------------------------------------------------------
13468: @node exception-idef, , The optional Exception word set, The optional Exception word set
13469: @subsection Implementation Defined Options
13470: @c ---------------------------------------------------------------------
13471: @cindex implementation-defined options, exception words
13472: @cindex exception words, implementation-defined options
13473:
13474: @table @i
13475: @item @code{THROW}-codes used in the system:
13476: @cindex @code{THROW}-codes used in the system
13477: The codes -256@minus{}-511 are used for reporting signals. The mapping
1.29 crook 13478: from OS signal numbers to throw codes is -256@minus{}@i{signal}. The
1.1 anton 13479: codes -512@minus{}-2047 are used for OS errors (for file and memory
13480: allocation operations). The mapping from OS error numbers to throw codes
13481: is -512@minus{}@code{errno}. One side effect of this mapping is that
13482: undefined OS errors produce a message with a strange number; e.g.,
13483: @code{-1000 THROW} results in @code{Unknown error 488} on my system.
13484: @end table
13485:
13486: @c =====================================================================
13487: @node The optional Facility word set, The optional File-Access word set, The optional Exception word set, ANS conformance
13488: @section The optional Facility word set
13489: @c =====================================================================
13490: @cindex system documentation, facility words
13491: @cindex facility words, system documentation
13492:
13493: @menu
13494: * facility-idef:: Implementation Defined Options
13495: * facility-ambcond:: Ambiguous Conditions
13496: @end menu
13497:
13498:
13499: @c ---------------------------------------------------------------------
13500: @node facility-idef, facility-ambcond, The optional Facility word set, The optional Facility word set
13501: @subsection Implementation Defined Options
13502: @c ---------------------------------------------------------------------
13503: @cindex implementation-defined options, facility words
13504: @cindex facility words, implementation-defined options
13505:
13506: @table @i
13507: @item encoding of keyboard events (@code{EKEY}):
13508: @cindex keyboard events, encoding in @code{EKEY}
13509: @cindex @code{EKEY}, encoding of keyboard events
1.40 anton 13510: Keys corresponding to ASCII characters are encoded as ASCII characters.
1.41 anton 13511: Other keys are encoded with the constants @code{k-left}, @code{k-right},
13512: @code{k-up}, @code{k-down}, @code{k-home}, @code{k-end}, @code{k1},
13513: @code{k2}, @code{k3}, @code{k4}, @code{k5}, @code{k6}, @code{k7},
13514: @code{k8}, @code{k9}, @code{k10}, @code{k11}, @code{k12}.
1.40 anton 13515:
1.1 anton 13516:
13517: @item duration of a system clock tick:
13518: @cindex duration of a system clock tick
13519: @cindex clock tick duration
13520: System dependent. With respect to @code{MS}, the time is specified in
13521: microseconds. How well the OS and the hardware implement this, is
13522: another question.
13523:
13524: @item repeatability to be expected from the execution of @code{MS}:
13525: @cindex repeatability to be expected from the execution of @code{MS}
13526: @cindex @code{MS}, repeatability to be expected
13527: System dependent. On Unix, a lot depends on load. If the system is
13528: lightly loaded, and the delay is short enough that Gforth does not get
13529: swapped out, the performance should be acceptable. Under MS-DOS and
13530: other single-tasking systems, it should be good.
13531:
13532: @end table
13533:
13534:
13535: @c ---------------------------------------------------------------------
13536: @node facility-ambcond, , facility-idef, The optional Facility word set
13537: @subsection Ambiguous conditions
13538: @c ---------------------------------------------------------------------
13539: @cindex facility words, ambiguous conditions
13540: @cindex ambiguous conditions, facility words
13541:
13542: @table @i
13543: @item @code{AT-XY} can't be performed on user output device:
13544: @cindex @code{AT-XY} can't be performed on user output device
13545: Largely terminal dependent. No range checks are done on the arguments.
13546: No errors are reported. You may see some garbage appearing, you may see
13547: simply nothing happen.
13548:
13549: @end table
13550:
13551:
13552: @c =====================================================================
13553: @node The optional File-Access word set, The optional Floating-Point word set, The optional Facility word set, ANS conformance
13554: @section The optional File-Access word set
13555: @c =====================================================================
13556: @cindex system documentation, file words
13557: @cindex file words, system documentation
13558:
13559: @menu
13560: * file-idef:: Implementation Defined Options
13561: * file-ambcond:: Ambiguous Conditions
13562: @end menu
13563:
13564: @c ---------------------------------------------------------------------
13565: @node file-idef, file-ambcond, The optional File-Access word set, The optional File-Access word set
13566: @subsection Implementation Defined Options
13567: @c ---------------------------------------------------------------------
13568: @cindex implementation-defined options, file words
13569: @cindex file words, implementation-defined options
13570:
13571: @table @i
13572: @item file access methods used:
13573: @cindex file access methods used
13574: @code{R/O}, @code{R/W} and @code{BIN} work as you would
13575: expect. @code{W/O} translates into the C file opening mode @code{w} (or
13576: @code{wb}): The file is cleared, if it exists, and created, if it does
13577: not (with both @code{open-file} and @code{create-file}). Under Unix
13578: @code{create-file} creates a file with 666 permissions modified by your
13579: umask.
13580:
13581: @item file exceptions:
13582: @cindex file exceptions
13583: The file words do not raise exceptions (except, perhaps, memory access
13584: faults when you pass illegal addresses or file-ids).
13585:
13586: @item file line terminator:
13587: @cindex file line terminator
13588: System-dependent. Gforth uses C's newline character as line
13589: terminator. What the actual character code(s) of this are is
13590: system-dependent.
13591:
13592: @item file name format:
13593: @cindex file name format
13594: System dependent. Gforth just uses the file name format of your OS.
13595:
13596: @item information returned by @code{FILE-STATUS}:
13597: @cindex @code{FILE-STATUS}, returned information
13598: @code{FILE-STATUS} returns the most powerful file access mode allowed
13599: for the file: Either @code{R/O}, @code{W/O} or @code{R/W}. If the file
13600: cannot be accessed, @code{R/O BIN} is returned. @code{BIN} is applicable
13601: along with the returned mode.
13602:
13603: @item input file state after an exception when including source:
13604: @cindex exception when including source
13605: All files that are left via the exception are closed.
13606:
1.29 crook 13607: @item @i{ior} values and meaning:
13608: @cindex @i{ior} values and meaning
1.68 anton 13609: @cindex @i{wior} values and meaning
1.29 crook 13610: The @i{ior}s returned by the file and memory allocation words are
1.1 anton 13611: intended as throw codes. They typically are in the range
13612: -512@minus{}-2047 of OS errors. The mapping from OS error numbers to
1.29 crook 13613: @i{ior}s is -512@minus{}@i{errno}.
1.1 anton 13614:
13615: @item maximum depth of file input nesting:
13616: @cindex maximum depth of file input nesting
13617: @cindex file input nesting, maximum depth
13618: limited by the amount of return stack, locals/TIB stack, and the number
13619: of open files available. This should not give you troubles.
13620:
13621: @item maximum size of input line:
13622: @cindex maximum size of input line
13623: @cindex input line size, maximum
13624: @code{/line}. Currently 255.
13625:
13626: @item methods of mapping block ranges to files:
13627: @cindex mapping block ranges to files
13628: @cindex files containing blocks
13629: @cindex blocks in files
13630: By default, blocks are accessed in the file @file{blocks.fb} in the
13631: current working directory. The file can be switched with @code{USE}.
13632:
13633: @item number of string buffers provided by @code{S"}:
13634: @cindex @code{S"}, number of string buffers
13635: 1
13636:
13637: @item size of string buffer used by @code{S"}:
13638: @cindex @code{S"}, size of string buffer
13639: @code{/line}. currently 255.
13640:
13641: @end table
13642:
13643: @c ---------------------------------------------------------------------
13644: @node file-ambcond, , file-idef, The optional File-Access word set
13645: @subsection Ambiguous conditions
13646: @c ---------------------------------------------------------------------
13647: @cindex file words, ambiguous conditions
13648: @cindex ambiguous conditions, file words
13649:
13650: @table @i
13651: @item attempting to position a file outside its boundaries:
13652: @cindex @code{REPOSITION-FILE}, outside the file's boundaries
13653: @code{REPOSITION-FILE} is performed as usual: Afterwards,
13654: @code{FILE-POSITION} returns the value given to @code{REPOSITION-FILE}.
13655:
13656: @item attempting to read from file positions not yet written:
13657: @cindex reading from file positions not yet written
13658: End-of-file, i.e., zero characters are read and no error is reported.
13659:
1.29 crook 13660: @item @i{file-id} is invalid (@code{INCLUDE-FILE}):
13661: @cindex @code{INCLUDE-FILE}, @i{file-id} is invalid
1.1 anton 13662: An appropriate exception may be thrown, but a memory fault or other
13663: problem is more probable.
13664:
1.29 crook 13665: @item I/O exception reading or closing @i{file-id} (@code{INCLUDE-FILE}, @code{INCLUDED}):
13666: @cindex @code{INCLUDE-FILE}, I/O exception reading or closing @i{file-id}
13667: @cindex @code{INCLUDED}, I/O exception reading or closing @i{file-id}
13668: The @i{ior} produced by the operation, that discovered the problem, is
1.1 anton 13669: thrown.
13670:
13671: @item named file cannot be opened (@code{INCLUDED}):
13672: @cindex @code{INCLUDED}, named file cannot be opened
1.29 crook 13673: The @i{ior} produced by @code{open-file} is thrown.
1.1 anton 13674:
13675: @item requesting an unmapped block number:
13676: @cindex unmapped block numbers
13677: There are no unmapped legal block numbers. On some operating systems,
13678: writing a block with a large number may overflow the file system and
13679: have an error message as consequence.
13680:
13681: @item using @code{source-id} when @code{blk} is non-zero:
13682: @cindex @code{SOURCE-ID}, behaviour when @code{BLK} is non-zero
13683: @code{source-id} performs its function. Typically it will give the id of
13684: the source which loaded the block. (Better ideas?)
13685:
13686: @end table
13687:
13688:
13689: @c =====================================================================
13690: @node The optional Floating-Point word set, The optional Locals word set, The optional File-Access word set, ANS conformance
13691: @section The optional Floating-Point word set
13692: @c =====================================================================
13693: @cindex system documentation, floating-point words
13694: @cindex floating-point words, system documentation
13695:
13696: @menu
13697: * floating-idef:: Implementation Defined Options
13698: * floating-ambcond:: Ambiguous Conditions
13699: @end menu
13700:
13701:
13702: @c ---------------------------------------------------------------------
13703: @node floating-idef, floating-ambcond, The optional Floating-Point word set, The optional Floating-Point word set
13704: @subsection Implementation Defined Options
13705: @c ---------------------------------------------------------------------
13706: @cindex implementation-defined options, floating-point words
13707: @cindex floating-point words, implementation-defined options
13708:
13709: @table @i
13710: @item format and range of floating point numbers:
13711: @cindex format and range of floating point numbers
13712: @cindex floating point numbers, format and range
13713: System-dependent; the @code{double} type of C.
13714:
1.29 crook 13715: @item results of @code{REPRESENT} when @i{float} is out of range:
13716: @cindex @code{REPRESENT}, results when @i{float} is out of range
1.1 anton 13717: System dependent; @code{REPRESENT} is implemented using the C library
13718: function @code{ecvt()} and inherits its behaviour in this respect.
13719:
13720: @item rounding or truncation of floating-point numbers:
13721: @cindex rounding of floating-point numbers
13722: @cindex truncation of floating-point numbers
13723: @cindex floating-point numbers, rounding or truncation
13724: System dependent; the rounding behaviour is inherited from the hosting C
13725: compiler. IEEE-FP-based (i.e., most) systems by default round to
13726: nearest, and break ties by rounding to even (i.e., such that the last
13727: bit of the mantissa is 0).
13728:
13729: @item size of floating-point stack:
13730: @cindex floating-point stack size
13731: @code{s" FLOATING-STACK" environment? drop .} gives the total size of
13732: the floating-point stack (in floats). You can specify this on startup
13733: with the command-line option @code{-f} (@pxref{Invoking Gforth}).
13734:
13735: @item width of floating-point stack:
13736: @cindex floating-point stack width
13737: @code{1 floats}.
13738:
13739: @end table
13740:
13741:
13742: @c ---------------------------------------------------------------------
13743: @node floating-ambcond, , floating-idef, The optional Floating-Point word set
13744: @subsection Ambiguous conditions
13745: @c ---------------------------------------------------------------------
13746: @cindex floating-point words, ambiguous conditions
13747: @cindex ambiguous conditions, floating-point words
13748:
13749: @table @i
13750: @item @code{df@@} or @code{df!} used with an address that is not double-float aligned:
13751: @cindex @code{df@@} or @code{df!} used with an address that is not double-float aligned
13752: System-dependent. Typically results in a @code{-23 THROW} like other
13753: alignment violations.
13754:
13755: @item @code{f@@} or @code{f!} used with an address that is not float aligned:
13756: @cindex @code{f@@} used with an address that is not float aligned
13757: @cindex @code{f!} used with an address that is not float aligned
13758: System-dependent. Typically results in a @code{-23 THROW} like other
13759: alignment violations.
13760:
13761: @item floating-point result out of range:
13762: @cindex floating-point result out of range
1.80 anton 13763: System-dependent. Can result in a @code{-43 throw} (floating point
13764: overflow), @code{-54 throw} (floating point underflow), @code{-41 throw}
13765: (floating point inexact result), @code{-55 THROW} (Floating-point
1.1 anton 13766: unidentified fault), or can produce a special value representing, e.g.,
13767: Infinity.
13768:
13769: @item @code{sf@@} or @code{sf!} used with an address that is not single-float aligned:
13770: @cindex @code{sf@@} or @code{sf!} used with an address that is not single-float aligned
13771: System-dependent. Typically results in an alignment fault like other
13772: alignment violations.
13773:
1.35 anton 13774: @item @code{base} is not decimal (@code{REPRESENT}, @code{F.}, @code{FE.}, @code{FS.}):
13775: @cindex @code{base} is not decimal (@code{REPRESENT}, @code{F.}, @code{FE.}, @code{FS.})
1.1 anton 13776: The floating-point number is converted into decimal nonetheless.
13777:
13778: @item Both arguments are equal to zero (@code{FATAN2}):
13779: @cindex @code{FATAN2}, both arguments are equal to zero
13780: System-dependent. @code{FATAN2} is implemented using the C library
13781: function @code{atan2()}.
13782:
1.29 crook 13783: @item Using @code{FTAN} on an argument @i{r1} where cos(@i{r1}) is zero:
13784: @cindex @code{FTAN} on an argument @i{r1} where cos(@i{r1}) is zero
13785: System-dependent. Anyway, typically the cos of @i{r1} will not be zero
1.1 anton 13786: because of small errors and the tan will be a very large (or very small)
13787: but finite number.
13788:
1.29 crook 13789: @item @i{d} cannot be presented precisely as a float in @code{D>F}:
13790: @cindex @code{D>F}, @i{d} cannot be presented precisely as a float
1.1 anton 13791: The result is rounded to the nearest float.
13792:
13793: @item dividing by zero:
13794: @cindex dividing by zero, floating-point
13795: @cindex floating-point dividing by zero
13796: @cindex floating-point unidentified fault, FP divide-by-zero
1.80 anton 13797: Platform-dependent; can produce an Infinity, NaN, @code{-42 throw}
13798: (floating point divide by zero) or @code{-55 throw} (Floating-point
13799: unidentified fault).
1.1 anton 13800:
13801: @item exponent too big for conversion (@code{DF!}, @code{DF@@}, @code{SF!}, @code{SF@@}):
13802: @cindex exponent too big for conversion (@code{DF!}, @code{DF@@}, @code{SF!}, @code{SF@@})
13803: System dependent. On IEEE-FP based systems the number is converted into
13804: an infinity.
13805:
1.29 crook 13806: @item @i{float}<1 (@code{FACOSH}):
13807: @cindex @code{FACOSH}, @i{float}<1
1.1 anton 13808: @cindex floating-point unidentified fault, @code{FACOSH}
1.80 anton 13809: Platform-dependent; on IEEE-FP systems typically produces a NaN.
1.1 anton 13810:
1.29 crook 13811: @item @i{float}=<-1 (@code{FLNP1}):
13812: @cindex @code{FLNP1}, @i{float}=<-1
1.1 anton 13813: @cindex floating-point unidentified fault, @code{FLNP1}
1.80 anton 13814: Platform-dependent; on IEEE-FP systems typically produces a NaN (or a
13815: negative infinity for @i{float}=-1).
1.1 anton 13816:
1.29 crook 13817: @item @i{float}=<0 (@code{FLN}, @code{FLOG}):
13818: @cindex @code{FLN}, @i{float}=<0
13819: @cindex @code{FLOG}, @i{float}=<0
1.1 anton 13820: @cindex floating-point unidentified fault, @code{FLN} or @code{FLOG}
1.80 anton 13821: Platform-dependent; on IEEE-FP systems typically produces a NaN (or a
13822: negative infinity for @i{float}=0).
1.1 anton 13823:
1.29 crook 13824: @item @i{float}<0 (@code{FASINH}, @code{FSQRT}):
13825: @cindex @code{FASINH}, @i{float}<0
13826: @cindex @code{FSQRT}, @i{float}<0
1.1 anton 13827: @cindex floating-point unidentified fault, @code{FASINH} or @code{FSQRT}
1.80 anton 13828: Platform-dependent; for @code{fsqrt} this typically gives a NaN, for
13829: @code{fasinh} some platforms produce a NaN, others a number (bug in the
13830: C library?).
1.1 anton 13831:
1.29 crook 13832: @item |@i{float}|>1 (@code{FACOS}, @code{FASIN}, @code{FATANH}):
13833: @cindex @code{FACOS}, |@i{float}|>1
13834: @cindex @code{FASIN}, |@i{float}|>1
13835: @cindex @code{FATANH}, |@i{float}|>1
1.1 anton 13836: @cindex floating-point unidentified fault, @code{FACOS}, @code{FASIN} or @code{FATANH}
1.80 anton 13837: Platform-dependent; IEEE-FP systems typically produce a NaN.
1.1 anton 13838:
1.29 crook 13839: @item integer part of float cannot be represented by @i{d} in @code{F>D}:
13840: @cindex @code{F>D}, integer part of float cannot be represented by @i{d}
1.1 anton 13841: @cindex floating-point unidentified fault, @code{F>D}
1.80 anton 13842: Platform-dependent; typically, some double number is produced and no
13843: error is reported.
1.1 anton 13844:
13845: @item string larger than pictured numeric output area (@code{f.}, @code{fe.}, @code{fs.}):
13846: @cindex string larger than pictured numeric output area (@code{f.}, @code{fe.}, @code{fs.})
1.80 anton 13847: @code{Precision} characters of the numeric output area are used. If
13848: @code{precision} is too high, these words will smash the data or code
13849: close to @code{here}.
1.1 anton 13850: @end table
13851:
13852: @c =====================================================================
13853: @node The optional Locals word set, The optional Memory-Allocation word set, The optional Floating-Point word set, ANS conformance
13854: @section The optional Locals word set
13855: @c =====================================================================
13856: @cindex system documentation, locals words
13857: @cindex locals words, system documentation
13858:
13859: @menu
13860: * locals-idef:: Implementation Defined Options
13861: * locals-ambcond:: Ambiguous Conditions
13862: @end menu
13863:
13864:
13865: @c ---------------------------------------------------------------------
13866: @node locals-idef, locals-ambcond, The optional Locals word set, The optional Locals word set
13867: @subsection Implementation Defined Options
13868: @c ---------------------------------------------------------------------
13869: @cindex implementation-defined options, locals words
13870: @cindex locals words, implementation-defined options
13871:
13872: @table @i
13873: @item maximum number of locals in a definition:
13874: @cindex maximum number of locals in a definition
13875: @cindex locals, maximum number in a definition
13876: @code{s" #locals" environment? drop .}. Currently 15. This is a lower
13877: bound, e.g., on a 32-bit machine there can be 41 locals of up to 8
13878: characters. The number of locals in a definition is bounded by the size
13879: of locals-buffer, which contains the names of the locals.
13880:
13881: @end table
13882:
13883:
13884: @c ---------------------------------------------------------------------
13885: @node locals-ambcond, , locals-idef, The optional Locals word set
13886: @subsection Ambiguous conditions
13887: @c ---------------------------------------------------------------------
13888: @cindex locals words, ambiguous conditions
13889: @cindex ambiguous conditions, locals words
13890:
13891: @table @i
13892: @item executing a named local in interpretation state:
13893: @cindex local in interpretation state
13894: @cindex Interpreting a compile-only word, for a local
13895: Locals have no interpretation semantics. If you try to perform the
13896: interpretation semantics, you will get a @code{-14 throw} somewhere
13897: (Interpreting a compile-only word). If you perform the compilation
13898: semantics, the locals access will be compiled (irrespective of state).
13899:
1.29 crook 13900: @item @i{name} not defined by @code{VALUE} or @code{(LOCAL)} (@code{TO}):
1.1 anton 13901: @cindex name not defined by @code{VALUE} or @code{(LOCAL)} used by @code{TO}
13902: @cindex @code{TO} on non-@code{VALUE}s and non-locals
13903: @cindex Invalid name argument, @code{TO}
13904: @code{-32 throw} (Invalid name argument)
13905:
13906: @end table
13907:
13908:
13909: @c =====================================================================
13910: @node The optional Memory-Allocation word set, The optional Programming-Tools word set, The optional Locals word set, ANS conformance
13911: @section The optional Memory-Allocation word set
13912: @c =====================================================================
13913: @cindex system documentation, memory-allocation words
13914: @cindex memory-allocation words, system documentation
13915:
13916: @menu
13917: * memory-idef:: Implementation Defined Options
13918: @end menu
13919:
13920:
13921: @c ---------------------------------------------------------------------
13922: @node memory-idef, , The optional Memory-Allocation word set, The optional Memory-Allocation word set
13923: @subsection Implementation Defined Options
13924: @c ---------------------------------------------------------------------
13925: @cindex implementation-defined options, memory-allocation words
13926: @cindex memory-allocation words, implementation-defined options
13927:
13928: @table @i
1.29 crook 13929: @item values and meaning of @i{ior}:
13930: @cindex @i{ior} values and meaning
13931: The @i{ior}s returned by the file and memory allocation words are
1.1 anton 13932: intended as throw codes. They typically are in the range
13933: -512@minus{}-2047 of OS errors. The mapping from OS error numbers to
1.29 crook 13934: @i{ior}s is -512@minus{}@i{errno}.
1.1 anton 13935:
13936: @end table
13937:
13938: @c =====================================================================
13939: @node The optional Programming-Tools word set, The optional Search-Order word set, The optional Memory-Allocation word set, ANS conformance
13940: @section The optional Programming-Tools word set
13941: @c =====================================================================
13942: @cindex system documentation, programming-tools words
13943: @cindex programming-tools words, system documentation
13944:
13945: @menu
13946: * programming-idef:: Implementation Defined Options
13947: * programming-ambcond:: Ambiguous Conditions
13948: @end menu
13949:
13950:
13951: @c ---------------------------------------------------------------------
13952: @node programming-idef, programming-ambcond, The optional Programming-Tools word set, The optional Programming-Tools word set
13953: @subsection Implementation Defined Options
13954: @c ---------------------------------------------------------------------
13955: @cindex implementation-defined options, programming-tools words
13956: @cindex programming-tools words, implementation-defined options
13957:
13958: @table @i
13959: @item ending sequence for input following @code{;CODE} and @code{CODE}:
13960: @cindex @code{;CODE} ending sequence
13961: @cindex @code{CODE} ending sequence
13962: @code{END-CODE}
13963:
13964: @item manner of processing input following @code{;CODE} and @code{CODE}:
13965: @cindex @code{;CODE}, processing input
13966: @cindex @code{CODE}, processing input
13967: The @code{ASSEMBLER} vocabulary is pushed on the search order stack, and
13968: the input is processed by the text interpreter, (starting) in interpret
13969: state.
13970:
13971: @item search order capability for @code{EDITOR} and @code{ASSEMBLER}:
13972: @cindex @code{ASSEMBLER}, search order capability
13973: The ANS Forth search order word set.
13974:
13975: @item source and format of display by @code{SEE}:
13976: @cindex @code{SEE}, source and format of output
1.80 anton 13977: The source for @code{see} is the executable code used by the inner
1.1 anton 13978: interpreter. The current @code{see} tries to output Forth source code
1.80 anton 13979: (and on some platforms, assembly code for primitives) as well as
13980: possible.
1.1 anton 13981:
13982: @end table
13983:
13984: @c ---------------------------------------------------------------------
13985: @node programming-ambcond, , programming-idef, The optional Programming-Tools word set
13986: @subsection Ambiguous conditions
13987: @c ---------------------------------------------------------------------
13988: @cindex programming-tools words, ambiguous conditions
13989: @cindex ambiguous conditions, programming-tools words
13990:
13991: @table @i
13992:
1.21 crook 13993: @item deleting the compilation word list (@code{FORGET}):
13994: @cindex @code{FORGET}, deleting the compilation word list
1.1 anton 13995: Not implemented (yet).
13996:
1.29 crook 13997: @item fewer than @i{u}+1 items on the control-flow stack (@code{CS-PICK}, @code{CS-ROLL}):
13998: @cindex @code{CS-PICK}, fewer than @i{u}+1 items on the control flow-stack
13999: @cindex @code{CS-ROLL}, fewer than @i{u}+1 items on the control flow-stack
1.1 anton 14000: @cindex control-flow stack underflow
14001: This typically results in an @code{abort"} with a descriptive error
14002: message (may change into a @code{-22 throw} (Control structure mismatch)
14003: in the future). You may also get a memory access error. If you are
14004: unlucky, this ambiguous condition is not caught.
14005:
1.29 crook 14006: @item @i{name} can't be found (@code{FORGET}):
14007: @cindex @code{FORGET}, @i{name} can't be found
1.1 anton 14008: Not implemented (yet).
14009:
1.29 crook 14010: @item @i{name} not defined via @code{CREATE}:
14011: @cindex @code{;CODE}, @i{name} not defined via @code{CREATE}
1.1 anton 14012: @code{;CODE} behaves like @code{DOES>} in this respect, i.e., it changes
14013: the execution semantics of the last defined word no matter how it was
14014: defined.
14015:
14016: @item @code{POSTPONE} applied to @code{[IF]}:
14017: @cindex @code{POSTPONE} applied to @code{[IF]}
14018: @cindex @code{[IF]} and @code{POSTPONE}
14019: After defining @code{: X POSTPONE [IF] ; IMMEDIATE}. @code{X} is
14020: equivalent to @code{[IF]}.
14021:
14022: @item reaching the end of the input source before matching @code{[ELSE]} or @code{[THEN]}:
14023: @cindex @code{[IF]}, end of the input source before matching @code{[ELSE]} or @code{[THEN]}
14024: Continue in the same state of conditional compilation in the next outer
14025: input source. Currently there is no warning to the user about this.
14026:
14027: @item removing a needed definition (@code{FORGET}):
14028: @cindex @code{FORGET}, removing a needed definition
14029: Not implemented (yet).
14030:
14031: @end table
14032:
14033:
14034: @c =====================================================================
14035: @node The optional Search-Order word set, , The optional Programming-Tools word set, ANS conformance
14036: @section The optional Search-Order word set
14037: @c =====================================================================
14038: @cindex system documentation, search-order words
14039: @cindex search-order words, system documentation
14040:
14041: @menu
14042: * search-idef:: Implementation Defined Options
14043: * search-ambcond:: Ambiguous Conditions
14044: @end menu
14045:
14046:
14047: @c ---------------------------------------------------------------------
14048: @node search-idef, search-ambcond, The optional Search-Order word set, The optional Search-Order word set
14049: @subsection Implementation Defined Options
14050: @c ---------------------------------------------------------------------
14051: @cindex implementation-defined options, search-order words
14052: @cindex search-order words, implementation-defined options
14053:
14054: @table @i
14055: @item maximum number of word lists in search order:
14056: @cindex maximum number of word lists in search order
14057: @cindex search order, maximum depth
14058: @code{s" wordlists" environment? drop .}. Currently 16.
14059:
14060: @item minimum search order:
14061: @cindex minimum search order
14062: @cindex search order, minimum
14063: @code{root root}.
14064:
14065: @end table
14066:
14067: @c ---------------------------------------------------------------------
14068: @node search-ambcond, , search-idef, The optional Search-Order word set
14069: @subsection Ambiguous conditions
14070: @c ---------------------------------------------------------------------
14071: @cindex search-order words, ambiguous conditions
14072: @cindex ambiguous conditions, search-order words
14073:
14074: @table @i
1.21 crook 14075: @item changing the compilation word list (during compilation):
14076: @cindex changing the compilation word list (during compilation)
14077: @cindex compilation word list, change before definition ends
14078: The word is entered into the word list that was the compilation word list
1.1 anton 14079: at the start of the definition. Any changes to the name field (e.g.,
14080: @code{immediate}) or the code field (e.g., when executing @code{DOES>})
1.116 anton 14081: are applied to the latest defined word (as reported by @code{latest} or
14082: @code{latestxt}), if possible, irrespective of the compilation word list.
1.1 anton 14083:
14084: @item search order empty (@code{previous}):
14085: @cindex @code{previous}, search order empty
1.26 crook 14086: @cindex vocstack empty, @code{previous}
1.1 anton 14087: @code{abort" Vocstack empty"}.
14088:
14089: @item too many word lists in search order (@code{also}):
14090: @cindex @code{also}, too many word lists in search order
1.26 crook 14091: @cindex vocstack full, @code{also}
1.1 anton 14092: @code{abort" Vocstack full"}.
14093:
14094: @end table
14095:
14096: @c ***************************************************************
1.65 anton 14097: @node Standard vs Extensions, Model, ANS conformance, Top
14098: @chapter Should I use Gforth extensions?
14099: @cindex Gforth extensions
14100:
14101: As you read through the rest of this manual, you will see documentation
14102: for @i{Standard} words, and documentation for some appealing Gforth
14103: @i{extensions}. You might ask yourself the question: @i{``Should I
14104: restrict myself to the standard, or should I use the extensions?''}
14105:
14106: The answer depends on the goals you have for the program you are working
14107: on:
14108:
14109: @itemize @bullet
14110:
14111: @item Is it just for yourself or do you want to share it with others?
14112:
14113: @item
14114: If you want to share it, do the others all use Gforth?
14115:
14116: @item
14117: If it is just for yourself, do you want to restrict yourself to Gforth?
14118:
14119: @end itemize
14120:
14121: If restricting the program to Gforth is ok, then there is no reason not
14122: to use extensions. It is still a good idea to keep to the standard
14123: where it is easy, in case you want to reuse these parts in another
14124: program that you want to be portable.
14125:
14126: If you want to be able to port the program to other Forth systems, there
14127: are the following points to consider:
14128:
14129: @itemize @bullet
14130:
14131: @item
14132: Most Forth systems that are being maintained support the ANS Forth
14133: standard. So if your program complies with the standard, it will be
14134: portable among many systems.
14135:
14136: @item
14137: A number of the Gforth extensions can be implemented in ANS Forth using
14138: public-domain files provided in the @file{compat/} directory. These are
14139: mentioned in the text in passing. There is no reason not to use these
14140: extensions, your program will still be ANS Forth compliant; just include
14141: the appropriate compat files with your program.
14142:
14143: @item
14144: The tool @file{ans-report.fs} (@pxref{ANS Report}) makes it easy to
14145: analyse your program and determine what non-Standard words it relies
14146: upon. However, it does not check whether you use standard words in a
14147: non-standard way.
14148:
14149: @item
14150: Some techniques are not standardized by ANS Forth, and are hard or
14151: impossible to implement in a standard way, but can be implemented in
14152: most Forth systems easily, and usually in similar ways (e.g., accessing
14153: word headers). Forth has a rich historical precedent for programmers
14154: taking advantage of implementation-dependent features of their tools
14155: (for example, relying on a knowledge of the dictionary
14156: structure). Sometimes these techniques are necessary to extract every
14157: last bit of performance from the hardware, sometimes they are just a
14158: programming shorthand.
14159:
14160: @item
14161: Does using a Gforth extension save more work than the porting this part
14162: to other Forth systems (if any) will cost?
14163:
14164: @item
14165: Is the additional functionality worth the reduction in portability and
14166: the additional porting problems?
14167:
14168: @end itemize
14169:
14170: In order to perform these consideratios, you need to know what's
14171: standard and what's not. This manual generally states if something is
1.81 anton 14172: non-standard, but the authoritative source is the
14173: @uref{http://www.taygeta.com/forth/dpans.html,standard document}.
1.65 anton 14174: Appendix A of the Standard (@var{Rationale}) provides a valuable insight
14175: into the thought processes of the technical committee.
14176:
14177: Note also that portability between Forth systems is not the only
14178: portability issue; there is also the issue of portability between
14179: different platforms (processor/OS combinations).
14180:
14181: @c ***************************************************************
14182: @node Model, Integrating Gforth, Standard vs Extensions, Top
1.1 anton 14183: @chapter Model
14184:
14185: This chapter has yet to be written. It will contain information, on
14186: which internal structures you can rely.
14187:
14188: @c ***************************************************************
14189: @node Integrating Gforth, Emacs and Gforth, Model, Top
14190: @chapter Integrating Gforth into C programs
14191:
14192: This is not yet implemented.
14193:
14194: Several people like to use Forth as scripting language for applications
14195: that are otherwise written in C, C++, or some other language.
14196:
14197: The Forth system ATLAST provides facilities for embedding it into
14198: applications; unfortunately it has several disadvantages: most
14199: importantly, it is not based on ANS Forth, and it is apparently dead
14200: (i.e., not developed further and not supported). The facilities
1.21 crook 14201: provided by Gforth in this area are inspired by ATLAST's facilities, so
1.1 anton 14202: making the switch should not be hard.
14203:
14204: We also tried to design the interface such that it can easily be
14205: implemented by other Forth systems, so that we may one day arrive at a
14206: standardized interface. Such a standard interface would allow you to
14207: replace the Forth system without having to rewrite C code.
14208:
14209: You embed the Gforth interpreter by linking with the library
14210: @code{libgforth.a} (give the compiler the option @code{-lgforth}). All
14211: global symbols in this library that belong to the interface, have the
14212: prefix @code{forth_}. (Global symbols that are used internally have the
14213: prefix @code{gforth_}).
14214:
14215: You can include the declarations of Forth types and the functions and
14216: variables of the interface with @code{#include <forth.h>}.
14217:
14218: Types.
14219:
14220: Variables.
14221:
14222: Data and FP Stack pointer. Area sizes.
14223:
14224: functions.
14225:
14226: forth_init(imagefile)
14227: forth_evaluate(string) exceptions?
14228: forth_goto(address) (or forth_execute(xt)?)
14229: forth_continue() (a corountining mechanism)
14230:
14231: Adding primitives.
14232:
14233: No checking.
14234:
14235: Signals?
14236:
14237: Accessing the Stacks
14238:
1.26 crook 14239: @c ******************************************************************
1.1 anton 14240: @node Emacs and Gforth, Image Files, Integrating Gforth, Top
14241: @chapter Emacs and Gforth
14242: @cindex Emacs and Gforth
14243:
14244: @cindex @file{gforth.el}
14245: @cindex @file{forth.el}
14246: @cindex Rydqvist, Goran
1.107 dvdkhlng 14247: @cindex Kuehling, David
1.1 anton 14248: @cindex comment editing commands
14249: @cindex @code{\}, editing with Emacs
14250: @cindex debug tracer editing commands
14251: @cindex @code{~~}, removal with Emacs
14252: @cindex Forth mode in Emacs
1.107 dvdkhlng 14253:
1.1 anton 14254: Gforth comes with @file{gforth.el}, an improved version of
14255: @file{forth.el} by Goran Rydqvist (included in the TILE package). The
1.26 crook 14256: improvements are:
14257:
14258: @itemize @bullet
14259: @item
1.107 dvdkhlng 14260: A better handling of indentation.
14261: @item
14262: A custom hilighting engine for Forth-code.
1.26 crook 14263: @item
14264: Comment paragraph filling (@kbd{M-q})
14265: @item
14266: Commenting (@kbd{C-x \}) and uncommenting (@kbd{C-u C-x \}) of regions
14267: @item
14268: Removal of debugging tracers (@kbd{C-x ~}, @pxref{Debugging}).
1.41 anton 14269: @item
14270: Support of the @code{info-lookup} feature for looking up the
14271: documentation of a word.
1.107 dvdkhlng 14272: @item
14273: Support for reading and writing blocks files.
1.26 crook 14274: @end itemize
14275:
1.107 dvdkhlng 14276: To get a basic description of these features, enter Forth mode and
14277: type @kbd{C-h m}.
1.1 anton 14278:
14279: @cindex source location of error or debugging output in Emacs
14280: @cindex error output, finding the source location in Emacs
14281: @cindex debugging output, finding the source location in Emacs
14282: In addition, Gforth supports Emacs quite well: The source code locations
14283: given in error messages, debugging output (from @code{~~}) and failed
14284: assertion messages are in the right format for Emacs' compilation mode
14285: (@pxref{Compilation, , Running Compilations under Emacs, emacs, Emacs
14286: Manual}) so the source location corresponding to an error or other
14287: message is only a few keystrokes away (@kbd{C-x `} for the next error,
14288: @kbd{C-c C-c} for the error under the cursor).
14289:
1.107 dvdkhlng 14290: @cindex viewing the documentation of a word in Emacs
14291: @cindex context-sensitive help
14292: Moreover, for words documented in this manual, you can look up the
14293: glossary entry quickly by using @kbd{C-h TAB}
14294: (@code{info-lookup-symbol}, @pxref{Documentation, ,Documentation
14295: Commands, emacs, Emacs Manual}). This feature requires Emacs 20.3 or
14296: later and does not work for words containing @code{:}.
14297:
14298: @menu
14299: * Installing gforth.el:: Making Emacs aware of Forth.
14300: * Emacs Tags:: Viewing the source of a word in Emacs.
14301: * Hilighting:: Making Forth code look prettier.
14302: * Auto-Indentation:: Customizing auto-indentation.
14303: * Blocks Files:: Reading and writing blocks files.
14304: @end menu
14305:
14306: @c ----------------------------------
1.109 anton 14307: @node Installing gforth.el, Emacs Tags, Emacs and Gforth, Emacs and Gforth
1.107 dvdkhlng 14308: @section Installing gforth.el
14309: @cindex @file{.emacs}
14310: @cindex @file{gforth.el}, installation
14311: To make the features from @file{gforth.el} available in Emacs, add
14312: the following lines to your @file{.emacs} file:
14313:
14314: @example
14315: (autoload 'forth-mode "gforth.el")
14316: (setq auto-mode-alist (cons '("\\.fs\\'" . forth-mode)
14317: auto-mode-alist))
14318: (autoload 'forth-block-mode "gforth.el")
14319: (setq auto-mode-alist (cons '("\\.fb\\'" . forth-block-mode)
14320: auto-mode-alist))
14321: (add-hook 'forth-mode-hook (function (lambda ()
14322: ;; customize variables here:
14323: (setq forth-indent-level 4)
14324: (setq forth-minor-indent-level 2)
14325: (setq forth-hilight-level 3)
14326: ;;; ...
14327: )))
14328: @end example
14329:
14330: @c ----------------------------------
14331: @node Emacs Tags, Hilighting, Installing gforth.el, Emacs and Gforth
14332: @section Emacs Tags
1.1 anton 14333: @cindex @file{TAGS} file
14334: @cindex @file{etags.fs}
14335: @cindex viewing the source of a word in Emacs
1.43 anton 14336: @cindex @code{require}, placement in files
14337: @cindex @code{include}, placement in files
1.107 dvdkhlng 14338: If you @code{require} @file{etags.fs}, a new @file{TAGS} file will be
14339: produced (@pxref{Tags, , Tags Tables, emacs, Emacs Manual}) that
1.1 anton 14340: contains the definitions of all words defined afterwards. You can then
1.107 dvdkhlng 14341: find the source for a word using @kbd{M-.}. Note that Emacs can use
1.1 anton 14342: several tags files at the same time (e.g., one for the Gforth sources
14343: and one for your program, @pxref{Select Tags Table,,Selecting a Tags
14344: Table,emacs, Emacs Manual}). The TAGS file for the preloaded words is
14345: @file{$(datadir)/gforth/$(VERSION)/TAGS} (e.g.,
1.43 anton 14346: @file{/usr/local/share/gforth/0.2.0/TAGS}). To get the best behaviour
14347: with @file{etags.fs}, you should avoid putting definitions both before
14348: and after @code{require} etc., otherwise you will see the same file
14349: visited several times by commands like @code{tags-search}.
1.1 anton 14350:
1.107 dvdkhlng 14351: @c ----------------------------------
14352: @node Hilighting, Auto-Indentation, Emacs Tags, Emacs and Gforth
14353: @section Hilighting
14354: @cindex hilighting Forth code in Emacs
14355: @cindex highlighting Forth code in Emacs
14356: @file{gforth.el} comes with a custom source hilighting engine. When
14357: you open a file in @code{forth-mode}, it will be completely parsed,
14358: assigning faces to keywords, comments, strings etc. While you edit
14359: the file, modified regions get parsed and updated on-the-fly.
14360:
14361: Use the variable `forth-hilight-level' to change the level of
14362: decoration from 0 (no hilighting at all) to 3 (the default). Even if
14363: you set the hilighting level to 0, the parser will still work in the
14364: background, collecting information about whether regions of text are
14365: ``compiled'' or ``interpreted''. Those information are required for
14366: auto-indentation to work properly. Set `forth-disable-parser' to
14367: non-nil if your computer is too slow to handle parsing. This will
14368: have an impact on the smartness of the auto-indentation engine,
14369: though.
14370:
14371: Sometimes Forth sources define new features that should be hilighted,
14372: new control structures, defining-words etc. You can use the variable
14373: `forth-custom-words' to make @code{forth-mode} hilight additional
14374: words and constructs. See the docstring of `forth-words' for details
14375: (in Emacs, type @kbd{C-h v forth-words}).
14376:
14377: `forth-custom-words' is meant to be customized in your
14378: @file{.emacs} file. To customize hilighing in a file-specific manner,
14379: set `forth-local-words' in a local-variables section at the end of
14380: your source file (@pxref{Local Variables in Files,, Variables, emacs, Emacs Manual}).
14381:
14382: Example:
14383: @example
14384: 0 [IF]
14385: Local Variables:
14386: forth-local-words:
14387: ((("t:") definition-starter (font-lock-keyword-face . 1)
14388: "[ \t\n]" t name (font-lock-function-name-face . 3))
14389: ((";t") definition-ender (font-lock-keyword-face . 1)))
14390: End:
14391: [THEN]
14392: @end example
14393:
14394: @c ----------------------------------
14395: @node Auto-Indentation, Blocks Files, Hilighting, Emacs and Gforth
14396: @section Auto-Indentation
14397: @cindex auto-indentation of Forth code in Emacs
14398: @cindex indentation of Forth code in Emacs
14399: @code{forth-mode} automatically tries to indent lines in a smart way,
14400: whenever you type @key{TAB} or break a line with @kbd{C-m}.
14401:
14402: Simple customization can be achieved by setting
14403: `forth-indent-level' and `forth-minor-indent-level' in your
14404: @file{.emacs} file. For historical reasons @file{gforth.el} indents
14405: per default by multiples of 4 columns. To use the more traditional
14406: 3-column indentation, add the following lines to your @file{.emacs}:
14407:
14408: @example
14409: (add-hook 'forth-mode-hook (function (lambda ()
14410: ;; customize variables here:
14411: (setq forth-indent-level 3)
14412: (setq forth-minor-indent-level 1)
14413: )))
14414: @end example
14415:
14416: If you want indentation to recognize non-default words, customize it
14417: by setting `forth-custom-indent-words' in your @file{.emacs}. See the
14418: docstring of `forth-indent-words' for details (in Emacs, type @kbd{C-h
14419: v forth-indent-words}).
14420:
14421: To customize indentation in a file-specific manner, set
14422: `forth-local-indent-words' in a local-variables section at the end of
14423: your source file (@pxref{Local Variables in Files, Variables,,emacs,
14424: Emacs Manual}).
14425:
14426: Example:
14427: @example
14428: 0 [IF]
14429: Local Variables:
14430: forth-local-indent-words:
14431: ((("t:") (0 . 2) (0 . 2))
14432: ((";t") (-2 . 0) (0 . -2)))
14433: End:
14434: [THEN]
14435: @end example
14436:
14437: @c ----------------------------------
1.109 anton 14438: @node Blocks Files, , Auto-Indentation, Emacs and Gforth
1.107 dvdkhlng 14439: @section Blocks Files
14440: @cindex blocks files, use with Emacs
14441: @code{forth-mode} Autodetects blocks files by checking whether the
14442: length of the first line exceeds 1023 characters. It then tries to
14443: convert the file into normal text format. When you save the file, it
14444: will be written to disk as normal stream-source file.
14445:
14446: If you want to write blocks files, use @code{forth-blocks-mode}. It
14447: inherits all the features from @code{forth-mode}, plus some additions:
1.41 anton 14448:
1.107 dvdkhlng 14449: @itemize @bullet
14450: @item
14451: Files are written to disk in blocks file format.
14452: @item
14453: Screen numbers are displayed in the mode line (enumerated beginning
14454: with the value of `forth-block-base')
14455: @item
14456: Warnings are displayed when lines exceed 64 characters.
14457: @item
14458: The beginning of the currently edited block is marked with an
14459: overlay-arrow.
14460: @end itemize
1.41 anton 14461:
1.107 dvdkhlng 14462: There are some restrictions you should be aware of. When you open a
14463: blocks file that contains tabulator or newline characters, these
14464: characters will be translated into spaces when the file is written
14465: back to disk. If tabs or newlines are encountered during blocks file
14466: reading, an error is output to the echo area. So have a look at the
14467: `*Messages*' buffer, when Emacs' bell rings during reading.
1.1 anton 14468:
1.107 dvdkhlng 14469: Please consult the docstring of @code{forth-blocks-mode} for more
14470: information by typing @kbd{C-h v forth-blocks-mode}).
1.1 anton 14471:
1.26 crook 14472: @c ******************************************************************
1.1 anton 14473: @node Image Files, Engine, Emacs and Gforth, Top
14474: @chapter Image Files
1.26 crook 14475: @cindex image file
14476: @cindex @file{.fi} files
1.1 anton 14477: @cindex precompiled Forth code
14478: @cindex dictionary in persistent form
14479: @cindex persistent form of dictionary
14480:
14481: An image file is a file containing an image of the Forth dictionary,
14482: i.e., compiled Forth code and data residing in the dictionary. By
14483: convention, we use the extension @code{.fi} for image files.
14484:
14485: @menu
1.18 anton 14486: * Image Licensing Issues:: Distribution terms for images.
14487: * Image File Background:: Why have image files?
1.67 anton 14488: * Non-Relocatable Image Files:: don't always work.
1.18 anton 14489: * Data-Relocatable Image Files:: are better.
1.67 anton 14490: * Fully Relocatable Image Files:: better yet.
1.18 anton 14491: * Stack and Dictionary Sizes:: Setting the default sizes for an image.
1.29 crook 14492: * Running Image Files:: @code{gforth -i @i{file}} or @i{file}.
1.18 anton 14493: * Modifying the Startup Sequence:: and turnkey applications.
1.1 anton 14494: @end menu
14495:
1.18 anton 14496: @node Image Licensing Issues, Image File Background, Image Files, Image Files
14497: @section Image Licensing Issues
14498: @cindex license for images
14499: @cindex image license
14500:
14501: An image created with @code{gforthmi} (@pxref{gforthmi}) or
14502: @code{savesystem} (@pxref{Non-Relocatable Image Files}) includes the
14503: original image; i.e., according to copyright law it is a derived work of
14504: the original image.
14505:
14506: Since Gforth is distributed under the GNU GPL, the newly created image
14507: falls under the GNU GPL, too. In particular, this means that if you
14508: distribute the image, you have to make all of the sources for the image
1.113 anton 14509: available, including those you wrote. For details see @ref{Copying, ,
1.18 anton 14510: GNU General Public License (Section 3)}.
14511:
14512: If you create an image with @code{cross} (@pxref{cross.fs}), the image
14513: contains only code compiled from the sources you gave it; if none of
14514: these sources is under the GPL, the terms discussed above do not apply
14515: to the image. However, if your image needs an engine (a gforth binary)
14516: that is under the GPL, you should make sure that you distribute both in
14517: a way that is at most a @emph{mere aggregation}, if you don't want the
14518: terms of the GPL to apply to the image.
14519:
14520: @node Image File Background, Non-Relocatable Image Files, Image Licensing Issues, Image Files
1.1 anton 14521: @section Image File Background
14522: @cindex image file background
14523:
1.80 anton 14524: Gforth consists not only of primitives (in the engine), but also of
1.1 anton 14525: definitions written in Forth. Since the Forth compiler itself belongs to
14526: those definitions, it is not possible to start the system with the
1.80 anton 14527: engine and the Forth source alone. Therefore we provide the Forth
1.26 crook 14528: code as an image file in nearly executable form. When Gforth starts up,
14529: a C routine loads the image file into memory, optionally relocates the
14530: addresses, then sets up the memory (stacks etc.) according to
14531: information in the image file, and (finally) starts executing Forth
14532: code.
1.1 anton 14533:
14534: The image file variants represent different compromises between the
14535: goals of making it easy to generate image files and making them
14536: portable.
14537:
14538: @cindex relocation at run-time
1.26 crook 14539: Win32Forth 3.4 and Mitch Bradley's @code{cforth} use relocation at
1.1 anton 14540: run-time. This avoids many of the complications discussed below (image
14541: files are data relocatable without further ado), but costs performance
14542: (one addition per memory access).
14543:
14544: @cindex relocation at load-time
1.26 crook 14545: By contrast, the Gforth loader performs relocation at image load time. The
14546: loader also has to replace tokens that represent primitive calls with the
1.1 anton 14547: appropriate code-field addresses (or code addresses in the case of
14548: direct threading).
14549:
14550: There are three kinds of image files, with different degrees of
14551: relocatability: non-relocatable, data-relocatable, and fully relocatable
14552: image files.
14553:
14554: @cindex image file loader
14555: @cindex relocating loader
14556: @cindex loader for image files
14557: These image file variants have several restrictions in common; they are
14558: caused by the design of the image file loader:
14559:
14560: @itemize @bullet
14561: @item
14562: There is only one segment; in particular, this means, that an image file
14563: cannot represent @code{ALLOCATE}d memory chunks (and pointers to
1.26 crook 14564: them). The contents of the stacks are not represented, either.
1.1 anton 14565:
14566: @item
14567: The only kinds of relocation supported are: adding the same offset to
14568: all cells that represent data addresses; and replacing special tokens
14569: with code addresses or with pieces of machine code.
14570:
14571: If any complex computations involving addresses are performed, the
14572: results cannot be represented in the image file. Several applications that
14573: use such computations come to mind:
14574: @itemize @minus
14575: @item
14576: Hashing addresses (or data structures which contain addresses) for table
14577: lookup. If you use Gforth's @code{table}s or @code{wordlist}s for this
14578: purpose, you will have no problem, because the hash tables are
14579: recomputed automatically when the system is started. If you use your own
14580: hash tables, you will have to do something similar.
14581:
14582: @item
14583: There's a cute implementation of doubly-linked lists that uses
14584: @code{XOR}ed addresses. You could represent such lists as singly-linked
14585: in the image file, and restore the doubly-linked representation on
14586: startup.@footnote{In my opinion, though, you should think thrice before
14587: using a doubly-linked list (whatever implementation).}
14588:
14589: @item
14590: The code addresses of run-time routines like @code{docol:} cannot be
14591: represented in the image file (because their tokens would be replaced by
14592: machine code in direct threaded implementations). As a workaround,
14593: compute these addresses at run-time with @code{>code-address} from the
14594: executions tokens of appropriate words (see the definitions of
1.80 anton 14595: @code{docol:} and friends in @file{kernel/getdoers.fs}).
1.1 anton 14596:
14597: @item
14598: On many architectures addresses are represented in machine code in some
14599: shifted or mangled form. You cannot put @code{CODE} words that contain
14600: absolute addresses in this form in a relocatable image file. Workarounds
14601: are representing the address in some relative form (e.g., relative to
14602: the CFA, which is present in some register), or loading the address from
14603: a place where it is stored in a non-mangled form.
14604: @end itemize
14605: @end itemize
14606:
14607: @node Non-Relocatable Image Files, Data-Relocatable Image Files, Image File Background, Image Files
14608: @section Non-Relocatable Image Files
14609: @cindex non-relocatable image files
1.26 crook 14610: @cindex image file, non-relocatable
1.1 anton 14611:
14612: These files are simple memory dumps of the dictionary. They are specific
14613: to the executable (i.e., @file{gforth} file) they were created
14614: with. What's worse, they are specific to the place on which the
14615: dictionary resided when the image was created. Now, there is no
14616: guarantee that the dictionary will reside at the same place the next
14617: time you start Gforth, so there's no guarantee that a non-relocatable
14618: image will work the next time (Gforth will complain instead of crashing,
14619: though).
14620:
14621: You can create a non-relocatable image file with
14622:
1.44 crook 14623:
1.1 anton 14624: doc-savesystem
14625:
1.44 crook 14626:
1.1 anton 14627: @node Data-Relocatable Image Files, Fully Relocatable Image Files, Non-Relocatable Image Files, Image Files
14628: @section Data-Relocatable Image Files
14629: @cindex data-relocatable image files
1.26 crook 14630: @cindex image file, data-relocatable
1.1 anton 14631:
14632: These files contain relocatable data addresses, but fixed code addresses
14633: (instead of tokens). They are specific to the executable (i.e.,
14634: @file{gforth} file) they were created with. For direct threading on some
14635: architectures (e.g., the i386), data-relocatable images do not work. You
14636: get a data-relocatable image, if you use @file{gforthmi} with a
14637: Gforth binary that is not doubly indirect threaded (@pxref{Fully
14638: Relocatable Image Files}).
14639:
14640: @node Fully Relocatable Image Files, Stack and Dictionary Sizes, Data-Relocatable Image Files, Image Files
14641: @section Fully Relocatable Image Files
14642: @cindex fully relocatable image files
1.26 crook 14643: @cindex image file, fully relocatable
1.1 anton 14644:
14645: @cindex @file{kern*.fi}, relocatability
14646: @cindex @file{gforth.fi}, relocatability
14647: These image files have relocatable data addresses, and tokens for code
14648: addresses. They can be used with different binaries (e.g., with and
14649: without debugging) on the same machine, and even across machines with
14650: the same data formats (byte order, cell size, floating point
14651: format). However, they are usually specific to the version of Gforth
14652: they were created with. The files @file{gforth.fi} and @file{kernl*.fi}
14653: are fully relocatable.
14654:
14655: There are two ways to create a fully relocatable image file:
14656:
14657: @menu
1.29 crook 14658: * gforthmi:: The normal way
1.1 anton 14659: * cross.fs:: The hard way
14660: @end menu
14661:
14662: @node gforthmi, cross.fs, Fully Relocatable Image Files, Fully Relocatable Image Files
14663: @subsection @file{gforthmi}
14664: @cindex @file{comp-i.fs}
14665: @cindex @file{gforthmi}
14666:
14667: You will usually use @file{gforthmi}. If you want to create an
1.29 crook 14668: image @i{file} that contains everything you would load by invoking
14669: Gforth with @code{gforth @i{options}}, you simply say:
1.1 anton 14670: @example
1.29 crook 14671: gforthmi @i{file} @i{options}
1.1 anton 14672: @end example
14673:
14674: E.g., if you want to create an image @file{asm.fi} that has the file
14675: @file{asm.fs} loaded in addition to the usual stuff, you could do it
14676: like this:
14677:
14678: @example
14679: gforthmi asm.fi asm.fs
14680: @end example
14681:
1.27 crook 14682: @file{gforthmi} is implemented as a sh script and works like this: It
14683: produces two non-relocatable images for different addresses and then
14684: compares them. Its output reflects this: first you see the output (if
1.62 crook 14685: any) of the two Gforth invocations that produce the non-relocatable image
1.27 crook 14686: files, then you see the output of the comparing program: It displays the
14687: offset used for data addresses and the offset used for code addresses;
1.1 anton 14688: moreover, for each cell that cannot be represented correctly in the
1.44 crook 14689: image files, it displays a line like this:
1.1 anton 14690:
14691: @example
14692: 78DC BFFFFA50 BFFFFA40
14693: @end example
14694:
14695: This means that at offset $78dc from @code{forthstart}, one input image
14696: contains $bffffa50, and the other contains $bffffa40. Since these cells
14697: cannot be represented correctly in the output image, you should examine
14698: these places in the dictionary and verify that these cells are dead
14699: (i.e., not read before they are written).
1.39 anton 14700:
14701: @cindex --application, @code{gforthmi} option
14702: If you insert the option @code{--application} in front of the image file
14703: name, you will get an image that uses the @code{--appl-image} option
14704: instead of the @code{--image-file} option (@pxref{Invoking
14705: Gforth}). When you execute such an image on Unix (by typing the image
14706: name as command), the Gforth engine will pass all options to the image
14707: instead of trying to interpret them as engine options.
1.1 anton 14708:
1.27 crook 14709: If you type @file{gforthmi} with no arguments, it prints some usage
14710: instructions.
14711:
1.1 anton 14712: @cindex @code{savesystem} during @file{gforthmi}
14713: @cindex @code{bye} during @file{gforthmi}
14714: @cindex doubly indirect threaded code
1.44 crook 14715: @cindex environment variables
14716: @cindex @code{GFORTHD} -- environment variable
14717: @cindex @code{GFORTH} -- environment variable
1.1 anton 14718: @cindex @code{gforth-ditc}
1.29 crook 14719: There are a few wrinkles: After processing the passed @i{options}, the
1.1 anton 14720: words @code{savesystem} and @code{bye} must be visible. A special doubly
14721: indirect threaded version of the @file{gforth} executable is used for
1.62 crook 14722: creating the non-relocatable images; you can pass the exact filename of
1.1 anton 14723: this executable through the environment variable @code{GFORTHD}
14724: (default: @file{gforth-ditc}); if you pass a version that is not doubly
14725: indirect threaded, you will not get a fully relocatable image, but a
1.27 crook 14726: data-relocatable image (because there is no code address offset). The
14727: normal @file{gforth} executable is used for creating the relocatable
14728: image; you can pass the exact filename of this executable through the
14729: environment variable @code{GFORTH}.
1.1 anton 14730:
14731: @node cross.fs, , gforthmi, Fully Relocatable Image Files
14732: @subsection @file{cross.fs}
14733: @cindex @file{cross.fs}
14734: @cindex cross-compiler
14735: @cindex metacompiler
1.47 crook 14736: @cindex target compiler
1.1 anton 14737:
14738: You can also use @code{cross}, a batch compiler that accepts a Forth-like
1.47 crook 14739: programming language (@pxref{Cross Compiler}).
1.1 anton 14740:
1.47 crook 14741: @code{cross} allows you to create image files for machines with
1.1 anton 14742: different data sizes and data formats than the one used for generating
14743: the image file. You can also use it to create an application image that
14744: does not contain a Forth compiler. These features are bought with
14745: restrictions and inconveniences in programming. E.g., addresses have to
14746: be stored in memory with special words (@code{A!}, @code{A,}, etc.) in
14747: order to make the code relocatable.
14748:
14749:
14750: @node Stack and Dictionary Sizes, Running Image Files, Fully Relocatable Image Files, Image Files
14751: @section Stack and Dictionary Sizes
14752: @cindex image file, stack and dictionary sizes
14753: @cindex dictionary size default
14754: @cindex stack size default
14755:
14756: If you invoke Gforth with a command line flag for the size
14757: (@pxref{Invoking Gforth}), the size you specify is stored in the
14758: dictionary. If you save the dictionary with @code{savesystem} or create
14759: an image with @file{gforthmi}, this size will become the default
14760: for the resulting image file. E.g., the following will create a
1.21 crook 14761: fully relocatable version of @file{gforth.fi} with a 1MB dictionary:
1.1 anton 14762:
14763: @example
14764: gforthmi gforth.fi -m 1M
14765: @end example
14766:
14767: In other words, if you want to set the default size for the dictionary
14768: and the stacks of an image, just invoke @file{gforthmi} with the
14769: appropriate options when creating the image.
14770:
14771: @cindex stack size, cache-friendly
14772: Note: For cache-friendly behaviour (i.e., good performance), you should
14773: make the sizes of the stacks modulo, say, 2K, somewhat different. E.g.,
14774: the default stack sizes are: data: 16k (mod 2k=0); fp: 15.5k (mod
14775: 2k=1.5k); return: 15k(mod 2k=1k); locals: 14.5k (mod 2k=0.5k).
14776:
14777: @node Running Image Files, Modifying the Startup Sequence, Stack and Dictionary Sizes, Image Files
14778: @section Running Image Files
14779: @cindex running image files
14780: @cindex invoking image files
14781: @cindex image file invocation
14782:
14783: @cindex -i, invoke image file
14784: @cindex --image file, invoke image file
1.29 crook 14785: You can invoke Gforth with an image file @i{image} instead of the
1.1 anton 14786: default @file{gforth.fi} with the @code{-i} flag (@pxref{Invoking Gforth}):
14787: @example
1.29 crook 14788: gforth -i @i{image}
1.1 anton 14789: @end example
14790:
14791: @cindex executable image file
1.26 crook 14792: @cindex image file, executable
1.1 anton 14793: If your operating system supports starting scripts with a line of the
14794: form @code{#! ...}, you just have to type the image file name to start
14795: Gforth with this image file (note that the file extension @code{.fi} is
1.29 crook 14796: just a convention). I.e., to run Gforth with the image file @i{image},
14797: you can just type @i{image} instead of @code{gforth -i @i{image}}.
1.27 crook 14798: This works because every @code{.fi} file starts with a line of this
14799: format:
14800:
14801: @example
14802: #! /usr/local/bin/gforth-0.4.0 -i
14803: @end example
14804:
14805: The file and pathname for the Gforth engine specified on this line is
14806: the specific Gforth executable that it was built against; i.e. the value
14807: of the environment variable @code{GFORTH} at the time that
14808: @file{gforthmi} was executed.
1.1 anton 14809:
1.27 crook 14810: You can make use of the same shell capability to make a Forth source
14811: file into an executable. For example, if you place this text in a file:
1.26 crook 14812:
14813: @example
14814: #! /usr/local/bin/gforth
14815:
14816: ." Hello, world" CR
14817: bye
14818: @end example
14819:
14820: @noindent
1.27 crook 14821: and then make the file executable (chmod +x in Unix), you can run it
1.26 crook 14822: directly from the command line. The sequence @code{#!} is used in two
14823: ways; firstly, it is recognised as a ``magic sequence'' by the operating
1.29 crook 14824: system@footnote{The Unix kernel actually recognises two types of files:
14825: executable files and files of data, where the data is processed by an
14826: interpreter that is specified on the ``interpreter line'' -- the first
14827: line of the file, starting with the sequence #!. There may be a small
14828: limit (e.g., 32) on the number of characters that may be specified on
14829: the interpreter line.} secondly it is treated as a comment character by
14830: Gforth. Because of the second usage, a space is required between
1.80 anton 14831: @code{#!} and the path to the executable (moreover, some Unixes
14832: require the sequence @code{#! /}).
1.27 crook 14833:
14834: The disadvantage of this latter technique, compared with using
1.80 anton 14835: @file{gforthmi}, is that it is slightly slower; the Forth source code is
14836: compiled on-the-fly, each time the program is invoked.
1.26 crook 14837:
1.1 anton 14838: doc-#!
14839:
1.44 crook 14840:
1.1 anton 14841: @node Modifying the Startup Sequence, , Running Image Files, Image Files
14842: @section Modifying the Startup Sequence
14843: @cindex startup sequence for image file
14844: @cindex image file initialization sequence
14845: @cindex initialization sequence of image file
14846:
1.120 anton 14847: You can add your own initialization to the startup sequence of an image
14848: through the deferred word @code{'cold}. @code{'cold} is invoked just
14849: before the image-specific command line processing (i.e., loading files
14850: and evaluating (@code{-e}) strings) starts.
1.1 anton 14851:
14852: A sequence for adding your initialization usually looks like this:
14853:
14854: @example
14855: :noname
14856: Defers 'cold \ do other initialization stuff (e.g., rehashing wordlists)
14857: ... \ your stuff
14858: ; IS 'cold
14859: @end example
14860:
1.157 anton 14861: After @code{'cold}, Gforth processes the image options
14862: (@pxref{Invoking Gforth}), and then it performs @code{bootmessage},
14863: another deferred word. This normally prints Gforth's startup message
14864: and does nothing else.
14865:
1.1 anton 14866: @cindex turnkey image files
1.26 crook 14867: @cindex image file, turnkey applications
1.157 anton 14868: So, if you want to make a turnkey image (i.e., an image for an
14869: application instead of an extended Forth system), you can do this in
14870: two ways:
14871:
14872: @itemize @bullet
14873:
14874: @item
14875: If you want to do your interpretation of the OS command-line
14876: arguments, hook into @code{'cold}. In that case you probably also
14877: want to build the image with @code{gforthmi --application}
14878: (@pxref{gforthmi}) to keep the engine from processing OS command line
14879: options. You can then do your own command-line processing with
14880: @code{next-arg}
14881:
14882: @item
14883: If you want to have the normal Gforth processing of OS command-line
14884: arguments, hook into @code{bootmessage}.
14885:
14886: @end itemize
14887:
14888: In either case, you probably do not want the word that you execute in
14889: these hooks to exit normally, but use @code{bye} or @code{throw}.
14890: Otherwise the Gforth startup process would continue and eventually
14891: present the Forth command line to the user.
1.26 crook 14892:
14893: doc-'cold
1.157 anton 14894: doc-bootmessage
1.44 crook 14895:
1.1 anton 14896: @c ******************************************************************
1.113 anton 14897: @node Engine, Cross Compiler, Image Files, Top
1.1 anton 14898: @chapter Engine
14899: @cindex engine
14900: @cindex virtual machine
14901:
1.26 crook 14902: Reading this chapter is not necessary for programming with Gforth. It
1.1 anton 14903: may be helpful for finding your way in the Gforth sources.
14904:
1.109 anton 14905: The ideas in this section have also been published in the following
14906: papers: Bernd Paysan, @cite{ANS fig/GNU/??? Forth} (in German),
14907: Forth-Tagung '93; M. Anton Ertl,
14908: @cite{@uref{http://www.complang.tuwien.ac.at/papers/ertl93.ps.Z, A
14909: Portable Forth Engine}}, EuroForth '93; M. Anton Ertl,
14910: @cite{@uref{http://www.complang.tuwien.ac.at/papers/ertl02.ps.gz,
14911: Threaded code variations and optimizations (extended version)}},
14912: Forth-Tagung '02.
1.1 anton 14913:
14914: @menu
14915: * Portability::
14916: * Threading::
14917: * Primitives::
14918: * Performance::
14919: @end menu
14920:
14921: @node Portability, Threading, Engine, Engine
14922: @section Portability
14923: @cindex engine portability
14924:
1.26 crook 14925: An important goal of the Gforth Project is availability across a wide
14926: range of personal machines. fig-Forth, and, to a lesser extent, F83,
14927: achieved this goal by manually coding the engine in assembly language
14928: for several then-popular processors. This approach is very
14929: labor-intensive and the results are short-lived due to progress in
14930: computer architecture.
1.1 anton 14931:
14932: @cindex C, using C for the engine
14933: Others have avoided this problem by coding in C, e.g., Mitch Bradley
14934: (cforth), Mikael Patel (TILE) and Dirk Zoller (pfe). This approach is
14935: particularly popular for UNIX-based Forths due to the large variety of
14936: architectures of UNIX machines. Unfortunately an implementation in C
14937: does not mix well with the goals of efficiency and with using
14938: traditional techniques: Indirect or direct threading cannot be expressed
14939: in C, and switch threading, the fastest technique available in C, is
14940: significantly slower. Another problem with C is that it is very
14941: cumbersome to express double integer arithmetic.
14942:
14943: @cindex GNU C for the engine
14944: @cindex long long
14945: Fortunately, there is a portable language that does not have these
14946: limitations: GNU C, the version of C processed by the GNU C compiler
14947: (@pxref{C Extensions, , Extensions to the C Language Family, gcc.info,
14948: GNU C Manual}). Its labels as values feature (@pxref{Labels as Values, ,
14949: Labels as Values, gcc.info, GNU C Manual}) makes direct and indirect
14950: threading possible, its @code{long long} type (@pxref{Long Long, ,
14951: Double-Word Integers, gcc.info, GNU C Manual}) corresponds to Forth's
1.109 anton 14952: double numbers on many systems. GNU C is freely available on all
1.1 anton 14953: important (and many unimportant) UNIX machines, VMS, 80386s running
14954: MS-DOS, the Amiga, and the Atari ST, so a Forth written in GNU C can run
14955: on all these machines.
14956:
14957: Writing in a portable language has the reputation of producing code that
14958: is slower than assembly. For our Forth engine we repeatedly looked at
14959: the code produced by the compiler and eliminated most compiler-induced
14960: inefficiencies by appropriate changes in the source code.
14961:
14962: @cindex explicit register declarations
14963: @cindex --enable-force-reg, configuration flag
14964: @cindex -DFORCE_REG
14965: However, register allocation cannot be portably influenced by the
14966: programmer, leading to some inefficiencies on register-starved
14967: machines. We use explicit register declarations (@pxref{Explicit Reg
14968: Vars, , Variables in Specified Registers, gcc.info, GNU C Manual}) to
14969: improve the speed on some machines. They are turned on by using the
14970: configuration flag @code{--enable-force-reg} (@code{gcc} switch
14971: @code{-DFORCE_REG}). Unfortunately, this feature not only depends on the
14972: machine, but also on the compiler version: On some machines some
14973: compiler versions produce incorrect code when certain explicit register
14974: declarations are used. So by default @code{-DFORCE_REG} is not used.
14975:
14976: @node Threading, Primitives, Portability, Engine
14977: @section Threading
14978: @cindex inner interpreter implementation
14979: @cindex threaded code implementation
14980:
14981: @cindex labels as values
14982: GNU C's labels as values extension (available since @code{gcc-2.0},
14983: @pxref{Labels as Values, , Labels as Values, gcc.info, GNU C Manual})
1.29 crook 14984: makes it possible to take the address of @i{label} by writing
14985: @code{&&@i{label}}. This address can then be used in a statement like
14986: @code{goto *@i{address}}. I.e., @code{goto *&&x} is the same as
1.1 anton 14987: @code{goto x}.
14988:
1.26 crook 14989: @cindex @code{NEXT}, indirect threaded
1.1 anton 14990: @cindex indirect threaded inner interpreter
14991: @cindex inner interpreter, indirect threaded
1.26 crook 14992: With this feature an indirect threaded @code{NEXT} looks like:
1.1 anton 14993: @example
14994: cfa = *ip++;
14995: ca = *cfa;
14996: goto *ca;
14997: @end example
14998: @cindex instruction pointer
14999: For those unfamiliar with the names: @code{ip} is the Forth instruction
15000: pointer; the @code{cfa} (code-field address) corresponds to ANS Forths
15001: execution token and points to the code field of the next word to be
15002: executed; The @code{ca} (code address) fetched from there points to some
15003: executable code, e.g., a primitive or the colon definition handler
15004: @code{docol}.
15005:
1.26 crook 15006: @cindex @code{NEXT}, direct threaded
1.1 anton 15007: @cindex direct threaded inner interpreter
15008: @cindex inner interpreter, direct threaded
15009: Direct threading is even simpler:
15010: @example
15011: ca = *ip++;
15012: goto *ca;
15013: @end example
15014:
15015: Of course we have packaged the whole thing neatly in macros called
1.26 crook 15016: @code{NEXT} and @code{NEXT1} (the part of @code{NEXT} after fetching the cfa).
1.1 anton 15017:
15018: @menu
15019: * Scheduling::
15020: * Direct or Indirect Threaded?::
1.109 anton 15021: * Dynamic Superinstructions::
1.1 anton 15022: * DOES>::
15023: @end menu
15024:
15025: @node Scheduling, Direct or Indirect Threaded?, Threading, Threading
15026: @subsection Scheduling
15027: @cindex inner interpreter optimization
15028:
15029: There is a little complication: Pipelined and superscalar processors,
15030: i.e., RISC and some modern CISC machines can process independent
15031: instructions while waiting for the results of an instruction. The
15032: compiler usually reorders (schedules) the instructions in a way that
15033: achieves good usage of these delay slots. However, on our first tries
15034: the compiler did not do well on scheduling primitives. E.g., for
15035: @code{+} implemented as
15036: @example
15037: n=sp[0]+sp[1];
15038: sp++;
15039: sp[0]=n;
15040: NEXT;
15041: @end example
1.81 anton 15042: the @code{NEXT} comes strictly after the other code, i.e., there is
15043: nearly no scheduling. After a little thought the problem becomes clear:
15044: The compiler cannot know that @code{sp} and @code{ip} point to different
1.21 crook 15045: addresses (and the version of @code{gcc} we used would not know it even
15046: if it was possible), so it could not move the load of the cfa above the
15047: store to the TOS. Indeed the pointers could be the same, if code on or
15048: very near the top of stack were executed. In the interest of speed we
15049: chose to forbid this probably unused ``feature'' and helped the compiler
1.81 anton 15050: in scheduling: @code{NEXT} is divided into several parts:
15051: @code{NEXT_P0}, @code{NEXT_P1} and @code{NEXT_P2}). @code{+} now looks
15052: like:
1.1 anton 15053: @example
1.81 anton 15054: NEXT_P0;
1.1 anton 15055: n=sp[0]+sp[1];
15056: sp++;
15057: NEXT_P1;
15058: sp[0]=n;
15059: NEXT_P2;
15060: @end example
15061:
1.81 anton 15062: There are various schemes that distribute the different operations of
15063: NEXT between these parts in several ways; in general, different schemes
15064: perform best on different processors. We use a scheme for most
15065: architectures that performs well for most processors of this
1.109 anton 15066: architecture; in the future we may switch to benchmarking and chosing
1.81 anton 15067: the scheme on installation time.
15068:
1.1 anton 15069:
1.109 anton 15070: @node Direct or Indirect Threaded?, Dynamic Superinstructions, Scheduling, Threading
1.1 anton 15071: @subsection Direct or Indirect Threaded?
15072: @cindex threading, direct or indirect?
15073:
1.109 anton 15074: Threaded forth code consists of references to primitives (simple machine
15075: code routines like @code{+}) and to non-primitives (e.g., colon
15076: definitions, variables, constants); for a specific class of
15077: non-primitives (e.g., variables) there is one code routine (e.g.,
15078: @code{dovar}), but each variable needs a separate reference to its data.
15079:
15080: Traditionally Forth has been implemented as indirect threaded code,
15081: because this allows to use only one cell to reference a non-primitive
15082: (basically you point to the data, and find the code address there).
15083:
15084: @cindex primitive-centric threaded code
15085: However, threaded code in Gforth (since 0.6.0) uses two cells for
15086: non-primitives, one for the code address, and one for the data address;
15087: the data pointer is an immediate argument for the virtual machine
15088: instruction represented by the code address. We call this
15089: @emph{primitive-centric} threaded code, because all code addresses point
15090: to simple primitives. E.g., for a variable, the code address is for
15091: @code{lit} (also used for integer literals like @code{99}).
15092:
15093: Primitive-centric threaded code allows us to use (faster) direct
15094: threading as dispatch method, completely portably (direct threaded code
15095: in Gforth before 0.6.0 required architecture-specific code). It also
15096: eliminates the performance problems related to I-cache consistency that
15097: 386 implementations have with direct threaded code, and allows
15098: additional optimizations.
15099:
15100: @cindex hybrid direct/indirect threaded code
15101: There is a catch, however: the @var{xt} parameter of @code{execute} can
15102: occupy only one cell, so how do we pass non-primitives with their code
15103: @emph{and} data addresses to them? Our answer is to use indirect
15104: threaded dispatch for @code{execute} and other words that use a
15105: single-cell xt. So, normal threaded code in colon definitions uses
15106: direct threading, and @code{execute} and similar words, which dispatch
15107: to xts on the data stack, use indirect threaded code. We call this
15108: @emph{hybrid direct/indirect} threaded code.
15109:
15110: @cindex engines, gforth vs. gforth-fast vs. gforth-itc
15111: @cindex gforth engine
15112: @cindex gforth-fast engine
15113: The engines @command{gforth} and @command{gforth-fast} use hybrid
15114: direct/indirect threaded code. This means that with these engines you
15115: cannot use @code{,} to compile an xt. Instead, you have to use
15116: @code{compile,}.
15117:
15118: @cindex gforth-itc engine
1.115 anton 15119: If you want to compile xts with @code{,}, use @command{gforth-itc}.
15120: This engine uses plain old indirect threaded code. It still compiles in
15121: a primitive-centric style, so you cannot use @code{compile,} instead of
1.109 anton 15122: @code{,} (e.g., for producing tables of xts with @code{] word1 word2
1.115 anton 15123: ... [}). If you want to do that, you have to use @command{gforth-itc}
1.109 anton 15124: and execute @code{' , is compile,}. Your program can check if it is
15125: running on a hybrid direct/indirect threaded engine or a pure indirect
15126: threaded engine with @code{threading-method} (@pxref{Threading Words}).
15127:
15128:
15129: @node Dynamic Superinstructions, DOES>, Direct or Indirect Threaded?, Threading
15130: @subsection Dynamic Superinstructions
15131: @cindex Dynamic superinstructions with replication
15132: @cindex Superinstructions
15133: @cindex Replication
15134:
15135: The engines @command{gforth} and @command{gforth-fast} use another
15136: optimization: Dynamic superinstructions with replication. As an
15137: example, consider the following colon definition:
15138:
15139: @example
15140: : squared ( n1 -- n2 )
15141: dup * ;
15142: @end example
15143:
15144: Gforth compiles this into the threaded code sequence
15145:
15146: @example
15147: dup
15148: *
15149: ;s
15150: @end example
15151:
15152: In normal direct threaded code there is a code address occupying one
15153: cell for each of these primitives. Each code address points to a
15154: machine code routine, and the interpreter jumps to this machine code in
15155: order to execute the primitive. The routines for these three
15156: primitives are (in @command{gforth-fast} on the 386):
15157:
15158: @example
15159: Code dup
15160: ( $804B950 ) add esi , # -4 \ $83 $C6 $FC
15161: ( $804B953 ) add ebx , # 4 \ $83 $C3 $4
15162: ( $804B956 ) mov dword ptr 4 [esi] , ecx \ $89 $4E $4
15163: ( $804B959 ) jmp dword ptr FC [ebx] \ $FF $63 $FC
15164: end-code
15165: Code *
15166: ( $804ACC4 ) mov eax , dword ptr 4 [esi] \ $8B $46 $4
15167: ( $804ACC7 ) add esi , # 4 \ $83 $C6 $4
15168: ( $804ACCA ) add ebx , # 4 \ $83 $C3 $4
15169: ( $804ACCD ) imul ecx , eax \ $F $AF $C8
15170: ( $804ACD0 ) jmp dword ptr FC [ebx] \ $FF $63 $FC
15171: end-code
15172: Code ;s
15173: ( $804A693 ) mov eax , dword ptr [edi] \ $8B $7
15174: ( $804A695 ) add edi , # 4 \ $83 $C7 $4
15175: ( $804A698 ) lea ebx , dword ptr 4 [eax] \ $8D $58 $4
15176: ( $804A69B ) jmp dword ptr FC [ebx] \ $FF $63 $FC
15177: end-code
15178: @end example
15179:
15180: With dynamic superinstructions and replication the compiler does not
15181: just lay down the threaded code, but also copies the machine code
15182: fragments, usually without the jump at the end.
15183:
15184: @example
15185: ( $4057D27D ) add esi , # -4 \ $83 $C6 $FC
15186: ( $4057D280 ) add ebx , # 4 \ $83 $C3 $4
15187: ( $4057D283 ) mov dword ptr 4 [esi] , ecx \ $89 $4E $4
15188: ( $4057D286 ) mov eax , dword ptr 4 [esi] \ $8B $46 $4
15189: ( $4057D289 ) add esi , # 4 \ $83 $C6 $4
15190: ( $4057D28C ) add ebx , # 4 \ $83 $C3 $4
15191: ( $4057D28F ) imul ecx , eax \ $F $AF $C8
15192: ( $4057D292 ) mov eax , dword ptr [edi] \ $8B $7
15193: ( $4057D294 ) add edi , # 4 \ $83 $C7 $4
15194: ( $4057D297 ) lea ebx , dword ptr 4 [eax] \ $8D $58 $4
15195: ( $4057D29A ) jmp dword ptr FC [ebx] \ $FF $63 $FC
15196: @end example
15197:
15198: Only when a threaded-code control-flow change happens (e.g., in
15199: @code{;s}), the jump is appended. This optimization eliminates many of
15200: these jumps and makes the rest much more predictable. The speedup
15201: depends on the processor and the application; on the Athlon and Pentium
15202: III this optimization typically produces a speedup by a factor of 2.
15203:
15204: The code addresses in the direct-threaded code are set to point to the
15205: appropriate points in the copied machine code, in this example like
15206: this:
1.1 anton 15207:
1.109 anton 15208: @example
15209: primitive code address
15210: dup $4057D27D
15211: * $4057D286
15212: ;s $4057D292
15213: @end example
15214:
15215: Thus there can be threaded-code jumps to any place in this piece of
15216: code. This also simplifies decompilation quite a bit.
15217:
15218: @cindex --no-dynamic command-line option
15219: @cindex --no-super command-line option
15220: You can disable this optimization with @option{--no-dynamic}. You can
15221: use the copying without eliminating the jumps (i.e., dynamic
15222: replication, but without superinstructions) with @option{--no-super};
15223: this gives the branch prediction benefit alone; the effect on
1.110 anton 15224: performance depends on the CPU; on the Athlon and Pentium III the
15225: speedup is a little less than for dynamic superinstructions with
15226: replication.
15227:
15228: @cindex patching threaded code
15229: One use of these options is if you want to patch the threaded code.
15230: With superinstructions, many of the dispatch jumps are eliminated, so
15231: patching often has no effect. These options preserve all the dispatch
15232: jumps.
1.109 anton 15233:
15234: @cindex --dynamic command-line option
1.110 anton 15235: On some machines dynamic superinstructions are disabled by default,
15236: because it is unsafe on these machines. However, if you feel
15237: adventurous, you can enable it with @option{--dynamic}.
1.109 anton 15238:
15239: @node DOES>, , Dynamic Superinstructions, Threading
1.1 anton 15240: @subsection DOES>
15241: @cindex @code{DOES>} implementation
15242:
1.26 crook 15243: @cindex @code{dodoes} routine
15244: @cindex @code{DOES>}-code
1.1 anton 15245: One of the most complex parts of a Forth engine is @code{dodoes}, i.e.,
15246: the chunk of code executed by every word defined by a
1.109 anton 15247: @code{CREATE}...@code{DOES>} pair; actually with primitive-centric code,
15248: this is only needed if the xt of the word is @code{execute}d. The main
15249: problem here is: How to find the Forth code to be executed, i.e. the
15250: code after the @code{DOES>} (the @code{DOES>}-code)? There are two
15251: solutions:
1.1 anton 15252:
1.21 crook 15253: In fig-Forth the code field points directly to the @code{dodoes} and the
1.109 anton 15254: @code{DOES>}-code address is stored in the cell after the code address
15255: (i.e. at @code{@i{CFA} cell+}). It may seem that this solution is
15256: illegal in the Forth-79 and all later standards, because in fig-Forth
15257: this address lies in the body (which is illegal in these
15258: standards). However, by making the code field larger for all words this
15259: solution becomes legal again. We use this approach. Leaving a cell
15260: unused in most words is a bit wasteful, but on the machines we are
15261: targeting this is hardly a problem.
15262:
1.1 anton 15263:
15264: @node Primitives, Performance, Threading, Engine
15265: @section Primitives
15266: @cindex primitives, implementation
15267: @cindex virtual machine instructions, implementation
15268:
15269: @menu
15270: * Automatic Generation::
15271: * TOS Optimization::
15272: * Produced code::
15273: @end menu
15274:
15275: @node Automatic Generation, TOS Optimization, Primitives, Primitives
15276: @subsection Automatic Generation
15277: @cindex primitives, automatic generation
15278:
15279: @cindex @file{prims2x.fs}
1.109 anton 15280:
1.1 anton 15281: Since the primitives are implemented in a portable language, there is no
15282: longer any need to minimize the number of primitives. On the contrary,
15283: having many primitives has an advantage: speed. In order to reduce the
15284: number of errors in primitives and to make programming them easier, we
1.109 anton 15285: provide a tool, the primitive generator (@file{prims2x.fs} aka Vmgen,
15286: @pxref{Top, Vmgen, Introduction, vmgen, Vmgen}), that automatically
15287: generates most (and sometimes all) of the C code for a primitive from
15288: the stack effect notation. The source for a primitive has the following
15289: form:
1.1 anton 15290:
15291: @cindex primitive source format
15292: @format
1.58 anton 15293: @i{Forth-name} ( @i{stack-effect} ) @i{category} [@i{pronounc.}]
1.29 crook 15294: [@code{""}@i{glossary entry}@code{""}]
15295: @i{C code}
1.1 anton 15296: [@code{:}
1.29 crook 15297: @i{Forth code}]
1.1 anton 15298: @end format
15299:
15300: The items in brackets are optional. The category and glossary fields
15301: are there for generating the documentation, the Forth code is there
15302: for manual implementations on machines without GNU C. E.g., the source
15303: for the primitive @code{+} is:
15304: @example
1.58 anton 15305: + ( n1 n2 -- n ) core plus
1.1 anton 15306: n = n1+n2;
15307: @end example
15308:
15309: This looks like a specification, but in fact @code{n = n1+n2} is C
15310: code. Our primitive generation tool extracts a lot of information from
15311: the stack effect notations@footnote{We use a one-stack notation, even
15312: though we have separate data and floating-point stacks; The separate
15313: notation can be generated easily from the unified notation.}: The number
15314: of items popped from and pushed on the stack, their type, and by what
15315: name they are referred to in the C code. It then generates a C code
15316: prelude and postlude for each primitive. The final C code for @code{+}
15317: looks like this:
15318:
15319: @example
1.46 pazsan 15320: I_plus: /* + ( n1 n2 -- n ) */ /* label, stack effect */
1.1 anton 15321: /* */ /* documentation */
1.81 anton 15322: NAME("+") /* debugging output (with -DDEBUG) */
1.1 anton 15323: @{
15324: DEF_CA /* definition of variable ca (indirect threading) */
15325: Cell n1; /* definitions of variables */
15326: Cell n2;
15327: Cell n;
1.81 anton 15328: NEXT_P0; /* NEXT part 0 */
1.1 anton 15329: n1 = (Cell) sp[1]; /* input */
15330: n2 = (Cell) TOS;
15331: sp += 1; /* stack adjustment */
15332: @{
15333: n = n1+n2; /* C code taken from the source */
15334: @}
15335: NEXT_P1; /* NEXT part 1 */
15336: TOS = (Cell)n; /* output */
15337: NEXT_P2; /* NEXT part 2 */
15338: @}
15339: @end example
15340:
15341: This looks long and inefficient, but the GNU C compiler optimizes quite
15342: well and produces optimal code for @code{+} on, e.g., the R3000 and the
15343: HP RISC machines: Defining the @code{n}s does not produce any code, and
15344: using them as intermediate storage also adds no cost.
15345:
1.26 crook 15346: There are also other optimizations that are not illustrated by this
15347: example: assignments between simple variables are usually for free (copy
1.1 anton 15348: propagation). If one of the stack items is not used by the primitive
15349: (e.g. in @code{drop}), the compiler eliminates the load from the stack
15350: (dead code elimination). On the other hand, there are some things that
15351: the compiler does not do, therefore they are performed by
15352: @file{prims2x.fs}: The compiler does not optimize code away that stores
15353: a stack item to the place where it just came from (e.g., @code{over}).
15354:
15355: While programming a primitive is usually easy, there are a few cases
15356: where the programmer has to take the actions of the generator into
15357: account, most notably @code{?dup}, but also words that do not (always)
1.26 crook 15358: fall through to @code{NEXT}.
1.109 anton 15359:
15360: For more information
1.1 anton 15361:
15362: @node TOS Optimization, Produced code, Automatic Generation, Primitives
15363: @subsection TOS Optimization
15364: @cindex TOS optimization for primitives
15365: @cindex primitives, keeping the TOS in a register
15366:
15367: An important optimization for stack machine emulators, e.g., Forth
15368: engines, is keeping one or more of the top stack items in
1.29 crook 15369: registers. If a word has the stack effect @i{in1}...@i{inx} @code{--}
15370: @i{out1}...@i{outy}, keeping the top @i{n} items in registers
1.1 anton 15371: @itemize @bullet
15372: @item
1.29 crook 15373: is better than keeping @i{n-1} items, if @i{x>=n} and @i{y>=n},
1.1 anton 15374: due to fewer loads from and stores to the stack.
1.29 crook 15375: @item is slower than keeping @i{n-1} items, if @i{x<>y} and @i{x<n} and
15376: @i{y<n}, due to additional moves between registers.
1.1 anton 15377: @end itemize
15378:
15379: @cindex -DUSE_TOS
15380: @cindex -DUSE_NO_TOS
15381: In particular, keeping one item in a register is never a disadvantage,
15382: if there are enough registers. Keeping two items in registers is a
15383: disadvantage for frequent words like @code{?branch}, constants,
15384: variables, literals and @code{i}. Therefore our generator only produces
15385: code that keeps zero or one items in registers. The generated C code
15386: covers both cases; the selection between these alternatives is made at
15387: C-compile time using the switch @code{-DUSE_TOS}. @code{TOS} in the C
15388: code for @code{+} is just a simple variable name in the one-item case,
15389: otherwise it is a macro that expands into @code{sp[0]}. Note that the
15390: GNU C compiler tries to keep simple variables like @code{TOS} in
15391: registers, and it usually succeeds, if there are enough registers.
15392:
15393: @cindex -DUSE_FTOS
15394: @cindex -DUSE_NO_FTOS
15395: The primitive generator performs the TOS optimization for the
15396: floating-point stack, too (@code{-DUSE_FTOS}). For floating-point
15397: operations the benefit of this optimization is even larger:
15398: floating-point operations take quite long on most processors, but can be
15399: performed in parallel with other operations as long as their results are
15400: not used. If the FP-TOS is kept in a register, this works. If
15401: it is kept on the stack, i.e., in memory, the store into memory has to
15402: wait for the result of the floating-point operation, lengthening the
15403: execution time of the primitive considerably.
15404:
15405: The TOS optimization makes the automatic generation of primitives a
15406: bit more complicated. Just replacing all occurrences of @code{sp[0]} by
15407: @code{TOS} is not sufficient. There are some special cases to
15408: consider:
15409: @itemize @bullet
15410: @item In the case of @code{dup ( w -- w w )} the generator must not
15411: eliminate the store to the original location of the item on the stack,
15412: if the TOS optimization is turned on.
15413: @item Primitives with stack effects of the form @code{--}
1.29 crook 15414: @i{out1}...@i{outy} must store the TOS to the stack at the start.
15415: Likewise, primitives with the stack effect @i{in1}...@i{inx} @code{--}
1.1 anton 15416: must load the TOS from the stack at the end. But for the null stack
15417: effect @code{--} no stores or loads should be generated.
15418: @end itemize
15419:
15420: @node Produced code, , TOS Optimization, Primitives
15421: @subsection Produced code
15422: @cindex primitives, assembly code listing
15423:
15424: @cindex @file{engine.s}
15425: To see what assembly code is produced for the primitives on your machine
15426: with your compiler and your flag settings, type @code{make engine.s} and
1.81 anton 15427: look at the resulting file @file{engine.s}. Alternatively, you can also
15428: disassemble the code of primitives with @code{see} on some architectures.
1.1 anton 15429:
15430: @node Performance, , Primitives, Engine
15431: @section Performance
15432: @cindex performance of some Forth interpreters
15433: @cindex engine performance
15434: @cindex benchmarking Forth systems
15435: @cindex Gforth performance
15436:
15437: On RISCs the Gforth engine is very close to optimal; i.e., it is usually
1.112 anton 15438: impossible to write a significantly faster threaded-code engine.
1.1 anton 15439:
15440: On register-starved machines like the 386 architecture processors
15441: improvements are possible, because @code{gcc} does not utilize the
15442: registers as well as a human, even with explicit register declarations;
15443: e.g., Bernd Beuster wrote a Forth system fragment in assembly language
15444: and hand-tuned it for the 486; this system is 1.19 times faster on the
15445: Sieve benchmark on a 486DX2/66 than Gforth compiled with
1.40 anton 15446: @code{gcc-2.6.3} with @code{-DFORCE_REG}. The situation has improved
15447: with gcc-2.95 and gforth-0.4.9; now the most important virtual machine
15448: registers fit in real registers (and we can even afford to use the TOS
15449: optimization), resulting in a speedup of 1.14 on the sieve over the
1.112 anton 15450: earlier results. And dynamic superinstructions provide another speedup
15451: (but only around a factor 1.2 on the 486).
1.1 anton 15452:
15453: @cindex Win32Forth performance
15454: @cindex NT Forth performance
15455: @cindex eforth performance
15456: @cindex ThisForth performance
15457: @cindex PFE performance
15458: @cindex TILE performance
1.81 anton 15459: The potential advantage of assembly language implementations is not
1.112 anton 15460: necessarily realized in complete Forth systems: We compared Gforth-0.5.9
1.81 anton 15461: (direct threaded, compiled with @code{gcc-2.95.1} and
15462: @code{-DFORCE_REG}) with Win32Forth 1.2093 (newer versions are
15463: reportedly much faster), LMI's NT Forth (Beta, May 1994) and Eforth
15464: (with and without peephole (aka pinhole) optimization of the threaded
15465: code); all these systems were written in assembly language. We also
15466: compared Gforth with three systems written in C: PFE-0.9.14 (compiled
15467: with @code{gcc-2.6.3} with the default configuration for Linux:
15468: @code{-O2 -fomit-frame-pointer -DUSE_REGS -DUNROLL_NEXT}), ThisForth
15469: Beta (compiled with @code{gcc-2.6.3 -O3 -fomit-frame-pointer}; ThisForth
15470: employs peephole optimization of the threaded code) and TILE (compiled
15471: with @code{make opt}). We benchmarked Gforth, PFE, ThisForth and TILE on
15472: a 486DX2/66 under Linux. Kenneth O'Heskin kindly provided the results
15473: for Win32Forth and NT Forth on a 486DX2/66 with similar memory
15474: performance under Windows NT. Marcel Hendrix ported Eforth to Linux,
15475: then extended it to run the benchmarks, added the peephole optimizer,
15476: ran the benchmarks and reported the results.
1.40 anton 15477:
1.1 anton 15478: We used four small benchmarks: the ubiquitous Sieve; bubble-sorting and
15479: matrix multiplication come from the Stanford integer benchmarks and have
15480: been translated into Forth by Martin Fraeman; we used the versions
15481: included in the TILE Forth package, but with bigger data set sizes; and
15482: a recursive Fibonacci number computation for benchmarking calling
15483: performance. The following table shows the time taken for the benchmarks
15484: scaled by the time taken by Gforth (in other words, it shows the speedup
15485: factor that Gforth achieved over the other systems).
15486:
15487: @example
1.112 anton 15488: relative Win32- NT eforth This-
15489: time Gforth Forth Forth eforth +opt PFE Forth TILE
15490: sieve 1.00 2.16 1.78 2.16 1.32 2.46 4.96 13.37
15491: bubble 1.00 1.93 2.07 2.18 1.29 2.21 5.70
15492: matmul 1.00 1.92 1.76 1.90 0.96 2.06 5.32
15493: fib 1.00 2.32 2.03 1.86 1.31 2.64 4.55 6.54
1.1 anton 15494: @end example
15495:
1.26 crook 15496: You may be quite surprised by the good performance of Gforth when
15497: compared with systems written in assembly language. One important reason
15498: for the disappointing performance of these other systems is probably
15499: that they are not written optimally for the 486 (e.g., they use the
15500: @code{lods} instruction). In addition, Win32Forth uses a comfortable,
15501: but costly method for relocating the Forth image: like @code{cforth}, it
15502: computes the actual addresses at run time, resulting in two address
15503: computations per @code{NEXT} (@pxref{Image File Background}).
15504:
1.1 anton 15505: The speedup of Gforth over PFE, ThisForth and TILE can be easily
15506: explained with the self-imposed restriction of the latter systems to
15507: standard C, which makes efficient threading impossible (however, the
1.4 anton 15508: measured implementation of PFE uses a GNU C extension: @pxref{Global Reg
1.1 anton 15509: Vars, , Defining Global Register Variables, gcc.info, GNU C Manual}).
15510: Moreover, current C compilers have a hard time optimizing other aspects
15511: of the ThisForth and the TILE source.
15512:
1.26 crook 15513: The performance of Gforth on 386 architecture processors varies widely
15514: with the version of @code{gcc} used. E.g., @code{gcc-2.5.8} failed to
15515: allocate any of the virtual machine registers into real machine
15516: registers by itself and would not work correctly with explicit register
1.112 anton 15517: declarations, giving a significantly slower engine (on a 486DX2/66
15518: running the Sieve) than the one measured above.
1.1 anton 15519:
1.26 crook 15520: Note that there have been several releases of Win32Forth since the
15521: release presented here, so the results presented above may have little
1.40 anton 15522: predictive value for the performance of Win32Forth today (results for
15523: the current release on an i486DX2/66 are welcome).
1.1 anton 15524:
15525: @cindex @file{Benchres}
1.66 anton 15526: In
15527: @cite{@uref{http://www.complang.tuwien.ac.at/papers/ertl&maierhofer95.ps.gz,
15528: Translating Forth to Efficient C}} by M. Anton Ertl and Martin
1.1 anton 15529: Maierhofer (presented at EuroForth '95), an indirect threaded version of
1.66 anton 15530: Gforth is compared with Win32Forth, NT Forth, PFE, ThisForth, and
15531: several native code systems; that version of Gforth is slower on a 486
1.112 anton 15532: than the version used here. You can find a newer version of these
15533: measurements at
1.47 crook 15534: @uref{http://www.complang.tuwien.ac.at/forth/performance.html}. You can
1.1 anton 15535: find numbers for Gforth on various machines in @file{Benchres}.
15536:
1.26 crook 15537: @c ******************************************************************
1.113 anton 15538: @c @node Binding to System Library, Cross Compiler, Engine, Top
15539: @c @chapter Binding to System Library
1.13 pazsan 15540:
1.113 anton 15541: @c ****************************************************************
15542: @node Cross Compiler, Bugs, Engine, Top
1.14 pazsan 15543: @chapter Cross Compiler
1.47 crook 15544: @cindex @file{cross.fs}
15545: @cindex cross-compiler
15546: @cindex metacompiler
15547: @cindex target compiler
1.13 pazsan 15548:
1.46 pazsan 15549: The cross compiler is used to bootstrap a Forth kernel. Since Gforth is
15550: mostly written in Forth, including crucial parts like the outer
15551: interpreter and compiler, it needs compiled Forth code to get
15552: started. The cross compiler allows to create new images for other
15553: architectures, even running under another Forth system.
1.13 pazsan 15554:
15555: @menu
1.67 anton 15556: * Using the Cross Compiler::
15557: * How the Cross Compiler Works::
1.13 pazsan 15558: @end menu
15559:
1.21 crook 15560: @node Using the Cross Compiler, How the Cross Compiler Works, Cross Compiler, Cross Compiler
1.14 pazsan 15561: @section Using the Cross Compiler
1.46 pazsan 15562:
15563: The cross compiler uses a language that resembles Forth, but isn't. The
15564: main difference is that you can execute Forth code after definition,
15565: while you usually can't execute the code compiled by cross, because the
15566: code you are compiling is typically for a different computer than the
15567: one you are compiling on.
15568:
1.81 anton 15569: @c anton: This chapter is somewhat different from waht I would expect: I
15570: @c would expect an explanation of the cross language and how to create an
15571: @c application image with it. The section explains some aspects of
15572: @c creating a Gforth kernel.
15573:
1.46 pazsan 15574: The Makefile is already set up to allow you to create kernels for new
15575: architectures with a simple make command. The generic kernels using the
15576: GCC compiled virtual machine are created in the normal build process
15577: with @code{make}. To create a embedded Gforth executable for e.g. the
15578: 8086 processor (running on a DOS machine), type
15579:
15580: @example
15581: make kernl-8086.fi
15582: @end example
15583:
15584: This will use the machine description from the @file{arch/8086}
15585: directory to create a new kernel. A machine file may look like that:
15586:
15587: @example
15588: \ Parameter for target systems 06oct92py
15589:
15590: 4 Constant cell \ cell size in bytes
15591: 2 Constant cell<< \ cell shift to bytes
15592: 5 Constant cell>bit \ cell shift to bits
15593: 8 Constant bits/char \ bits per character
15594: 8 Constant bits/byte \ bits per byte [default: 8]
15595: 8 Constant float \ bytes per float
15596: 8 Constant /maxalign \ maximum alignment in bytes
15597: false Constant bigendian \ byte order
15598: ( true=big, false=little )
15599:
15600: include machpc.fs \ feature list
15601: @end example
15602:
15603: This part is obligatory for the cross compiler itself, the feature list
15604: is used by the kernel to conditionally compile some features in and out,
15605: depending on whether the target supports these features.
15606:
15607: There are some optional features, if you define your own primitives,
15608: have an assembler, or need special, nonstandard preparation to make the
1.81 anton 15609: boot process work. @code{asm-include} includes an assembler,
1.46 pazsan 15610: @code{prims-include} includes primitives, and @code{>boot} prepares for
15611: booting.
15612:
15613: @example
15614: : asm-include ." Include assembler" cr
15615: s" arch/8086/asm.fs" included ;
15616:
15617: : prims-include ." Include primitives" cr
15618: s" arch/8086/prim.fs" included ;
15619:
15620: : >boot ." Prepare booting" cr
15621: s" ' boot >body into-forth 1+ !" evaluate ;
15622: @end example
15623:
15624: These words are used as sort of macro during the cross compilation in
1.81 anton 15625: the file @file{kernel/main.fs}. Instead of using these macros, it would
1.46 pazsan 15626: be possible --- but more complicated --- to write a new kernel project
15627: file, too.
15628:
15629: @file{kernel/main.fs} expects the machine description file name on the
15630: stack; the cross compiler itself (@file{cross.fs}) assumes that either
15631: @code{mach-file} leaves a counted string on the stack, or
15632: @code{machine-file} leaves an address, count pair of the filename on the
15633: stack.
15634:
15635: The feature list is typically controlled using @code{SetValue}, generic
15636: files that are used by several projects can use @code{DefaultValue}
15637: instead. Both functions work like @code{Value}, when the value isn't
15638: defined, but @code{SetValue} works like @code{to} if the value is
15639: defined, and @code{DefaultValue} doesn't set anything, if the value is
15640: defined.
15641:
15642: @example
15643: \ generic mach file for pc gforth 03sep97jaw
15644:
15645: true DefaultValue NIL \ relocating
15646:
15647: >ENVIRON
15648:
15649: true DefaultValue file \ controls the presence of the
15650: \ file access wordset
15651: true DefaultValue OS \ flag to indicate a operating system
15652:
15653: true DefaultValue prims \ true: primitives are c-code
15654:
15655: true DefaultValue floating \ floating point wordset is present
15656:
15657: true DefaultValue glocals \ gforth locals are present
15658: \ will be loaded
15659: true DefaultValue dcomps \ double number comparisons
15660:
15661: true DefaultValue hash \ hashing primitives are loaded/present
15662:
15663: true DefaultValue xconds \ used together with glocals,
15664: \ special conditionals supporting gforths'
15665: \ local variables
15666: true DefaultValue header \ save a header information
15667:
15668: true DefaultValue backtrace \ enables backtrace code
15669:
15670: false DefaultValue ec
15671: false DefaultValue crlf
15672:
15673: cell 2 = [IF] &32 [ELSE] &256 [THEN] KB DefaultValue kernel-size
15674:
15675: &16 KB DefaultValue stack-size
15676: &15 KB &512 + DefaultValue fstack-size
15677: &15 KB DefaultValue rstack-size
15678: &14 KB &512 + DefaultValue lstack-size
15679: @end example
1.13 pazsan 15680:
1.48 anton 15681: @node How the Cross Compiler Works, , Using the Cross Compiler, Cross Compiler
1.14 pazsan 15682: @section How the Cross Compiler Works
1.13 pazsan 15683:
15684: @node Bugs, Origin, Cross Compiler, Top
1.21 crook 15685: @appendix Bugs
1.1 anton 15686: @cindex bug reporting
15687:
1.21 crook 15688: Known bugs are described in the file @file{BUGS} in the Gforth distribution.
1.1 anton 15689:
1.103 anton 15690: If you find a bug, please submit a bug report through
15691: @uref{https://savannah.gnu.org/bugs/?func=addbug&group=gforth}.
1.21 crook 15692:
15693: @itemize @bullet
15694: @item
1.81 anton 15695: A program (or a sequence of keyboard commands) that reproduces the bug.
15696: @item
15697: A description of what you think constitutes the buggy behaviour.
15698: @item
1.21 crook 15699: The Gforth version used (it is announced at the start of an
15700: interactive Gforth session).
15701: @item
15702: The machine and operating system (on Unix
15703: systems @code{uname -a} will report this information).
15704: @item
1.81 anton 15705: The installation options (you can find the configure options at the
15706: start of @file{config.status}) and configuration (@code{configure}
15707: output or @file{config.cache}).
1.21 crook 15708: @item
15709: A complete list of changes (if any) you (or your installer) have made to the
15710: Gforth sources.
15711: @end itemize
1.1 anton 15712:
15713: For a thorough guide on reporting bugs read @ref{Bug Reporting, , How
15714: to Report Bugs, gcc.info, GNU C Manual}.
15715:
15716:
1.21 crook 15717: @node Origin, Forth-related information, Bugs, Top
15718: @appendix Authors and Ancestors of Gforth
1.1 anton 15719:
15720: @section Authors and Contributors
15721: @cindex authors of Gforth
15722: @cindex contributors to Gforth
15723:
15724: The Gforth project was started in mid-1992 by Bernd Paysan and Anton
1.81 anton 15725: Ertl. The third major author was Jens Wilke. Neal Crook contributed a
15726: lot to the manual. Assemblers and disassemblers were contributed by
1.161 anton 15727: Andrew McKewan, Christian Pirker, Bernd Thallner, and Michal Revucky.
15728: Lennart Benschop (who was one of Gforth's first users, in mid-1993)
15729: and Stuart Ramsden inspired us with their continuous feedback. Lennart
15730: Benshop contributed @file{glosgen.fs}, while Stuart Ramsden has been
15731: working on automatic support for calling C libraries. Helpful comments
15732: also came from Paul Kleinrubatscher, Christian Pirker, Dirk Zoller,
15733: Marcel Hendrix, John Wavrik, Barrie Stott, Marc de Groot, Jorge
15734: Acerada, Bruce Hoyt, Robert Epprecht, Dennis Ruffer and David
15735: N. Williams. Since the release of Gforth-0.2.1 there were also helpful
15736: comments from many others; thank you all, sorry for not listing you
15737: here (but digging through my mailbox to extract your names is on my
15738: to-do list).
1.1 anton 15739:
15740: Gforth also owes a lot to the authors of the tools we used (GCC, CVS,
15741: and autoconf, among others), and to the creators of the Internet: Gforth
1.21 crook 15742: was developed across the Internet, and its authors did not meet
1.20 pazsan 15743: physically for the first 4 years of development.
1.1 anton 15744:
15745: @section Pedigree
1.26 crook 15746: @cindex pedigree of Gforth
1.1 anton 15747:
1.81 anton 15748: Gforth descends from bigFORTH (1993) and fig-Forth. Of course, a
15749: significant part of the design of Gforth was prescribed by ANS Forth.
1.1 anton 15750:
1.20 pazsan 15751: Bernd Paysan wrote bigFORTH, a descendent from TurboForth, an unreleased
1.1 anton 15752: 32 bit native code version of VolksForth for the Atari ST, written
15753: mostly by Dietrich Weineck.
15754:
1.81 anton 15755: VolksForth was written by Klaus Schleisiek, Bernd Pennemann, Georg
15756: Rehfeld and Dietrich Weineck for the C64 (called UltraForth there) in
1.147 anton 15757: the mid-80s and ported to the Atari ST in 1986. It descends from fig-Forth.
1.1 anton 15758:
1.147 anton 15759: @c Henry Laxen and Mike Perry wrote F83 as a model implementation of the
15760: @c Forth-83 standard. !! Pedigree? When?
1.1 anton 15761:
15762: A team led by Bill Ragsdale implemented fig-Forth on many processors in
15763: 1979. Robert Selzer and Bill Ragsdale developed the original
15764: implementation of fig-Forth for the 6502 based on microForth.
15765:
15766: The principal architect of microForth was Dean Sanderson. microForth was
15767: FORTH, Inc.'s first off-the-shelf product. It was developed in 1976 for
15768: the 1802, and subsequently implemented on the 8080, the 6800 and the
15769: Z80.
15770:
15771: All earlier Forth systems were custom-made, usually by Charles Moore,
15772: who discovered (as he puts it) Forth during the late 60s. The first full
15773: Forth existed in 1971.
15774:
1.81 anton 15775: A part of the information in this section comes from
15776: @cite{@uref{http://www.forth.com/Content/History/History1.htm,The
15777: Evolution of Forth}} by Elizabeth D. Rather, Donald R. Colburn and
1.147 anton 15778: Charles H. Moore, presented at the HOPL-II conference and preprinted
15779: in SIGPLAN Notices 28(3), 1993. You can find more historical and
15780: genealogical information about Forth there. For a more general (and
15781: graphical) Forth family tree look see
15782: @cite{@uref{http://www.complang.tuwien.ac.at/forth/family-tree/},
15783: Forth Family Tree and Timeline}.
1.1 anton 15784:
1.81 anton 15785: @c ------------------------------------------------------------------
1.113 anton 15786: @node Forth-related information, Licenses, Origin, Top
1.21 crook 15787: @appendix Other Forth-related information
15788: @cindex Forth-related information
15789:
1.81 anton 15790: @c anton: I threw most of this stuff out, because it can be found through
15791: @c the FAQ and the FAQ is more likely to be up-to-date.
1.21 crook 15792:
15793: @cindex comp.lang.forth
15794: @cindex frequently asked questions
1.81 anton 15795: There is an active news group (comp.lang.forth) discussing Forth
15796: (including Gforth) and Forth-related issues. Its
15797: @uref{http://www.complang.tuwien.ac.at/forth/faq/faq-general-2.html,FAQs}
15798: (frequently asked questions and their answers) contains a lot of
15799: information on Forth. You should read it before posting to
15800: comp.lang.forth.
1.21 crook 15801:
1.81 anton 15802: The ANS Forth standard is most usable in its
15803: @uref{http://www.taygeta.com/forth/dpans.html, HTML form}.
1.21 crook 15804:
1.113 anton 15805: @c ---------------------------------------------------
15806: @node Licenses, Word Index, Forth-related information, Top
15807: @appendix Licenses
15808:
15809: @menu
15810: * GNU Free Documentation License:: License for copying this manual.
15811: * Copying:: GPL (for copying this software).
15812: @end menu
15813:
15814: @include fdl.texi
15815:
15816: @include gpl.texi
15817:
15818:
15819:
1.81 anton 15820: @c ------------------------------------------------------------------
1.113 anton 15821: @node Word Index, Concept Index, Licenses, Top
1.1 anton 15822: @unnumbered Word Index
15823:
1.26 crook 15824: This index is a list of Forth words that have ``glossary'' entries
15825: within this manual. Each word is listed with its stack effect and
15826: wordset.
1.1 anton 15827:
15828: @printindex fn
15829:
1.81 anton 15830: @c anton: the name index seems superfluous given the word and concept indices.
15831:
15832: @c @node Name Index, Concept Index, Word Index, Top
15833: @c @unnumbered Name Index
1.41 anton 15834:
1.81 anton 15835: @c This index is a list of Forth words that have ``glossary'' entries
15836: @c within this manual.
1.41 anton 15837:
1.81 anton 15838: @c @printindex ky
1.41 anton 15839:
1.113 anton 15840: @c -------------------------------------------------------
1.81 anton 15841: @node Concept Index, , Word Index, Top
1.1 anton 15842: @unnumbered Concept and Word Index
15843:
1.26 crook 15844: Not all entries listed in this index are present verbatim in the
15845: text. This index also duplicates, in abbreviated form, all of the words
15846: listed in the Word Index (only the names are listed for the words here).
1.1 anton 15847:
15848: @printindex cp
15849:
15850: @bye
1.81 anton 15851:
15852:
1.1 anton 15853:
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>