1: \input texinfo @c -*-texinfo-*-
2: @comment The source is gforth.ds, from which gforth.texi is generated
3:
4: @comment TODO: nac29jan99 - a list of things to add in the next edit:
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.
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
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
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:
24: @c POSTPONE, COMPILE, [COMPILE], LITERAL should have their own section
25:
26: @comment %**start of header (This is for running Texinfo on a region.)
27: @setfilename gforth.info
28: @include version.texi
29: @settitle Gforth Manual
30: @c @syncodeindex pg cp
31:
32: @macro progstyle {}
33: Programming style note:
34: @end macro
35:
36: @macro assignment {}
37: @table @i
38: @item Assignment:
39: @end macro
40: @macro endassignment {}
41: @end table
42: @end macro
43:
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:
57: @comment %**end of header (This is for running Texinfo on a region.)
58: @copying
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.
63:
64: Copyright @copyright{} 1995, 1996, 1997, 1998, 2000, 2003, 2004,2005,2006,2007,2008,2009 Free Software Foundation, Inc.
65:
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
80:
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
91:
92: @titlepage
93: @title Gforth
94: @subtitle for version @value{VERSION}, @value{UPDATED}
95: @author Neal Crook
96: @author Anton Ertl
97: @author David Kuehling
98: @author Bernd Paysan
99: @author Jens Wilke
100: @page
101: @vskip 0pt plus 1filll
102: @insertcopying
103: @end titlepage
104:
105: @contents
106:
107: @ifnottex
108: @node Top, Goals, (dir), (dir)
109: @top Gforth
110:
111: @insertcopying
112: @end ifnottex
113:
114: @menu
115: * Goals:: About the Gforth Project
116: * Gforth Environment:: Starting (and exiting) Gforth
117: * Tutorial:: Hands-on Forth Tutorial
118: * Introduction:: An introduction to ANS Forth
119: * Words:: Forth words available in Gforth
120: * Error messages:: How to interpret them
121: * Tools:: Programming tools
122: * ANS conformance:: Implementation-defined options etc.
123: * Standard vs Extensions:: Should I use extensions?
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
129: * Cross Compiler:: The Cross Compiler
130: * Bugs:: How to report them
131: * Origin:: Authors and ancestors of Gforth
132: * Forth-related information:: Books and places to look on the WWW
133: * Licenses::
134: * Word Index:: An item for each Forth word
135: * Concept Index:: A menu covering many topics
136:
137: @detailmenu
138: --- The Detailed Node Listing ---
139:
140: Gforth Environment
141:
142: * Invoking Gforth:: Getting in
143: * Leaving Gforth:: Getting out
144: * Command-line editing::
145: * Environment variables:: that affect how Gforth starts up
146: * Gforth Files:: What gets installed and where
147: * Gforth in pipes::
148: * Startup speed:: When 14ms 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::
177: * Floating Point Tutorial::
178: * Files Tutorial::
179: * Interpretation and Compilation Semantics and Immediacy Tutorial::
180: * Execution Tokens Tutorial::
181: * Exceptions Tutorial::
182: * Defining Words Tutorial::
183: * Arrays and Records Tutorial::
184: * POSTPONE Tutorial::
185: * Literal Tutorial::
186: * Advanced macros Tutorial::
187: * Compilation Tokens Tutorial::
188: * Wordlists and Search Order Tutorial::
189:
190: An Introduction to ANS Forth
191:
192: * Introducing the Text Interpreter::
193: * Stacks and Postfix notation::
194: * Your first definition::
195: * How does that work?::
196: * Forth is written in Forth::
197: * Review - elements of a Forth system::
198: * Where to go next::
199: * Exercises::
200:
201: Forth Words
202:
203: * Notation::
204: * Case insensitivity::
205: * Comments::
206: * Boolean Flags::
207: * Arithmetic::
208: * Stack Manipulation::
209: * Memory::
210: * Control Structures::
211: * Defining Words::
212: * Interpretation and Compilation Semantics::
213: * Tokens for Words::
214: * Compiling words::
215: * The Text Interpreter::
216: * The Input Stream::
217: * Word Lists::
218: * Environmental Queries::
219: * Files::
220: * Blocks::
221: * Other I/O::
222: * OS command line arguments::
223: * Locals::
224: * Structures::
225: * Object-oriented Forth::
226: * Programming Tools::
227: * C Interface::
228: * Assembler and Code Words::
229: * Threading Words::
230: * Passing Commands to the OS::
231: * Keeping track of Time::
232: * Miscellaneous Words::
233:
234: Arithmetic
235:
236: * Single precision::
237: * Double precision:: Double-cell integer arithmetic
238: * Bitwise operations::
239: * Numeric comparison::
240: * Mixed precision:: Operations with single and double-cell integers
241: * Floating Point::
242:
243: Stack Manipulation
244:
245: * Data stack::
246: * Floating point stack::
247: * Return stack::
248: * Locals stack::
249: * Stack pointer manipulation::
250:
251: Memory
252:
253: * Memory model::
254: * Dictionary allocation::
255: * Heap Allocation::
256: * Memory Access::
257: * Address arithmetic::
258: * Memory Blocks::
259:
260: Control Structures
261:
262: * Selection:: IF ... ELSE ... ENDIF
263: * Simple Loops:: BEGIN ...
264: * Counted Loops:: DO
265: * Arbitrary control structures::
266: * Calls and returns::
267: * Exception Handling::
268:
269: Defining Words
270:
271: * CREATE::
272: * Variables:: Variables and user variables
273: * Constants::
274: * Values:: Initialised variables
275: * Colon Definitions::
276: * Anonymous Definitions:: Definitions without names
277: * Supplying names:: Passing definition names as strings
278: * User-defined Defining Words::
279: * Deferred Words:: Allow forward references
280: * Aliases::
281:
282: User-defined Defining Words
283:
284: * CREATE..DOES> applications::
285: * CREATE..DOES> details::
286: * Advanced does> usage example::
287: * Const-does>::
288:
289: Interpretation and Compilation Semantics
290:
291: * Combined words::
292:
293: Tokens for Words
294:
295: * Execution token:: represents execution/interpretation semantics
296: * Compilation token:: represents compilation semantics
297: * Name token:: represents named words
298:
299: Compiling words
300:
301: * Literals:: Compiling data values
302: * Macros:: Compiling words
303:
304: The Text Interpreter
305:
306: * Input Sources::
307: * Number Conversion::
308: * Interpret/Compile states::
309: * Interpreter Directives::
310:
311: Word Lists
312:
313: * Vocabularies::
314: * Why use word lists?::
315: * Word list example::
316:
317: Files
318:
319: * Forth source files::
320: * General files::
321: * Redirection::
322: * Search Paths::
323:
324: Search Paths
325:
326: * Source Search Paths::
327: * General Search Paths::
328:
329: Other I/O
330:
331: * Simple numeric output:: Predefined formats
332: * Formatted numeric output:: Formatted (pictured) output
333: * String Formats:: How Forth stores strings in memory
334: * Displaying characters and strings:: Other stuff
335: * Terminal output:: Cursor positioning etc.
336: * Single-key input::
337: * Line input and conversion::
338: * Pipes:: How to create your own pipes
339: * Xchars and Unicode:: Non-ASCII characters
340:
341: Locals
342:
343: * Gforth locals::
344: * ANS Forth locals::
345:
346: Gforth locals
347:
348: * Where are locals visible by name?::
349: * How long do locals live?::
350: * Locals programming style::
351: * Locals implementation::
352:
353: Structures
354:
355: * Why explicit structure support?::
356: * Structure Usage::
357: * Structure Naming Convention::
358: * Structure Implementation::
359: * Structure Glossary::
360: * Forth200x Structures::
361:
362: Object-oriented Forth
363:
364: * Why object-oriented programming?::
365: * Object-Oriented Terminology::
366: * Objects::
367: * OOF::
368: * Mini-OOF::
369: * Comparison with other object models::
370:
371: The @file{objects.fs} model
372:
373: * Properties of the Objects model::
374: * Basic Objects Usage::
375: * The Objects base class::
376: * Creating objects::
377: * Object-Oriented Programming Style::
378: * Class Binding::
379: * Method conveniences::
380: * Classes and Scoping::
381: * Dividing classes::
382: * Object Interfaces::
383: * Objects Implementation::
384: * Objects Glossary::
385:
386: The @file{oof.fs} model
387:
388: * Properties of the OOF model::
389: * Basic OOF Usage::
390: * The OOF base class::
391: * Class Declaration::
392: * Class Implementation::
393:
394: The @file{mini-oof.fs} model
395:
396: * Basic Mini-OOF Usage::
397: * Mini-OOF Example::
398: * Mini-OOF Implementation::
399:
400: Programming Tools
401:
402: * Examining:: Data and Code.
403: * Forgetting words:: Usually before reloading.
404: * Debugging:: Simple and quick.
405: * Assertions:: Making your programs self-checking.
406: * Singlestep Debugger:: Executing your program word by word.
407:
408: C Interface
409:
410: * Calling C Functions::
411: * Declaring C Functions::
412: * Calling C function pointers::
413: * Defining library interfaces::
414: * Declaring OS-level libraries::
415: * Callbacks::
416: * C interface internals::
417: * Low-Level C Interface Words::
418:
419: Assembler and Code Words
420:
421: * Code and ;code::
422: * Common Assembler:: Assembler Syntax
423: * Common Disassembler::
424: * 386 Assembler:: Deviations and special cases
425: * Alpha Assembler:: Deviations and special cases
426: * MIPS assembler:: Deviations and special cases
427: * PowerPC assembler:: Deviations and special cases
428: * ARM Assembler:: Deviations and special cases
429: * Other assemblers:: How to write them
430:
431: Tools
432:
433: * ANS Report:: Report the words used, sorted by wordset.
434: * Stack depth changes:: Where does this stack item come from?
435:
436: ANS conformance
437:
438: * The Core Words::
439: * The optional Block word set::
440: * The optional Double Number word set::
441: * The optional Exception word set::
442: * The optional Facility word set::
443: * The optional File-Access word set::
444: * The optional Floating-Point word set::
445: * The optional Locals word set::
446: * The optional Memory-Allocation word set::
447: * The optional Programming-Tools word set::
448: * The optional Search-Order word set::
449:
450: The Core Words
451:
452: * core-idef:: Implementation Defined Options
453: * core-ambcond:: Ambiguous Conditions
454: * core-other:: Other System Documentation
455:
456: The optional Block word set
457:
458: * block-idef:: Implementation Defined Options
459: * block-ambcond:: Ambiguous Conditions
460: * block-other:: Other System Documentation
461:
462: The optional Double Number word set
463:
464: * double-ambcond:: Ambiguous Conditions
465:
466: The optional Exception word set
467:
468: * exception-idef:: Implementation Defined Options
469:
470: The optional Facility word set
471:
472: * facility-idef:: Implementation Defined Options
473: * facility-ambcond:: Ambiguous Conditions
474:
475: The optional File-Access word set
476:
477: * file-idef:: Implementation Defined Options
478: * file-ambcond:: Ambiguous Conditions
479:
480: The optional Floating-Point word set
481:
482: * floating-idef:: Implementation Defined Options
483: * floating-ambcond:: Ambiguous Conditions
484:
485: The optional Locals word set
486:
487: * locals-idef:: Implementation Defined Options
488: * locals-ambcond:: Ambiguous Conditions
489:
490: The optional Memory-Allocation word set
491:
492: * memory-idef:: Implementation Defined Options
493:
494: The optional Programming-Tools word set
495:
496: * programming-idef:: Implementation Defined Options
497: * programming-ambcond:: Ambiguous Conditions
498:
499: The optional Search-Order word set
500:
501: * search-idef:: Implementation Defined Options
502: * search-ambcond:: Ambiguous Conditions
503:
504: Emacs and Gforth
505:
506: * Installing gforth.el:: Making Emacs aware of Forth.
507: * Emacs Tags:: Viewing the source of a word in Emacs.
508: * Hilighting:: Making Forth code look prettier.
509: * Auto-Indentation:: Customizing auto-indentation.
510: * Blocks Files:: Reading and writing blocks files.
511:
512: Image Files
513:
514: * Image Licensing Issues:: Distribution terms for images.
515: * Image File Background:: Why have image files?
516: * Non-Relocatable Image Files:: don't always work.
517: * Data-Relocatable Image Files:: are better.
518: * Fully Relocatable Image Files:: better yet.
519: * Stack and Dictionary Sizes:: Setting the default sizes for an image.
520: * Running Image Files:: @code{gforth -i @i{file}} or @i{file}.
521: * Modifying the Startup Sequence:: and turnkey applications.
522:
523: Fully Relocatable Image Files
524:
525: * gforthmi:: The normal way
526: * cross.fs:: The hard way
527:
528: Engine
529:
530: * Portability::
531: * Threading::
532: * Primitives::
533: * Performance::
534:
535: Threading
536:
537: * Scheduling::
538: * Direct or Indirect Threaded?::
539: * Dynamic Superinstructions::
540: * DOES>::
541:
542: Primitives
543:
544: * Automatic Generation::
545: * TOS Optimization::
546: * Produced code::
547:
548: Cross Compiler
549:
550: * Using the Cross Compiler::
551: * How the Cross Compiler Works::
552:
553: Licenses
554:
555: * GNU Free Documentation License:: License for copying this manual.
556: * Copying:: GPL (for copying this software).
557:
558: @end detailmenu
559: @end menu
560:
561: @c ----------------------------------------------------------
562: @iftex
563: @unnumbered Preface
564: @cindex Preface
565: This manual documents Gforth. Some introductory material is provided for
566: readers who are unfamiliar with Forth or who are migrating to Gforth
567: from other Forth compilers. However, this manual is primarily a
568: reference manual.
569: @end iftex
570:
571: @comment TODO much more blurb here.
572:
573: @c ******************************************************************
574: @node Goals, Gforth Environment, Top, Top
575: @comment node-name, next, previous, up
576: @chapter Goals of Gforth
577: @cindex goals of the Gforth project
578: The goal of the Gforth Project is to develop a standard model for
579: ANS Forth. This can be split into several subgoals:
580:
581: @itemize @bullet
582: @item
583: Gforth should conform to the ANS Forth Standard.
584: @item
585: It should be a model, i.e. it should define all the
586: implementation-dependent things.
587: @item
588: It should become standard, i.e. widely accepted and used. This goal
589: is the most difficult one.
590: @end itemize
591:
592: To achieve these goals Gforth should be
593: @itemize @bullet
594: @item
595: Similar to previous models (fig-Forth, F83)
596: @item
597: Powerful. It should provide for all the things that are considered
598: necessary today and even some that are not yet considered necessary.
599: @item
600: Efficient. It should not get the reputation of being exceptionally
601: slow.
602: @item
603: Free.
604: @item
605: Available on many machines/easy to port.
606: @end itemize
607:
608: Have we achieved these goals? Gforth conforms to the ANS Forth
609: standard. It may be considered a model, but we have not yet documented
610: which parts of the model are stable and which parts we are likely to
611: change. It certainly has not yet become a de facto standard, but it
612: appears to be quite popular. It has some similarities to and some
613: differences from previous models. It has some powerful features, but not
614: yet everything that we envisioned. We certainly have achieved our
615: execution speed goals (@pxref{Performance})@footnote{However, in 1998
616: the bar was raised when the major commercial Forth vendors switched to
617: native code compilers.}. It is free and available on many machines.
618:
619: @c ******************************************************************
620: @node Gforth Environment, Tutorial, Goals, Top
621: @chapter Gforth Environment
622: @cindex Gforth environment
623:
624: Note: ultimately, the Gforth man page will be auto-generated from the
625: material in this chapter.
626:
627: @menu
628: * Invoking Gforth:: Getting in
629: * Leaving Gforth:: Getting out
630: * Command-line editing::
631: * Environment variables:: that affect how Gforth starts up
632: * Gforth Files:: What gets installed and where
633: * Gforth in pipes::
634: * Startup speed:: When 14ms is not fast enough ...
635: @end menu
636:
637: For related information about the creation of images see @ref{Image Files}.
638:
639: @comment ----------------------------------------------
640: @node Invoking Gforth, Leaving Gforth, Gforth Environment, Gforth Environment
641: @section Invoking Gforth
642: @cindex invoking Gforth
643: @cindex running Gforth
644: @cindex command-line options
645: @cindex options on the command line
646: @cindex flags on the command line
647:
648: Gforth is made up of two parts; an executable ``engine'' (named
649: @command{gforth} or @command{gforth-fast}) and an image file. To start it, you
650: will usually just say @code{gforth} -- this automatically loads the
651: default image file @file{gforth.fi}. In many other cases the default
652: Gforth image will be invoked like this:
653: @example
654: gforth [file | -e forth-code] ...
655: @end example
656: @noindent
657: This interprets the contents of the files and the Forth code in the order they
658: are given.
659:
660: In addition to the @command{gforth} engine, there is also an engine
661: called @command{gforth-fast}, which is faster, but gives less
662: informative error messages (@pxref{Error messages}) and may catch some
663: errors (in particular, stack underflows and integer division errors)
664: later or not at all. You should use it for debugged,
665: performance-critical programs.
666:
667: Moreover, there is an engine called @command{gforth-itc}, which is
668: useful in some backwards-compatibility situations (@pxref{Direct or
669: Indirect Threaded?}).
670:
671: In general, the command line looks like this:
672:
673: @example
674: gforth[-fast] [engine options] [image options]
675: @end example
676:
677: The engine options must come before the rest of the command
678: line. They are:
679:
680: @table @code
681: @cindex -i, command-line option
682: @cindex --image-file, command-line option
683: @item --image-file @i{file}
684: @itemx -i @i{file}
685: Loads the Forth image @i{file} instead of the default
686: @file{gforth.fi} (@pxref{Image Files}).
687:
688: @cindex --appl-image, command-line option
689: @item --appl-image @i{file}
690: Loads the image @i{file} and leaves all further command-line arguments
691: to the image (instead of processing them as engine options). This is
692: useful for building executable application images on Unix, built with
693: @code{gforthmi --application ...}.
694:
695: @cindex --path, command-line option
696: @cindex -p, command-line option
697: @item --path @i{path}
698: @itemx -p @i{path}
699: Uses @i{path} for searching the image file and Forth source code files
700: instead of the default in the environment variable @code{GFORTHPATH} or
701: the path specified at installation time (e.g.,
702: @file{/usr/local/share/gforth/0.2.0:.}). A path is given as a list of
703: directories, separated by @samp{:} (on Unix) or @samp{;} (on other OSs).
704:
705: @cindex --dictionary-size, command-line option
706: @cindex -m, command-line option
707: @cindex @i{size} parameters for command-line options
708: @cindex size of the dictionary and the stacks
709: @item --dictionary-size @i{size}
710: @itemx -m @i{size}
711: Allocate @i{size} space for the Forth dictionary space instead of
712: using the default specified in the image (typically 256K). The
713: @i{size} specification for this and subsequent options consists of
714: an integer and a unit (e.g.,
715: @code{4M}). The unit can be one of @code{b} (bytes), @code{e} (element
716: size, in this case Cells), @code{k} (kilobytes), @code{M} (Megabytes),
717: @code{G} (Gigabytes), and @code{T} (Terabytes). If no unit is specified,
718: @code{e} is used.
719:
720: @cindex --data-stack-size, command-line option
721: @cindex -d, command-line option
722: @item --data-stack-size @i{size}
723: @itemx -d @i{size}
724: Allocate @i{size} space for the data stack instead of using the
725: default specified in the image (typically 16K).
726:
727: @cindex --return-stack-size, command-line option
728: @cindex -r, command-line option
729: @item --return-stack-size @i{size}
730: @itemx -r @i{size}
731: Allocate @i{size} space for the return stack instead of using the
732: default specified in the image (typically 15K).
733:
734: @cindex --fp-stack-size, command-line option
735: @cindex -f, command-line option
736: @item --fp-stack-size @i{size}
737: @itemx -f @i{size}
738: Allocate @i{size} space for the floating point stack instead of
739: using the default specified in the image (typically 15.5K). In this case
740: the unit specifier @code{e} refers to floating point numbers.
741:
742: @cindex --locals-stack-size, command-line option
743: @cindex -l, command-line option
744: @item --locals-stack-size @i{size}
745: @itemx -l @i{size}
746: Allocate @i{size} space for the locals stack instead of using the
747: default specified in the image (typically 14.5K).
748:
749: @cindex --vm-commit, command-line option
750: @cindex overcommit memory for dictionary and stacks
751: @cindex memory overcommit for dictionary and stacks
752: @item --vm-commit
753: Normally, Gforth tries to start up even if there is not enough virtual
754: memory for the dictionary and the stacks (using @code{MAP_NORESERVE}
755: on OSs that support it); so you can ask for a really big dictionary
756: and/or stacks, and as long as you don't use more virtual memory than
757: is available, everything will be fine (but if you use more, processes
758: get killed). With this option you just use the default allocation
759: policy of the OS; for OSs that don't overcommit (e.g., Solaris), this
760: means that you cannot and should not ask for as big dictionary and
761: stacks, but once Gforth successfully starts up, out-of-memory won't
762: kill it.
763:
764: @cindex -h, command-line option
765: @cindex --help, command-line option
766: @item --help
767: @itemx -h
768: Print a message about the command-line options
769:
770: @cindex -v, command-line option
771: @cindex --version, command-line option
772: @item --version
773: @itemx -v
774: Print version and exit
775:
776: @cindex --debug, command-line option
777: @item --debug
778: Print some information useful for debugging on startup.
779:
780: @cindex --offset-image, command-line option
781: @item --offset-image
782: Start the dictionary at a slightly different position than would be used
783: otherwise (useful for creating data-relocatable images,
784: @pxref{Data-Relocatable Image Files}).
785:
786: @cindex --no-offset-im, command-line option
787: @item --no-offset-im
788: Start the dictionary at the normal position.
789:
790: @cindex --clear-dictionary, command-line option
791: @item --clear-dictionary
792: Initialize all bytes in the dictionary to 0 before loading the image
793: (@pxref{Data-Relocatable Image Files}).
794:
795: @cindex --die-on-signal, command-line-option
796: @item --die-on-signal
797: Normally Gforth handles most signals (e.g., the user interrupt SIGINT,
798: or the segmentation violation SIGSEGV) by translating it into a Forth
799: @code{THROW}. With this option, Gforth exits if it receives such a
800: signal. This option is useful when the engine and/or the image might be
801: severely broken (such that it causes another signal before recovering
802: from the first); this option avoids endless loops in such cases.
803:
804: @cindex --no-dynamic, command-line option
805: @cindex --dynamic, command-line option
806: @item --no-dynamic
807: @item --dynamic
808: Disable or enable dynamic superinstructions with replication
809: (@pxref{Dynamic Superinstructions}).
810:
811: @cindex --no-super, command-line option
812: @item --no-super
813: Disable dynamic superinstructions, use just dynamic replication; this is
814: useful if you want to patch threaded code (@pxref{Dynamic
815: Superinstructions}).
816:
817: @cindex --ss-number, command-line option
818: @item --ss-number=@var{N}
819: Use only the first @var{N} static superinstructions compiled into the
820: engine (default: use them all; note that only @code{gforth-fast} has
821: any). This option is useful for measuring the performance impact of
822: static superinstructions.
823:
824: @cindex --ss-min-..., command-line options
825: @item --ss-min-codesize
826: @item --ss-min-ls
827: @item --ss-min-lsu
828: @item --ss-min-nexts
829: Use specified metric for determining the cost of a primitive or static
830: superinstruction for static superinstruction selection. @code{Codesize}
831: is the native code size of the primive or static superinstruction,
832: @code{ls} is the number of loads and stores, @code{lsu} is the number of
833: loads, stores, and updates, and @code{nexts} is the number of dispatches
834: (not taking dynamic superinstructions into account), i.e. every
835: primitive or static superinstruction has cost 1. Default:
836: @code{codesize} if you use dynamic code generation, otherwise
837: @code{nexts}.
838:
839: @cindex --ss-greedy, command-line option
840: @item --ss-greedy
841: This option is useful for measuring the performance impact of static
842: superinstructions. By default, an optimal shortest-path algorithm is
843: used for selecting static superinstructions. With @option{--ss-greedy}
844: this algorithm is modified to assume that anything after the static
845: superinstruction currently under consideration is not combined into
846: static superinstructions. With @option{--ss-min-nexts} this produces
847: the same result as a greedy algorithm that always selects the longest
848: superinstruction available at the moment. E.g., if there are
849: superinstructions AB and BCD, then for the sequence A B C D the optimal
850: algorithm will select A BCD and the greedy algorithm will select AB C D.
851:
852: @cindex --print-metrics, command-line option
853: @item --print-metrics
854: Prints some metrics used during static superinstruction selection:
855: @code{code size} is the actual size of the dynamically generated code.
856: @code{Metric codesize} is the sum of the codesize metrics as seen by
857: static superinstruction selection; there is a difference from @code{code
858: size}, because not all primitives and static superinstructions are
859: compiled into dynamically generated code, and because of markers. The
860: other metrics correspond to the @option{ss-min-...} options. This
861: option is useful for evaluating the effects of the @option{--ss-...}
862: options.
863:
864: @end table
865:
866: @cindex loading files at startup
867: @cindex executing code on startup
868: @cindex batch processing with Gforth
869: As explained above, the image-specific command-line arguments for the
870: default image @file{gforth.fi} consist of a sequence of filenames and
871: @code{-e @var{forth-code}} options that are interpreted in the sequence
872: in which they are given. The @code{-e @var{forth-code}} or
873: @code{--evaluate @var{forth-code}} option evaluates the Forth code. This
874: option takes only one argument; if you want to evaluate more Forth
875: words, you have to quote them or use @code{-e} several times. To exit
876: after processing the command line (instead of entering interactive mode)
877: append @code{-e bye} to the command line. You can also process the
878: command-line arguments with a Forth program (@pxref{OS command line
879: arguments}).
880:
881: @cindex versions, invoking other versions of Gforth
882: If you have several versions of Gforth installed, @code{gforth} will
883: invoke the version that was installed last. @code{gforth-@i{version}}
884: invokes a specific version. If your environment contains the variable
885: @code{GFORTHPATH}, you may want to override it by using the
886: @code{--path} option.
887:
888: Not yet implemented:
889: On startup the system first executes the system initialization file
890: (unless the option @code{--no-init-file} is given; note that the system
891: resulting from using this option may not be ANS Forth conformant). Then
892: the user initialization file @file{.gforth.fs} is executed, unless the
893: option @code{--no-rc} is given; this file is searched for in @file{.},
894: then in @file{~}, then in the normal path (see above).
895:
896:
897:
898: @comment ----------------------------------------------
899: @node Leaving Gforth, Command-line editing, Invoking Gforth, Gforth Environment
900: @section Leaving Gforth
901: @cindex Gforth - leaving
902: @cindex leaving Gforth
903:
904: You can leave Gforth by typing @code{bye} or @kbd{Ctrl-d} (at the start
905: of a line) or (if you invoked Gforth with the @code{--die-on-signal}
906: option) @kbd{Ctrl-c}. When you leave Gforth, all of your definitions and
907: data are discarded. For ways of saving the state of the system before
908: leaving Gforth see @ref{Image Files}.
909:
910: doc-bye
911:
912:
913: @comment ----------------------------------------------
914: @node Command-line editing, Environment variables, Leaving Gforth, Gforth Environment
915: @section Command-line editing
916: @cindex command-line editing
917:
918: Gforth maintains a history file that records every line that you type to
919: the text interpreter. This file is preserved between sessions, and is
920: used to provide a command-line recall facility; if you type @kbd{Ctrl-P}
921: repeatedly you can recall successively older commands from this (or
922: previous) session(s). The full list of command-line editing facilities is:
923:
924: @itemize @bullet
925: @item
926: @kbd{Ctrl-p} (``previous'') (or up-arrow) to recall successively older
927: commands from the history buffer.
928: @item
929: @kbd{Ctrl-n} (``next'') (or down-arrow) to recall successively newer commands
930: from the history buffer.
931: @item
932: @kbd{Ctrl-f} (or right-arrow) to move the cursor right, non-destructively.
933: @item
934: @kbd{Ctrl-b} (or left-arrow) to move the cursor left, non-destructively.
935: @item
936: @kbd{Ctrl-h} (backspace) to delete the character to the left of the cursor,
937: closing up the line.
938: @item
939: @kbd{Ctrl-k} to delete (``kill'') from the cursor to the end of the line.
940: @item
941: @kbd{Ctrl-a} to move the cursor to the start of the line.
942: @item
943: @kbd{Ctrl-e} to move the cursor to the end of the line.
944: @item
945: @key{RET} (@kbd{Ctrl-m}) or @key{LFD} (@kbd{Ctrl-j}) to submit the current
946: line.
947: @item
948: @key{TAB} to step through all possible full-word completions of the word
949: currently being typed.
950: @item
951: @kbd{Ctrl-d} on an empty line line to terminate Gforth (gracefully,
952: using @code{bye}).
953: @item
954: @kbd{Ctrl-x} (or @code{Ctrl-d} on a non-empty line) to delete the
955: character under the cursor.
956: @end itemize
957:
958: When editing, displayable characters are inserted to the left of the
959: cursor position; the line is always in ``insert'' (as opposed to
960: ``overstrike'') mode.
961:
962: @cindex history file
963: @cindex @file{.gforth-history}
964: On Unix systems, the history file is @file{~/.gforth-history} by
965: default@footnote{i.e. it is stored in the user's home directory.}. You
966: can find out the name and location of your history file using:
967:
968: @example
969: history-file type \ Unix-class systems
970:
971: history-file type \ Other systems
972: history-dir type
973: @end example
974:
975: If you enter long definitions by hand, you can use a text editor to
976: paste them out of the history file into a Forth source file for reuse at
977: a later time.
978:
979: Gforth never trims the size of the history file, so you should do this
980: periodically, if necessary.
981:
982: @comment this is all defined in history.fs
983: @comment NAC TODO the ctrl-D behaviour can either do a bye or a beep.. how is that option
984: @comment chosen?
985:
986:
987: @comment ----------------------------------------------
988: @node Environment variables, Gforth Files, Command-line editing, Gforth Environment
989: @section Environment variables
990: @cindex environment variables
991:
992: Gforth uses these environment variables:
993:
994: @itemize @bullet
995: @item
996: @cindex @code{GFORTHHIST} -- environment variable
997: @code{GFORTHHIST} -- (Unix systems only) specifies the directory in which to
998: open/create the history file, @file{.gforth-history}. Default:
999: @code{$HOME}.
1000:
1001: @item
1002: @cindex @code{GFORTHPATH} -- environment variable
1003: @code{GFORTHPATH} -- specifies the path used when searching for the gforth image file and
1004: for Forth source-code files.
1005:
1006: @item
1007: @cindex @code{LANG} -- environment variable
1008: @code{LANG} -- see @code{LC_CTYPE}
1009:
1010: @item
1011: @cindex @code{LC_ALL} -- environment variable
1012: @code{LC_ALL} -- see @code{LC_CTYPE}
1013:
1014: @item
1015: @cindex @code{LC_CTYPE} -- environment variable
1016: @code{LC_CTYPE} -- If this variable contains ``UTF-8'' on Gforth
1017: startup, Gforth uses the UTF-8 encoding for strings internally and
1018: expects its input and produces its output in UTF-8 encoding, otherwise
1019: the encoding is 8bit (see @pxref{Xchars and Unicode}). If this
1020: environment variable is unset, Gforth looks in @code{LC_ALL}, and if
1021: that is unset, in @code{LANG}.
1022:
1023: @item
1024: @cindex @code{GFORTHSYSTEMPREFIX} -- environment variable
1025:
1026: @code{GFORTHSYSTEMPREFIX} -- specifies what to prepend to the argument
1027: of @code{system} before passing it to C's @code{system()}. Default:
1028: @code{"./$COMSPEC /c "} on Windows, @code{""} on other OSs. The prefix
1029: and the command are directly concatenated, so if a space between them is
1030: necessary, append it to the prefix.
1031:
1032: @item
1033: @cindex @code{GFORTH} -- environment variable
1034: @code{GFORTH} -- used by @file{gforthmi}, @xref{gforthmi}.
1035:
1036: @item
1037: @cindex @code{GFORTHD} -- environment variable
1038: @code{GFORTHD} -- used by @file{gforthmi}, @xref{gforthmi}.
1039:
1040: @item
1041: @cindex @code{TMP}, @code{TEMP} - environment variable
1042: @code{TMP}, @code{TEMP} - (non-Unix systems only) used as a potential
1043: location for the history file.
1044: @end itemize
1045:
1046: @comment also POSIXELY_CORRECT LINES COLUMNS HOME but no interest in
1047: @comment mentioning these.
1048:
1049: All the Gforth environment variables default to sensible values if they
1050: are not set.
1051:
1052:
1053: @comment ----------------------------------------------
1054: @node Gforth Files, Gforth in pipes, Environment variables, Gforth Environment
1055: @section Gforth files
1056: @cindex Gforth files
1057:
1058: When you install Gforth on a Unix system, it installs files in these
1059: locations by default:
1060:
1061: @itemize @bullet
1062: @item
1063: @file{/usr/local/bin/gforth}
1064: @item
1065: @file{/usr/local/bin/gforthmi}
1066: @item
1067: @file{/usr/local/man/man1/gforth.1} - man page.
1068: @item
1069: @file{/usr/local/info} - the Info version of this manual.
1070: @item
1071: @file{/usr/local/lib/gforth/<version>/...} - Gforth @file{.fi} files.
1072: @item
1073: @file{/usr/local/share/gforth/<version>/TAGS} - Emacs TAGS file.
1074: @item
1075: @file{/usr/local/share/gforth/<version>/...} - Gforth source files.
1076: @item
1077: @file{.../emacs/site-lisp/gforth.el} - Emacs gforth mode.
1078: @end itemize
1079:
1080: You can select different places for installation by using
1081: @code{configure} options (listed with @code{configure --help}).
1082:
1083: @comment ----------------------------------------------
1084: @node Gforth in pipes, Startup speed, Gforth Files, Gforth Environment
1085: @section Gforth in pipes
1086: @cindex pipes, Gforth as part of
1087:
1088: Gforth can be used in pipes created elsewhere (described here). It can
1089: also create pipes on its own (@pxref{Pipes}).
1090:
1091: @cindex input from pipes
1092: If you pipe into Gforth, your program should read with @code{read-file}
1093: or @code{read-line} from @code{stdin} (@pxref{General files}).
1094: @code{Key} does not recognize the end of input. Words like
1095: @code{accept} echo the input and are therefore usually not useful for
1096: reading from a pipe. You have to invoke the Forth program with an OS
1097: command-line option, as you have no chance to use the Forth command line
1098: (the text interpreter would try to interpret the pipe input).
1099:
1100: @cindex output in pipes
1101: You can output to a pipe with @code{type}, @code{emit}, @code{cr} etc.
1102:
1103: @cindex silent exiting from Gforth
1104: When you write to a pipe that has been closed at the other end, Gforth
1105: receives a SIGPIPE signal (``pipe broken''). Gforth translates this
1106: into the exception @code{broken-pipe-error}. If your application does
1107: not catch that exception, the system catches it and exits, usually
1108: silently (unless you were working on the Forth command line; then it
1109: prints an error message and exits). This is usually the desired
1110: behaviour.
1111:
1112: If you do not like this behaviour, you have to catch the exception
1113: yourself, and react to it.
1114:
1115: Here's an example of an invocation of Gforth that is usable in a pipe:
1116:
1117: @example
1118: gforth -e ": foo begin pad dup 10 stdin read-file throw dup while \
1119: type repeat ; foo bye"
1120: @end example
1121:
1122: This example just copies the input verbatim to the output. A very
1123: simple pipe containing this example looks like this:
1124:
1125: @example
1126: cat startup.fs |
1127: gforth -e ": foo begin pad dup 80 stdin read-file throw dup while \
1128: type repeat ; foo bye"|
1129: head
1130: @end example
1131:
1132: @cindex stderr and pipes
1133: Pipes involving Gforth's @code{stderr} output do not work.
1134:
1135: @comment ----------------------------------------------
1136: @node Startup speed, , Gforth in pipes, Gforth Environment
1137: @section Startup speed
1138: @cindex Startup speed
1139: @cindex speed, startup
1140:
1141: If Gforth is used for CGI scripts or in shell scripts, its startup
1142: speed may become a problem. On a 3GHz Core 2 Duo E8400 under 64-bit
1143: Linux 2.6.27.8 with libc-2.7, @code{gforth-fast -e bye} takes 13.1ms
1144: user and 1.2ms system time (@code{gforth -e bye} is faster on startup
1145: with about 3.4ms user time and 1.2ms system time, because it subsumes
1146: some of the options discussed below).
1147:
1148: If startup speed is a problem, you may consider the following ways to
1149: improve it; or you may consider ways to reduce the number of startups
1150: (for example, by using Fast-CGI). Note that the first steps below
1151: improve the startup time at the cost of run-time (including
1152: compile-time), so whether they are profitable depends on the balance
1153: of these times in your application.
1154:
1155: An easy step that influences Gforth startup speed is the use of a
1156: number of options that increase run-time, but decrease image-loading
1157: time.
1158:
1159: The first of these that you should try is @code{--ss-number=0
1160: --ss-states=1} because this option buys relatively little run-time
1161: speedup and costs quite a bit of time at startup. @code{gforth-fast
1162: --ss-number=0 --ss-states=1 -e bye} takes about 2.8ms user and 1.5ms
1163: system time.
1164:
1165: The next option is @code{--no-dynamic} which has a substantial impact
1166: on run-time (about a factor of 2 on several platforms), but still
1167: makes startup speed a little faster: @code{gforth-fast --ss-number=0
1168: --ss-states=1 --no-dynamic -e bye} consumes about 2.6ms user and 1.2ms
1169: system time.
1170:
1171: The next step to improve startup speed is to use a data-relocatable
1172: image (@pxref{Data-Relocatable Image Files}). This avoids the
1173: relocation cost for the code in the image (but not for the data).
1174: Note that the image is then specific to the particular binary you are
1175: using (i.e., whether it is @code{gforth}, @code{gforth-fast}, and even
1176: the particular build). You create the data-relocatable image that
1177: works with @code{./gforth-fast} with @code{GFORTHD="./gforth-fast
1178: --no-dynamic" gforthmi gforthdr.fi} (the @code{--no-dynamic} is
1179: required here or the image will not work). And you run it with
1180: @code{gforth-fast -i gforthdr.fi ... -e bye} (the flags discussed
1181: above don't matter here, because they only come into play on
1182: relocatable code). @code{gforth-fast -i gforthdr.fi -e bye} takes
1183: about 1.1ms user and 1.2ms system time.
1184:
1185: One step further is to avoid all relocation cost and part of the
1186: copy-on-write cost through using a non-relocatable image
1187: (@pxref{Non-Relocatable Image Files}). However, this has the
1188: disadvantage that it does not work on operating systems with address
1189: space randomization (the default in, e.g., Linux nowadays), or if the
1190: dictionary moves for any other reason (e.g., because of a change of
1191: the OS kernel or an updated library), so we cannot really recommend
1192: it. You create a non-relocatable image with @code{gforth-fast
1193: --no-dynamic -e "savesystem gforthnr.fi bye"} (the @code{--no-dynamic}
1194: is required here, too). And you run it with @code{gforth-fast -i
1195: gforthnr.fi ... -e bye} (again the flags discussed above don't
1196: matter). @code{gforth-fast -i gforthdr.fi -e bye} takes
1197: about 0.9ms user and 0.9ms system time.
1198:
1199: If the script you want to execute contains a significant amount of
1200: code, it may be profitable to compile it into the image to avoid the
1201: cost of compiling it at startup time.
1202:
1203: @c ******************************************************************
1204: @node Tutorial, Introduction, Gforth Environment, Top
1205: @chapter Forth Tutorial
1206: @cindex Tutorial
1207: @cindex Forth Tutorial
1208:
1209: @c Topics from nac's Introduction that could be mentioned:
1210: @c press <ret> after each line
1211: @c Prompt
1212: @c numbers vs. words in dictionary on text interpretation
1213: @c what happens on redefinition
1214: @c parsing words (in particular, defining words)
1215:
1216: The difference of this chapter from the Introduction
1217: (@pxref{Introduction}) is that this tutorial is more fast-paced, should
1218: be used while sitting in front of a computer, and covers much more
1219: material, but does not explain how the Forth system works.
1220:
1221: This tutorial can be used with any ANS-compliant Forth; any
1222: Gforth-specific features are marked as such and you can skip them if
1223: you work with another Forth. This tutorial does not explain all
1224: features of Forth, just enough to get you started and give you some
1225: ideas about the facilities available in Forth. Read the rest of the
1226: manual when you are through this.
1227:
1228: The intended way to use this tutorial is that you work through it while
1229: sitting in front of the console, take a look at the examples and predict
1230: what they will do, then try them out; if the outcome is not as expected,
1231: find out why (e.g., by trying out variations of the example), so you
1232: understand what's going on. There are also some assignments that you
1233: should solve.
1234:
1235: This tutorial assumes that you have programmed before and know what,
1236: e.g., a loop is.
1237:
1238: @c !! explain compat library
1239:
1240: @menu
1241: * Starting Gforth Tutorial::
1242: * Syntax Tutorial::
1243: * Crash Course Tutorial::
1244: * Stack Tutorial::
1245: * Arithmetics Tutorial::
1246: * Stack Manipulation Tutorial::
1247: * Using files for Forth code Tutorial::
1248: * Comments Tutorial::
1249: * Colon Definitions Tutorial::
1250: * Decompilation Tutorial::
1251: * Stack-Effect Comments Tutorial::
1252: * Types Tutorial::
1253: * Factoring Tutorial::
1254: * Designing the stack effect Tutorial::
1255: * Local Variables Tutorial::
1256: * Conditional execution Tutorial::
1257: * Flags and Comparisons Tutorial::
1258: * General Loops Tutorial::
1259: * Counted loops Tutorial::
1260: * Recursion Tutorial::
1261: * Leaving definitions or loops Tutorial::
1262: * Return Stack Tutorial::
1263: * Memory Tutorial::
1264: * Characters and Strings Tutorial::
1265: * Alignment Tutorial::
1266: * Floating Point Tutorial::
1267: * Files Tutorial::
1268: * Interpretation and Compilation Semantics and Immediacy Tutorial::
1269: * Execution Tokens Tutorial::
1270: * Exceptions Tutorial::
1271: * Defining Words Tutorial::
1272: * Arrays and Records Tutorial::
1273: * POSTPONE Tutorial::
1274: * Literal Tutorial::
1275: * Advanced macros Tutorial::
1276: * Compilation Tokens Tutorial::
1277: * Wordlists and Search Order Tutorial::
1278: @end menu
1279:
1280: @node Starting Gforth Tutorial, Syntax Tutorial, Tutorial, Tutorial
1281: @section Starting Gforth
1282: @cindex starting Gforth tutorial
1283: You can start Gforth by typing its name:
1284:
1285: @example
1286: gforth
1287: @end example
1288:
1289: That puts you into interactive mode; you can leave Gforth by typing
1290: @code{bye}. While in Gforth, you can edit the command line and access
1291: the command line history with cursor keys, similar to bash.
1292:
1293:
1294: @node Syntax Tutorial, Crash Course Tutorial, Starting Gforth Tutorial, Tutorial
1295: @section Syntax
1296: @cindex syntax tutorial
1297:
1298: A @dfn{word} is a sequence of arbitrary characters (except white
1299: space). Words are separated by white space. E.g., each of the
1300: following lines contains exactly one word:
1301:
1302: @example
1303: word
1304: !@@#$%^&*()
1305: 1234567890
1306: 5!a
1307: @end example
1308:
1309: A frequent beginner's error is to leave out necessary white space,
1310: resulting in an error like @samp{Undefined word}; so if you see such an
1311: error, check if you have put spaces wherever necessary.
1312:
1313: @example
1314: ." hello, world" \ correct
1315: ."hello, world" \ gives an "Undefined word" error
1316: @end example
1317:
1318: Gforth and most other Forth systems ignore differences in case (they are
1319: case-insensitive), i.e., @samp{word} is the same as @samp{Word}. If
1320: your system is case-sensitive, you may have to type all the examples
1321: given here in upper case.
1322:
1323:
1324: @node Crash Course Tutorial, Stack Tutorial, Syntax Tutorial, Tutorial
1325: @section Crash Course
1326:
1327: Forth does not prevent you from shooting yourself in the foot. Let's
1328: try a few ways to crash Gforth:
1329:
1330: @example
1331: 0 0 !
1332: here execute
1333: ' catch >body 20 erase abort
1334: ' (quit) >body 20 erase
1335: @end example
1336:
1337: The last two examples are guaranteed to destroy important parts of
1338: Gforth (and most other systems), so you better leave Gforth afterwards
1339: (if it has not finished by itself). On some systems you may have to
1340: kill gforth from outside (e.g., in Unix with @code{kill}).
1341:
1342: You will find out later what these lines do and then you will get an
1343: idea why they produce crashes.
1344:
1345: Now that you know how to produce crashes (and that there's not much to
1346: them), let's learn how to produce meaningful programs.
1347:
1348:
1349: @node Stack Tutorial, Arithmetics Tutorial, Crash Course Tutorial, Tutorial
1350: @section Stack
1351: @cindex stack tutorial
1352:
1353: The most obvious feature of Forth is the stack. When you type in a
1354: number, it is pushed on the stack. You can display the contents of the
1355: stack with @code{.s}.
1356:
1357: @example
1358: 1 2 .s
1359: 3 .s
1360: @end example
1361:
1362: @code{.s} displays the top-of-stack to the right, i.e., the numbers
1363: appear in @code{.s} output as they appeared in the input.
1364:
1365: You can print the top element of the stack with @code{.}.
1366:
1367: @example
1368: 1 2 3 . . .
1369: @end example
1370:
1371: In general, words consume their stack arguments (@code{.s} is an
1372: exception).
1373:
1374: @quotation Assignment
1375: What does the stack contain after @code{5 6 7 .}?
1376: @end quotation
1377:
1378:
1379: @node Arithmetics Tutorial, Stack Manipulation Tutorial, Stack Tutorial, Tutorial
1380: @section Arithmetics
1381: @cindex arithmetics tutorial
1382:
1383: The words @code{+}, @code{-}, @code{*}, @code{/}, and @code{mod} always
1384: operate on the top two stack items:
1385:
1386: @example
1387: 2 2 .s
1388: + .s
1389: .
1390: 2 1 - .
1391: 7 3 mod .
1392: @end example
1393:
1394: The operands of @code{-}, @code{/}, and @code{mod} are in the same order
1395: as in the corresponding infix expression (this is generally the case in
1396: Forth).
1397:
1398: Parentheses are superfluous (and not available), because the order of
1399: the words unambiguously determines the order of evaluation and the
1400: operands:
1401:
1402: @example
1403: 3 4 + 5 * .
1404: 3 4 5 * + .
1405: @end example
1406:
1407: @quotation Assignment
1408: What are the infix expressions corresponding to the Forth code above?
1409: Write @code{6-7*8+9} in Forth notation@footnote{This notation is also
1410: known as Postfix or RPN (Reverse Polish Notation).}.
1411: @end quotation
1412:
1413: To change the sign, use @code{negate}:
1414:
1415: @example
1416: 2 negate .
1417: @end example
1418:
1419: @quotation Assignment
1420: Convert -(-3)*4-5 to Forth.
1421: @end quotation
1422:
1423: @code{/mod} performs both @code{/} and @code{mod}.
1424:
1425: @example
1426: 7 3 /mod . .
1427: @end example
1428:
1429: Reference: @ref{Arithmetic}.
1430:
1431:
1432: @node Stack Manipulation Tutorial, Using files for Forth code Tutorial, Arithmetics Tutorial, Tutorial
1433: @section Stack Manipulation
1434: @cindex stack manipulation tutorial
1435:
1436: Stack manipulation words rearrange the data on the stack.
1437:
1438: @example
1439: 1 .s drop .s
1440: 1 .s dup .s drop drop .s
1441: 1 2 .s over .s drop drop drop
1442: 1 2 .s swap .s drop drop
1443: 1 2 3 .s rot .s drop drop drop
1444: @end example
1445:
1446: These are the most important stack manipulation words. There are also
1447: variants that manipulate twice as many stack items:
1448:
1449: @example
1450: 1 2 3 4 .s 2swap .s 2drop 2drop
1451: @end example
1452:
1453: Two more stack manipulation words are:
1454:
1455: @example
1456: 1 2 .s nip .s drop
1457: 1 2 .s tuck .s 2drop drop
1458: @end example
1459:
1460: @quotation Assignment
1461: Replace @code{nip} and @code{tuck} with combinations of other stack
1462: manipulation words.
1463:
1464: @example
1465: Given: How do you get:
1466: 1 2 3 3 2 1
1467: 1 2 3 1 2 3 2
1468: 1 2 3 1 2 3 3
1469: 1 2 3 1 3 3
1470: 1 2 3 2 1 3
1471: 1 2 3 4 4 3 2 1
1472: 1 2 3 1 2 3 1 2 3
1473: 1 2 3 4 1 2 3 4 1 2
1474: 1 2 3
1475: 1 2 3 1 2 3 4
1476: 1 2 3 1 3
1477: @end example
1478: @end quotation
1479:
1480: @example
1481: 5 dup * .
1482: @end example
1483:
1484: @quotation Assignment
1485: Write 17^3 and 17^4 in Forth, without writing @code{17} more than once.
1486: Write a piece of Forth code that expects two numbers on the stack
1487: (@var{a} and @var{b}, with @var{b} on top) and computes
1488: @code{(a-b)(a+1)}.
1489: @end quotation
1490:
1491: Reference: @ref{Stack Manipulation}.
1492:
1493:
1494: @node Using files for Forth code Tutorial, Comments Tutorial, Stack Manipulation Tutorial, Tutorial
1495: @section Using files for Forth code
1496: @cindex loading Forth code, tutorial
1497: @cindex files containing Forth code, tutorial
1498:
1499: While working at the Forth command line is convenient for one-line
1500: examples and short one-off code, you probably want to store your source
1501: code in files for convenient editing and persistence. You can use your
1502: favourite editor (Gforth includes Emacs support, @pxref{Emacs and
1503: Gforth}) to create @var{file.fs} and use
1504:
1505: @example
1506: s" @var{file.fs}" included
1507: @end example
1508:
1509: to load it into your Forth system. The file name extension I use for
1510: Forth files is @samp{.fs}.
1511:
1512: You can easily start Gforth with some files loaded like this:
1513:
1514: @example
1515: gforth @var{file1.fs} @var{file2.fs}
1516: @end example
1517:
1518: If an error occurs during loading these files, Gforth terminates,
1519: whereas an error during @code{INCLUDED} within Gforth usually gives you
1520: a Gforth command line. Starting the Forth system every time gives you a
1521: clean start every time, without interference from the results of earlier
1522: tries.
1523:
1524: I often put all the tests in a file, then load the code and run the
1525: tests with
1526:
1527: @example
1528: gforth @var{code.fs} @var{tests.fs} -e bye
1529: @end example
1530:
1531: (often by performing this command with @kbd{C-x C-e} in Emacs). The
1532: @code{-e bye} ensures that Gforth terminates afterwards so that I can
1533: restart this command without ado.
1534:
1535: The advantage of this approach is that the tests can be repeated easily
1536: every time the program ist changed, making it easy to catch bugs
1537: introduced by the change.
1538:
1539: Reference: @ref{Forth source files}.
1540:
1541:
1542: @node Comments Tutorial, Colon Definitions Tutorial, Using files for Forth code Tutorial, Tutorial
1543: @section Comments
1544: @cindex comments tutorial
1545:
1546: @example
1547: \ That's a comment; it ends at the end of the line
1548: ( Another comment; it ends here: ) .s
1549: @end example
1550:
1551: @code{\} and @code{(} are ordinary Forth words and therefore have to be
1552: separated with white space from the following text.
1553:
1554: @example
1555: \This gives an "Undefined word" error
1556: @end example
1557:
1558: The first @code{)} ends a comment started with @code{(}, so you cannot
1559: nest @code{(}-comments; and you cannot comment out text containing a
1560: @code{)} with @code{( ... )}@footnote{therefore it's a good idea to
1561: avoid @code{)} in word names.}.
1562:
1563: I use @code{\}-comments for descriptive text and for commenting out code
1564: of one or more line; I use @code{(}-comments for describing the stack
1565: effect, the stack contents, or for commenting out sub-line pieces of
1566: code.
1567:
1568: The Emacs mode @file{gforth.el} (@pxref{Emacs and Gforth}) supports
1569: these uses by commenting out a region with @kbd{C-x \}, uncommenting a
1570: region with @kbd{C-u C-x \}, and filling a @code{\}-commented region
1571: with @kbd{M-q}.
1572:
1573: Reference: @ref{Comments}.
1574:
1575:
1576: @node Colon Definitions Tutorial, Decompilation Tutorial, Comments Tutorial, Tutorial
1577: @section Colon Definitions
1578: @cindex colon definitions, tutorial
1579: @cindex definitions, tutorial
1580: @cindex procedures, tutorial
1581: @cindex functions, tutorial
1582:
1583: are similar to procedures and functions in other programming languages.
1584:
1585: @example
1586: : squared ( n -- n^2 )
1587: dup * ;
1588: 5 squared .
1589: 7 squared .
1590: @end example
1591:
1592: @code{:} starts the colon definition; its name is @code{squared}. The
1593: following comment describes its stack effect. The words @code{dup *}
1594: are not executed, but compiled into the definition. @code{;} ends the
1595: colon definition.
1596:
1597: The newly-defined word can be used like any other word, including using
1598: it in other definitions:
1599:
1600: @example
1601: : cubed ( n -- n^3 )
1602: dup squared * ;
1603: -5 cubed .
1604: : fourth-power ( n -- n^4 )
1605: squared squared ;
1606: 3 fourth-power .
1607: @end example
1608:
1609: @quotation Assignment
1610: Write colon definitions for @code{nip}, @code{tuck}, @code{negate}, and
1611: @code{/mod} in terms of other Forth words, and check if they work (hint:
1612: test your tests on the originals first). Don't let the
1613: @samp{redefined}-Messages spook you, they are just warnings.
1614: @end quotation
1615:
1616: Reference: @ref{Colon Definitions}.
1617:
1618:
1619: @node Decompilation Tutorial, Stack-Effect Comments Tutorial, Colon Definitions Tutorial, Tutorial
1620: @section Decompilation
1621: @cindex decompilation tutorial
1622: @cindex see tutorial
1623:
1624: You can decompile colon definitions with @code{see}:
1625:
1626: @example
1627: see squared
1628: see cubed
1629: @end example
1630:
1631: In Gforth @code{see} shows you a reconstruction of the source code from
1632: the executable code. Informations that were present in the source, but
1633: not in the executable code, are lost (e.g., comments).
1634:
1635: You can also decompile the predefined words:
1636:
1637: @example
1638: see .
1639: see +
1640: @end example
1641:
1642:
1643: @node Stack-Effect Comments Tutorial, Types Tutorial, Decompilation Tutorial, Tutorial
1644: @section Stack-Effect Comments
1645: @cindex stack-effect comments, tutorial
1646: @cindex --, tutorial
1647: By convention the comment after the name of a definition describes the
1648: stack effect: The part in front of the @samp{--} describes the state of
1649: the stack before the execution of the definition, i.e., the parameters
1650: that are passed into the colon definition; the part behind the @samp{--}
1651: is the state of the stack after the execution of the definition, i.e.,
1652: the results of the definition. The stack comment only shows the top
1653: stack items that the definition accesses and/or changes.
1654:
1655: You should put a correct stack effect on every definition, even if it is
1656: just @code{( -- )}. You should also add some descriptive comment to
1657: more complicated words (I usually do this in the lines following
1658: @code{:}). If you don't do this, your code becomes unreadable (because
1659: you have to work through every definition before you can understand
1660: any).
1661:
1662: @quotation Assignment
1663: The stack effect of @code{swap} can be written like this: @code{x1 x2 --
1664: x2 x1}. Describe the stack effect of @code{-}, @code{drop}, @code{dup},
1665: @code{over}, @code{rot}, @code{nip}, and @code{tuck}. Hint: When you
1666: are done, you can compare your stack effects to those in this manual
1667: (@pxref{Word Index}).
1668: @end quotation
1669:
1670: Sometimes programmers put comments at various places in colon
1671: definitions that describe the contents of the stack at that place (stack
1672: comments); i.e., they are like the first part of a stack-effect
1673: comment. E.g.,
1674:
1675: @example
1676: : cubed ( n -- n^3 )
1677: dup squared ( n n^2 ) * ;
1678: @end example
1679:
1680: In this case the stack comment is pretty superfluous, because the word
1681: is simple enough. If you think it would be a good idea to add such a
1682: comment to increase readability, you should also consider factoring the
1683: word into several simpler words (@pxref{Factoring Tutorial,,
1684: Factoring}), which typically eliminates the need for the stack comment;
1685: however, if you decide not to refactor it, then having such a comment is
1686: better than not having it.
1687:
1688: The names of the stack items in stack-effect and stack comments in the
1689: standard, in this manual, and in many programs specify the type through
1690: a type prefix, similar to Fortran and Hungarian notation. The most
1691: frequent prefixes are:
1692:
1693: @table @code
1694: @item n
1695: signed integer
1696: @item u
1697: unsigned integer
1698: @item c
1699: character
1700: @item f
1701: Boolean flags, i.e. @code{false} or @code{true}.
1702: @item a-addr,a-
1703: Cell-aligned address
1704: @item c-addr,c-
1705: Char-aligned address (note that a Char may have two bytes in Windows NT)
1706: @item xt
1707: Execution token, same size as Cell
1708: @item w,x
1709: Cell, can contain an integer or an address. It usually takes 32, 64 or
1710: 16 bits (depending on your platform and Forth system). A cell is more
1711: commonly known as machine word, but the term @emph{word} already means
1712: something different in Forth.
1713: @item d
1714: signed double-cell integer
1715: @item ud
1716: unsigned double-cell integer
1717: @item r
1718: Float (on the FP stack)
1719: @end table
1720:
1721: You can find a more complete list in @ref{Notation}.
1722:
1723: @quotation Assignment
1724: Write stack-effect comments for all definitions you have written up to
1725: now.
1726: @end quotation
1727:
1728:
1729: @node Types Tutorial, Factoring Tutorial, Stack-Effect Comments Tutorial, Tutorial
1730: @section Types
1731: @cindex types tutorial
1732:
1733: In Forth the names of the operations are not overloaded; so similar
1734: operations on different types need different names; e.g., @code{+} adds
1735: integers, and you have to use @code{f+} to add floating-point numbers.
1736: The following prefixes are often used for related operations on
1737: different types:
1738:
1739: @table @code
1740: @item (none)
1741: signed integer
1742: @item u
1743: unsigned integer
1744: @item c
1745: character
1746: @item d
1747: signed double-cell integer
1748: @item ud, du
1749: unsigned double-cell integer
1750: @item 2
1751: two cells (not-necessarily double-cell numbers)
1752: @item m, um
1753: mixed single-cell and double-cell operations
1754: @item f
1755: floating-point (note that in stack comments @samp{f} represents flags,
1756: and @samp{r} represents FP numbers; also, you need to include the
1757: exponent part in literal FP numbers, @pxref{Floating Point Tutorial}).
1758: @end table
1759:
1760: If there are no differences between the signed and the unsigned variant
1761: (e.g., for @code{+}), there is only the prefix-less variant.
1762:
1763: Forth does not perform type checking, neither at compile time, nor at
1764: run time. If you use the wrong operation, the data are interpreted
1765: incorrectly:
1766:
1767: @example
1768: -1 u.
1769: @end example
1770:
1771: If you have only experience with type-checked languages until now, and
1772: have heard how important type-checking is, don't panic! In my
1773: experience (and that of other Forthers), type errors in Forth code are
1774: usually easy to find (once you get used to it), the increased vigilance
1775: of the programmer tends to catch some harder errors in addition to most
1776: type errors, and you never have to work around the type system, so in
1777: most situations the lack of type-checking seems to be a win (projects to
1778: add type checking to Forth have not caught on).
1779:
1780:
1781: @node Factoring Tutorial, Designing the stack effect Tutorial, Types Tutorial, Tutorial
1782: @section Factoring
1783: @cindex factoring tutorial
1784:
1785: If you try to write longer definitions, you will soon find it hard to
1786: keep track of the stack contents. Therefore, good Forth programmers
1787: tend to write only short definitions (e.g., three lines). The art of
1788: finding meaningful short definitions is known as factoring (as in
1789: factoring polynomials).
1790:
1791: Well-factored programs offer additional advantages: smaller, more
1792: general words, are easier to test and debug and can be reused more and
1793: better than larger, specialized words.
1794:
1795: So, if you run into difficulties with stack management, when writing
1796: code, try to define meaningful factors for the word, and define the word
1797: in terms of those. Even if a factor contains only two words, it is
1798: often helpful.
1799:
1800: Good factoring is not easy, and it takes some practice to get the knack
1801: for it; but even experienced Forth programmers often don't find the
1802: right solution right away, but only when rewriting the program. So, if
1803: you don't come up with a good solution immediately, keep trying, don't
1804: despair.
1805:
1806: @c example !!
1807:
1808:
1809: @node Designing the stack effect Tutorial, Local Variables Tutorial, Factoring Tutorial, Tutorial
1810: @section Designing the stack effect
1811: @cindex Stack effect design, tutorial
1812: @cindex design of stack effects, tutorial
1813:
1814: In other languages you can use an arbitrary order of parameters for a
1815: function; and since there is only one result, you don't have to deal with
1816: the order of results, either.
1817:
1818: In Forth (and other stack-based languages, e.g., PostScript) the
1819: parameter and result order of a definition is important and should be
1820: designed well. The general guideline is to design the stack effect such
1821: that the word is simple to use in most cases, even if that complicates
1822: the implementation of the word. Some concrete rules are:
1823:
1824: @itemize @bullet
1825:
1826: @item
1827: Words consume all of their parameters (e.g., @code{.}).
1828:
1829: @item
1830: If there is a convention on the order of parameters (e.g., from
1831: mathematics or another programming language), stick with it (e.g.,
1832: @code{-}).
1833:
1834: @item
1835: If one parameter usually requires only a short computation (e.g., it is
1836: a constant), pass it on the top of the stack. Conversely, parameters
1837: that usually require a long sequence of code to compute should be passed
1838: as the bottom (i.e., first) parameter. This makes the code easier to
1839: read, because the reader does not need to keep track of the bottom item
1840: through a long sequence of code (or, alternatively, through stack
1841: manipulations). E.g., @code{!} (store, @pxref{Memory}) expects the
1842: address on top of the stack because it is usually simpler to compute
1843: than the stored value (often the address is just a variable).
1844:
1845: @item
1846: Similarly, results that are usually consumed quickly should be returned
1847: on the top of stack, whereas a result that is often used in long
1848: computations should be passed as bottom result. E.g., the file words
1849: like @code{open-file} return the error code on the top of stack, because
1850: it is usually consumed quickly by @code{throw}; moreover, the error code
1851: has to be checked before doing anything with the other results.
1852:
1853: @end itemize
1854:
1855: These rules are just general guidelines, don't lose sight of the overall
1856: goal to make the words easy to use. E.g., if the convention rule
1857: conflicts with the computation-length rule, you might decide in favour
1858: of the convention if the word will be used rarely, and in favour of the
1859: computation-length rule if the word will be used frequently (because
1860: with frequent use the cost of breaking the computation-length rule would
1861: be quite high, and frequent use makes it easier to remember an
1862: unconventional order).
1863:
1864: @c example !! structure package
1865:
1866:
1867: @node Local Variables Tutorial, Conditional execution Tutorial, Designing the stack effect Tutorial, Tutorial
1868: @section Local Variables
1869: @cindex local variables, tutorial
1870:
1871: You can define local variables (@emph{locals}) in a colon definition:
1872:
1873: @example
1874: : swap @{ a b -- b a @}
1875: b a ;
1876: 1 2 swap .s 2drop
1877: @end example
1878:
1879: (If your Forth system does not support this syntax, include
1880: @file{compat/anslocal.fs} first).
1881:
1882: In this example @code{@{ a b -- b a @}} is the locals definition; it
1883: takes two cells from the stack, puts the top of stack in @code{b} and
1884: the next stack element in @code{a}. @code{--} starts a comment ending
1885: with @code{@}}. After the locals definition, using the name of the
1886: local will push its value on the stack. You can leave the comment
1887: part (@code{-- b a}) away:
1888:
1889: @example
1890: : swap ( x1 x2 -- x2 x1 )
1891: @{ a b @} b a ;
1892: @end example
1893:
1894: In Gforth you can have several locals definitions, anywhere in a colon
1895: definition; in contrast, in a standard program you can have only one
1896: locals definition per colon definition, and that locals definition must
1897: be outside any control structure.
1898:
1899: With locals you can write slightly longer definitions without running
1900: into stack trouble. However, I recommend trying to write colon
1901: definitions without locals for exercise purposes to help you gain the
1902: essential factoring skills.
1903:
1904: @quotation Assignment
1905: Rewrite your definitions until now with locals
1906: @end quotation
1907:
1908: Reference: @ref{Locals}.
1909:
1910:
1911: @node Conditional execution Tutorial, Flags and Comparisons Tutorial, Local Variables Tutorial, Tutorial
1912: @section Conditional execution
1913: @cindex conditionals, tutorial
1914: @cindex if, tutorial
1915:
1916: In Forth you can use control structures only inside colon definitions.
1917: An @code{if}-structure looks like this:
1918:
1919: @example
1920: : abs ( n1 -- +n2 )
1921: dup 0 < if
1922: negate
1923: endif ;
1924: 5 abs .
1925: -5 abs .
1926: @end example
1927:
1928: @code{if} takes a flag from the stack. If the flag is non-zero (true),
1929: the following code is performed, otherwise execution continues after the
1930: @code{endif} (or @code{else}). @code{<} compares the top two stack
1931: elements and produces a flag:
1932:
1933: @example
1934: 1 2 < .
1935: 2 1 < .
1936: 1 1 < .
1937: @end example
1938:
1939: Actually the standard name for @code{endif} is @code{then}. This
1940: tutorial presents the examples using @code{endif}, because this is often
1941: less confusing for people familiar with other programming languages
1942: where @code{then} has a different meaning. If your system does not have
1943: @code{endif}, define it with
1944:
1945: @example
1946: : endif postpone then ; immediate
1947: @end example
1948:
1949: You can optionally use an @code{else}-part:
1950:
1951: @example
1952: : min ( n1 n2 -- n )
1953: 2dup < if
1954: drop
1955: else
1956: nip
1957: endif ;
1958: 2 3 min .
1959: 3 2 min .
1960: @end example
1961:
1962: @quotation Assignment
1963: Write @code{min} without @code{else}-part (hint: what's the definition
1964: of @code{nip}?).
1965: @end quotation
1966:
1967: Reference: @ref{Selection}.
1968:
1969:
1970: @node Flags and Comparisons Tutorial, General Loops Tutorial, Conditional execution Tutorial, Tutorial
1971: @section Flags and Comparisons
1972: @cindex flags tutorial
1973: @cindex comparison tutorial
1974:
1975: In a false-flag all bits are clear (0 when interpreted as integer). In
1976: a canonical true-flag all bits are set (-1 as a twos-complement signed
1977: integer); in many contexts (e.g., @code{if}) any non-zero value is
1978: treated as true flag.
1979:
1980: @example
1981: false .
1982: true .
1983: true hex u. decimal
1984: @end example
1985:
1986: Comparison words produce canonical flags:
1987:
1988: @example
1989: 1 1 = .
1990: 1 0= .
1991: 0 1 < .
1992: 0 0 < .
1993: -1 1 u< . \ type error, u< interprets -1 as large unsigned number
1994: -1 1 < .
1995: @end example
1996:
1997: Gforth supports all combinations of the prefixes @code{0 u d d0 du f f0}
1998: (or none) and the comparisons @code{= <> < > <= >=}. Only a part of
1999: these combinations are standard (for details see the standard,
2000: @ref{Numeric comparison}, @ref{Floating Point} or @ref{Word Index}).
2001:
2002: You can use @code{and or xor invert} as operations on canonical flags.
2003: Actually they are bitwise operations:
2004:
2005: @example
2006: 1 2 and .
2007: 1 2 or .
2008: 1 3 xor .
2009: 1 invert .
2010: @end example
2011:
2012: You can convert a zero/non-zero flag into a canonical flag with
2013: @code{0<>} (and complement it on the way with @code{0=}).
2014:
2015: @example
2016: 1 0= .
2017: 1 0<> .
2018: @end example
2019:
2020: You can use the all-bits-set feature of canonical flags and the bitwise
2021: operation of the Boolean operations to avoid @code{if}s:
2022:
2023: @example
2024: : foo ( n1 -- n2 )
2025: 0= if
2026: 14
2027: else
2028: 0
2029: endif ;
2030: 0 foo .
2031: 1 foo .
2032:
2033: : foo ( n1 -- n2 )
2034: 0= 14 and ;
2035: 0 foo .
2036: 1 foo .
2037: @end example
2038:
2039: @quotation Assignment
2040: Write @code{min} without @code{if}.
2041: @end quotation
2042:
2043: For reference, see @ref{Boolean Flags}, @ref{Numeric comparison}, and
2044: @ref{Bitwise operations}.
2045:
2046:
2047: @node General Loops Tutorial, Counted loops Tutorial, Flags and Comparisons Tutorial, Tutorial
2048: @section General Loops
2049: @cindex loops, indefinite, tutorial
2050:
2051: The endless loop is the most simple one:
2052:
2053: @example
2054: : endless ( -- )
2055: 0 begin
2056: dup . 1+
2057: again ;
2058: endless
2059: @end example
2060:
2061: Terminate this loop by pressing @kbd{Ctrl-C} (in Gforth). @code{begin}
2062: does nothing at run-time, @code{again} jumps back to @code{begin}.
2063:
2064: A loop with one exit at any place looks like this:
2065:
2066: @example
2067: : log2 ( +n1 -- n2 )
2068: \ logarithmus dualis of n1>0, rounded down to the next integer
2069: assert( dup 0> )
2070: 2/ 0 begin
2071: over 0> while
2072: 1+ swap 2/ swap
2073: repeat
2074: nip ;
2075: 7 log2 .
2076: 8 log2 .
2077: @end example
2078:
2079: At run-time @code{while} consumes a flag; if it is 0, execution
2080: continues behind the @code{repeat}; if the flag is non-zero, execution
2081: continues behind the @code{while}. @code{Repeat} jumps back to
2082: @code{begin}, just like @code{again}.
2083:
2084: In Forth there are a number of combinations/abbreviations, like
2085: @code{1+}. However, @code{2/} is not one of them; it shifts its
2086: argument right by one bit (arithmetic shift right), and viewed as
2087: division that always rounds towards negative infinity (floored
2088: division). In contrast, @code{/} rounds towards zero on some systems
2089: (not on default installations of gforth (>=0.7.0), however).
2090:
2091: @example
2092: -5 2 / . \ -2 or -3
2093: -5 2/ . \ -3
2094: @end example
2095:
2096: @code{assert(} is no standard word, but you can get it on systems other
2097: than Gforth by including @file{compat/assert.fs}. You can see what it
2098: does by trying
2099:
2100: @example
2101: 0 log2 .
2102: @end example
2103:
2104: Here's a loop with an exit at the end:
2105:
2106: @example
2107: : log2 ( +n1 -- n2 )
2108: \ logarithmus dualis of n1>0, rounded down to the next integer
2109: assert( dup 0 > )
2110: -1 begin
2111: 1+ swap 2/ swap
2112: over 0 <=
2113: until
2114: nip ;
2115: @end example
2116:
2117: @code{Until} consumes a flag; if it is non-zero, execution continues at
2118: the @code{begin}, otherwise after the @code{until}.
2119:
2120: @quotation Assignment
2121: Write a definition for computing the greatest common divisor.
2122: @end quotation
2123:
2124: Reference: @ref{Simple Loops}.
2125:
2126:
2127: @node Counted loops Tutorial, Recursion Tutorial, General Loops Tutorial, Tutorial
2128: @section Counted loops
2129: @cindex loops, counted, tutorial
2130:
2131: @example
2132: : ^ ( n1 u -- n )
2133: \ n = the uth power of n1
2134: 1 swap 0 u+do
2135: over *
2136: loop
2137: nip ;
2138: 3 2 ^ .
2139: 4 3 ^ .
2140: @end example
2141:
2142: @code{U+do} (from @file{compat/loops.fs}, if your Forth system doesn't
2143: have it) takes two numbers of the stack @code{( u3 u4 -- )}, and then
2144: performs the code between @code{u+do} and @code{loop} for @code{u3-u4}
2145: times (or not at all, if @code{u3-u4<0}).
2146:
2147: You can see the stack effect design rules at work in the stack effect of
2148: the loop start words: Since the start value of the loop is more
2149: frequently constant than the end value, the start value is passed on
2150: the top-of-stack.
2151:
2152: You can access the counter of a counted loop with @code{i}:
2153:
2154: @example
2155: : fac ( u -- u! )
2156: 1 swap 1+ 1 u+do
2157: i *
2158: loop ;
2159: 5 fac .
2160: 7 fac .
2161: @end example
2162:
2163: There is also @code{+do}, which expects signed numbers (important for
2164: deciding whether to enter the loop).
2165:
2166: @quotation Assignment
2167: Write a definition for computing the nth Fibonacci number.
2168: @end quotation
2169:
2170: You can also use increments other than 1:
2171:
2172: @example
2173: : up2 ( n1 n2 -- )
2174: +do
2175: i .
2176: 2 +loop ;
2177: 10 0 up2
2178:
2179: : down2 ( n1 n2 -- )
2180: -do
2181: i .
2182: 2 -loop ;
2183: 0 10 down2
2184: @end example
2185:
2186: Reference: @ref{Counted Loops}.
2187:
2188:
2189: @node Recursion Tutorial, Leaving definitions or loops Tutorial, Counted loops Tutorial, Tutorial
2190: @section Recursion
2191: @cindex recursion tutorial
2192:
2193: Usually the name of a definition is not visible in the definition; but
2194: earlier definitions are usually visible:
2195:
2196: @example
2197: 1 0 / . \ "Floating-point unidentified fault" in Gforth on some platforms
2198: : / ( n1 n2 -- n )
2199: dup 0= if
2200: -10 throw \ report division by zero
2201: endif
2202: / \ old version
2203: ;
2204: 1 0 /
2205: @end example
2206:
2207: For recursive definitions you can use @code{recursive} (non-standard) or
2208: @code{recurse}:
2209:
2210: @example
2211: : fac1 ( n -- n! ) recursive
2212: dup 0> if
2213: dup 1- fac1 *
2214: else
2215: drop 1
2216: endif ;
2217: 7 fac1 .
2218:
2219: : fac2 ( n -- n! )
2220: dup 0> if
2221: dup 1- recurse *
2222: else
2223: drop 1
2224: endif ;
2225: 8 fac2 .
2226: @end example
2227:
2228: @quotation Assignment
2229: Write a recursive definition for computing the nth Fibonacci number.
2230: @end quotation
2231:
2232: Reference (including indirect recursion): @xref{Calls and returns}.
2233:
2234:
2235: @node Leaving definitions or loops Tutorial, Return Stack Tutorial, Recursion Tutorial, Tutorial
2236: @section Leaving definitions or loops
2237: @cindex leaving definitions, tutorial
2238: @cindex leaving loops, tutorial
2239:
2240: @code{EXIT} exits the current definition right away. For every counted
2241: loop that is left in this way, an @code{UNLOOP} has to be performed
2242: before the @code{EXIT}:
2243:
2244: @c !! real examples
2245: @example
2246: : ...
2247: ... u+do
2248: ... if
2249: ... unloop exit
2250: endif
2251: ...
2252: loop
2253: ... ;
2254: @end example
2255:
2256: @code{LEAVE} leaves the innermost counted loop right away:
2257:
2258: @example
2259: : ...
2260: ... u+do
2261: ... if
2262: ... leave
2263: endif
2264: ...
2265: loop
2266: ... ;
2267: @end example
2268:
2269: @c !! example
2270:
2271: Reference: @ref{Calls and returns}, @ref{Counted Loops}.
2272:
2273:
2274: @node Return Stack Tutorial, Memory Tutorial, Leaving definitions or loops Tutorial, Tutorial
2275: @section Return Stack
2276: @cindex return stack tutorial
2277:
2278: In addition to the data stack Forth also has a second stack, the return
2279: stack; most Forth systems store the return addresses of procedure calls
2280: there (thus its name). Programmers can also use this stack:
2281:
2282: @example
2283: : foo ( n1 n2 -- )
2284: .s
2285: >r .s
2286: r@@ .
2287: >r .s
2288: r@@ .
2289: r> .
2290: r@@ .
2291: r> . ;
2292: 1 2 foo
2293: @end example
2294:
2295: @code{>r} takes an element from the data stack and pushes it onto the
2296: return stack; conversely, @code{r>} moves an elementm from the return to
2297: the data stack; @code{r@@} pushes a copy of the top of the return stack
2298: on the data stack.
2299:
2300: Forth programmers usually use the return stack for storing data
2301: temporarily, if using the data stack alone would be too complex, and
2302: factoring and locals are not an option:
2303:
2304: @example
2305: : 2swap ( x1 x2 x3 x4 -- x3 x4 x1 x2 )
2306: rot >r rot r> ;
2307: @end example
2308:
2309: The return address of the definition and the loop control parameters of
2310: counted loops usually reside on the return stack, so you have to take
2311: all items, that you have pushed on the return stack in a colon
2312: definition or counted loop, from the return stack before the definition
2313: or loop ends. You cannot access items that you pushed on the return
2314: stack outside some definition or loop within the definition of loop.
2315:
2316: If you miscount the return stack items, this usually ends in a crash:
2317:
2318: @example
2319: : crash ( n -- )
2320: >r ;
2321: 5 crash
2322: @end example
2323:
2324: You cannot mix using locals and using the return stack (according to the
2325: standard; Gforth has no problem). However, they solve the same
2326: problems, so this shouldn't be an issue.
2327:
2328: @quotation Assignment
2329: Can you rewrite any of the definitions you wrote until now in a better
2330: way using the return stack?
2331: @end quotation
2332:
2333: Reference: @ref{Return stack}.
2334:
2335:
2336: @node Memory Tutorial, Characters and Strings Tutorial, Return Stack Tutorial, Tutorial
2337: @section Memory
2338: @cindex memory access/allocation tutorial
2339:
2340: You can create a global variable @code{v} with
2341:
2342: @example
2343: variable v ( -- addr )
2344: @end example
2345:
2346: @code{v} pushes the address of a cell in memory on the stack. This cell
2347: was reserved by @code{variable}. You can use @code{!} (store) to store
2348: values into this cell and @code{@@} (fetch) to load the value from the
2349: stack into memory:
2350:
2351: @example
2352: v .
2353: 5 v ! .s
2354: v @@ .
2355: @end example
2356:
2357: You can see a raw dump of memory with @code{dump}:
2358:
2359: @example
2360: v 1 cells .s dump
2361: @end example
2362:
2363: @code{Cells ( n1 -- n2 )} gives you the number of bytes (or, more
2364: generally, address units (aus)) that @code{n1 cells} occupy. You can
2365: also reserve more memory:
2366:
2367: @example
2368: create v2 20 cells allot
2369: v2 20 cells dump
2370: @end example
2371:
2372: creates a variable-like word @code{v2} and reserves 20 uninitialized
2373: cells; the address pushed by @code{v2} points to the start of these 20
2374: cells (@pxref{CREATE}). You can use address arithmetic to access
2375: these cells:
2376:
2377: @example
2378: 3 v2 5 cells + !
2379: v2 20 cells dump
2380: @end example
2381:
2382: You can reserve and initialize memory with @code{,}:
2383:
2384: @example
2385: create v3
2386: 5 , 4 , 3 , 2 , 1 ,
2387: v3 @@ .
2388: v3 cell+ @@ .
2389: v3 2 cells + @@ .
2390: v3 5 cells dump
2391: @end example
2392:
2393: @quotation Assignment
2394: Write a definition @code{vsum ( addr u -- n )} that computes the sum of
2395: @code{u} cells, with the first of these cells at @code{addr}, the next
2396: one at @code{addr cell+} etc.
2397: @end quotation
2398:
2399: You can also reserve memory without creating a new word:
2400:
2401: @example
2402: here 10 cells allot .
2403: here .
2404: @end example
2405:
2406: The first @code{here} pushes the start address of the memory area, the
2407: second @code{here} the address after the dictionary area. You should
2408: store the start address somewhere, or you will have a hard time
2409: finding the memory area again.
2410:
2411: @code{Allot} manages dictionary memory. The dictionary memory contains
2412: the system's data structures for words etc. on Gforth and most other
2413: Forth systems. It is managed like a stack: You can free the memory that
2414: you have just @code{allot}ed with
2415:
2416: @example
2417: -10 cells allot
2418: here .
2419: @end example
2420:
2421: Note that you cannot do this if you have created a new word in the
2422: meantime (because then your @code{allot}ed memory is no longer on the
2423: top of the dictionary ``stack'').
2424:
2425: Alternatively, you can use @code{allocate} and @code{free} which allow
2426: freeing memory in any order:
2427:
2428: @example
2429: 10 cells allocate throw .s
2430: 20 cells allocate throw .s
2431: swap
2432: free throw
2433: free throw
2434: @end example
2435:
2436: The @code{throw}s deal with errors (e.g., out of memory).
2437:
2438: And there is also a
2439: @uref{http://www.complang.tuwien.ac.at/forth/garbage-collection.zip,
2440: garbage collector}, which eliminates the need to @code{free} memory
2441: explicitly.
2442:
2443: Reference: @ref{Memory}.
2444:
2445:
2446: @node Characters and Strings Tutorial, Alignment Tutorial, Memory Tutorial, Tutorial
2447: @section Characters and Strings
2448: @cindex strings tutorial
2449: @cindex characters tutorial
2450:
2451: On the stack characters take up a cell, like numbers. In memory they
2452: have their own size (one 8-bit byte on most systems), and therefore
2453: require their own words for memory access:
2454:
2455: @example
2456: create v4
2457: 104 c, 97 c, 108 c, 108 c, 111 c,
2458: v4 4 chars + c@@ .
2459: v4 5 chars dump
2460: @end example
2461:
2462: The preferred representation of strings on the stack is @code{addr
2463: u-count}, where @code{addr} is the address of the first character and
2464: @code{u-count} is the number of characters in the string.
2465:
2466: @example
2467: v4 5 type
2468: @end example
2469:
2470: You get a string constant with
2471:
2472: @example
2473: s" hello, world" .s
2474: type
2475: @end example
2476:
2477: Make sure you have a space between @code{s"} and the string; @code{s"}
2478: is a normal Forth word and must be delimited with white space (try what
2479: happens when you remove the space).
2480:
2481: However, this interpretive use of @code{s"} is quite restricted: the
2482: string exists only until the next call of @code{s"} (some Forth systems
2483: keep more than one of these strings, but usually they still have a
2484: limited lifetime).
2485:
2486: @example
2487: s" hello," s" world" .s
2488: type
2489: type
2490: @end example
2491:
2492: You can also use @code{s"} in a definition, and the resulting
2493: strings then live forever (well, for as long as the definition):
2494:
2495: @example
2496: : foo s" hello," s" world" ;
2497: foo .s
2498: type
2499: type
2500: @end example
2501:
2502: @quotation Assignment
2503: @code{Emit ( c -- )} types @code{c} as character (not a number).
2504: Implement @code{type ( addr u -- )}.
2505: @end quotation
2506:
2507: Reference: @ref{Memory Blocks}.
2508:
2509:
2510: @node Alignment Tutorial, Floating Point Tutorial, Characters and Strings Tutorial, Tutorial
2511: @section Alignment
2512: @cindex alignment tutorial
2513: @cindex memory alignment tutorial
2514:
2515: On many processors cells have to be aligned in memory, if you want to
2516: access them with @code{@@} and @code{!} (and even if the processor does
2517: not require alignment, access to aligned cells is faster).
2518:
2519: @code{Create} aligns @code{here} (i.e., the place where the next
2520: allocation will occur, and that the @code{create}d word points to).
2521: Likewise, the memory produced by @code{allocate} starts at an aligned
2522: address. Adding a number of @code{cells} to an aligned address produces
2523: another aligned address.
2524:
2525: However, address arithmetic involving @code{char+} and @code{chars} can
2526: create an address that is not cell-aligned. @code{Aligned ( addr --
2527: a-addr )} produces the next aligned address:
2528:
2529: @example
2530: v3 char+ aligned .s @@ .
2531: v3 char+ .s @@ .
2532: @end example
2533:
2534: Similarly, @code{align} advances @code{here} to the next aligned
2535: address:
2536:
2537: @example
2538: create v5 97 c,
2539: here .
2540: align here .
2541: 1000 ,
2542: @end example
2543:
2544: Note that you should use aligned addresses even if your processor does
2545: not require them, if you want your program to be portable.
2546:
2547: Reference: @ref{Address arithmetic}.
2548:
2549: @node Floating Point Tutorial, Files Tutorial, Alignment Tutorial, Tutorial
2550: @section Floating Point
2551: @cindex floating point tutorial
2552: @cindex FP tutorial
2553:
2554: Floating-point (FP) numbers and arithmetic in Forth works mostly as one
2555: might expect, but there are a few things worth noting:
2556:
2557: The first point is not specific to Forth, but so important and yet not
2558: universally known that I mention it here: FP numbers are not reals.
2559: Many properties (e.g., arithmetic laws) that reals have and that one
2560: expects of all kinds of numbers do not hold for FP numbers. If you
2561: want to use FP computations, you should learn about their problems and
2562: how to avoid them; a good starting point is @cite{David Goldberg,
2563: @uref{http://docs.sun.com/source/806-3568/ncg_goldberg.html,What Every
2564: Computer Scientist Should Know About Floating-Point Arithmetic}, ACM
2565: Computing Surveys 23(1):5@minus{}48, March 1991}.
2566:
2567: In Forth source code literal FP numbers need an exponent, e.g.,
2568: @code{1e0}; this can also be written shorter as @code{1e}, longer as
2569: @code{+1.0e+0}, and many variations in between. The reason for this is
2570: that, for historical reasons, Forth interprets a decimal point alone
2571: (e.g., @code{1.}) as indicating a double-cell integer. Examples:
2572:
2573: @example
2574: 2e 2e f+ f.
2575: @end example
2576:
2577: Another requirement for literal FP numbers is that the current base is
2578: decimal; with a hex base @code{1e} is interpreted as an integer.
2579:
2580: Forth has a separate stack for FP numbers.@footnote{Theoretically, an
2581: ANS Forth system may implement the FP stack on the data stack, but
2582: virtually all systems implement a separate FP stack; and programming
2583: in a way that accommodates all models is so cumbersome that nobody
2584: does it.} One advantage of this model is that cells are not in the
2585: way when accessing FP values, and vice versa. Forth has a set of
2586: words for manipulating the FP stack: @code{fdup fswap fdrop fover
2587: frot} and (non-standard) @code{fnip ftuck fpick}.
2588:
2589: FP arithmetic words are prefixed with @code{F}. There is the usual
2590: set @code{f+ f- f* f/ f** fnegate} as well as a number of words for
2591: other functions, e.g., @code{fsqrt fsin fln fmin}. One word that you
2592: might expect is @code{f=}; but @code{f=} is non-standard, because FP
2593: computation results are usually inaccurate, so exact comparison is
2594: usually a mistake, and one should use approximate comparison.
2595: Unfortunately, @code{f~}, the standard word for that purpose, is not
2596: well designed, so Gforth provides @code{f~abs} and @code{f~rel} as
2597: well.
2598:
2599: And of course there are words for accessing FP numbers in memory
2600: (@code{f@@ f!}), and for address arithmetic (@code{floats float+
2601: faligned}). There are also variants of these words with an @code{sf}
2602: and @code{df} prefix for accessing IEEE format single-precision and
2603: double-precision numbers in memory; their main purpose is for
2604: accessing external FP data (e.g., that has been read from or will be
2605: written to a file).
2606:
2607: Here is an example of a dot-product word and its use:
2608:
2609: @example
2610: : v* ( f_addr1 nstride1 f_addr2 nstride2 ucount -- r )
2611: >r swap 2swap swap 0e r> 0 ?DO
2612: dup f@@ over + 2swap dup f@@ f* f+ over + 2swap
2613: LOOP
2614: 2drop 2drop ;
2615:
2616: create v 1.23e f, 4.56e f, 7.89e f,
2617:
2618: v 1 floats v 1 floats 3 v* f.
2619: @end example
2620:
2621: @quotation Assignment
2622: Write a program to solve a quadratic equation. Then read @cite{Henry
2623: G. Baker,
2624: @uref{http://home.pipeline.com/~hbaker1/sigplannotices/sigcol05.ps.gz,You
2625: Could Learn a Lot from a Quadratic}, ACM SIGPLAN Notices,
2626: 33(1):30@minus{}39, January 1998}, and see if you can improve your
2627: program. Finally, find a test case where the original and the
2628: improved version produce different results.
2629: @end quotation
2630:
2631: Reference: @ref{Floating Point}; @ref{Floating point stack};
2632: @ref{Number Conversion}; @ref{Memory Access}; @ref{Address
2633: arithmetic}.
2634:
2635: @node Files Tutorial, Interpretation and Compilation Semantics and Immediacy Tutorial, Floating Point Tutorial, Tutorial
2636: @section Files
2637: @cindex files tutorial
2638:
2639: This section gives a short introduction into how to use files inside
2640: Forth. It's broken up into five easy steps:
2641:
2642: @enumerate 1
2643: @item Opened an ASCII text file for input
2644: @item Opened a file for output
2645: @item Read input file until string matched (or some other condition matched)
2646: @item Wrote some lines from input ( modified or not) to output
2647: @item Closed the files.
2648: @end enumerate
2649:
2650: Reference: @ref{General files}.
2651:
2652: @subsection Open file for input
2653:
2654: @example
2655: s" foo.in" r/o open-file throw Value fd-in
2656: @end example
2657:
2658: @subsection Create file for output
2659:
2660: @example
2661: s" foo.out" w/o create-file throw Value fd-out
2662: @end example
2663:
2664: The available file modes are r/o for read-only access, r/w for
2665: read-write access, and w/o for write-only access. You could open both
2666: files with r/w, too, if you like. All file words return error codes; for
2667: most applications, it's best to pass there error codes with @code{throw}
2668: to the outer error handler.
2669:
2670: If you want words for opening and assigning, define them as follows:
2671:
2672: @example
2673: 0 Value fd-in
2674: 0 Value fd-out
2675: : open-input ( addr u -- ) r/o open-file throw to fd-in ;
2676: : open-output ( addr u -- ) w/o create-file throw to fd-out ;
2677: @end example
2678:
2679: Usage example:
2680:
2681: @example
2682: s" foo.in" open-input
2683: s" foo.out" open-output
2684: @end example
2685:
2686: @subsection Scan file for a particular line
2687:
2688: @example
2689: 256 Constant max-line
2690: Create line-buffer max-line 2 + allot
2691:
2692: : scan-file ( addr u -- )
2693: begin
2694: line-buffer max-line fd-in read-line throw
2695: while
2696: >r 2dup line-buffer r> compare 0=
2697: until
2698: else
2699: drop
2700: then
2701: 2drop ;
2702: @end example
2703:
2704: @code{read-line ( addr u1 fd -- u2 flag ior )} reads up to u1 bytes into
2705: the buffer at addr, and returns the number of bytes read, a flag that is
2706: false when the end of file is reached, and an error code.
2707:
2708: @code{compare ( addr1 u1 addr2 u2 -- n )} compares two strings and
2709: returns zero if both strings are equal. It returns a positive number if
2710: the first string is lexically greater, a negative if the second string
2711: is lexically greater.
2712:
2713: We haven't seen this loop here; it has two exits. Since the @code{while}
2714: exits with the number of bytes read on the stack, we have to clean up
2715: that separately; that's after the @code{else}.
2716:
2717: Usage example:
2718:
2719: @example
2720: s" The text I search is here" scan-file
2721: @end example
2722:
2723: @subsection Copy input to output
2724:
2725: @example
2726: : copy-file ( -- )
2727: begin
2728: line-buffer max-line fd-in read-line throw
2729: while
2730: line-buffer swap fd-out write-line throw
2731: repeat ;
2732: @end example
2733: @c !! does not handle long lines, no newline at end of file
2734:
2735: @subsection Close files
2736:
2737: @example
2738: fd-in close-file throw
2739: fd-out close-file throw
2740: @end example
2741:
2742: Likewise, you can put that into definitions, too:
2743:
2744: @example
2745: : close-input ( -- ) fd-in close-file throw ;
2746: : close-output ( -- ) fd-out close-file throw ;
2747: @end example
2748:
2749: @quotation Assignment
2750: How could you modify @code{copy-file} so that it copies until a second line is
2751: matched? Can you write a program that extracts a section of a text file,
2752: given the line that starts and the line that terminates that section?
2753: @end quotation
2754:
2755: @node Interpretation and Compilation Semantics and Immediacy Tutorial, Execution Tokens Tutorial, Files Tutorial, Tutorial
2756: @section Interpretation and Compilation Semantics and Immediacy
2757: @cindex semantics tutorial
2758: @cindex interpretation semantics tutorial
2759: @cindex compilation semantics tutorial
2760: @cindex immediate, tutorial
2761:
2762: When a word is compiled, it behaves differently from being interpreted.
2763: E.g., consider @code{+}:
2764:
2765: @example
2766: 1 2 + .
2767: : foo + ;
2768: @end example
2769:
2770: These two behaviours are known as compilation and interpretation
2771: semantics. For normal words (e.g., @code{+}), the compilation semantics
2772: is to append the interpretation semantics to the currently defined word
2773: (@code{foo} in the example above). I.e., when @code{foo} is executed
2774: later, the interpretation semantics of @code{+} (i.e., adding two
2775: numbers) will be performed.
2776:
2777: However, there are words with non-default compilation semantics, e.g.,
2778: the control-flow words like @code{if}. You can use @code{immediate} to
2779: change the compilation semantics of the last defined word to be equal to
2780: the interpretation semantics:
2781:
2782: @example
2783: : [FOO] ( -- )
2784: 5 . ; immediate
2785:
2786: [FOO]
2787: : bar ( -- )
2788: [FOO] ;
2789: bar
2790: see bar
2791: @end example
2792:
2793: Two conventions to mark words with non-default compilation semantics are
2794: names with brackets (more frequently used) and to write them all in
2795: upper case (less frequently used).
2796:
2797: In Gforth (and many other systems) you can also remove the
2798: interpretation semantics with @code{compile-only} (the compilation
2799: semantics is derived from the original interpretation semantics):
2800:
2801: @example
2802: : flip ( -- )
2803: 6 . ; compile-only \ but not immediate
2804: flip
2805:
2806: : flop ( -- )
2807: flip ;
2808: flop
2809: @end example
2810:
2811: In this example the interpretation semantics of @code{flop} is equal to
2812: the original interpretation semantics of @code{flip}.
2813:
2814: The text interpreter has two states: in interpret state, it performs the
2815: interpretation semantics of words it encounters; in compile state, it
2816: performs the compilation semantics of these words.
2817:
2818: Among other things, @code{:} switches into compile state, and @code{;}
2819: switches back to interpret state. They contain the factors @code{]}
2820: (switch to compile state) and @code{[} (switch to interpret state), that
2821: do nothing but switch the state.
2822:
2823: @example
2824: : xxx ( -- )
2825: [ 5 . ]
2826: ;
2827:
2828: xxx
2829: see xxx
2830: @end example
2831:
2832: These brackets are also the source of the naming convention mentioned
2833: above.
2834:
2835: Reference: @ref{Interpretation and Compilation Semantics}.
2836:
2837:
2838: @node Execution Tokens Tutorial, Exceptions Tutorial, Interpretation and Compilation Semantics and Immediacy Tutorial, Tutorial
2839: @section Execution Tokens
2840: @cindex execution tokens tutorial
2841: @cindex XT tutorial
2842:
2843: @code{' word} gives you the execution token (XT) of a word. The XT is a
2844: cell representing the interpretation semantics of a word. You can
2845: execute this semantics with @code{execute}:
2846:
2847: @example
2848: ' + .s
2849: 1 2 rot execute .
2850: @end example
2851:
2852: The XT is similar to a function pointer in C. However, parameter
2853: passing through the stack makes it a little more flexible:
2854:
2855: @example
2856: : map-array ( ... addr u xt -- ... )
2857: \ executes xt ( ... x -- ... ) for every element of the array starting
2858: \ at addr and containing u elements
2859: @{ xt @}
2860: cells over + swap ?do
2861: i @@ xt execute
2862: 1 cells +loop ;
2863:
2864: create a 3 , 4 , 2 , -1 , 4 ,
2865: a 5 ' . map-array .s
2866: 0 a 5 ' + map-array .
2867: s" max-n" environment? drop .s
2868: a 5 ' min map-array .
2869: @end example
2870:
2871: You can use map-array with the XTs of words that consume one element
2872: more than they produce. In theory you can also use it with other XTs,
2873: but the stack effect then depends on the size of the array, which is
2874: hard to understand.
2875:
2876: Since XTs are cell-sized, you can store them in memory and manipulate
2877: them on the stack like other cells. You can also compile the XT into a
2878: word with @code{compile,}:
2879:
2880: @example
2881: : foo1 ( n1 n2 -- n )
2882: [ ' + compile, ] ;
2883: see foo
2884: @end example
2885:
2886: This is non-standard, because @code{compile,} has no compilation
2887: semantics in the standard, but it works in good Forth systems. For the
2888: broken ones, use
2889:
2890: @example
2891: : [compile,] compile, ; immediate
2892:
2893: : foo1 ( n1 n2 -- n )
2894: [ ' + ] [compile,] ;
2895: see foo
2896: @end example
2897:
2898: @code{'} is a word with default compilation semantics; it parses the
2899: next word when its interpretation semantics are executed, not during
2900: compilation:
2901:
2902: @example
2903: : foo ( -- xt )
2904: ' ;
2905: see foo
2906: : bar ( ... "word" -- ... )
2907: ' execute ;
2908: see bar
2909: 1 2 bar + .
2910: @end example
2911:
2912: You often want to parse a word during compilation and compile its XT so
2913: it will be pushed on the stack at run-time. @code{[']} does this:
2914:
2915: @example
2916: : xt-+ ( -- xt )
2917: ['] + ;
2918: see xt-+
2919: 1 2 xt-+ execute .
2920: @end example
2921:
2922: Many programmers tend to see @code{'} and the word it parses as one
2923: unit, and expect it to behave like @code{[']} when compiled, and are
2924: confused by the actual behaviour. If you are, just remember that the
2925: Forth system just takes @code{'} as one unit and has no idea that it is
2926: a parsing word (attempts to convenience programmers in this issue have
2927: usually resulted in even worse pitfalls, see
2928: @uref{http://www.complang.tuwien.ac.at/papers/ertl98.ps.gz,
2929: @code{State}-smartness---Why it is evil and How to Exorcise it}).
2930:
2931: Note that the state of the interpreter does not come into play when
2932: creating and executing XTs. I.e., even when you execute @code{'} in
2933: compile state, it still gives you the interpretation semantics. And
2934: whatever that state is, @code{execute} performs the semantics
2935: represented by the XT (i.e., for XTs produced with @code{'} the
2936: interpretation semantics).
2937:
2938: Reference: @ref{Tokens for Words}.
2939:
2940:
2941: @node Exceptions Tutorial, Defining Words Tutorial, Execution Tokens Tutorial, Tutorial
2942: @section Exceptions
2943: @cindex exceptions tutorial
2944:
2945: @code{throw ( n -- )} causes an exception unless n is zero.
2946:
2947: @example
2948: 100 throw .s
2949: 0 throw .s
2950: @end example
2951:
2952: @code{catch ( ... xt -- ... n )} behaves similar to @code{execute}, but
2953: it catches exceptions and pushes the number of the exception on the
2954: stack (or 0, if the xt executed without exception). If there was an
2955: exception, the stacks have the same depth as when entering @code{catch}:
2956:
2957: @example
2958: .s
2959: 3 0 ' / catch .s
2960: 3 2 ' / catch .s
2961: @end example
2962:
2963: @quotation Assignment
2964: Try the same with @code{execute} instead of @code{catch}.
2965: @end quotation
2966:
2967: @code{Throw} always jumps to the dynamically next enclosing
2968: @code{catch}, even if it has to leave several call levels to achieve
2969: this:
2970:
2971: @example
2972: : foo 100 throw ;
2973: : foo1 foo ." after foo" ;
2974: : bar ['] foo1 catch ;
2975: bar .
2976: @end example
2977:
2978: It is often important to restore a value upon leaving a definition, even
2979: if the definition is left through an exception. You can ensure this
2980: like this:
2981:
2982: @example
2983: : ...
2984: save-x
2985: ['] word-changing-x catch ( ... n )
2986: restore-x
2987: ( ... n ) throw ;
2988: @end example
2989:
2990: However, this is still not safe against, e.g., the user pressing
2991: @kbd{Ctrl-C} when execution is between the @code{catch} and
2992: @code{restore-x}.
2993:
2994: Gforth provides an alternative exception handling syntax that is safe
2995: against such cases: @code{try ... restore ... endtry}. If the code
2996: between @code{try} and @code{endtry} has an exception, the stack
2997: depths are restored, the exception number is pushed on the stack, and
2998: the execution continues right after @code{restore}.
2999:
3000: The safer equivalent to the restoration code above is
3001:
3002: @example
3003: : ...
3004: save-x
3005: try
3006: word-changing-x 0
3007: restore
3008: restore-x
3009: endtry
3010: throw ;
3011: @end example
3012:
3013: Reference: @ref{Exception Handling}.
3014:
3015:
3016: @node Defining Words Tutorial, Arrays and Records Tutorial, Exceptions Tutorial, Tutorial
3017: @section Defining Words
3018: @cindex defining words tutorial
3019: @cindex does> tutorial
3020: @cindex create...does> tutorial
3021:
3022: @c before semantics?
3023:
3024: @code{:}, @code{create}, and @code{variable} are definition words: They
3025: define other words. @code{Constant} is another definition word:
3026:
3027: @example
3028: 5 constant foo
3029: foo .
3030: @end example
3031:
3032: You can also use the prefixes @code{2} (double-cell) and @code{f}
3033: (floating point) with @code{variable} and @code{constant}.
3034:
3035: You can also define your own defining words. E.g.:
3036:
3037: @example
3038: : variable ( "name" -- )
3039: create 0 , ;
3040: @end example
3041:
3042: You can also define defining words that create words that do something
3043: other than just producing their address:
3044:
3045: @example
3046: : constant ( n "name" -- )
3047: create ,
3048: does> ( -- n )
3049: ( addr ) @@ ;
3050:
3051: 5 constant foo
3052: foo .
3053: @end example
3054:
3055: The definition of @code{constant} above ends at the @code{does>}; i.e.,
3056: @code{does>} replaces @code{;}, but it also does something else: It
3057: changes the last defined word such that it pushes the address of the
3058: body of the word and then performs the code after the @code{does>}
3059: whenever it is called.
3060:
3061: In the example above, @code{constant} uses @code{,} to store 5 into the
3062: body of @code{foo}. When @code{foo} executes, it pushes the address of
3063: the body onto the stack, then (in the code after the @code{does>})
3064: fetches the 5 from there.
3065:
3066: The stack comment near the @code{does>} reflects the stack effect of the
3067: defined word, not the stack effect of the code after the @code{does>}
3068: (the difference is that the code expects the address of the body that
3069: the stack comment does not show).
3070:
3071: You can use these definition words to do factoring in cases that involve
3072: (other) definition words. E.g., a field offset is always added to an
3073: address. Instead of defining
3074:
3075: @example
3076: 2 cells constant offset-field1
3077: @end example
3078:
3079: and using this like
3080:
3081: @example
3082: ( addr ) offset-field1 +
3083: @end example
3084:
3085: you can define a definition word
3086:
3087: @example
3088: : simple-field ( n "name" -- )
3089: create ,
3090: does> ( n1 -- n1+n )
3091: ( addr ) @@ + ;
3092: @end example
3093:
3094: Definition and use of field offsets now look like this:
3095:
3096: @example
3097: 2 cells simple-field field1
3098: create mystruct 4 cells allot
3099: mystruct .s field1 .s drop
3100: @end example
3101:
3102: If you want to do something with the word without performing the code
3103: after the @code{does>}, you can access the body of a @code{create}d word
3104: with @code{>body ( xt -- addr )}:
3105:
3106: @example
3107: : value ( n "name" -- )
3108: create ,
3109: does> ( -- n1 )
3110: @@ ;
3111: : to ( n "name" -- )
3112: ' >body ! ;
3113:
3114: 5 value foo
3115: foo .
3116: 7 to foo
3117: foo .
3118: @end example
3119:
3120: @quotation Assignment
3121: Define @code{defer ( "name" -- )}, which creates a word that stores an
3122: XT (at the start the XT of @code{abort}), and upon execution
3123: @code{execute}s the XT. Define @code{is ( xt "name" -- )} that stores
3124: @code{xt} into @code{name}, a word defined with @code{defer}. Indirect
3125: recursion is one application of @code{defer}.
3126: @end quotation
3127:
3128: Reference: @ref{User-defined Defining Words}.
3129:
3130:
3131: @node Arrays and Records Tutorial, POSTPONE Tutorial, Defining Words Tutorial, Tutorial
3132: @section Arrays and Records
3133: @cindex arrays tutorial
3134: @cindex records tutorial
3135: @cindex structs tutorial
3136:
3137: Forth has no standard words for defining data structures such as arrays
3138: and records (structs in C terminology), but you can build them yourself
3139: based on address arithmetic. You can also define words for defining
3140: arrays and records (@pxref{Defining Words Tutorial,, Defining Words}).
3141:
3142: One of the first projects a Forth newcomer sets out upon when learning
3143: about defining words is an array defining word (possibly for
3144: n-dimensional arrays). Go ahead and do it, I did it, too; you will
3145: learn something from it. However, don't be disappointed when you later
3146: learn that you have little use for these words (inappropriate use would
3147: be even worse). I have not found a set of useful array words yet;
3148: the needs are just too diverse, and named, global arrays (the result of
3149: naive use of defining words) are often not flexible enough (e.g.,
3150: consider how to pass them as parameters). Another such project is a set
3151: of words to help dealing with strings.
3152:
3153: On the other hand, there is a useful set of record words, and it has
3154: been defined in @file{compat/struct.fs}; these words are predefined in
3155: Gforth. They are explained in depth elsewhere in this manual (see
3156: @pxref{Structures}). The @code{simple-field} example above is
3157: simplified variant of fields in this package.
3158:
3159:
3160: @node POSTPONE Tutorial, Literal Tutorial, Arrays and Records Tutorial, Tutorial
3161: @section @code{POSTPONE}
3162: @cindex postpone tutorial
3163:
3164: You can compile the compilation semantics (instead of compiling the
3165: interpretation semantics) of a word with @code{POSTPONE}:
3166:
3167: @example
3168: : MY-+ ( Compilation: -- ; Run-time of compiled code: n1 n2 -- n )
3169: POSTPONE + ; immediate
3170: : foo ( n1 n2 -- n )
3171: MY-+ ;
3172: 1 2 foo .
3173: see foo
3174: @end example
3175:
3176: During the definition of @code{foo} the text interpreter performs the
3177: compilation semantics of @code{MY-+}, which performs the compilation
3178: semantics of @code{+}, i.e., it compiles @code{+} into @code{foo}.
3179:
3180: This example also displays separate stack comments for the compilation
3181: semantics and for the stack effect of the compiled code. For words with
3182: default compilation semantics these stack effects are usually not
3183: displayed; the stack effect of the compilation semantics is always
3184: @code{( -- )} for these words, the stack effect for the compiled code is
3185: the stack effect of the interpretation semantics.
3186:
3187: Note that the state of the interpreter does not come into play when
3188: performing the compilation semantics in this way. You can also perform
3189: it interpretively, e.g.:
3190:
3191: @example
3192: : foo2 ( n1 n2 -- n )
3193: [ MY-+ ] ;
3194: 1 2 foo .
3195: see foo
3196: @end example
3197:
3198: However, there are some broken Forth systems where this does not always
3199: work, and therefore this practice was been declared non-standard in
3200: 1999.
3201: @c !! repair.fs
3202:
3203: Here is another example for using @code{POSTPONE}:
3204:
3205: @example
3206: : MY-- ( Compilation: -- ; Run-time of compiled code: n1 n2 -- n )
3207: POSTPONE negate POSTPONE + ; immediate compile-only
3208: : bar ( n1 n2 -- n )
3209: MY-- ;
3210: 2 1 bar .
3211: see bar
3212: @end example
3213:
3214: You can define @code{ENDIF} in this way:
3215:
3216: @example
3217: : ENDIF ( Compilation: orig -- )
3218: POSTPONE then ; immediate
3219: @end example
3220:
3221: @quotation Assignment
3222: Write @code{MY-2DUP} that has compilation semantics equivalent to
3223: @code{2dup}, but compiles @code{over over}.
3224: @end quotation
3225:
3226: @c !! @xref{Macros} for reference
3227:
3228:
3229: @node Literal Tutorial, Advanced macros Tutorial, POSTPONE Tutorial, Tutorial
3230: @section @code{Literal}
3231: @cindex literal tutorial
3232:
3233: You cannot @code{POSTPONE} numbers:
3234:
3235: @example
3236: : [FOO] POSTPONE 500 ; immediate
3237: @end example
3238:
3239: Instead, you can use @code{LITERAL (compilation: n --; run-time: -- n )}:
3240:
3241: @example
3242: : [FOO] ( compilation: --; run-time: -- n )
3243: 500 POSTPONE literal ; immediate
3244:
3245: : flip [FOO] ;
3246: flip .
3247: see flip
3248: @end example
3249:
3250: @code{LITERAL} consumes a number at compile-time (when it's compilation
3251: semantics are executed) and pushes it at run-time (when the code it
3252: compiled is executed). A frequent use of @code{LITERAL} is to compile a
3253: number computed at compile time into the current word:
3254:
3255: @example
3256: : bar ( -- n )
3257: [ 2 2 + ] literal ;
3258: see bar
3259: @end example
3260:
3261: @quotation Assignment
3262: Write @code{]L} which allows writing the example above as @code{: bar (
3263: -- n ) [ 2 2 + ]L ;}
3264: @end quotation
3265:
3266: @c !! @xref{Macros} for reference
3267:
3268:
3269: @node Advanced macros Tutorial, Compilation Tokens Tutorial, Literal Tutorial, Tutorial
3270: @section Advanced macros
3271: @cindex macros, advanced tutorial
3272: @cindex run-time code generation, tutorial
3273:
3274: Reconsider @code{map-array} from @ref{Execution Tokens Tutorial,,
3275: Execution Tokens}. It frequently performs @code{execute}, a relatively
3276: expensive operation in some Forth implementations. You can use
3277: @code{compile,} and @code{POSTPONE} to eliminate these @code{execute}s
3278: and produce a word that contains the word to be performed directly:
3279:
3280: @c use ]] ... [[
3281: @example
3282: : compile-map-array ( compilation: xt -- ; run-time: ... addr u -- ... )
3283: \ at run-time, execute xt ( ... x -- ... ) for each element of the
3284: \ array beginning at addr and containing u elements
3285: @{ xt @}
3286: POSTPONE cells POSTPONE over POSTPONE + POSTPONE swap POSTPONE ?do
3287: POSTPONE i POSTPONE @@ xt compile,
3288: 1 cells POSTPONE literal POSTPONE +loop ;
3289:
3290: : sum-array ( addr u -- n )
3291: 0 rot rot [ ' + compile-map-array ] ;
3292: see sum-array
3293: a 5 sum-array .
3294: @end example
3295:
3296: You can use the full power of Forth for generating the code; here's an
3297: example where the code is generated in a loop:
3298:
3299: @example
3300: : compile-vmul-step ( compilation: n --; run-time: n1 addr1 -- n2 addr2 )
3301: \ n2=n1+(addr1)*n, addr2=addr1+cell
3302: POSTPONE tuck POSTPONE @@
3303: POSTPONE literal POSTPONE * POSTPONE +
3304: POSTPONE swap POSTPONE cell+ ;
3305:
3306: : compile-vmul ( compilation: addr1 u -- ; run-time: addr2 -- n )
3307: \ n=v1*v2 (inner product), where the v_i are represented as addr_i u
3308: 0 postpone literal postpone swap
3309: [ ' compile-vmul-step compile-map-array ]
3310: postpone drop ;
3311: see compile-vmul
3312:
3313: : a-vmul ( addr -- n )
3314: \ n=a*v, where v is a vector that's as long as a and starts at addr
3315: [ a 5 compile-vmul ] ;
3316: see a-vmul
3317: a a-vmul .
3318: @end example
3319:
3320: This example uses @code{compile-map-array} to show off, but you could
3321: also use @code{map-array} instead (try it now!).
3322:
3323: You can use this technique for efficient multiplication of large
3324: matrices. In matrix multiplication, you multiply every line of one
3325: matrix with every column of the other matrix. You can generate the code
3326: for one line once, and use it for every column. The only downside of
3327: this technique is that it is cumbersome to recover the memory consumed
3328: by the generated code when you are done (and in more complicated cases
3329: it is not possible portably).
3330:
3331: @c !! @xref{Macros} for reference
3332:
3333:
3334: @node Compilation Tokens Tutorial, Wordlists and Search Order Tutorial, Advanced macros Tutorial, Tutorial
3335: @section Compilation Tokens
3336: @cindex compilation tokens, tutorial
3337: @cindex CT, tutorial
3338:
3339: This section is Gforth-specific. You can skip it.
3340:
3341: @code{' word compile,} compiles the interpretation semantics. For words
3342: with default compilation semantics this is the same as performing the
3343: compilation semantics. To represent the compilation semantics of other
3344: words (e.g., words like @code{if} that have no interpretation
3345: semantics), Gforth has the concept of a compilation token (CT,
3346: consisting of two cells), and words @code{comp'} and @code{[comp']}.
3347: You can perform the compilation semantics represented by a CT with
3348: @code{execute}:
3349:
3350: @example
3351: : foo2 ( n1 n2 -- n )
3352: [ comp' + execute ] ;
3353: see foo
3354: @end example
3355:
3356: You can compile the compilation semantics represented by a CT with
3357: @code{postpone,}:
3358:
3359: @example
3360: : foo3 ( -- )
3361: [ comp' + postpone, ] ;
3362: see foo3
3363: @end example
3364:
3365: @code{[ comp' word postpone, ]} is equivalent to @code{POSTPONE word}.
3366: @code{comp'} is particularly useful for words that have no
3367: interpretation semantics:
3368:
3369: @example
3370: ' if
3371: comp' if .s 2drop
3372: @end example
3373:
3374: Reference: @ref{Tokens for Words}.
3375:
3376:
3377: @node Wordlists and Search Order Tutorial, , Compilation Tokens Tutorial, Tutorial
3378: @section Wordlists and Search Order
3379: @cindex wordlists tutorial
3380: @cindex search order, tutorial
3381:
3382: The dictionary is not just a memory area that allows you to allocate
3383: memory with @code{allot}, it also contains the Forth words, arranged in
3384: several wordlists. When searching for a word in a wordlist,
3385: conceptually you start searching at the youngest and proceed towards
3386: older words (in reality most systems nowadays use hash-tables); i.e., if
3387: you define a word with the same name as an older word, the new word
3388: shadows the older word.
3389:
3390: Which wordlists are searched in which order is determined by the search
3391: order. You can display the search order with @code{order}. It displays
3392: first the search order, starting with the wordlist searched first, then
3393: it displays the wordlist that will contain newly defined words.
3394:
3395: You can create a new, empty wordlist with @code{wordlist ( -- wid )}:
3396:
3397: @example
3398: wordlist constant mywords
3399: @end example
3400:
3401: @code{Set-current ( wid -- )} sets the wordlist that will contain newly
3402: defined words (the @emph{current} wordlist):
3403:
3404: @example
3405: mywords set-current
3406: order
3407: @end example
3408:
3409: Gforth does not display a name for the wordlist in @code{mywords}
3410: because this wordlist was created anonymously with @code{wordlist}.
3411:
3412: You can get the current wordlist with @code{get-current ( -- wid)}. If
3413: you want to put something into a specific wordlist without overall
3414: effect on the current wordlist, this typically looks like this:
3415:
3416: @example
3417: get-current mywords set-current ( wid )
3418: create someword
3419: ( wid ) set-current
3420: @end example
3421:
3422: You can write the search order with @code{set-order ( wid1 .. widn n --
3423: )} and read it with @code{get-order ( -- wid1 .. widn n )}. The first
3424: searched wordlist is topmost.
3425:
3426: @example
3427: get-order mywords swap 1+ set-order
3428: order
3429: @end example
3430:
3431: Yes, the order of wordlists in the output of @code{order} is reversed
3432: from stack comments and the output of @code{.s} and thus unintuitive.
3433:
3434: @quotation Assignment
3435: Define @code{>order ( wid -- )} with adds @code{wid} as first searched
3436: wordlist to the search order. Define @code{previous ( -- )}, which
3437: removes the first searched wordlist from the search order. Experiment
3438: with boundary conditions (you will see some crashes or situations that
3439: are hard or impossible to leave).
3440: @end quotation
3441:
3442: The search order is a powerful foundation for providing features similar
3443: to Modula-2 modules and C++ namespaces. However, trying to modularize
3444: programs in this way has disadvantages for debugging and reuse/factoring
3445: that overcome the advantages in my experience (I don't do huge projects,
3446: though). These disadvantages are not so clear in other
3447: languages/programming environments, because these languages are not so
3448: strong in debugging and reuse.
3449:
3450: @c !! example
3451:
3452: Reference: @ref{Word Lists}.
3453:
3454: @c ******************************************************************
3455: @node Introduction, Words, Tutorial, Top
3456: @comment node-name, next, previous, up
3457: @chapter An Introduction to ANS Forth
3458: @cindex Forth - an introduction
3459:
3460: The difference of this chapter from the Tutorial (@pxref{Tutorial}) is
3461: that it is slower-paced in its examples, but uses them to dive deep into
3462: explaining Forth internals (not covered by the Tutorial). Apart from
3463: that, this chapter covers far less material. It is suitable for reading
3464: without using a computer.
3465:
3466: The primary purpose of this manual is to document Gforth. However, since
3467: Forth is not a widely-known language and there is a lack of up-to-date
3468: teaching material, it seems worthwhile to provide some introductory
3469: material. For other sources of Forth-related
3470: information, see @ref{Forth-related information}.
3471:
3472: The examples in this section should work on any ANS Forth; the
3473: output shown was produced using Gforth. Each example attempts to
3474: reproduce the exact output that Gforth produces. If you try out the
3475: examples (and you should), what you should type is shown @kbd{like this}
3476: and Gforth's response is shown @code{like this}. The single exception is
3477: that, where the example shows @key{RET} it means that you should
3478: press the ``carriage return'' key. Unfortunately, some output formats for
3479: this manual cannot show the difference between @kbd{this} and
3480: @code{this} which will make trying out the examples harder (but not
3481: impossible).
3482:
3483: Forth is an unusual language. It provides an interactive development
3484: environment which includes both an interpreter and compiler. Forth
3485: programming style encourages you to break a problem down into many
3486: @cindex factoring
3487: small fragments (@dfn{factoring}), and then to develop and test each
3488: fragment interactively. Forth advocates assert that breaking the
3489: edit-compile-test cycle used by conventional programming languages can
3490: lead to great productivity improvements.
3491:
3492: @menu
3493: * Introducing the Text Interpreter::
3494: * Stacks and Postfix notation::
3495: * Your first definition::
3496: * How does that work?::
3497: * Forth is written in Forth::
3498: * Review - elements of a Forth system::
3499: * Where to go next::
3500: * Exercises::
3501: @end menu
3502:
3503: @comment ----------------------------------------------
3504: @node Introducing the Text Interpreter, Stacks and Postfix notation, Introduction, Introduction
3505: @section Introducing the Text Interpreter
3506: @cindex text interpreter
3507: @cindex outer interpreter
3508:
3509: @c IMO this is too detailed and the pace is too slow for
3510: @c an introduction. If you know German, take a look at
3511: @c http://www.complang.tuwien.ac.at/anton/lvas/skriptum-stack.html
3512: @c to see how I do it - anton
3513:
3514: @c nac-> Where I have accepted your comments 100% and modified the text
3515: @c accordingly, I have deleted your comments. Elsewhere I have added a
3516: @c response like this to attempt to rationalise what I have done. Of
3517: @c course, this is a very clumsy mechanism for something that would be
3518: @c done far more efficiently over a beer. Please delete any dialogue
3519: @c you consider closed.
3520:
3521: When you invoke the Forth image, you will see a startup banner printed
3522: and nothing else (if you have Gforth installed on your system, try
3523: invoking it now, by typing @kbd{gforth@key{RET}}). Forth is now running
3524: its command line interpreter, which is called the @dfn{Text Interpreter}
3525: (also known as the @dfn{Outer Interpreter}). (You will learn a lot
3526: about the text interpreter as you read through this chapter, for more
3527: detail @pxref{The Text Interpreter}).
3528:
3529: Although it's not obvious, Forth is actually waiting for your
3530: input. Type a number and press the @key{RET} key:
3531:
3532: @example
3533: @kbd{45@key{RET}} ok
3534: @end example
3535:
3536: Rather than give you a prompt to invite you to input something, the text
3537: interpreter prints a status message @i{after} it has processed a line
3538: of input. The status message in this case (``@code{ ok}'' followed by
3539: carriage-return) indicates that the text interpreter was able to process
3540: all of your input successfully. Now type something illegal:
3541:
3542: @example
3543: @kbd{qwer341@key{RET}}
3544: *the terminal*:2: Undefined word
3545: >>>qwer341<<<
3546: Backtrace:
3547: $2A95B42A20 throw
3548: $2A95B57FB8 no.extensions
3549: @end example
3550:
3551: The exact text, other than the ``Undefined word'' may differ slightly
3552: on your system, but the effect is the same; when the text interpreter
3553: detects an error, it discards any remaining text on a line, resets
3554: certain internal state and prints an error message. For a detailed
3555: description of error messages see @ref{Error messages}.
3556:
3557: The text interpreter waits for you to press carriage-return, and then
3558: processes your input line. Starting at the beginning of the line, it
3559: breaks the line into groups of characters separated by spaces. For each
3560: group of characters in turn, it makes two attempts to do something:
3561:
3562: @itemize @bullet
3563: @item
3564: @cindex name dictionary
3565: It tries to treat it as a command. It does this by searching a @dfn{name
3566: dictionary}. If the group of characters matches an entry in the name
3567: dictionary, the name dictionary provides the text interpreter with
3568: information that allows the text interpreter perform some actions. In
3569: Forth jargon, we say that the group
3570: @cindex word
3571: @cindex definition
3572: @cindex execution token
3573: @cindex xt
3574: of characters names a @dfn{word}, that the dictionary search returns an
3575: @dfn{execution token (xt)} corresponding to the @dfn{definition} of the
3576: word, and that the text interpreter executes the xt. Often, the terms
3577: @dfn{word} and @dfn{definition} are used interchangeably.
3578: @item
3579: If the text interpreter fails to find a match in the name dictionary, it
3580: tries to treat the group of characters as a number in the current number
3581: base (when you start up Forth, the current number base is base 10). If
3582: the group of characters legitimately represents a number, the text
3583: interpreter pushes the number onto a stack (we'll learn more about that
3584: in the next section).
3585: @end itemize
3586:
3587: If the text interpreter is unable to do either of these things with any
3588: group of characters, it discards the group of characters and the rest of
3589: the line, then prints an error message. If the text interpreter reaches
3590: the end of the line without error, it prints the status message ``@code{ ok}''
3591: followed by carriage-return.
3592:
3593: This is the simplest command we can give to the text interpreter:
3594:
3595: @example
3596: @key{RET} ok
3597: @end example
3598:
3599: The text interpreter did everything we asked it to do (nothing) without
3600: an error, so it said that everything is ``@code{ ok}''. Try a slightly longer
3601: command:
3602:
3603: @example
3604: @kbd{12 dup fred dup@key{RET}}
3605: *the terminal*:3: Undefined word
3606: 12 dup >>>fred<<< dup
3607: Backtrace:
3608: $2A95B42A20 throw
3609: $2A95B57FB8 no.extensions
3610: @end example
3611:
3612: When you press the carriage-return key, the text interpreter starts to
3613: work its way along the line:
3614:
3615: @itemize @bullet
3616: @item
3617: When it gets to the space after the @code{2}, it takes the group of
3618: characters @code{12} and looks them up in the name
3619: dictionary@footnote{We can't tell if it found them or not, but assume
3620: for now that it did not}. There is no match for this group of characters
3621: in the name dictionary, so it tries to treat them as a number. It is
3622: able to do this successfully, so it puts the number, 12, ``on the stack''
3623: (whatever that means).
3624: @item
3625: The text interpreter resumes scanning the line and gets the next group
3626: of characters, @code{dup}. It looks it up in the name dictionary and
3627: (you'll have to take my word for this) finds it, and executes the word
3628: @code{dup} (whatever that means).
3629: @item
3630: Once again, the text interpreter resumes scanning the line and gets the
3631: group of characters @code{fred}. It looks them up in the name
3632: dictionary, but can't find them. It tries to treat them as a number, but
3633: they don't represent any legal number.
3634: @end itemize
3635:
3636: At this point, the text interpreter gives up and prints an error
3637: message. The error message shows exactly how far the text interpreter
3638: got in processing the line. In particular, it shows that the text
3639: interpreter made no attempt to do anything with the final character
3640: group, @code{dup}, even though we have good reason to believe that the
3641: text interpreter would have no problem looking that word up and
3642: executing it a second time.
3643:
3644:
3645: @comment ----------------------------------------------
3646: @node Stacks and Postfix notation, Your first definition, Introducing the Text Interpreter, Introduction
3647: @section Stacks, postfix notation and parameter passing
3648: @cindex text interpreter
3649: @cindex outer interpreter
3650:
3651: In procedural programming languages (like C and Pascal), the
3652: building-block of programs is the @dfn{function} or @dfn{procedure}. These
3653: functions or procedures are called with @dfn{explicit parameters}. For
3654: example, in C we might write:
3655:
3656: @example
3657: total = total + new_volume(length,height,depth);
3658: @end example
3659:
3660: @noindent
3661: where new_volume is a function-call to another piece of code, and total,
3662: length, height and depth are all variables. length, height and depth are
3663: parameters to the function-call.
3664:
3665: In Forth, the equivalent of the function or procedure is the
3666: @dfn{definition} and parameters are implicitly passed between
3667: definitions using a shared stack that is visible to the
3668: programmer. Although Forth does support variables, the existence of the
3669: stack means that they are used far less often than in most other
3670: programming languages. When the text interpreter encounters a number, it
3671: will place (@dfn{push}) it on the stack. There are several stacks (the
3672: actual number is implementation-dependent ...) and the particular stack
3673: used for any operation is implied unambiguously by the operation being
3674: performed. The stack used for all integer operations is called the @dfn{data
3675: stack} and, since this is the stack used most commonly, references to
3676: ``the data stack'' are often abbreviated to ``the stack''.
3677:
3678: The stacks have a last-in, first-out (LIFO) organisation. If you type:
3679:
3680: @example
3681: @kbd{1 2 3@key{RET}} ok
3682: @end example
3683:
3684: Then this instructs the text interpreter to placed three numbers on the
3685: (data) stack. An analogy for the behaviour of the stack is to take a
3686: pack of playing cards and deal out the ace (1), 2 and 3 into a pile on
3687: the table. The 3 was the last card onto the pile (``last-in'') and if
3688: you take a card off the pile then, unless you're prepared to fiddle a
3689: bit, the card that you take off will be the 3 (``first-out''). The
3690: number that will be first-out of the stack is called the @dfn{top of
3691: stack}, which
3692: @cindex TOS definition
3693: is often abbreviated to @dfn{TOS}.
3694:
3695: To understand how parameters are passed in Forth, consider the
3696: behaviour of the definition @code{+} (pronounced ``plus''). You will not
3697: be surprised to learn that this definition performs addition. More
3698: precisely, it adds two number together and produces a result. Where does
3699: it get the two numbers from? It takes the top two numbers off the
3700: stack. Where does it place the result? On the stack. You can act-out the
3701: behaviour of @code{+} with your playing cards like this:
3702:
3703: @itemize @bullet
3704: @item
3705: Pick up two cards from the stack on the table
3706: @item
3707: Stare at them intently and ask yourself ``what @i{is} the sum of these two
3708: numbers''
3709: @item
3710: Decide that the answer is 5
3711: @item
3712: Shuffle the two cards back into the pack and find a 5
3713: @item
3714: Put a 5 on the remaining ace that's on the table.
3715: @end itemize
3716:
3717: If you don't have a pack of cards handy but you do have Forth running,
3718: you can use the definition @code{.s} to show the current state of the stack,
3719: without affecting the stack. Type:
3720:
3721: @example
3722: @kbd{clearstacks 1 2 3@key{RET}} ok
3723: @kbd{.s@key{RET}} <3> 1 2 3 ok
3724: @end example
3725:
3726: The text interpreter looks up the word @code{clearstacks} and executes
3727: it; it tidies up the stacks and removes any entries that may have been
3728: left on it by earlier examples. The text interpreter pushes each of the
3729: three numbers in turn onto the stack. Finally, the text interpreter
3730: looks up the word @code{.s} and executes it. The effect of executing
3731: @code{.s} is to print the ``<3>'' (the total number of items on the stack)
3732: followed by a list of all the items on the stack; the item on the far
3733: right-hand side is the TOS.
3734:
3735: You can now type:
3736:
3737: @example
3738: @kbd{+ .s@key{RET}} <2> 1 5 ok
3739: @end example
3740:
3741: @noindent
3742: which is correct; there are now 2 items on the stack and the result of
3743: the addition is 5.
3744:
3745: If you're playing with cards, try doing a second addition: pick up the
3746: two cards, work out that their sum is 6, shuffle them into the pack,
3747: look for a 6 and place that on the table. You now have just one item on
3748: the stack. What happens if you try to do a third addition? Pick up the
3749: first card, pick up the second card -- ah! There is no second card. This
3750: is called a @dfn{stack underflow} and consitutes an error. If you try to
3751: do the same thing with Forth it often reports an error (probably a Stack
3752: Underflow or an Invalid Memory Address error).
3753:
3754: The opposite situation to a stack underflow is a @dfn{stack overflow},
3755: which simply accepts that there is a finite amount of storage space
3756: reserved for the stack. To stretch the playing card analogy, if you had
3757: enough packs of cards and you piled the cards up on the table, you would
3758: eventually be unable to add another card; you'd hit the ceiling. Gforth
3759: allows you to set the maximum size of the stacks. In general, the only
3760: time that you will get a stack overflow is because a definition has a
3761: bug in it and is generating data on the stack uncontrollably.
3762:
3763: There's one final use for the playing card analogy. If you model your
3764: stack using a pack of playing cards, the maximum number of items on
3765: your stack will be 52 (I assume you didn't use the Joker). The maximum
3766: @i{value} of any item on the stack is 13 (the King). In fact, the only
3767: possible numbers are positive integer numbers 1 through 13; you can't
3768: have (for example) 0 or 27 or 3.52 or -2. If you change the way you
3769: think about some of the cards, you can accommodate different
3770: numbers. For example, you could think of the Jack as representing 0,
3771: the Queen as representing -1 and the King as representing -2. Your
3772: @i{range} remains unchanged (you can still only represent a total of 13
3773: numbers) but the numbers that you can represent are -2 through 10.
3774:
3775: In that analogy, the limit was the amount of information that a single
3776: stack entry could hold, and Forth has a similar limit. In Forth, the
3777: size of a stack entry is called a @dfn{cell}. The actual size of a cell is
3778: implementation dependent and affects the maximum value that a stack
3779: entry can hold. A Standard Forth provides a cell size of at least
3780: 16-bits, and most desktop systems use a cell size of 32-bits.
3781:
3782: Forth does not do any type checking for you, so you are free to
3783: manipulate and combine stack items in any way you wish. A convenient way
3784: of treating stack items is as 2's complement signed integers, and that
3785: is what Standard words like @code{+} do. Therefore you can type:
3786:
3787: @example
3788: @kbd{-5 12 + .s@key{RET}} <1> 7 ok
3789: @end example
3790:
3791: If you use numbers and definitions like @code{+} in order to turn Forth
3792: into a great big pocket calculator, you will realise that it's rather
3793: different from a normal calculator. Rather than typing 2 + 3 = you had
3794: to type 2 3 + (ignore the fact that you had to use @code{.s} to see the
3795: result). The terminology used to describe this difference is to say that
3796: your calculator uses @dfn{Infix Notation} (parameters and operators are
3797: mixed) whilst Forth uses @dfn{Postfix Notation} (parameters and
3798: operators are separate), also called @dfn{Reverse Polish Notation}.
3799:
3800: Whilst postfix notation might look confusing to begin with, it has
3801: several important advantages:
3802:
3803: @itemize @bullet
3804: @item
3805: it is unambiguous
3806: @item
3807: it is more concise
3808: @item
3809: it fits naturally with a stack-based system
3810: @end itemize
3811:
3812: To examine these claims in more detail, consider these sums:
3813:
3814: @example
3815: 6 + 5 * 4 =
3816: 4 * 5 + 6 =
3817: @end example
3818:
3819: If you're just learning maths or your maths is very rusty, you will
3820: probably come up with the answer 44 for the first and 26 for the
3821: second. If you are a bit of a whizz at maths you will remember the
3822: @i{convention} that multiplication takes precendence over addition, and
3823: you'd come up with the answer 26 both times. To explain the answer 26
3824: to someone who got the answer 44, you'd probably rewrite the first sum
3825: like this:
3826:
3827: @example
3828: 6 + (5 * 4) =
3829: @end example
3830:
3831: If what you really wanted was to perform the addition before the
3832: multiplication, you would have to use parentheses to force it.
3833:
3834: If you did the first two sums on a pocket calculator you would probably
3835: get the right answers, unless you were very cautious and entered them using
3836: these keystroke sequences:
3837:
3838: 6 + 5 = * 4 =
3839: 4 * 5 = + 6 =
3840:
3841: Postfix notation is unambiguous because the order that the operators
3842: are applied is always explicit; that also means that parentheses are
3843: never required. The operators are @i{active} (the act of quoting the
3844: operator makes the operation occur) which removes the need for ``=''.
3845:
3846: The sum 6 + 5 * 4 can be written (in postfix notation) in two
3847: equivalent ways:
3848:
3849: @example
3850: 6 5 4 * + or:
3851: 5 4 * 6 +
3852: @end example
3853:
3854: An important thing that you should notice about this notation is that
3855: the @i{order} of the numbers does not change; if you want to subtract
3856: 2 from 10 you type @code{10 2 -}.
3857:
3858: The reason that Forth uses postfix notation is very simple to explain: it
3859: makes the implementation extremely simple, and it follows naturally from
3860: using the stack as a mechanism for passing parameters. Another way of
3861: thinking about this is to realise that all Forth definitions are
3862: @i{active}; they execute as they are encountered by the text
3863: interpreter. The result of this is that the syntax of Forth is trivially
3864: simple.
3865:
3866:
3867:
3868: @comment ----------------------------------------------
3869: @node Your first definition, How does that work?, Stacks and Postfix notation, Introduction
3870: @section Your first Forth definition
3871: @cindex first definition
3872:
3873: Until now, the examples we've seen have been trivial; we've just been
3874: using Forth as a bigger-than-pocket calculator. Also, each calculation
3875: we've shown has been a ``one-off'' -- to repeat it we'd need to type it in
3876: again@footnote{That's not quite true. If you press the up-arrow key on
3877: your keyboard you should be able to scroll back to any earlier command,
3878: edit it and re-enter it.} In this section we'll see how to add new
3879: words to Forth's vocabulary.
3880:
3881: The easiest way to create a new word is to use a @dfn{colon
3882: definition}. We'll define a few and try them out before worrying too
3883: much about how they work. Try typing in these examples; be careful to
3884: copy the spaces accurately:
3885:
3886: @example
3887: : add-two 2 + . ;
3888: : greet ." Hello and welcome" ;
3889: : demo 5 add-two ;
3890: @end example
3891:
3892: @noindent
3893: Now try them out:
3894:
3895: @example
3896: @kbd{greet@key{RET}} Hello and welcome ok
3897: @kbd{greet greet@key{RET}} Hello and welcomeHello and welcome ok
3898: @kbd{4 add-two@key{RET}} 6 ok
3899: @kbd{demo@key{RET}} 7 ok
3900: @kbd{9 greet demo add-two@key{RET}} Hello and welcome7 11 ok
3901: @end example
3902:
3903: The first new thing that we've introduced here is the pair of words
3904: @code{:} and @code{;}. These are used to start and terminate a new
3905: definition, respectively. The first word after the @code{:} is the name
3906: for the new definition.
3907:
3908: As you can see from the examples, a definition is built up of words that
3909: have already been defined; Forth makes no distinction between
3910: definitions that existed when you started the system up, and those that
3911: you define yourself.
3912:
3913: The examples also introduce the words @code{.} (dot), @code{."}
3914: (dot-quote) and @code{dup} (dewp). Dot takes the value from the top of
3915: the stack and displays it. It's like @code{.s} except that it only
3916: displays the top item of the stack and it is destructive; after it has
3917: executed, the number is no longer on the stack. There is always one
3918: space printed after the number, and no spaces before it. Dot-quote
3919: defines a string (a sequence of characters) that will be printed when
3920: the word is executed. The string can contain any printable characters
3921: except @code{"}. A @code{"} has a special function; it is not a Forth
3922: word but it acts as a delimiter (the way that delimiters work is
3923: described in the next section). Finally, @code{dup} duplicates the value
3924: at the top of the stack. Try typing @code{5 dup .s} to see what it does.
3925:
3926: We already know that the text interpreter searches through the
3927: dictionary to locate names. If you've followed the examples earlier, you
3928: will already have a definition called @code{add-two}. Lets try modifying
3929: it by typing in a new definition:
3930:
3931: @example
3932: @kbd{: add-two dup . ." + 2 =" 2 + . ;@key{RET}} redefined add-two ok
3933: @end example
3934:
3935: Forth recognised that we were defining a word that already exists, and
3936: printed a message to warn us of that fact. Let's try out the new
3937: definition:
3938:
3939: @example
3940: @kbd{9 add-two@key{RET}} 9 + 2 =11 ok
3941: @end example
3942:
3943: @noindent
3944: All that we've actually done here, though, is to create a new
3945: definition, with a particular name. The fact that there was already a
3946: definition with the same name did not make any difference to the way
3947: that the new definition was created (except that Forth printed a warning
3948: message). The old definition of add-two still exists (try @code{demo}
3949: again to see that this is true). Any new definition will use the new
3950: definition of @code{add-two}, but old definitions continue to use the
3951: version that already existed at the time that they were @code{compiled}.
3952:
3953: Before you go on to the next section, try defining and redefining some
3954: words of your own.
3955:
3956: @comment ----------------------------------------------
3957: @node How does that work?, Forth is written in Forth, Your first definition, Introduction
3958: @section How does that work?
3959: @cindex parsing words
3960:
3961: @c That's pretty deep (IMO way too deep) for an introduction. - anton
3962:
3963: @c Is it a good idea to talk about the interpretation semantics of a
3964: @c number? We don't have an xt to go along with it. - anton
3965:
3966: @c Now that I have eliminated execution semantics, I wonder if it would not
3967: @c be better to keep them (or add run-time semantics), to make it easier to
3968: @c explain what compilation semantics usually does. - anton
3969:
3970: @c nac-> I removed the term ``default compilation sematics'' from the
3971: @c introductory chapter. Removing ``execution semantics'' was making
3972: @c everything simpler to explain, then I think the use of this term made
3973: @c everything more complex again. I replaced it with ``default
3974: @c semantics'' (which is used elsewhere in the manual) by which I mean
3975: @c ``a definition that has neither the immediate nor the compile-only
3976: @c flag set''.
3977:
3978: @c anton: I have eliminated default semantics (except in one place where it
3979: @c means "default interpretation and compilation semantics"), because it
3980: @c makes no sense in the presence of combined words. I reverted to
3981: @c "execution semantics" where necessary.
3982:
3983: @c nac-> I reworded big chunks of the ``how does that work''
3984: @c section (and, unusually for me, I think I even made it shorter!). See
3985: @c what you think -- I know I have not addressed your primary concern
3986: @c that it is too heavy-going for an introduction. From what I understood
3987: @c of your course notes it looks as though they might be a good framework.
3988: @c Things that I've tried to capture here are some things that came as a
3989: @c great revelation here when I first understood them. Also, I like the
3990: @c fact that a very simple code example shows up almost all of the issues
3991: @c that you need to understand to see how Forth works. That's unique and
3992: @c worthwhile to emphasise.
3993:
3994: @c anton: I think it's a good idea to present the details, especially those
3995: @c that you found to be a revelation, and probably the tutorial tries to be
3996: @c too superficial and does not get some of the things across that make
3997: @c Forth special. I do believe that most of the time these things should
3998: @c be discussed at the end of a section or in separate sections instead of
3999: @c in the middle of a section (e.g., the stuff you added in "User-defined
4000: @c defining words" leads in a completely different direction from the rest
4001: @c of the section).
4002:
4003: Now we're going to take another look at the definition of @code{add-two}
4004: from the previous section. From our knowledge of the way that the text
4005: interpreter works, we would have expected this result when we tried to
4006: define @code{add-two}:
4007:
4008: @example
4009: @kbd{: add-two 2 + . ;@key{RET}}
4010: *the terminal*:4: Undefined word
4011: : >>>add-two<<< 2 + . ;
4012: @end example
4013:
4014: The reason that this didn't happen is bound up in the way that @code{:}
4015: works. The word @code{:} does two special things. The first special
4016: thing that it does prevents the text interpreter from ever seeing the
4017: characters @code{add-two}. The text interpreter uses a variable called
4018: @cindex modifying >IN
4019: @code{>IN} (pronounced ``to-in'') to keep track of where it is in the
4020: input line. When it encounters the word @code{:} it behaves in exactly
4021: the same way as it does for any other word; it looks it up in the name
4022: dictionary, finds its xt and executes it. When @code{:} executes, it
4023: looks at the input buffer, finds the word @code{add-two} and advances the
4024: value of @code{>IN} to point past it. It then does some other stuff
4025: associated with creating the new definition (including creating an entry
4026: for @code{add-two} in the name dictionary). When the execution of @code{:}
4027: completes, control returns to the text interpreter, which is oblivious
4028: to the fact that it has been tricked into ignoring part of the input
4029: line.
4030:
4031: @cindex parsing words
4032: Words like @code{:} -- words that advance the value of @code{>IN} and so
4033: prevent the text interpreter from acting on the whole of the input line
4034: -- are called @dfn{parsing words}.
4035:
4036: @cindex @code{state} - effect on the text interpreter
4037: @cindex text interpreter - effect of state
4038: The second special thing that @code{:} does is change the value of a
4039: variable called @code{state}, which affects the way that the text
4040: interpreter behaves. When Gforth starts up, @code{state} has the value
4041: 0, and the text interpreter is said to be @dfn{interpreting}. During a
4042: colon definition (started with @code{:}), @code{state} is set to -1 and
4043: the text interpreter is said to be @dfn{compiling}.
4044:
4045: In this example, the text interpreter is compiling when it processes the
4046: string ``@code{2 + . ;}''. It still breaks the string down into
4047: character sequences in the same way. However, instead of pushing the
4048: number @code{2} onto the stack, it lays down (@dfn{compiles}) some magic
4049: into the definition of @code{add-two} that will make the number @code{2} get
4050: pushed onto the stack when @code{add-two} is @dfn{executed}. Similarly,
4051: the behaviours of @code{+} and @code{.} are also compiled into the
4052: definition.
4053:
4054: One category of words don't get compiled. These so-called @dfn{immediate
4055: words} get executed (performed @i{now}) regardless of whether the text
4056: interpreter is interpreting or compiling. The word @code{;} is an
4057: immediate word. Rather than being compiled into the definition, it
4058: executes. Its effect is to terminate the current definition, which
4059: includes changing the value of @code{state} back to 0.
4060:
4061: When you execute @code{add-two}, it has a @dfn{run-time effect} that is
4062: exactly the same as if you had typed @code{2 + . @key{RET}} outside of a
4063: definition.
4064:
4065: In Forth, every word or number can be described in terms of two
4066: properties:
4067:
4068: @itemize @bullet
4069: @item
4070: @cindex interpretation semantics
4071: Its @dfn{interpretation semantics} describe how it will behave when the
4072: text interpreter encounters it in @dfn{interpret} state. The
4073: interpretation semantics of a word are represented by an @dfn{execution
4074: token}.
4075: @item
4076: @cindex compilation semantics
4077: Its @dfn{compilation semantics} describe how it will behave when the
4078: text interpreter encounters it in @dfn{compile} state. The compilation
4079: semantics of a word are represented in an implementation-dependent way;
4080: Gforth uses a @dfn{compilation token}.
4081: @end itemize
4082:
4083: @noindent
4084: Numbers are always treated in a fixed way:
4085:
4086: @itemize @bullet
4087: @item
4088: When the number is @dfn{interpreted}, its behaviour is to push the
4089: number onto the stack.
4090: @item
4091: When the number is @dfn{compiled}, a piece of code is appended to the
4092: current definition that pushes the number when it runs. (In other words,
4093: the compilation semantics of a number are to postpone its interpretation
4094: semantics until the run-time of the definition that it is being compiled
4095: into.)
4096: @end itemize
4097:
4098: Words don't behave in such a regular way, but most have @i{default
4099: semantics} which means that they behave like this:
4100:
4101: @itemize @bullet
4102: @item
4103: The @dfn{interpretation semantics} of the word are to do something useful.
4104: @item
4105: The @dfn{compilation semantics} of the word are to append its
4106: @dfn{interpretation semantics} to the current definition (so that its
4107: run-time behaviour is to do something useful).
4108: @end itemize
4109:
4110: @cindex immediate words
4111: The actual behaviour of any particular word can be controlled by using
4112: the words @code{immediate} and @code{compile-only} when the word is
4113: defined. These words set flags in the name dictionary entry of the most
4114: recently defined word, and these flags are retrieved by the text
4115: interpreter when it finds the word in the name dictionary.
4116:
4117: A word that is marked as @dfn{immediate} has compilation semantics that
4118: are identical to its interpretation semantics. In other words, it
4119: behaves like this:
4120:
4121: @itemize @bullet
4122: @item
4123: The @dfn{interpretation semantics} of the word are to do something useful.
4124: @item
4125: The @dfn{compilation semantics} of the word are to do something useful
4126: (and actually the same thing); i.e., it is executed during compilation.
4127: @end itemize
4128:
4129: Marking a word as @dfn{compile-only} prohibits the text interpreter from
4130: performing the interpretation semantics of the word directly; an attempt
4131: to do so will generate an error. It is never necessary to use
4132: @code{compile-only} (and it is not even part of ANS Forth, though it is
4133: provided by many implementations) but it is good etiquette to apply it
4134: to a word that will not behave correctly (and might have unexpected
4135: side-effects) in interpret state. For example, it is only legal to use
4136: the conditional word @code{IF} within a definition. If you forget this
4137: and try to use it elsewhere, the fact that (in Gforth) it is marked as
4138: @code{compile-only} allows the text interpreter to generate a helpful
4139: error message rather than subjecting you to the consequences of your
4140: folly.
4141:
4142: This example shows the difference between an immediate and a
4143: non-immediate word:
4144:
4145: @example
4146: : show-state state @@ . ;
4147: : show-state-now show-state ; immediate
4148: : word1 show-state ;
4149: : word2 show-state-now ;
4150: @end example
4151:
4152: The word @code{immediate} after the definition of @code{show-state-now}
4153: makes that word an immediate word. These definitions introduce a new
4154: word: @code{@@} (pronounced ``fetch''). This word fetches the value of a
4155: variable, and leaves it on the stack. Therefore, the behaviour of
4156: @code{show-state} is to print a number that represents the current value
4157: of @code{state}.
4158:
4159: When you execute @code{word1}, it prints the number 0, indicating that
4160: the system is interpreting. When the text interpreter compiled the
4161: definition of @code{word1}, it encountered @code{show-state} whose
4162: compilation semantics are to append its interpretation semantics to the
4163: current definition. When you execute @code{word1}, it performs the
4164: interpretation semantics of @code{show-state}. At the time that @code{word1}
4165: (and therefore @code{show-state}) are executed, the system is
4166: interpreting.
4167:
4168: When you pressed @key{RET} after entering the definition of @code{word2},
4169: you should have seen the number -1 printed, followed by ``@code{
4170: ok}''. When the text interpreter compiled the definition of
4171: @code{word2}, it encountered @code{show-state-now}, an immediate word,
4172: whose compilation semantics are therefore to perform its interpretation
4173: semantics. It is executed straight away (even before the text
4174: interpreter has moved on to process another group of characters; the
4175: @code{;} in this example). The effect of executing it are to display the
4176: value of @code{state} @i{at the time that the definition of}
4177: @code{word2} @i{is being defined}. Printing -1 demonstrates that the
4178: system is compiling at this time. If you execute @code{word2} it does
4179: nothing at all.
4180:
4181: @cindex @code{."}, how it works
4182: Before leaving the subject of immediate words, consider the behaviour of
4183: @code{."} in the definition of @code{greet}, in the previous
4184: section. This word is both a parsing word and an immediate word. Notice
4185: that there is a space between @code{."} and the start of the text
4186: @code{Hello and welcome}, but that there is no space between the last
4187: letter of @code{welcome} and the @code{"} character. The reason for this
4188: is that @code{."} is a Forth word; it must have a space after it so that
4189: the text interpreter can identify it. The @code{"} is not a Forth word;
4190: it is a @dfn{delimiter}. The examples earlier show that, when the string
4191: is displayed, there is neither a space before the @code{H} nor after the
4192: @code{e}. Since @code{."} is an immediate word, it executes at the time
4193: that @code{greet} is defined. When it executes, its behaviour is to
4194: search forward in the input line looking for the delimiter. When it
4195: finds the delimiter, it updates @code{>IN} to point past the
4196: delimiter. It also compiles some magic code into the definition of
4197: @code{greet}; the xt of a run-time routine that prints a text string. It
4198: compiles the string @code{Hello and welcome} into memory so that it is
4199: available to be printed later. When the text interpreter gains control,
4200: the next word it finds in the input stream is @code{;} and so it
4201: terminates the definition of @code{greet}.
4202:
4203:
4204: @comment ----------------------------------------------
4205: @node Forth is written in Forth, Review - elements of a Forth system, How does that work?, Introduction
4206: @section Forth is written in Forth
4207: @cindex structure of Forth programs
4208:
4209: When you start up a Forth compiler, a large number of definitions
4210: already exist. In Forth, you develop a new application using bottom-up
4211: programming techniques to create new definitions that are defined in
4212: terms of existing definitions. As you create each definition you can
4213: test and debug it interactively.
4214:
4215: If you have tried out the examples in this section, you will probably
4216: have typed them in by hand; when you leave Gforth, your definitions will
4217: be lost. You can avoid this by using a text editor to enter Forth source
4218: code into a file, and then loading code from the file using
4219: @code{include} (@pxref{Forth source files}). A Forth source file is
4220: processed by the text interpreter, just as though you had typed it in by
4221: hand@footnote{Actually, there are some subtle differences -- see
4222: @ref{The Text Interpreter}.}.
4223:
4224: Gforth also supports the traditional Forth alternative to using text
4225: files for program entry (@pxref{Blocks}).
4226:
4227: In common with many, if not most, Forth compilers, most of Gforth is
4228: actually written in Forth. All of the @file{.fs} files in the
4229: installation directory@footnote{For example,
4230: @file{/usr/local/share/gforth...}} are Forth source files, which you can
4231: study to see examples of Forth programming.
4232:
4233: Gforth maintains a history file that records every line that you type to
4234: the text interpreter. This file is preserved between sessions, and is
4235: used to provide a command-line recall facility. If you enter long
4236: definitions by hand, you can use a text editor to paste them out of the
4237: history file into a Forth source file for reuse at a later time
4238: (for more information @pxref{Command-line editing}).
4239:
4240:
4241: @comment ----------------------------------------------
4242: @node Review - elements of a Forth system, Where to go next, Forth is written in Forth, Introduction
4243: @section Review - elements of a Forth system
4244: @cindex elements of a Forth system
4245:
4246: To summarise this chapter:
4247:
4248: @itemize @bullet
4249: @item
4250: Forth programs use @dfn{factoring} to break a problem down into small
4251: fragments called @dfn{words} or @dfn{definitions}.
4252: @item
4253: Forth program development is an interactive process.
4254: @item
4255: The main command loop that accepts input, and controls both
4256: interpretation and compilation, is called the @dfn{text interpreter}
4257: (also known as the @dfn{outer interpreter}).
4258: @item
4259: Forth has a very simple syntax, consisting of words and numbers
4260: separated by spaces or carriage-return characters. Any additional syntax
4261: is imposed by @dfn{parsing words}.
4262: @item
4263: Forth uses a stack to pass parameters between words. As a result, it
4264: uses postfix notation.
4265: @item
4266: To use a word that has previously been defined, the text interpreter
4267: searches for the word in the @dfn{name dictionary}.
4268: @item
4269: Words have @dfn{interpretation semantics} and @dfn{compilation semantics}.
4270: @item
4271: The text interpreter uses the value of @code{state} to select between
4272: the use of the @dfn{interpretation semantics} and the @dfn{compilation
4273: semantics} of a word that it encounters.
4274: @item
4275: The relationship between the @dfn{interpretation semantics} and
4276: @dfn{compilation semantics} for a word
4277: depend upon the way in which the word was defined (for example, whether
4278: it is an @dfn{immediate} word).
4279: @item
4280: Forth definitions can be implemented in Forth (called @dfn{high-level
4281: definitions}) or in some other way (usually a lower-level language and
4282: as a result often called @dfn{low-level definitions}, @dfn{code
4283: definitions} or @dfn{primitives}).
4284: @item
4285: Many Forth systems are implemented mainly in Forth.
4286: @end itemize
4287:
4288:
4289: @comment ----------------------------------------------
4290: @node Where to go next, Exercises, Review - elements of a Forth system, Introduction
4291: @section Where To Go Next
4292: @cindex where to go next
4293:
4294: Amazing as it may seem, if you have read (and understood) this far, you
4295: know almost all the fundamentals about the inner workings of a Forth
4296: system. You certainly know enough to be able to read and understand the
4297: rest of this manual and the ANS Forth document, to learn more about the
4298: facilities that Forth in general and Gforth in particular provide. Even
4299: scarier, you know almost enough to implement your own Forth system.
4300: However, that's not a good idea just yet... better to try writing some
4301: programs in Gforth.
4302:
4303: Forth has such a rich vocabulary that it can be hard to know where to
4304: start in learning it. This section suggests a few sets of words that are
4305: enough to write small but useful programs. Use the word index in this
4306: document to learn more about each word, then try it out and try to write
4307: small definitions using it. Start by experimenting with these words:
4308:
4309: @itemize @bullet
4310: @item
4311: Arithmetic: @code{+ - * / /MOD */ ABS INVERT}
4312: @item
4313: Comparison: @code{MIN MAX =}
4314: @item
4315: Logic: @code{AND OR XOR NOT}
4316: @item
4317: Stack manipulation: @code{DUP DROP SWAP OVER}
4318: @item
4319: Loops and decisions: @code{IF ELSE ENDIF ?DO I LOOP}
4320: @item
4321: Input/Output: @code{. ." EMIT CR KEY}
4322: @item
4323: Defining words: @code{: ; CREATE}
4324: @item
4325: Memory allocation words: @code{ALLOT ,}
4326: @item
4327: Tools: @code{SEE WORDS .S MARKER}
4328: @end itemize
4329:
4330: When you have mastered those, go on to:
4331:
4332: @itemize @bullet
4333: @item
4334: More defining words: @code{VARIABLE CONSTANT VALUE TO CREATE DOES>}
4335: @item
4336: Memory access: @code{@@ !}
4337: @end itemize
4338:
4339: When you have mastered these, there's nothing for it but to read through
4340: the whole of this manual and find out what you've missed.
4341:
4342: @comment ----------------------------------------------
4343: @node Exercises, , Where to go next, Introduction
4344: @section Exercises
4345: @cindex exercises
4346:
4347: TODO: provide a set of programming excercises linked into the stuff done
4348: already and into other sections of the manual. Provide solutions to all
4349: the exercises in a .fs file in the distribution.
4350:
4351: @c Get some inspiration from Starting Forth and Kelly&Spies.
4352:
4353: @c excercises:
4354: @c 1. take inches and convert to feet and inches.
4355: @c 2. take temperature and convert from fahrenheight to celcius;
4356: @c may need to care about symmetric vs floored??
4357: @c 3. take input line and do character substitution
4358: @c to encipher or decipher
4359: @c 4. as above but work on a file for in and out
4360: @c 5. take input line and convert to pig-latin
4361: @c
4362: @c thing of sets of things to exercise then come up with
4363: @c problems that need those things.
4364:
4365:
4366: @c ******************************************************************
4367: @node Words, Error messages, Introduction, Top
4368: @chapter Forth Words
4369: @cindex words
4370:
4371: @menu
4372: * Notation::
4373: * Case insensitivity::
4374: * Comments::
4375: * Boolean Flags::
4376: * Arithmetic::
4377: * Stack Manipulation::
4378: * Memory::
4379: * Control Structures::
4380: * Defining Words::
4381: * Interpretation and Compilation Semantics::
4382: * Tokens for Words::
4383: * Compiling words::
4384: * The Text Interpreter::
4385: * The Input Stream::
4386: * Word Lists::
4387: * Environmental Queries::
4388: * Files::
4389: * Blocks::
4390: * Other I/O::
4391: * OS command line arguments::
4392: * Locals::
4393: * Structures::
4394: * Object-oriented Forth::
4395: * Programming Tools::
4396: * C Interface::
4397: * Assembler and Code Words::
4398: * Threading Words::
4399: * Passing Commands to the OS::
4400: * Keeping track of Time::
4401: * Miscellaneous Words::
4402: @end menu
4403:
4404: @node Notation, Case insensitivity, Words, Words
4405: @section Notation
4406: @cindex notation of glossary entries
4407: @cindex format of glossary entries
4408: @cindex glossary notation format
4409: @cindex word glossary entry format
4410:
4411: The Forth words are described in this section in the glossary notation
4412: that has become a de-facto standard for Forth texts:
4413:
4414: @format
4415: @i{word} @i{Stack effect} @i{wordset} @i{pronunciation}
4416: @end format
4417: @i{Description}
4418:
4419: @table @var
4420: @item word
4421: The name of the word.
4422:
4423: @item Stack effect
4424: @cindex stack effect
4425: The stack effect is written in the notation @code{@i{before} --
4426: @i{after}}, where @i{before} and @i{after} describe the top of
4427: stack entries before and after the execution of the word. The rest of
4428: the stack is not touched by the word. The top of stack is rightmost,
4429: i.e., a stack sequence is written as it is typed in. Note that Gforth
4430: uses a separate floating point stack, but a unified stack
4431: notation. Also, return stack effects are not shown in @i{stack
4432: effect}, but in @i{Description}. The name of a stack item describes
4433: the type and/or the function of the item. See below for a discussion of
4434: the types.
4435:
4436: All words have two stack effects: A compile-time stack effect and a
4437: run-time stack effect. The compile-time stack-effect of most words is
4438: @i{ -- }. If the compile-time stack-effect of a word deviates from
4439: this standard behaviour, or the word does other unusual things at
4440: compile time, both stack effects are shown; otherwise only the run-time
4441: stack effect is shown.
4442:
4443: Also note that in code templates or examples there can be comments in
4444: parentheses that display the stack picture at this point; there is no
4445: @code{--} in these places, because there is no before-after situation.
4446:
4447: @cindex pronounciation of words
4448: @item pronunciation
4449: How the word is pronounced.
4450:
4451: @cindex wordset
4452: @cindex environment wordset
4453: @item wordset
4454: The ANS Forth standard is divided into several word sets. A standard
4455: system need not support all of them. Therefore, in theory, the fewer
4456: word sets your program uses the more portable it will be. However, we
4457: suspect that most ANS Forth systems on personal machines will feature
4458: all word sets. Words that are not defined in ANS Forth have
4459: @code{gforth} or @code{gforth-internal} as word set. @code{gforth}
4460: describes words that will work in future releases of Gforth;
4461: @code{gforth-internal} words are more volatile. Environmental query
4462: strings are also displayed like words; you can recognize them by the
4463: @code{environment} in the word set field.
4464:
4465: @item Description
4466: A description of the behaviour of the word.
4467: @end table
4468:
4469: @cindex types of stack items
4470: @cindex stack item types
4471: The type of a stack item is specified by the character(s) the name
4472: starts with:
4473:
4474: @table @code
4475: @item f
4476: @cindex @code{f}, stack item type
4477: Boolean flags, i.e. @code{false} or @code{true}.
4478: @item c
4479: @cindex @code{c}, stack item type
4480: Char
4481: @item w
4482: @cindex @code{w}, stack item type
4483: Cell, can contain an integer or an address
4484: @item n
4485: @cindex @code{n}, stack item type
4486: signed integer
4487: @item u
4488: @cindex @code{u}, stack item type
4489: unsigned integer
4490: @item d
4491: @cindex @code{d}, stack item type
4492: double sized signed integer
4493: @item ud
4494: @cindex @code{ud}, stack item type
4495: double sized unsigned integer
4496: @item r
4497: @cindex @code{r}, stack item type
4498: Float (on the FP stack)
4499: @item a-
4500: @cindex @code{a_}, stack item type
4501: Cell-aligned address
4502: @item c-
4503: @cindex @code{c_}, stack item type
4504: Char-aligned address (note that a Char may have two bytes in Windows NT)
4505: @item f-
4506: @cindex @code{f_}, stack item type
4507: Float-aligned address
4508: @item df-
4509: @cindex @code{df_}, stack item type
4510: Address aligned for IEEE double precision float
4511: @item sf-
4512: @cindex @code{sf_}, stack item type
4513: Address aligned for IEEE single precision float
4514: @item xt
4515: @cindex @code{xt}, stack item type
4516: Execution token, same size as Cell
4517: @item wid
4518: @cindex @code{wid}, stack item type
4519: Word list ID, same size as Cell
4520: @item ior, wior
4521: @cindex ior type description
4522: @cindex wior type description
4523: I/O result code, cell-sized. In Gforth, you can @code{throw} iors.
4524: @item f83name
4525: @cindex @code{f83name}, stack item type
4526: Pointer to a name structure
4527: @item "
4528: @cindex @code{"}, stack item type
4529: string in the input stream (not on the stack). The terminating character
4530: is a blank by default. If it is not a blank, it is shown in @code{<>}
4531: quotes.
4532: @end table
4533:
4534: @comment ----------------------------------------------
4535: @node Case insensitivity, Comments, Notation, Words
4536: @section Case insensitivity
4537: @cindex case sensitivity
4538: @cindex upper and lower case
4539:
4540: Gforth is case-insensitive; you can enter definitions and invoke
4541: Standard words using upper, lower or mixed case (however,
4542: @pxref{core-idef, Implementation-defined options, Implementation-defined
4543: options}).
4544:
4545: ANS Forth only @i{requires} implementations to recognise Standard words
4546: when they are typed entirely in upper case. Therefore, a Standard
4547: program must use upper case for all Standard words. You can use whatever
4548: case you like for words that you define, but in a Standard program you
4549: have to use the words in the same case that you defined them.
4550:
4551: Gforth supports case sensitivity through @code{table}s (case-sensitive
4552: wordlists, @pxref{Word Lists}).
4553:
4554: Two people have asked how to convert Gforth to be case-sensitive; while
4555: we think this is a bad idea, you can change all wordlists into tables
4556: like this:
4557:
4558: @example
4559: ' table-find forth-wordlist wordlist-map @ !
4560: @end example
4561:
4562: Note that you now have to type the predefined words in the same case
4563: that we defined them, which are varying. You may want to convert them
4564: to your favourite case before doing this operation (I won't explain how,
4565: because if you are even contemplating doing this, you'd better have
4566: enough knowledge of Forth systems to know this already).
4567:
4568: @node Comments, Boolean Flags, Case insensitivity, Words
4569: @section Comments
4570: @cindex comments
4571:
4572: Forth supports two styles of comment; the traditional @i{in-line} comment,
4573: @code{(} and its modern cousin, the @i{comment to end of line}; @code{\}.
4574:
4575:
4576: doc-(
4577: doc-\
4578: doc-\G
4579:
4580:
4581: @node Boolean Flags, Arithmetic, Comments, Words
4582: @section Boolean Flags
4583: @cindex Boolean flags
4584:
4585: A Boolean flag is cell-sized. A cell with all bits clear represents the
4586: flag @code{false} and a flag with all bits set represents the flag
4587: @code{true}. Words that check a flag (for example, @code{IF}) will treat
4588: a cell that has @i{any} bit set as @code{true}.
4589: @c on and off to Memory?
4590: @c true and false to "Bitwise operations" or "Numeric comparison"?
4591:
4592: doc-true
4593: doc-false
4594: doc-on
4595: doc-off
4596:
4597:
4598: @node Arithmetic, Stack Manipulation, Boolean Flags, Words
4599: @section Arithmetic
4600: @cindex arithmetic words
4601:
4602: @cindex division with potentially negative operands
4603: Forth arithmetic is not checked, i.e., you will not hear about integer
4604: overflow on addition or multiplication, you may hear about division by
4605: zero if you are lucky. The operator is written after the operands, but
4606: the operands are still in the original order. I.e., the infix @code{2-1}
4607: corresponds to @code{2 1 -}. Forth offers a variety of division
4608: operators. If you perform division with potentially negative operands,
4609: you do not want to use @code{/} or @code{/mod} with its undefined
4610: behaviour, but rather @code{fm/mod} or @code{sm/mod} (probably the
4611: former, @pxref{Mixed precision}).
4612: @comment TODO discuss the different division forms and the std approach
4613:
4614: @menu
4615: * Single precision::
4616: * Double precision:: Double-cell integer arithmetic
4617: * Bitwise operations::
4618: * Numeric comparison::
4619: * Mixed precision:: Operations with single and double-cell integers
4620: * Floating Point::
4621: @end menu
4622:
4623: @node Single precision, Double precision, Arithmetic, Arithmetic
4624: @subsection Single precision
4625: @cindex single precision arithmetic words
4626:
4627: @c !! cell undefined
4628:
4629: By default, numbers in Forth are single-precision integers that are one
4630: cell in size. They can be signed or unsigned, depending upon how you
4631: treat them. For the rules used by the text interpreter for recognising
4632: single-precision integers see @ref{Number Conversion}.
4633:
4634: These words are all defined for signed operands, but some of them also
4635: work for unsigned numbers: @code{+}, @code{1+}, @code{-}, @code{1-},
4636: @code{*}.
4637:
4638: doc-+
4639: doc-1+
4640: doc-under+
4641: doc--
4642: doc-1-
4643: doc-*
4644: doc-/
4645: doc-mod
4646: doc-/mod
4647: doc-negate
4648: doc-abs
4649: doc-min
4650: doc-max
4651: doc-floored
4652:
4653:
4654: @node Double precision, Bitwise operations, Single precision, Arithmetic
4655: @subsection Double precision
4656: @cindex double precision arithmetic words
4657:
4658: For the rules used by the text interpreter for
4659: recognising double-precision integers, see @ref{Number Conversion}.
4660:
4661: A double precision number is represented by a cell pair, with the most
4662: significant cell at the TOS. It is trivial to convert an unsigned single
4663: to a double: simply push a @code{0} onto the TOS. Since numbers are
4664: represented by Gforth using 2's complement arithmetic, converting a
4665: signed single to a (signed) double requires sign-extension across the
4666: most significant cell. This can be achieved using @code{s>d}. The moral
4667: of the story is that you cannot convert a number without knowing whether
4668: it represents an unsigned or a signed number.
4669:
4670: These words are all defined for signed operands, but some of them also
4671: work for unsigned numbers: @code{d+}, @code{d-}.
4672:
4673: doc-s>d
4674: doc-d>s
4675: doc-d+
4676: doc-d-
4677: doc-dnegate
4678: doc-dabs
4679: doc-dmin
4680: doc-dmax
4681:
4682:
4683: @node Bitwise operations, Numeric comparison, Double precision, Arithmetic
4684: @subsection Bitwise operations
4685: @cindex bitwise operation words
4686:
4687:
4688: doc-and
4689: doc-or
4690: doc-xor
4691: doc-invert
4692: doc-lshift
4693: doc-rshift
4694: doc-2*
4695: doc-d2*
4696: doc-2/
4697: doc-d2/
4698:
4699:
4700: @node Numeric comparison, Mixed precision, Bitwise operations, Arithmetic
4701: @subsection Numeric comparison
4702: @cindex numeric comparison words
4703:
4704: Note that the words that compare for equality (@code{= <> 0= 0<> d= d<>
4705: d0= d0<>}) work for for both signed and unsigned numbers.
4706:
4707: doc-<
4708: doc-<=
4709: doc-<>
4710: doc-=
4711: doc->
4712: doc->=
4713:
4714: doc-0<
4715: doc-0<=
4716: doc-0<>
4717: doc-0=
4718: doc-0>
4719: doc-0>=
4720:
4721: doc-u<
4722: doc-u<=
4723: @c u<> and u= exist but are the same as <> and =
4724: @c doc-u<>
4725: @c doc-u=
4726: doc-u>
4727: doc-u>=
4728:
4729: doc-within
4730:
4731: doc-d<
4732: doc-d<=
4733: doc-d<>
4734: doc-d=
4735: doc-d>
4736: doc-d>=
4737:
4738: doc-d0<
4739: doc-d0<=
4740: doc-d0<>
4741: doc-d0=
4742: doc-d0>
4743: doc-d0>=
4744:
4745: doc-du<
4746: doc-du<=
4747: @c du<> and du= exist but are the same as d<> and d=
4748: @c doc-du<>
4749: @c doc-du=
4750: doc-du>
4751: doc-du>=
4752:
4753:
4754: @node Mixed precision, Floating Point, Numeric comparison, Arithmetic
4755: @subsection Mixed precision
4756: @cindex mixed precision arithmetic words
4757:
4758:
4759: doc-m+
4760: doc-*/
4761: doc-*/mod
4762: doc-m*
4763: doc-um*
4764: doc-m*/
4765: doc-um/mod
4766: doc-fm/mod
4767: doc-sm/rem
4768:
4769:
4770: @node Floating Point, , Mixed precision, Arithmetic
4771: @subsection Floating Point
4772: @cindex floating point arithmetic words
4773:
4774: For the rules used by the text interpreter for
4775: recognising floating-point numbers see @ref{Number Conversion}.
4776:
4777: Gforth has a separate floating point stack, but the documentation uses
4778: the unified notation.@footnote{It's easy to generate the separate
4779: notation from that by just separating the floating-point numbers out:
4780: e.g. @code{( n r1 u r2 -- r3 )} becomes @code{( n u -- ) ( F: r1 r2 --
4781: r3 )}.}
4782:
4783: @cindex floating-point arithmetic, pitfalls
4784: Floating point numbers have a number of unpleasant surprises for the
4785: unwary (e.g., floating point addition is not associative) and even a
4786: few for the wary. You should not use them unless you know what you are
4787: doing or you don't care that the results you get are totally bogus. If
4788: you want to learn about the problems of floating point numbers (and
4789: how to avoid them), you might start with @cite{David Goldberg,
4790: @uref{http://docs.sun.com/source/806-3568/ncg_goldberg.html,What Every
4791: Computer Scientist Should Know About Floating-Point Arithmetic}, ACM
4792: Computing Surveys 23(1):5@minus{}48, March 1991}.
4793:
4794:
4795: doc-d>f
4796: doc-f>d
4797: doc-f+
4798: doc-f-
4799: doc-f*
4800: doc-f/
4801: doc-fnegate
4802: doc-fabs
4803: doc-fmax
4804: doc-fmin
4805: doc-floor
4806: doc-fround
4807: doc-f**
4808: doc-fsqrt
4809: doc-fexp
4810: doc-fexpm1
4811: doc-fln
4812: doc-flnp1
4813: doc-flog
4814: doc-falog
4815: doc-f2*
4816: doc-f2/
4817: doc-1/f
4818: doc-precision
4819: doc-set-precision
4820:
4821: @cindex angles in trigonometric operations
4822: @cindex trigonometric operations
4823: Angles in floating point operations are given in radians (a full circle
4824: has 2 pi radians).
4825:
4826: doc-fsin
4827: doc-fcos
4828: doc-fsincos
4829: doc-ftan
4830: doc-fasin
4831: doc-facos
4832: doc-fatan
4833: doc-fatan2
4834: doc-fsinh
4835: doc-fcosh
4836: doc-ftanh
4837: doc-fasinh
4838: doc-facosh
4839: doc-fatanh
4840: doc-pi
4841:
4842: @cindex equality of floats
4843: @cindex floating-point comparisons
4844: One particular problem with floating-point arithmetic is that comparison
4845: for equality often fails when you would expect it to succeed. For this
4846: reason approximate equality is often preferred (but you still have to
4847: know what you are doing). Also note that IEEE NaNs may compare
4848: differently from what you might expect. The comparison words are:
4849:
4850: doc-f~rel
4851: doc-f~abs
4852: doc-f~
4853: doc-f=
4854: doc-f<>
4855:
4856: doc-f<
4857: doc-f<=
4858: doc-f>
4859: doc-f>=
4860:
4861: doc-f0<
4862: doc-f0<=
4863: doc-f0<>
4864: doc-f0=
4865: doc-f0>
4866: doc-f0>=
4867:
4868:
4869: @node Stack Manipulation, Memory, Arithmetic, Words
4870: @section Stack Manipulation
4871: @cindex stack manipulation words
4872:
4873: @cindex floating-point stack in the standard
4874: Gforth maintains a number of separate stacks:
4875:
4876: @cindex data stack
4877: @cindex parameter stack
4878: @itemize @bullet
4879: @item
4880: A data stack (also known as the @dfn{parameter stack}) -- for
4881: characters, cells, addresses, and double cells.
4882:
4883: @cindex floating-point stack
4884: @item
4885: A floating point stack -- for holding floating point (FP) numbers.
4886:
4887: @cindex return stack
4888: @item
4889: A return stack -- for holding the return addresses of colon
4890: definitions and other (non-FP) data.
4891:
4892: @cindex locals stack
4893: @item
4894: A locals stack -- for holding local variables.
4895: @end itemize
4896:
4897: @menu
4898: * Data stack::
4899: * Floating point stack::
4900: * Return stack::
4901: * Locals stack::
4902: * Stack pointer manipulation::
4903: @end menu
4904:
4905: @node Data stack, Floating point stack, Stack Manipulation, Stack Manipulation
4906: @subsection Data stack
4907: @cindex data stack manipulation words
4908: @cindex stack manipulations words, data stack
4909:
4910:
4911: doc-drop
4912: doc-nip
4913: doc-dup
4914: doc-over
4915: doc-tuck
4916: doc-swap
4917: doc-pick
4918: doc-rot
4919: doc--rot
4920: doc-?dup
4921: doc-roll
4922: doc-2drop
4923: doc-2nip
4924: doc-2dup
4925: doc-2over
4926: doc-2tuck
4927: doc-2swap
4928: doc-2rot
4929:
4930:
4931: @node Floating point stack, Return stack, Data stack, Stack Manipulation
4932: @subsection Floating point stack
4933: @cindex floating-point stack manipulation words
4934: @cindex stack manipulation words, floating-point stack
4935:
4936: Whilst every sane Forth has a separate floating-point stack, it is not
4937: strictly required; an ANS Forth system could theoretically keep
4938: floating-point numbers on the data stack. As an additional difficulty,
4939: you don't know how many cells a floating-point number takes. It is
4940: reportedly possible to write words in a way that they work also for a
4941: unified stack model, but we do not recommend trying it. Instead, just
4942: say that your program has an environmental dependency on a separate
4943: floating-point stack.
4944:
4945: doc-floating-stack
4946:
4947: doc-fdrop
4948: doc-fnip
4949: doc-fdup
4950: doc-fover
4951: doc-ftuck
4952: doc-fswap
4953: doc-fpick
4954: doc-frot
4955:
4956:
4957: @node Return stack, Locals stack, Floating point stack, Stack Manipulation
4958: @subsection Return stack
4959: @cindex return stack manipulation words
4960: @cindex stack manipulation words, return stack
4961:
4962: @cindex return stack and locals
4963: @cindex locals and return stack
4964: A Forth system is allowed to keep local variables on the
4965: return stack. This is reasonable, as local variables usually eliminate
4966: the need to use the return stack explicitly. So, if you want to produce
4967: a standard compliant program and you are using local variables in a
4968: word, forget about return stack manipulations in that word (refer to the
4969: standard document for the exact rules).
4970:
4971: doc->r
4972: doc-r>
4973: doc-r@
4974: doc-rdrop
4975: doc-2>r
4976: doc-2r>
4977: doc-2r@
4978: doc-2rdrop
4979:
4980:
4981: @node Locals stack, Stack pointer manipulation, Return stack, Stack Manipulation
4982: @subsection Locals stack
4983:
4984: Gforth uses an extra locals stack. It is described, along with the
4985: reasons for its existence, in @ref{Locals implementation}.
4986:
4987: @node Stack pointer manipulation, , Locals stack, Stack Manipulation
4988: @subsection Stack pointer manipulation
4989: @cindex stack pointer manipulation words
4990:
4991: @c removed s0 r0 l0 -- they are obsolete aliases for sp0 rp0 lp0
4992: doc-sp0
4993: doc-sp@
4994: doc-sp!
4995: doc-fp0
4996: doc-fp@
4997: doc-fp!
4998: doc-rp0
4999: doc-rp@
5000: doc-rp!
5001: doc-lp0
5002: doc-lp@
5003: doc-lp!
5004:
5005:
5006: @node Memory, Control Structures, Stack Manipulation, Words
5007: @section Memory
5008: @cindex memory words
5009:
5010: @menu
5011: * Memory model::
5012: * Dictionary allocation::
5013: * Heap Allocation::
5014: * Memory Access::
5015: * Address arithmetic::
5016: * Memory Blocks::
5017: @end menu
5018:
5019: In addition to the standard Forth memory allocation words, there is also
5020: a @uref{http://www.complang.tuwien.ac.at/forth/garbage-collection.zip,
5021: garbage collector}.
5022:
5023: @node Memory model, Dictionary allocation, Memory, Memory
5024: @subsection ANS Forth and Gforth memory models
5025:
5026: @c The ANS Forth description is a mess (e.g., is the heap part of
5027: @c the dictionary?), so let's not stick to closely with it.
5028:
5029: ANS Forth considers a Forth system as consisting of several address
5030: spaces, of which only @dfn{data space} is managed and accessible with
5031: the memory words. Memory not necessarily in data space includes the
5032: stacks, the code (called code space) and the headers (called name
5033: space). In Gforth everything is in data space, but the code for the
5034: primitives is usually read-only.
5035:
5036: Data space is divided into a number of areas: The (data space portion of
5037: the) dictionary@footnote{Sometimes, the term @dfn{dictionary} is used to
5038: refer to the search data structure embodied in word lists and headers,
5039: because it is used for looking up names, just as you would in a
5040: conventional dictionary.}, the heap, and a number of system-allocated
5041: buffers.
5042:
5043: @cindex address arithmetic restrictions, ANS vs. Gforth
5044: @cindex contiguous regions, ANS vs. Gforth
5045: In ANS Forth data space is also divided into contiguous regions. You
5046: can only use address arithmetic within a contiguous region, not between
5047: them. Usually each allocation gives you one contiguous region, but the
5048: dictionary allocation words have additional rules (@pxref{Dictionary
5049: allocation}).
5050:
5051: Gforth provides one big address space, and address arithmetic can be
5052: performed between any addresses. However, in the dictionary headers or
5053: code are interleaved with data, so almost the only contiguous data space
5054: regions there are those described by ANS Forth as contiguous; but you
5055: can be sure that the dictionary is allocated towards increasing
5056: addresses even between contiguous regions. The memory order of
5057: allocations in the heap is platform-dependent (and possibly different
5058: from one run to the next).
5059:
5060:
5061: @node Dictionary allocation, Heap Allocation, Memory model, Memory
5062: @subsection Dictionary allocation
5063: @cindex reserving data space
5064: @cindex data space - reserving some
5065:
5066: Dictionary allocation is a stack-oriented allocation scheme, i.e., if
5067: you want to deallocate X, you also deallocate everything
5068: allocated after X.
5069:
5070: @cindex contiguous regions in dictionary allocation
5071: The allocations using the words below are contiguous and grow the region
5072: towards increasing addresses. Other words that allocate dictionary
5073: memory of any kind (i.e., defining words including @code{:noname}) end
5074: the contiguous region and start a new one.
5075:
5076: In ANS Forth only @code{create}d words are guaranteed to produce an
5077: address that is the start of the following contiguous region. In
5078: particular, the cell allocated by @code{variable} is not guaranteed to
5079: be contiguous with following @code{allot}ed memory.
5080:
5081: You can deallocate memory by using @code{allot} with a negative argument
5082: (with some restrictions, see @code{allot}). For larger deallocations use
5083: @code{marker}.
5084:
5085:
5086: doc-here
5087: doc-unused
5088: doc-allot
5089: doc-c,
5090: doc-f,
5091: doc-,
5092: doc-2,
5093:
5094: Memory accesses have to be aligned (@pxref{Address arithmetic}). So of
5095: course you should allocate memory in an aligned way, too. I.e., before
5096: allocating allocating a cell, @code{here} must be cell-aligned, etc.
5097: The words below align @code{here} if it is not already. Basically it is
5098: only already aligned for a type, if the last allocation was a multiple
5099: of the size of this type and if @code{here} was aligned for this type
5100: before.
5101:
5102: After freshly @code{create}ing a word, @code{here} is @code{align}ed in
5103: ANS Forth (@code{maxalign}ed in Gforth).
5104:
5105: doc-align
5106: doc-falign
5107: doc-sfalign
5108: doc-dfalign
5109: doc-maxalign
5110: doc-cfalign
5111:
5112:
5113: @node Heap Allocation, Memory Access, Dictionary allocation, Memory
5114: @subsection Heap allocation
5115: @cindex heap allocation
5116: @cindex dynamic allocation of memory
5117: @cindex memory-allocation word set
5118:
5119: @cindex contiguous regions and heap allocation
5120: Heap allocation supports deallocation of allocated memory in any
5121: order. Dictionary allocation is not affected by it (i.e., it does not
5122: end a contiguous region). In Gforth, these words are implemented using
5123: the standard C library calls malloc(), free() and resize().
5124:
5125: The memory region produced by one invocation of @code{allocate} or
5126: @code{resize} is internally contiguous. There is no contiguity between
5127: such a region and any other region (including others allocated from the
5128: heap).
5129:
5130: doc-allocate
5131: doc-free
5132: doc-resize
5133:
5134:
5135: @node Memory Access, Address arithmetic, Heap Allocation, Memory
5136: @subsection Memory Access
5137: @cindex memory access words
5138:
5139: doc-@
5140: doc-!
5141: doc-+!
5142: doc-c@
5143: doc-c!
5144: doc-2@
5145: doc-2!
5146: doc-f@
5147: doc-f!
5148: doc-sf@
5149: doc-sf!
5150: doc-df@
5151: doc-df!
5152: doc-sw@
5153: doc-uw@
5154: doc-w!
5155: doc-sl@
5156: doc-ul@
5157: doc-l!
5158:
5159: @node Address arithmetic, Memory Blocks, Memory Access, Memory
5160: @subsection Address arithmetic
5161: @cindex address arithmetic words
5162:
5163: Address arithmetic is the foundation on which you can build data
5164: structures like arrays, records (@pxref{Structures}) and objects
5165: (@pxref{Object-oriented Forth}).
5166:
5167: @cindex address unit
5168: @cindex au (address unit)
5169: ANS Forth does not specify the sizes of the data types. Instead, it
5170: offers a number of words for computing sizes and doing address
5171: arithmetic. Address arithmetic is performed in terms of address units
5172: (aus); on most systems the address unit is one byte. Note that a
5173: character may have more than one au, so @code{chars} is no noop (on
5174: platforms where it is a noop, it compiles to nothing).
5175:
5176: The basic address arithmetic words are @code{+} and @code{-}. E.g., if
5177: you have the address of a cell, perform @code{1 cells +}, and you will
5178: have the address of the next cell.
5179:
5180: @cindex contiguous regions and address arithmetic
5181: In ANS Forth you can perform address arithmetic only within a contiguous
5182: region, i.e., if you have an address into one region, you can only add
5183: and subtract such that the result is still within the region; you can
5184: only subtract or compare addresses from within the same contiguous
5185: region. Reasons: several contiguous regions can be arranged in memory
5186: in any way; on segmented systems addresses may have unusual
5187: representations, such that address arithmetic only works within a
5188: region. Gforth provides a few more guarantees (linear address space,
5189: dictionary grows upwards), but in general I have found it easy to stay
5190: within contiguous regions (exception: computing and comparing to the
5191: address just beyond the end of an array).
5192:
5193: @cindex alignment of addresses for types
5194: ANS Forth also defines words for aligning addresses for specific
5195: types. Many computers require that accesses to specific data types
5196: must only occur at specific addresses; e.g., that cells may only be
5197: accessed at addresses divisible by 4. Even if a machine allows unaligned
5198: accesses, it can usually perform aligned accesses faster.
5199:
5200: For the performance-conscious: alignment operations are usually only
5201: necessary during the definition of a data structure, not during the
5202: (more frequent) accesses to it.
5203:
5204: ANS Forth defines no words for character-aligning addresses. This is not
5205: an oversight, but reflects the fact that addresses that are not
5206: char-aligned have no use in the standard and therefore will not be
5207: created.
5208:
5209: @cindex @code{CREATE} and alignment
5210: ANS Forth guarantees that addresses returned by @code{CREATE}d words
5211: are cell-aligned; in addition, Gforth guarantees that these addresses
5212: are aligned for all purposes.
5213:
5214: Note that the ANS Forth word @code{char} has nothing to do with address
5215: arithmetic.
5216:
5217:
5218: doc-chars
5219: doc-char+
5220: doc-cells
5221: doc-cell+
5222: doc-cell
5223: doc-aligned
5224: doc-floats
5225: doc-float+
5226: doc-float
5227: doc-faligned
5228: doc-sfloats
5229: doc-sfloat+
5230: doc-sfaligned
5231: doc-dfloats
5232: doc-dfloat+
5233: doc-dfaligned
5234: doc-maxaligned
5235: doc-cfaligned
5236: doc-address-unit-bits
5237: doc-/w
5238: doc-/l
5239:
5240: @node Memory Blocks, , Address arithmetic, Memory
5241: @subsection Memory Blocks
5242: @cindex memory block words
5243: @cindex character strings - moving and copying
5244:
5245: Memory blocks often represent character strings; For ways of storing
5246: character strings in memory see @ref{String Formats}. For other
5247: string-processing words see @ref{Displaying characters and strings}.
5248:
5249: A few of these words work on address unit blocks. In that case, you
5250: usually have to insert @code{CHARS} before the word when working on
5251: character strings. Most words work on character blocks, and expect a
5252: char-aligned address.
5253:
5254: When copying characters between overlapping memory regions, use
5255: @code{chars move} or choose carefully between @code{cmove} and
5256: @code{cmove>}.
5257:
5258: doc-move
5259: doc-erase
5260: doc-cmove
5261: doc-cmove>
5262: doc-fill
5263: doc-blank
5264: doc-compare
5265: doc-str=
5266: doc-str<
5267: doc-string-prefix?
5268: doc-search
5269: doc--trailing
5270: doc-/string
5271: doc-bounds
5272: doc-pad
5273:
5274: @comment TODO examples
5275:
5276:
5277: @node Control Structures, Defining Words, Memory, Words
5278: @section Control Structures
5279: @cindex control structures
5280:
5281: Control structures in Forth cannot be used interpretively, only in a
5282: colon definition@footnote{To be precise, they have no interpretation
5283: semantics (@pxref{Interpretation and Compilation Semantics}).}. We do
5284: not like this limitation, but have not seen a satisfying way around it
5285: yet, although many schemes have been proposed.
5286:
5287: @menu
5288: * Selection:: IF ... ELSE ... ENDIF
5289: * Simple Loops:: BEGIN ...
5290: * Counted Loops:: DO
5291: * Arbitrary control structures::
5292: * Calls and returns::
5293: * Exception Handling::
5294: @end menu
5295:
5296: @node Selection, Simple Loops, Control Structures, Control Structures
5297: @subsection Selection
5298: @cindex selection control structures
5299: @cindex control structures for selection
5300:
5301: @cindex @code{IF} control structure
5302: @example
5303: @i{flag}
5304: IF
5305: @i{code}
5306: ENDIF
5307: @end example
5308: @noindent
5309:
5310: If @i{flag} is non-zero (as far as @code{IF} etc. are concerned, a cell
5311: with any bit set represents truth) @i{code} is executed.
5312:
5313: @example
5314: @i{flag}
5315: IF
5316: @i{code1}
5317: ELSE
5318: @i{code2}
5319: ENDIF
5320: @end example
5321:
5322: If @var{flag} is true, @i{code1} is executed, otherwise @i{code2} is
5323: executed.
5324:
5325: You can use @code{THEN} instead of @code{ENDIF}. Indeed, @code{THEN} is
5326: standard, and @code{ENDIF} is not, although it is quite popular. We
5327: recommend using @code{ENDIF}, because it is less confusing for people
5328: who also know other languages (and is not prone to reinforcing negative
5329: prejudices against Forth in these people). Adding @code{ENDIF} to a
5330: system that only supplies @code{THEN} is simple:
5331: @example
5332: : ENDIF POSTPONE then ; immediate
5333: @end example
5334:
5335: [According to @cite{Webster's New Encyclopedic Dictionary}, @dfn{then
5336: (adv.)} has the following meanings:
5337: @quotation
5338: ... 2b: following next after in order ... 3d: as a necessary consequence
5339: (if you were there, then you saw them).
5340: @end quotation
5341: Forth's @code{THEN} has the meaning 2b, whereas @code{THEN} in Pascal
5342: and many other programming languages has the meaning 3d.]
5343:
5344: Gforth also provides the words @code{?DUP-IF} and @code{?DUP-0=-IF}, so
5345: you can avoid using @code{?dup}. Using these alternatives is also more
5346: efficient than using @code{?dup}. Definitions in ANS Forth
5347: for @code{ENDIF}, @code{?DUP-IF} and @code{?DUP-0=-IF} are provided in
5348: @file{compat/control.fs}.
5349:
5350: @cindex @code{CASE} control structure
5351: @example
5352: @i{x}
5353: CASE
5354: @i{x1} OF @i{code1} ENDOF
5355: @i{x2} OF @i{code2} ENDOF
5356: @dots{}
5357: ( x ) @i{default-code} ( x )
5358: ENDCASE ( )
5359: @end example
5360:
5361: Executes the first @i{codei}, where the @i{xi} is equal to @i{x}. If no
5362: @i{xi} matches, the optional @i{default-code} is executed. The optional
5363: default case can be added by simply writing the code after the last
5364: @code{ENDOF}. It may use @i{x}, which is on top of the stack, but must
5365: not consume it. The value @i{x} is consumed by this construction
5366: (either by an @code{OF} that matches, or by the @code{ENDCASE}, if no OF
5367: matches). Example:
5368:
5369: @example
5370: : num-name ( n -- c-addr u )
5371: case
5372: 0 of s" zero " endof
5373: 1 of s" one " endof
5374: 2 of s" two " endof
5375: \ default case:
5376: s" other number"
5377: rot \ get n on top so ENDCASE can drop it
5378: endcase ;
5379: @end example
5380:
5381: @progstyle
5382: To keep the code understandable, you should ensure that you change the
5383: stack in the same way (wrt. number and types of stack items consumed
5384: and pushed) on all paths through a selection construct.
5385:
5386: @node Simple Loops, Counted Loops, Selection, Control Structures
5387: @subsection Simple Loops
5388: @cindex simple loops
5389: @cindex loops without count
5390:
5391: @cindex @code{WHILE} loop
5392: @example
5393: BEGIN
5394: @i{code1}
5395: @i{flag}
5396: WHILE
5397: @i{code2}
5398: REPEAT
5399: @end example
5400:
5401: @i{code1} is executed and @i{flag} is computed. If it is true,
5402: @i{code2} is executed and the loop is restarted; If @i{flag} is
5403: false, execution continues after the @code{REPEAT}.
5404:
5405: @cindex @code{UNTIL} loop
5406: @example
5407: BEGIN
5408: @i{code}
5409: @i{flag}
5410: UNTIL
5411: @end example
5412:
5413: @i{code} is executed. The loop is restarted if @code{flag} is false.
5414:
5415: @progstyle
5416: To keep the code understandable, a complete iteration of the loop should
5417: not change the number and types of the items on the stacks.
5418:
5419: @cindex endless loop
5420: @cindex loops, endless
5421: @example
5422: BEGIN
5423: @i{code}
5424: AGAIN
5425: @end example
5426:
5427: This is an endless loop.
5428:
5429: @node Counted Loops, Arbitrary control structures, Simple Loops, Control Structures
5430: @subsection Counted Loops
5431: @cindex counted loops
5432: @cindex loops, counted
5433: @cindex @code{DO} loops
5434:
5435: The basic counted loop is:
5436: @example
5437: @i{limit} @i{start}
5438: ?DO
5439: @i{body}
5440: LOOP
5441: @end example
5442:
5443: This performs one iteration for every integer, starting from @i{start}
5444: and up to, but excluding @i{limit}. The counter, or @i{index}, can be
5445: accessed with @code{i}. For example, the loop:
5446: @example
5447: 10 0 ?DO
5448: i .
5449: LOOP
5450: @end example
5451: @noindent
5452: prints @code{0 1 2 3 4 5 6 7 8 9}
5453:
5454: The index of the innermost loop can be accessed with @code{i}, the index
5455: of the next loop with @code{j}, and the index of the third loop with
5456: @code{k}.
5457:
5458:
5459: doc-i
5460: doc-j
5461: doc-k
5462:
5463:
5464: The loop control data are kept on the return stack, so there are some
5465: restrictions on mixing return stack accesses and counted loop words. In
5466: particuler, if you put values on the return stack outside the loop, you
5467: cannot read them inside the loop@footnote{well, not in a way that is
5468: portable.}. If you put values on the return stack within a loop, you
5469: have to remove them before the end of the loop and before accessing the
5470: index of the loop.
5471:
5472: There are several variations on the counted loop:
5473:
5474: @itemize @bullet
5475: @item
5476: @code{LEAVE} leaves the innermost counted loop immediately; execution
5477: continues after the associated @code{LOOP} or @code{NEXT}. For example:
5478:
5479: @example
5480: 10 0 ?DO i DUP . 3 = IF LEAVE THEN LOOP
5481: @end example
5482: prints @code{0 1 2 3}
5483:
5484:
5485: @item
5486: @code{UNLOOP} prepares for an abnormal loop exit, e.g., via
5487: @code{EXIT}. @code{UNLOOP} removes the loop control parameters from the
5488: return stack so @code{EXIT} can get to its return address. For example:
5489:
5490: @example
5491: : demo 10 0 ?DO i DUP . 3 = IF UNLOOP EXIT THEN LOOP ." Done" ;
5492: @end example
5493: prints @code{0 1 2 3}
5494:
5495:
5496: @item
5497: If @i{start} is greater than @i{limit}, a @code{?DO} loop is entered
5498: (and @code{LOOP} iterates until they become equal by wrap-around
5499: arithmetic). This behaviour is usually not what you want. Therefore,
5500: Gforth offers @code{+DO} and @code{U+DO} (as replacements for
5501: @code{?DO}), which do not enter the loop if @i{start} is greater than
5502: @i{limit}; @code{+DO} is for signed loop parameters, @code{U+DO} for
5503: unsigned loop parameters.
5504:
5505: @item
5506: @code{?DO} can be replaced by @code{DO}. @code{DO} always enters
5507: the loop, independent of the loop parameters. Do not use @code{DO}, even
5508: if you know that the loop is entered in any case. Such knowledge tends
5509: to become invalid during maintenance of a program, and then the
5510: @code{DO} will make trouble.
5511:
5512: @item
5513: @code{LOOP} can be replaced with @code{@i{n} +LOOP}; this updates the
5514: index by @i{n} instead of by 1. The loop is terminated when the border
5515: between @i{limit-1} and @i{limit} is crossed. E.g.:
5516:
5517: @example
5518: 4 0 +DO i . 2 +LOOP
5519: @end example
5520: @noindent
5521: prints @code{0 2}
5522:
5523: @example
5524: 4 1 +DO i . 2 +LOOP
5525: @end example
5526: @noindent
5527: prints @code{1 3}
5528:
5529: @item
5530: @cindex negative increment for counted loops
5531: @cindex counted loops with negative increment
5532: The behaviour of @code{@i{n} +LOOP} is peculiar when @i{n} is negative:
5533:
5534: @example
5535: -1 0 ?DO i . -1 +LOOP
5536: @end example
5537: @noindent
5538: prints @code{0 -1}
5539:
5540: @example
5541: 0 0 ?DO i . -1 +LOOP
5542: @end example
5543: prints nothing.
5544:
5545: Therefore we recommend avoiding @code{@i{n} +LOOP} with negative
5546: @i{n}. One alternative is @code{@i{u} -LOOP}, which reduces the
5547: index by @i{u} each iteration. The loop is terminated when the border
5548: between @i{limit+1} and @i{limit} is crossed. Gforth also provides
5549: @code{-DO} and @code{U-DO} for down-counting loops. E.g.:
5550:
5551: @example
5552: -2 0 -DO i . 1 -LOOP
5553: @end example
5554: @noindent
5555: prints @code{0 -1}
5556:
5557: @example
5558: -1 0 -DO i . 1 -LOOP
5559: @end example
5560: @noindent
5561: prints @code{0}
5562:
5563: @example
5564: 0 0 -DO i . 1 -LOOP
5565: @end example
5566: @noindent
5567: prints nothing.
5568:
5569: @end itemize
5570:
5571: Unfortunately, @code{+DO}, @code{U+DO}, @code{-DO}, @code{U-DO} and
5572: @code{-LOOP} are not defined in ANS Forth. However, an implementation
5573: for these words that uses only standard words is provided in
5574: @file{compat/loops.fs}.
5575:
5576:
5577: @cindex @code{FOR} loops
5578: Another counted loop is:
5579: @example
5580: @i{n}
5581: FOR
5582: @i{body}
5583: NEXT
5584: @end example
5585: This is the preferred loop of native code compiler writers who are too
5586: lazy to optimize @code{?DO} loops properly. This loop structure is not
5587: defined in ANS Forth. In Gforth, this loop iterates @i{n+1} times;
5588: @code{i} produces values starting with @i{n} and ending with 0. Other
5589: Forth systems may behave differently, even if they support @code{FOR}
5590: loops. To avoid problems, don't use @code{FOR} loops.
5591:
5592: @node Arbitrary control structures, Calls and returns, Counted Loops, Control Structures
5593: @subsection Arbitrary control structures
5594: @cindex control structures, user-defined
5595:
5596: @cindex control-flow stack
5597: ANS Forth permits and supports using control structures in a non-nested
5598: way. Information about incomplete control structures is stored on the
5599: control-flow stack. This stack may be implemented on the Forth data
5600: stack, and this is what we have done in Gforth.
5601:
5602: @cindex @code{orig}, control-flow stack item
5603: @cindex @code{dest}, control-flow stack item
5604: An @i{orig} entry represents an unresolved forward branch, a @i{dest}
5605: entry represents a backward branch target. A few words are the basis for
5606: building any control structure possible (except control structures that
5607: need storage, like calls, coroutines, and backtracking).
5608:
5609:
5610: doc-if
5611: doc-ahead
5612: doc-then
5613: doc-begin
5614: doc-until
5615: doc-again
5616: doc-cs-pick
5617: doc-cs-roll
5618:
5619:
5620: The Standard words @code{CS-PICK} and @code{CS-ROLL} allow you to
5621: manipulate the control-flow stack in a portable way. Without them, you
5622: would need to know how many stack items are occupied by a control-flow
5623: entry (many systems use one cell. In Gforth they currently take three,
5624: but this may change in the future).
5625:
5626: Some standard control structure words are built from these words:
5627:
5628:
5629: doc-else
5630: doc-while
5631: doc-repeat
5632:
5633:
5634: @noindent
5635: Gforth adds some more control-structure words:
5636:
5637:
5638: doc-endif
5639: doc-?dup-if
5640: doc-?dup-0=-if
5641:
5642:
5643: @noindent
5644: Counted loop words constitute a separate group of words:
5645:
5646:
5647: doc-?do
5648: doc-+do
5649: doc-u+do
5650: doc--do
5651: doc-u-do
5652: doc-do
5653: doc-for
5654: doc-loop
5655: doc-+loop
5656: doc--loop
5657: doc-next
5658: doc-leave
5659: doc-?leave
5660: doc-unloop
5661: doc-done
5662:
5663:
5664: The standard does not allow using @code{CS-PICK} and @code{CS-ROLL} on
5665: @i{do-sys}. Gforth allows it, but it's your job to ensure that for
5666: every @code{?DO} etc. there is exactly one @code{UNLOOP} on any path
5667: through the definition (@code{LOOP} etc. compile an @code{UNLOOP} on the
5668: fall-through path). Also, you have to ensure that all @code{LEAVE}s are
5669: resolved (by using one of the loop-ending words or @code{DONE}).
5670:
5671: @noindent
5672: Another group of control structure words are:
5673:
5674:
5675: doc-case
5676: doc-endcase
5677: doc-of
5678: doc-endof
5679:
5680:
5681: @i{case-sys} and @i{of-sys} cannot be processed using @code{CS-PICK} and
5682: @code{CS-ROLL}.
5683:
5684: @subsubsection Programming Style
5685: @cindex control structures programming style
5686: @cindex programming style, arbitrary control structures
5687:
5688: In order to ensure readability we recommend that you do not create
5689: arbitrary control structures directly, but define new control structure
5690: words for the control structure you want and use these words in your
5691: program. For example, instead of writing:
5692:
5693: @example
5694: BEGIN
5695: ...
5696: IF [ 1 CS-ROLL ]
5697: ...
5698: AGAIN THEN
5699: @end example
5700:
5701: @noindent
5702: we recommend defining control structure words, e.g.,
5703:
5704: @example
5705: : WHILE ( DEST -- ORIG DEST )
5706: POSTPONE IF
5707: 1 CS-ROLL ; immediate
5708:
5709: : REPEAT ( orig dest -- )
5710: POSTPONE AGAIN
5711: POSTPONE THEN ; immediate
5712: @end example
5713:
5714: @noindent
5715: and then using these to create the control structure:
5716:
5717: @example
5718: BEGIN
5719: ...
5720: WHILE
5721: ...
5722: REPEAT
5723: @end example
5724:
5725: That's much easier to read, isn't it? Of course, @code{REPEAT} and
5726: @code{WHILE} are predefined, so in this example it would not be
5727: necessary to define them.
5728:
5729: @node Calls and returns, Exception Handling, Arbitrary control structures, Control Structures
5730: @subsection Calls and returns
5731: @cindex calling a definition
5732: @cindex returning from a definition
5733:
5734: @cindex recursive definitions
5735: A definition can be called simply be writing the name of the definition
5736: to be called. Normally a definition is invisible during its own
5737: definition. If you want to write a directly recursive definition, you
5738: can use @code{recursive} to make the current definition visible, or
5739: @code{recurse} to call the current definition directly.
5740:
5741:
5742: doc-recursive
5743: doc-recurse
5744:
5745:
5746: @comment TODO add example of the two recursion methods
5747: @quotation
5748: @progstyle
5749: I prefer using @code{recursive} to @code{recurse}, because calling the
5750: definition by name is more descriptive (if the name is well-chosen) than
5751: the somewhat cryptic @code{recurse}. E.g., in a quicksort
5752: implementation, it is much better to read (and think) ``now sort the
5753: partitions'' than to read ``now do a recursive call''.
5754: @end quotation
5755:
5756: For mutual recursion, use @code{Defer}red words, like this:
5757:
5758: @example
5759: Defer foo
5760:
5761: : bar ( ... -- ... )
5762: ... foo ... ;
5763:
5764: :noname ( ... -- ... )
5765: ... bar ... ;
5766: IS foo
5767: @end example
5768:
5769: Deferred words are discussed in more detail in @ref{Deferred Words}.
5770:
5771: The current definition returns control to the calling definition when
5772: the end of the definition is reached or @code{EXIT} is encountered.
5773:
5774: doc-exit
5775: doc-;s
5776:
5777:
5778: @node Exception Handling, , Calls and returns, Control Structures
5779: @subsection Exception Handling
5780: @cindex exceptions
5781:
5782: @c quit is a very bad idea for error handling,
5783: @c because it does not translate into a THROW
5784: @c it also does not belong into this chapter
5785:
5786: If a word detects an error condition that it cannot handle, it can
5787: @code{throw} an exception. In the simplest case, this will terminate
5788: your program, and report an appropriate error.
5789:
5790: doc-throw
5791:
5792: @code{Throw} consumes a cell-sized error number on the stack. There are
5793: some predefined error numbers in ANS Forth (see @file{errors.fs}). In
5794: Gforth (and most other systems) you can use the iors produced by various
5795: words as error numbers (e.g., a typical use of @code{allocate} is
5796: @code{allocate throw}). Gforth also provides the word @code{exception}
5797: to define your own error numbers (with decent error reporting); an ANS
5798: Forth version of this word (but without the error messages) is available
5799: in @code{compat/except.fs}. And finally, you can use your own error
5800: numbers (anything outside the range -4095..0), but won't get nice error
5801: messages, only numbers. For example, try:
5802:
5803: @example
5804: -10 throw \ ANS defined
5805: -267 throw \ system defined
5806: s" my error" exception throw \ user defined
5807: 7 throw \ arbitrary number
5808: @end example
5809:
5810: doc---exception-exception
5811:
5812: A common idiom to @code{THROW} a specific error if a flag is true is
5813: this:
5814:
5815: @example
5816: @code{( flag ) 0<> @i{errno} and throw}
5817: @end example
5818:
5819: Your program can provide exception handlers to catch exceptions. An
5820: exception handler can be used to correct the problem, or to clean up
5821: some data structures and just throw the exception to the next exception
5822: handler. Note that @code{throw} jumps to the dynamically innermost
5823: exception handler. The system's exception handler is outermost, and just
5824: prints an error and restarts command-line interpretation (or, in batch
5825: mode (i.e., while processing the shell command line), leaves Gforth).
5826:
5827: The ANS Forth way to catch exceptions is @code{catch}:
5828:
5829: doc-catch
5830: doc-nothrow
5831:
5832: The most common use of exception handlers is to clean up the state when
5833: an error happens. E.g.,
5834:
5835: @example
5836: base @ >r hex \ actually the hex should be inside foo, or we h
5837: ['] foo catch ( nerror|0 )
5838: r> base !
5839: ( nerror|0 ) throw \ pass it on
5840: @end example
5841:
5842: A use of @code{catch} for handling the error @code{myerror} might look
5843: like this:
5844:
5845: @example
5846: ['] foo catch
5847: CASE
5848: myerror OF ... ( do something about it ) nothrow ENDOF
5849: dup throw \ default: pass other errors on, do nothing on non-errors
5850: ENDCASE
5851: @end example
5852:
5853: Having to wrap the code into a separate word is often cumbersome,
5854: therefore Gforth provides an alternative syntax:
5855:
5856: @example
5857: TRY
5858: @i{code1}
5859: IFERROR
5860: @i{code2}
5861: THEN
5862: @i{code3}
5863: ENDTRY
5864: @end example
5865:
5866: This performs @i{code1}. If @i{code1} completes normally, execution
5867: continues with @i{code3}. If there is an exception in @i{code1} or
5868: before @code{endtry}, the stacks are reset to the depth during
5869: @code{try}, the throw value is pushed on the data stack, and execution
5870: constinues at @i{code2}, and finally falls through to @i{code3}.
5871:
5872: doc-try
5873: doc-endtry
5874: doc-iferror
5875:
5876: If you don't need @i{code2}, you can write @code{restore} instead of
5877: @code{iferror then}:
5878:
5879: @example
5880: TRY
5881: @i{code1}
5882: RESTORE
5883: @i{code3}
5884: ENDTRY
5885: @end example
5886:
5887: @cindex unwind-protect
5888: The cleanup example from above in this syntax:
5889:
5890: @example
5891: base @@ @{ oldbase @}
5892: TRY
5893: hex foo \ now the hex is placed correctly
5894: 0 \ value for throw
5895: RESTORE
5896: oldbase base !
5897: ENDTRY
5898: throw
5899: @end example
5900:
5901: An additional advantage of this variant is that an exception between
5902: @code{restore} and @code{endtry} (e.g., from the user pressing
5903: @kbd{Ctrl-C}) restarts the execution of the code after @code{restore},
5904: so the base will be restored under all circumstances.
5905:
5906: However, you have to ensure that this code does not cause an exception
5907: itself, otherwise the @code{iferror}/@code{restore} code will loop.
5908: Moreover, you should also make sure that the stack contents needed by
5909: the @code{iferror}/@code{restore} code exist everywhere between
5910: @code{try} and @code{endtry}; in our example this is achived by
5911: putting the data in a local before the @code{try} (you cannot use the
5912: return stack because the exception frame (@i{sys1}) is in the way
5913: there).
5914:
5915: This kind of usage corresponds to Lisp's @code{unwind-protect}.
5916:
5917: @cindex @code{recover} (old Gforth versions)
5918: If you do not want this exception-restarting behaviour, you achieve
5919: this as follows:
5920:
5921: @example
5922: TRY
5923: @i{code1}
5924: ENDTRY-IFERROR
5925: @i{code2}
5926: THEN
5927: @end example
5928:
5929: If there is an exception in @i{code1}, then @i{code2} is executed,
5930: otherwise execution continues behind the @code{then} (or in a possible
5931: @code{else} branch). This corresponds to the construct
5932:
5933: @example
5934: TRY
5935: @i{code1}
5936: RECOVER
5937: @i{code2}
5938: ENDTRY
5939: @end example
5940:
5941: in Gforth before version 0.7. So you can directly replace
5942: @code{recover}-using code; however, we recommend that you check if it
5943: would not be better to use one of the other @code{try} variants while
5944: you are at it.
5945:
5946: To ease the transition, Gforth provides two compatibility files:
5947: @file{endtry-iferror.fs} provides the @code{try ... endtry-iferror
5948: ... then} syntax (but not @code{iferror} or @code{restore}) for old
5949: systems; @file{recover-endtry.fs} provides the @code{try ... recover
5950: ... endtry} syntax on new systems, so you can use that file as a
5951: stopgap to run old programs. Both files work on any system (they just
5952: do nothing if the system already has the syntax it implements), so you
5953: can unconditionally @code{require} one of these files, even if you use
5954: a mix old and new systems.
5955:
5956: doc-restore
5957: doc-endtry-iferror
5958:
5959: Here's the error handling example:
5960:
5961: @example
5962: TRY
5963: foo
5964: ENDTRY-IFERROR
5965: CASE
5966: myerror OF ... ( do something about it ) nothrow ENDOF
5967: throw \ pass other errors on
5968: ENDCASE
5969: THEN
5970: @end example
5971:
5972: @progstyle
5973: As usual, you should ensure that the stack depth is statically known at
5974: the end: either after the @code{throw} for passing on errors, or after
5975: the @code{ENDTRY} (or, if you use @code{catch}, after the end of the
5976: selection construct for handling the error).
5977:
5978: There are two alternatives to @code{throw}: @code{Abort"} is conditional
5979: and you can provide an error message. @code{Abort} just produces an
5980: ``Aborted'' error.
5981:
5982: The problem with these words is that exception handlers cannot
5983: differentiate between different @code{abort"}s; they just look like
5984: @code{-2 throw} to them (the error message cannot be accessed by
5985: standard programs). Similar @code{abort} looks like @code{-1 throw} to
5986: exception handlers.
5987:
5988: doc-abort"
5989: doc-abort
5990:
5991:
5992:
5993: @c -------------------------------------------------------------
5994: @node Defining Words, Interpretation and Compilation Semantics, Control Structures, Words
5995: @section Defining Words
5996: @cindex defining words
5997:
5998: Defining words are used to extend Forth by creating new entries in the dictionary.
5999:
6000: @menu
6001: * CREATE::
6002: * Variables:: Variables and user variables
6003: * Constants::
6004: * Values:: Initialised variables
6005: * Colon Definitions::
6006: * Anonymous Definitions:: Definitions without names
6007: * Supplying names:: Passing definition names as strings
6008: * User-defined Defining Words::
6009: * Deferred Words:: Allow forward references
6010: * Aliases::
6011: @end menu
6012:
6013: @node CREATE, Variables, Defining Words, Defining Words
6014: @subsection @code{CREATE}
6015: @cindex simple defining words
6016: @cindex defining words, simple
6017:
6018: Defining words are used to create new entries in the dictionary. The
6019: simplest defining word is @code{CREATE}. @code{CREATE} is used like
6020: this:
6021:
6022: @example
6023: CREATE new-word1
6024: @end example
6025:
6026: @code{CREATE} is a parsing word, i.e., it takes an argument from the
6027: input stream (@code{new-word1} in our example). It generates a
6028: dictionary entry for @code{new-word1}. When @code{new-word1} is
6029: executed, all that it does is leave an address on the stack. The address
6030: represents the value of the data space pointer (@code{HERE}) at the time
6031: that @code{new-word1} was defined. Therefore, @code{CREATE} is a way of
6032: associating a name with the address of a region of memory.
6033:
6034: doc-create
6035:
6036: Note that in ANS Forth guarantees only for @code{create} that its body
6037: is in dictionary data space (i.e., where @code{here}, @code{allot}
6038: etc. work, @pxref{Dictionary allocation}). Also, in ANS Forth only
6039: @code{create}d words can be modified with @code{does>}
6040: (@pxref{User-defined Defining Words}). And in ANS Forth @code{>body}
6041: can only be applied to @code{create}d words.
6042:
6043: By extending this example to reserve some memory in data space, we end
6044: up with something like a @i{variable}. Here are two different ways to do
6045: it:
6046:
6047: @example
6048: CREATE new-word2 1 cells allot \ reserve 1 cell - initial value undefined
6049: CREATE new-word3 4 , \ reserve 1 cell and initialise it (to 4)
6050: @end example
6051:
6052: The variable can be examined and modified using @code{@@} (``fetch'') and
6053: @code{!} (``store'') like this:
6054:
6055: @example
6056: new-word2 @@ . \ get address, fetch from it and display
6057: 1234 new-word2 ! \ new value, get address, store to it
6058: @end example
6059:
6060: @cindex arrays
6061: A similar mechanism can be used to create arrays. For example, an
6062: 80-character text input buffer:
6063:
6064: @example
6065: CREATE text-buf 80 chars allot
6066:
6067: text-buf 0 chars + c@@ \ the 1st character (offset 0)
6068: text-buf 3 chars + c@@ \ the 4th character (offset 3)
6069: @end example
6070:
6071: You can build arbitrarily complex data structures by allocating
6072: appropriate areas of memory. For further discussions of this, and to
6073: learn about some Gforth tools that make it easier,
6074: @xref{Structures}.
6075:
6076:
6077: @node Variables, Constants, CREATE, Defining Words
6078: @subsection Variables
6079: @cindex variables
6080:
6081: The previous section showed how a sequence of commands could be used to
6082: generate a variable. As a final refinement, the whole code sequence can
6083: be wrapped up in a defining word (pre-empting the subject of the next
6084: section), making it easier to create new variables:
6085:
6086: @example
6087: : myvariableX ( "name" -- a-addr ) CREATE 1 cells allot ;
6088: : myvariable0 ( "name" -- a-addr ) CREATE 0 , ;
6089:
6090: myvariableX foo \ variable foo starts off with an unknown value
6091: myvariable0 joe \ whilst joe is initialised to 0
6092:
6093: 45 3 * foo ! \ set foo to 135
6094: 1234 joe ! \ set joe to 1234
6095: 3 joe +! \ increment joe by 3.. to 1237
6096: @end example
6097:
6098: Not surprisingly, there is no need to define @code{myvariable}, since
6099: Forth already has a definition @code{Variable}. ANS Forth does not
6100: guarantee that a @code{Variable} is initialised when it is created
6101: (i.e., it may behave like @code{myvariableX}). In contrast, Gforth's
6102: @code{Variable} initialises the variable to 0 (i.e., it behaves exactly
6103: like @code{myvariable0}). Forth also provides @code{2Variable} and
6104: @code{fvariable} for double and floating-point variables, respectively
6105: -- they are initialised to 0. and 0e in Gforth. If you use a @code{Variable} to
6106: store a boolean, you can use @code{on} and @code{off} to toggle its
6107: state.
6108:
6109: doc-variable
6110: doc-2variable
6111: doc-fvariable
6112:
6113: @cindex user variables
6114: @cindex user space
6115: The defining word @code{User} behaves in the same way as @code{Variable}.
6116: The difference is that it reserves space in @i{user (data) space} rather
6117: than normal data space. In a Forth system that has a multi-tasker, each
6118: task has its own set of user variables.
6119:
6120: doc-user
6121: @c doc-udp
6122: @c doc-uallot
6123:
6124: @comment TODO is that stuff about user variables strictly correct? Is it
6125: @comment just terminal tasks that have user variables?
6126: @comment should document tasker.fs (with some examples) elsewhere
6127: @comment in this manual, then expand on user space and user variables.
6128:
6129: @node Constants, Values, Variables, Defining Words
6130: @subsection Constants
6131: @cindex constants
6132:
6133: @code{Constant} allows you to declare a fixed value and refer to it by
6134: name. For example:
6135:
6136: @example
6137: 12 Constant INCHES-PER-FOOT
6138: 3E+08 fconstant SPEED-O-LIGHT
6139: @end example
6140:
6141: A @code{Variable} can be both read and written, so its run-time
6142: behaviour is to supply an address through which its current value can be
6143: manipulated. In contrast, the value of a @code{Constant} cannot be
6144: changed once it has been declared@footnote{Well, often it can be -- but
6145: not in a Standard, portable way. It's safer to use a @code{Value} (read
6146: on).} so it's not necessary to supply the address -- it is more
6147: efficient to return the value of the constant directly. That's exactly
6148: what happens; the run-time effect of a constant is to put its value on
6149: the top of the stack (You can find one
6150: way of implementing @code{Constant} in @ref{User-defined Defining Words}).
6151:
6152: Forth also provides @code{2Constant} and @code{fconstant} for defining
6153: double and floating-point constants, respectively.
6154:
6155: doc-constant
6156: doc-2constant
6157: doc-fconstant
6158:
6159: @c that's too deep, and it's not necessarily true for all ANS Forths. - anton
6160: @c nac-> How could that not be true in an ANS Forth? You can't define a
6161: @c constant, use it and then delete the definition of the constant..
6162:
6163: @c anton->An ANS Forth system can compile a constant to a literal; On
6164: @c decompilation you would see only the number, just as if it had been used
6165: @c in the first place. The word will stay, of course, but it will only be
6166: @c used by the text interpreter (no run-time duties, except when it is
6167: @c POSTPONEd or somesuch).
6168:
6169: @c nac:
6170: @c I agree that it's rather deep, but IMO it is an important difference
6171: @c relative to other programming languages.. often it's annoying: it
6172: @c certainly changes my programming style relative to C.
6173:
6174: @c anton: In what way?
6175:
6176: Constants in Forth behave differently from their equivalents in other
6177: programming languages. In other languages, a constant (such as an EQU in
6178: assembler or a #define in C) only exists at compile-time; in the
6179: executable program the constant has been translated into an absolute
6180: number and, unless you are using a symbolic debugger, it's impossible to
6181: know what abstract thing that number represents. In Forth a constant has
6182: an entry in the header space and remains there after the code that uses
6183: it has been defined. In fact, it must remain in the dictionary since it
6184: has run-time duties to perform. For example:
6185:
6186: @example
6187: 12 Constant INCHES-PER-FOOT
6188: : FEET-TO-INCHES ( n1 -- n2 ) INCHES-PER-FOOT * ;
6189: @end example
6190:
6191: @cindex in-lining of constants
6192: When @code{FEET-TO-INCHES} is executed, it will in turn execute the xt
6193: associated with the constant @code{INCHES-PER-FOOT}. If you use
6194: @code{see} to decompile the definition of @code{FEET-TO-INCHES}, you can
6195: see that it makes a call to @code{INCHES-PER-FOOT}. Some Forth compilers
6196: attempt to optimise constants by in-lining them where they are used. You
6197: can force Gforth to in-line a constant like this:
6198:
6199: @example
6200: : FEET-TO-INCHES ( n1 -- n2 ) [ INCHES-PER-FOOT ] LITERAL * ;
6201: @end example
6202:
6203: If you use @code{see} to decompile @i{this} version of
6204: @code{FEET-TO-INCHES}, you can see that @code{INCHES-PER-FOOT} is no
6205: longer present. To understand how this works, read
6206: @ref{Interpret/Compile states}, and @ref{Literals}.
6207:
6208: In-lining constants in this way might improve execution time
6209: fractionally, and can ensure that a constant is now only referenced at
6210: compile-time. However, the definition of the constant still remains in
6211: the dictionary. Some Forth compilers provide a mechanism for controlling
6212: a second dictionary for holding transient words such that this second
6213: dictionary can be deleted later in order to recover memory
6214: space. However, there is no standard way of doing this.
6215:
6216:
6217: @node Values, Colon Definitions, Constants, Defining Words
6218: @subsection Values
6219: @cindex values
6220:
6221: A @code{Value} behaves like a @code{Constant}, but it can be changed.
6222: @code{TO} is a parsing word that changes a @code{Values}. In Gforth
6223: (not in ANS Forth) you can access (and change) a @code{value} also with
6224: @code{>body}.
6225:
6226: Here are some
6227: examples:
6228:
6229: @example
6230: 12 Value APPLES \ Define APPLES with an initial value of 12
6231: 34 TO APPLES \ Change the value of APPLES. TO is a parsing word
6232: 1 ' APPLES >body +! \ Increment APPLES. Non-standard usage.
6233: APPLES \ puts 35 on the top of the stack.
6234: @end example
6235:
6236: doc-value
6237: doc-to
6238:
6239:
6240:
6241: @node Colon Definitions, Anonymous Definitions, Values, Defining Words
6242: @subsection Colon Definitions
6243: @cindex colon definitions
6244:
6245: @example
6246: : name ( ... -- ... )
6247: word1 word2 word3 ;
6248: @end example
6249:
6250: @noindent
6251: Creates a word called @code{name} that, upon execution, executes
6252: @code{word1 word2 word3}. @code{name} is a @dfn{(colon) definition}.
6253:
6254: The explanation above is somewhat superficial. For simple examples of
6255: colon definitions see @ref{Your first definition}. For an in-depth
6256: discussion of some of the issues involved, @xref{Interpretation and
6257: Compilation Semantics}.
6258:
6259: doc-:
6260: doc-;
6261:
6262:
6263: @node Anonymous Definitions, Supplying names, Colon Definitions, Defining Words
6264: @subsection Anonymous Definitions
6265: @cindex colon definitions
6266: @cindex defining words without name
6267:
6268: Sometimes you want to define an @dfn{anonymous word}; a word without a
6269: name. You can do this with:
6270:
6271: doc-:noname
6272:
6273: This leaves the execution token for the word on the stack after the
6274: closing @code{;}. Here's an example in which a deferred word is
6275: initialised with an @code{xt} from an anonymous colon definition:
6276:
6277: @example
6278: Defer deferred
6279: :noname ( ... -- ... )
6280: ... ;
6281: IS deferred
6282: @end example
6283:
6284: @noindent
6285: Gforth provides an alternative way of doing this, using two separate
6286: words:
6287:
6288: doc-noname
6289: @cindex execution token of last defined word
6290: doc-latestxt
6291:
6292: @noindent
6293: The previous example can be rewritten using @code{noname} and
6294: @code{latestxt}:
6295:
6296: @example
6297: Defer deferred
6298: noname : ( ... -- ... )
6299: ... ;
6300: latestxt IS deferred
6301: @end example
6302:
6303: @noindent
6304: @code{noname} works with any defining word, not just @code{:}.
6305:
6306: @code{latestxt} also works when the last word was not defined as
6307: @code{noname}. It does not work for combined words, though. It also has
6308: the useful property that is is valid as soon as the header for a
6309: definition has been built. Thus:
6310:
6311: @example
6312: latestxt . : foo [ latestxt . ] ; ' foo .
6313: @end example
6314:
6315: @noindent
6316: prints 3 numbers; the last two are the same.
6317:
6318: @node Supplying names, User-defined Defining Words, Anonymous Definitions, Defining Words
6319: @subsection Supplying the name of a defined word
6320: @cindex names for defined words
6321: @cindex defining words, name given in a string
6322:
6323: By default, a defining word takes the name for the defined word from the
6324: input stream. Sometimes you want to supply the name from a string. You
6325: can do this with:
6326:
6327: doc-nextname
6328:
6329: For example:
6330:
6331: @example
6332: s" foo" nextname create
6333: @end example
6334:
6335: @noindent
6336: is equivalent to:
6337:
6338: @example
6339: create foo
6340: @end example
6341:
6342: @noindent
6343: @code{nextname} works with any defining word.
6344:
6345:
6346: @node User-defined Defining Words, Deferred Words, Supplying names, Defining Words
6347: @subsection User-defined Defining Words
6348: @cindex user-defined defining words
6349: @cindex defining words, user-defined
6350:
6351: You can create a new defining word by wrapping defining-time code around
6352: an existing defining word and putting the sequence in a colon
6353: definition.
6354:
6355: @c anton: This example is very complex and leads in a quite different
6356: @c direction from the CREATE-DOES> stuff that follows. It should probably
6357: @c be done elsewhere, or as a subsubsection of this subsection (or as a
6358: @c subsection of Defining Words)
6359:
6360: For example, suppose that you have a word @code{stats} that
6361: gathers statistics about colon definitions given the @i{xt} of the
6362: definition, and you want every colon definition in your application to
6363: make a call to @code{stats}. You can define and use a new version of
6364: @code{:} like this:
6365:
6366: @example
6367: : stats ( xt -- ) DUP ." (Gathering statistics for " . ." )"
6368: ... ; \ other code
6369:
6370: : my: : latestxt postpone literal ['] stats compile, ;
6371:
6372: my: foo + - ;
6373: @end example
6374:
6375: When @code{foo} is defined using @code{my:} these steps occur:
6376:
6377: @itemize @bullet
6378: @item
6379: @code{my:} is executed.
6380: @item
6381: The @code{:} within the definition (the one between @code{my:} and
6382: @code{latestxt}) is executed, and does just what it always does; it parses
6383: the input stream for a name, builds a dictionary header for the name
6384: @code{foo} and switches @code{state} from interpret to compile.
6385: @item
6386: The word @code{latestxt} is executed. It puts the @i{xt} for the word that is
6387: being defined -- @code{foo} -- onto the stack.
6388: @item
6389: The code that was produced by @code{postpone literal} is executed; this
6390: causes the value on the stack to be compiled as a literal in the code
6391: area of @code{foo}.
6392: @item
6393: The code @code{['] stats} compiles a literal into the definition of
6394: @code{my:}. When @code{compile,} is executed, that literal -- the
6395: execution token for @code{stats} -- is layed down in the code area of
6396: @code{foo} , following the literal@footnote{Strictly speaking, the
6397: mechanism that @code{compile,} uses to convert an @i{xt} into something
6398: in the code area is implementation-dependent. A threaded implementation
6399: might spit out the execution token directly whilst another
6400: implementation might spit out a native code sequence.}.
6401: @item
6402: At this point, the execution of @code{my:} is complete, and control
6403: returns to the text interpreter. The text interpreter is in compile
6404: state, so subsequent text @code{+ -} is compiled into the definition of
6405: @code{foo} and the @code{;} terminates the definition as always.
6406: @end itemize
6407:
6408: You can use @code{see} to decompile a word that was defined using
6409: @code{my:} and see how it is different from a normal @code{:}
6410: definition. For example:
6411:
6412: @example
6413: : bar + - ; \ like foo but using : rather than my:
6414: see bar
6415: : bar
6416: + - ;
6417: see foo
6418: : foo
6419: 107645672 stats + - ;
6420:
6421: \ use ' foo . to show that 107645672 is the xt for foo
6422: @end example
6423:
6424: You can use techniques like this to make new defining words in terms of
6425: @i{any} existing defining word.
6426:
6427:
6428: @cindex defining defining words
6429: @cindex @code{CREATE} ... @code{DOES>}
6430: If you want the words defined with your defining words to behave
6431: differently from words defined with standard defining words, you can
6432: write your defining word like this:
6433:
6434: @example
6435: : def-word ( "name" -- )
6436: CREATE @i{code1}
6437: DOES> ( ... -- ... )
6438: @i{code2} ;
6439:
6440: def-word name
6441: @end example
6442:
6443: @cindex child words
6444: This fragment defines a @dfn{defining word} @code{def-word} and then
6445: executes it. When @code{def-word} executes, it @code{CREATE}s a new
6446: word, @code{name}, and executes the code @i{code1}. The code @i{code2}
6447: is not executed at this time. The word @code{name} is sometimes called a
6448: @dfn{child} of @code{def-word}.
6449:
6450: When you execute @code{name}, the address of the body of @code{name} is
6451: put on the data stack and @i{code2} is executed (the address of the body
6452: of @code{name} is the address @code{HERE} returns immediately after the
6453: @code{CREATE}, i.e., the address a @code{create}d word returns by
6454: default).
6455:
6456: @c anton:
6457: @c www.dictionary.com says:
6458: @c at·a·vism: 1.The reappearance of a characteristic in an organism after
6459: @c several generations of absence, usually caused by the chance
6460: @c recombination of genes. 2.An individual or a part that exhibits
6461: @c atavism. Also called throwback. 3.The return of a trait or recurrence
6462: @c of previous behavior after a period of absence.
6463: @c
6464: @c Doesn't seem to fit.
6465:
6466: @c @cindex atavism in child words
6467: You can use @code{def-word} to define a set of child words that behave
6468: similarly; they all have a common run-time behaviour determined by
6469: @i{code2}. Typically, the @i{code1} sequence builds a data area in the
6470: body of the child word. The structure of the data is common to all
6471: children of @code{def-word}, but the data values are specific -- and
6472: private -- to each child word. When a child word is executed, the
6473: address of its private data area is passed as a parameter on TOS to be
6474: used and manipulated@footnote{It is legitimate both to read and write to
6475: this data area.} by @i{code2}.
6476:
6477: The two fragments of code that make up the defining words act (are
6478: executed) at two completely separate times:
6479:
6480: @itemize @bullet
6481: @item
6482: At @i{define time}, the defining word executes @i{code1} to generate a
6483: child word
6484: @item
6485: At @i{child execution time}, when a child word is invoked, @i{code2}
6486: is executed, using parameters (data) that are private and specific to
6487: the child word.
6488: @end itemize
6489:
6490: Another way of understanding the behaviour of @code{def-word} and
6491: @code{name} is to say that, if you make the following definitions:
6492: @example
6493: : def-word1 ( "name" -- )
6494: CREATE @i{code1} ;
6495:
6496: : action1 ( ... -- ... )
6497: @i{code2} ;
6498:
6499: def-word1 name1
6500: @end example
6501:
6502: @noindent
6503: Then using @code{name1 action1} is equivalent to using @code{name}.
6504:
6505: The classic example is that you can define @code{CONSTANT} in this way:
6506:
6507: @example
6508: : CONSTANT ( w "name" -- )
6509: CREATE ,
6510: DOES> ( -- w )
6511: @@ ;
6512: @end example
6513:
6514: @comment There is a beautiful description of how this works and what
6515: @comment it does in the Forthwrite 100th edition.. as well as an elegant
6516: @comment commentary on the Counting Fruits problem.
6517:
6518: When you create a constant with @code{5 CONSTANT five}, a set of
6519: define-time actions take place; first a new word @code{five} is created,
6520: then the value 5 is laid down in the body of @code{five} with
6521: @code{,}. When @code{five} is executed, the address of the body is put on
6522: the stack, and @code{@@} retrieves the value 5. The word @code{five} has
6523: no code of its own; it simply contains a data field and a pointer to the
6524: code that follows @code{DOES>} in its defining word. That makes words
6525: created in this way very compact.
6526:
6527: The final example in this section is intended to remind you that space
6528: reserved in @code{CREATE}d words is @i{data} space and therefore can be
6529: both read and written by a Standard program@footnote{Exercise: use this
6530: example as a starting point for your own implementation of @code{Value}
6531: and @code{TO} -- if you get stuck, investigate the behaviour of @code{'} and
6532: @code{[']}.}:
6533:
6534: @example
6535: : foo ( "name" -- )
6536: CREATE -1 ,
6537: DOES> ( -- )
6538: @@ . ;
6539:
6540: foo first-word
6541: foo second-word
6542:
6543: 123 ' first-word >BODY !
6544: @end example
6545:
6546: If @code{first-word} had been a @code{CREATE}d word, we could simply
6547: have executed it to get the address of its data field. However, since it
6548: was defined to have @code{DOES>} actions, its execution semantics are to
6549: perform those @code{DOES>} actions. To get the address of its data field
6550: it's necessary to use @code{'} to get its xt, then @code{>BODY} to
6551: translate the xt into the address of the data field. When you execute
6552: @code{first-word}, it will display @code{123}. When you execute
6553: @code{second-word} it will display @code{-1}.
6554:
6555: @cindex stack effect of @code{DOES>}-parts
6556: @cindex @code{DOES>}-parts, stack effect
6557: In the examples above the stack comment after the @code{DOES>} specifies
6558: the stack effect of the defined words, not the stack effect of the
6559: following code (the following code expects the address of the body on
6560: the top of stack, which is not reflected in the stack comment). This is
6561: the convention that I use and recommend (it clashes a bit with using
6562: locals declarations for stack effect specification, though).
6563:
6564: @menu
6565: * CREATE..DOES> applications::
6566: * CREATE..DOES> details::
6567: * Advanced does> usage example::
6568: * Const-does>::
6569: @end menu
6570:
6571: @node CREATE..DOES> applications, CREATE..DOES> details, User-defined Defining Words, User-defined Defining Words
6572: @subsubsection Applications of @code{CREATE..DOES>}
6573: @cindex @code{CREATE} ... @code{DOES>}, applications
6574:
6575: You may wonder how to use this feature. Here are some usage patterns:
6576:
6577: @cindex factoring similar colon definitions
6578: When you see a sequence of code occurring several times, and you can
6579: identify a meaning, you will factor it out as a colon definition. When
6580: you see similar colon definitions, you can factor them using
6581: @code{CREATE..DOES>}. E.g., an assembler usually defines several words
6582: that look very similar:
6583: @example
6584: : ori, ( reg-target reg-source n -- )
6585: 0 asm-reg-reg-imm ;
6586: : andi, ( reg-target reg-source n -- )
6587: 1 asm-reg-reg-imm ;
6588: @end example
6589:
6590: @noindent
6591: This could be factored with:
6592: @example
6593: : reg-reg-imm ( op-code -- )
6594: CREATE ,
6595: DOES> ( reg-target reg-source n -- )
6596: @@ asm-reg-reg-imm ;
6597:
6598: 0 reg-reg-imm ori,
6599: 1 reg-reg-imm andi,
6600: @end example
6601:
6602: @cindex currying
6603: Another view of @code{CREATE..DOES>} is to consider it as a crude way to
6604: supply a part of the parameters for a word (known as @dfn{currying} in
6605: the functional language community). E.g., @code{+} needs two
6606: parameters. Creating versions of @code{+} with one parameter fixed can
6607: be done like this:
6608:
6609: @example
6610: : curry+ ( n1 "name" -- )
6611: CREATE ,
6612: DOES> ( n2 -- n1+n2 )
6613: @@ + ;
6614:
6615: 3 curry+ 3+
6616: -2 curry+ 2-
6617: @end example
6618:
6619:
6620: @node CREATE..DOES> details, Advanced does> usage example, CREATE..DOES> applications, User-defined Defining Words
6621: @subsubsection The gory details of @code{CREATE..DOES>}
6622: @cindex @code{CREATE} ... @code{DOES>}, details
6623:
6624: doc-does>
6625:
6626: @cindex @code{DOES>} in a separate definition
6627: This means that you need not use @code{CREATE} and @code{DOES>} in the
6628: same definition; you can put the @code{DOES>}-part in a separate
6629: definition. This allows us to, e.g., select among different @code{DOES>}-parts:
6630: @example
6631: : does1
6632: DOES> ( ... -- ... )
6633: ... ;
6634:
6635: : does2
6636: DOES> ( ... -- ... )
6637: ... ;
6638:
6639: : def-word ( ... -- ... )
6640: create ...
6641: IF
6642: does1
6643: ELSE
6644: does2
6645: ENDIF ;
6646: @end example
6647:
6648: In this example, the selection of whether to use @code{does1} or
6649: @code{does2} is made at definition-time; at the time that the child word is
6650: @code{CREATE}d.
6651:
6652: @cindex @code{DOES>} in interpretation state
6653: In a standard program you can apply a @code{DOES>}-part only if the last
6654: word was defined with @code{CREATE}. In Gforth, the @code{DOES>}-part
6655: will override the behaviour of the last word defined in any case. In a
6656: standard program, you can use @code{DOES>} only in a colon
6657: definition. In Gforth, you can also use it in interpretation state, in a
6658: kind of one-shot mode; for example:
6659: @example
6660: CREATE name ( ... -- ... )
6661: @i{initialization}
6662: DOES>
6663: @i{code} ;
6664: @end example
6665:
6666: @noindent
6667: is equivalent to the standard:
6668: @example
6669: :noname
6670: DOES>
6671: @i{code} ;
6672: CREATE name EXECUTE ( ... -- ... )
6673: @i{initialization}
6674: @end example
6675:
6676: doc->body
6677:
6678: @node Advanced does> usage example, Const-does>, CREATE..DOES> details, User-defined Defining Words
6679: @subsubsection Advanced does> usage example
6680:
6681: The MIPS disassembler (@file{arch/mips/disasm.fs}) contains many words
6682: for disassembling instructions, that follow a very repetetive scheme:
6683:
6684: @example
6685: :noname @var{disasm-operands} s" @var{inst-name}" type ;
6686: @var{entry-num} cells @var{table} + !
6687: @end example
6688:
6689: Of course, this inspires the idea to factor out the commonalities to
6690: allow a definition like
6691:
6692: @example
6693: @var{disasm-operands} @var{entry-num} @var{table} define-inst @var{inst-name}
6694: @end example
6695:
6696: The parameters @var{disasm-operands} and @var{table} are usually
6697: correlated. Moreover, before I wrote the disassembler, there already
6698: existed code that defines instructions like this:
6699:
6700: @example
6701: @var{entry-num} @var{inst-format} @var{inst-name}
6702: @end example
6703:
6704: This code comes from the assembler and resides in
6705: @file{arch/mips/insts.fs}.
6706:
6707: So I had to define the @var{inst-format} words that performed the scheme
6708: above when executed. At first I chose to use run-time code-generation:
6709:
6710: @example
6711: : @var{inst-format} ( entry-num "name" -- ; compiled code: addr w -- )
6712: :noname Postpone @var{disasm-operands}
6713: name Postpone sliteral Postpone type Postpone ;
6714: swap cells @var{table} + ! ;
6715: @end example
6716:
6717: Note that this supplies the other two parameters of the scheme above.
6718:
6719: An alternative would have been to write this using
6720: @code{create}/@code{does>}:
6721:
6722: @example
6723: : @var{inst-format} ( entry-num "name" -- )
6724: here name string, ( entry-num c-addr ) \ parse and save "name"
6725: noname create , ( entry-num )
6726: latestxt swap cells @var{table} + !
6727: does> ( addr w -- )
6728: \ disassemble instruction w at addr
6729: @@ >r
6730: @var{disasm-operands}
6731: r> count type ;
6732: @end example
6733:
6734: Somehow the first solution is simpler, mainly because it's simpler to
6735: shift a string from definition-time to use-time with @code{sliteral}
6736: than with @code{string,} and friends.
6737:
6738: I wrote a lot of words following this scheme and soon thought about
6739: factoring out the commonalities among them. Note that this uses a
6740: two-level defining word, i.e., a word that defines ordinary defining
6741: words.
6742:
6743: This time a solution involving @code{postpone} and friends seemed more
6744: difficult (try it as an exercise), so I decided to use a
6745: @code{create}/@code{does>} word; since I was already at it, I also used
6746: @code{create}/@code{does>} for the lower level (try using
6747: @code{postpone} etc. as an exercise), resulting in the following
6748: definition:
6749:
6750: @example
6751: : define-format ( disasm-xt table-xt -- )
6752: \ define an instruction format that uses disasm-xt for
6753: \ disassembling and enters the defined instructions into table
6754: \ table-xt
6755: create 2,
6756: does> ( u "inst" -- )
6757: \ defines an anonymous word for disassembling instruction inst,
6758: \ and enters it as u-th entry into table-xt
6759: 2@@ swap here name string, ( u table-xt disasm-xt c-addr ) \ remember string
6760: noname create 2, \ define anonymous word
6761: execute latestxt swap ! \ enter xt of defined word into table-xt
6762: does> ( addr w -- )
6763: \ disassemble instruction w at addr
6764: 2@@ >r ( addr w disasm-xt R: c-addr )
6765: execute ( R: c-addr ) \ disassemble operands
6766: r> count type ; \ print name
6767: @end example
6768:
6769: Note that the tables here (in contrast to above) do the @code{cells +}
6770: by themselves (that's why you have to pass an xt). This word is used in
6771: the following way:
6772:
6773: @example
6774: ' @var{disasm-operands} ' @var{table} define-format @var{inst-format}
6775: @end example
6776:
6777: As shown above, the defined instruction format is then used like this:
6778:
6779: @example
6780: @var{entry-num} @var{inst-format} @var{inst-name}
6781: @end example
6782:
6783: In terms of currying, this kind of two-level defining word provides the
6784: parameters in three stages: first @var{disasm-operands} and @var{table},
6785: then @var{entry-num} and @var{inst-name}, finally @code{addr w}, i.e.,
6786: the instruction to be disassembled.
6787:
6788: Of course this did not quite fit all the instruction format names used
6789: in @file{insts.fs}, so I had to define a few wrappers that conditioned
6790: the parameters into the right form.
6791:
6792: If you have trouble following this section, don't worry. First, this is
6793: involved and takes time (and probably some playing around) to
6794: understand; second, this is the first two-level
6795: @code{create}/@code{does>} word I have written in seventeen years of
6796: Forth; and if I did not have @file{insts.fs} to start with, I may well
6797: have elected to use just a one-level defining word (with some repeating
6798: of parameters when using the defining word). So it is not necessary to
6799: understand this, but it may improve your understanding of Forth.
6800:
6801:
6802: @node Const-does>, , Advanced does> usage example, User-defined Defining Words
6803: @subsubsection @code{Const-does>}
6804:
6805: A frequent use of @code{create}...@code{does>} is for transferring some
6806: values from definition-time to run-time. Gforth supports this use with
6807:
6808: doc-const-does>
6809:
6810: A typical use of this word is:
6811:
6812: @example
6813: : curry+ ( n1 "name" -- )
6814: 1 0 CONST-DOES> ( n2 -- n1+n2 )
6815: + ;
6816:
6817: 3 curry+ 3+
6818: @end example
6819:
6820: Here the @code{1 0} means that 1 cell and 0 floats are transferred from
6821: definition to run-time.
6822:
6823: The advantages of using @code{const-does>} are:
6824:
6825: @itemize
6826:
6827: @item
6828: You don't have to deal with storing and retrieving the values, i.e.,
6829: your program becomes more writable and readable.
6830:
6831: @item
6832: When using @code{does>}, you have to introduce a @code{@@} that cannot
6833: be optimized away (because you could change the data using
6834: @code{>body}...@code{!}); @code{const-does>} avoids this problem.
6835:
6836: @end itemize
6837:
6838: An ANS Forth implementation of @code{const-does>} is available in
6839: @file{compat/const-does.fs}.
6840:
6841:
6842: @node Deferred Words, Aliases, User-defined Defining Words, Defining Words
6843: @subsection Deferred Words
6844: @cindex deferred words
6845:
6846: The defining word @code{Defer} allows you to define a word by name
6847: without defining its behaviour; the definition of its behaviour is
6848: deferred. Here are two situation where this can be useful:
6849:
6850: @itemize @bullet
6851: @item
6852: Where you want to allow the behaviour of a word to be altered later, and
6853: for all precompiled references to the word to change when its behaviour
6854: is changed.
6855: @item
6856: For mutual recursion; @xref{Calls and returns}.
6857: @end itemize
6858:
6859: In the following example, @code{foo} always invokes the version of
6860: @code{greet} that prints ``@code{Good morning}'' whilst @code{bar}
6861: always invokes the version that prints ``@code{Hello}''. There is no way
6862: of getting @code{foo} to use the later version without re-ordering the
6863: source code and recompiling it.
6864:
6865: @example
6866: : greet ." Good morning" ;
6867: : foo ... greet ... ;
6868: : greet ." Hello" ;
6869: : bar ... greet ... ;
6870: @end example
6871:
6872: This problem can be solved by defining @code{greet} as a @code{Defer}red
6873: word. The behaviour of a @code{Defer}red word can be defined and
6874: redefined at any time by using @code{IS} to associate the xt of a
6875: previously-defined word with it. The previous example becomes:
6876:
6877: @example
6878: Defer greet ( -- )
6879: : foo ... greet ... ;
6880: : bar ... greet ... ;
6881: : greet1 ( -- ) ." Good morning" ;
6882: : greet2 ( -- ) ." Hello" ;
6883: ' greet2 IS greet \ make greet behave like greet2
6884: @end example
6885:
6886: @progstyle
6887: You should write a stack comment for every deferred word, and put only
6888: XTs into deferred words that conform to this stack effect. Otherwise
6889: it's too difficult to use the deferred word.
6890:
6891: A deferred word can be used to improve the statistics-gathering example
6892: from @ref{User-defined Defining Words}; rather than edit the
6893: application's source code to change every @code{:} to a @code{my:}, do
6894: this:
6895:
6896: @example
6897: : real: : ; \ retain access to the original
6898: defer : \ redefine as a deferred word
6899: ' my: IS : \ use special version of :
6900: \
6901: \ load application here
6902: \
6903: ' real: IS : \ go back to the original
6904: @end example
6905:
6906:
6907: One thing to note is that @code{IS} has special compilation semantics,
6908: such that it parses the name at compile time (like @code{TO}):
6909:
6910: @example
6911: : set-greet ( xt -- )
6912: IS greet ;
6913:
6914: ' greet1 set-greet
6915: @end example
6916:
6917: In situations where @code{IS} does not fit, use @code{defer!} instead.
6918:
6919: A deferred word can only inherit execution semantics from the xt
6920: (because that is all that an xt can represent -- for more discussion of
6921: this @pxref{Tokens for Words}); by default it will have default
6922: interpretation and compilation semantics deriving from this execution
6923: semantics. However, you can change the interpretation and compilation
6924: semantics of the deferred word in the usual ways:
6925:
6926: @example
6927: : bar .... ; immediate
6928: Defer fred immediate
6929: Defer jim
6930:
6931: ' bar IS jim \ jim has default semantics
6932: ' bar IS fred \ fred is immediate
6933: @end example
6934:
6935: doc-defer
6936: doc-defer!
6937: doc-is
6938: doc-defer@
6939: doc-action-of
6940: @comment TODO document these: what's defers [is]
6941: doc-defers
6942:
6943: @c Use @code{words-deferred} to see a list of deferred words.
6944:
6945: Definitions of these words (except @code{defers}) in ANS Forth are
6946: provided in @file{compat/defer.fs}.
6947:
6948:
6949: @node Aliases, , Deferred Words, Defining Words
6950: @subsection Aliases
6951: @cindex aliases
6952:
6953: The defining word @code{Alias} allows you to define a word by name that
6954: has the same behaviour as some other word. Here are two situation where
6955: this can be useful:
6956:
6957: @itemize @bullet
6958: @item
6959: When you want access to a word's definition from a different word list
6960: (for an example of this, see the definition of the @code{Root} word list
6961: in the Gforth source).
6962: @item
6963: When you want to create a synonym; a definition that can be known by
6964: either of two names (for example, @code{THEN} and @code{ENDIF} are
6965: aliases).
6966: @end itemize
6967:
6968: Like deferred words, an alias has default compilation and interpretation
6969: semantics at the beginning (not the modifications of the other word),
6970: but you can change them in the usual ways (@code{immediate},
6971: @code{compile-only}). For example:
6972:
6973: @example
6974: : foo ... ; immediate
6975:
6976: ' foo Alias bar \ bar is not an immediate word
6977: ' foo Alias fooby immediate \ fooby is an immediate word
6978: @end example
6979:
6980: Words that are aliases have the same xt, different headers in the
6981: dictionary, and consequently different name tokens (@pxref{Tokens for
6982: Words}) and possibly different immediate flags. An alias can only have
6983: default or immediate compilation semantics; you can define aliases for
6984: combined words with @code{interpret/compile:} -- see @ref{Combined words}.
6985:
6986: doc-alias
6987:
6988:
6989: @node Interpretation and Compilation Semantics, Tokens for Words, Defining Words, Words
6990: @section Interpretation and Compilation Semantics
6991: @cindex semantics, interpretation and compilation
6992:
6993: @c !! state and ' are used without explanation
6994: @c example for immediate/compile-only? or is the tutorial enough
6995:
6996: @cindex interpretation semantics
6997: The @dfn{interpretation semantics} of a (named) word are what the text
6998: interpreter does when it encounters the word in interpret state. It also
6999: appears in some other contexts, e.g., the execution token returned by
7000: @code{' @i{word}} identifies the interpretation semantics of @i{word}
7001: (in other words, @code{' @i{word} execute} is equivalent to
7002: interpret-state text interpretation of @code{@i{word}}).
7003:
7004: @cindex compilation semantics
7005: The @dfn{compilation semantics} of a (named) word are what the text
7006: interpreter does when it encounters the word in compile state. It also
7007: appears in other contexts, e.g, @code{POSTPONE @i{word}}
7008: compiles@footnote{In standard terminology, ``appends to the current
7009: definition''.} the compilation semantics of @i{word}.
7010:
7011: @cindex execution semantics
7012: The standard also talks about @dfn{execution semantics}. They are used
7013: only for defining the interpretation and compilation semantics of many
7014: words. By default, the interpretation semantics of a word are to
7015: @code{execute} its execution semantics, and the compilation semantics of
7016: a word are to @code{compile,} its execution semantics.@footnote{In
7017: standard terminology: The default interpretation semantics are its
7018: execution semantics; the default compilation semantics are to append its
7019: execution semantics to the execution semantics of the current
7020: definition.}
7021:
7022: Unnamed words (@pxref{Anonymous Definitions}) cannot be encountered by
7023: the text interpreter, ticked, or @code{postpone}d, so they have no
7024: interpretation or compilation semantics. Their behaviour is represented
7025: by their XT (@pxref{Tokens for Words}), and we call it execution
7026: semantics, too.
7027:
7028: @comment TODO expand, make it co-operate with new sections on text interpreter.
7029:
7030: @cindex immediate words
7031: @cindex compile-only words
7032: You can change the semantics of the most-recently defined word:
7033:
7034:
7035: doc-immediate
7036: doc-compile-only
7037: doc-restrict
7038:
7039: By convention, words with non-default compilation semantics (e.g.,
7040: immediate words) often have names surrounded with brackets (e.g.,
7041: @code{[']}, @pxref{Execution token}).
7042:
7043: Note that ticking (@code{'}) a compile-only word gives an error
7044: (``Interpreting a compile-only word'').
7045:
7046: @menu
7047: * Combined words::
7048: @end menu
7049:
7050:
7051: @node Combined words, , Interpretation and Compilation Semantics, Interpretation and Compilation Semantics
7052: @subsection Combined Words
7053: @cindex combined words
7054:
7055: Gforth allows you to define @dfn{combined words} -- words that have an
7056: arbitrary combination of interpretation and compilation semantics.
7057:
7058: doc-interpret/compile:
7059:
7060: This feature was introduced for implementing @code{TO} and @code{S"}. I
7061: recommend that you do not define such words, as cute as they may be:
7062: they make it hard to get at both parts of the word in some contexts.
7063: E.g., assume you want to get an execution token for the compilation
7064: part. Instead, define two words, one that embodies the interpretation
7065: part, and one that embodies the compilation part. Once you have done
7066: that, you can define a combined word with @code{interpret/compile:} for
7067: the convenience of your users.
7068:
7069: You might try to use this feature to provide an optimizing
7070: implementation of the default compilation semantics of a word. For
7071: example, by defining:
7072: @example
7073: :noname
7074: foo bar ;
7075: :noname
7076: POSTPONE foo POSTPONE bar ;
7077: interpret/compile: opti-foobar
7078: @end example
7079:
7080: @noindent
7081: as an optimizing version of:
7082:
7083: @example
7084: : foobar
7085: foo bar ;
7086: @end example
7087:
7088: Unfortunately, this does not work correctly with @code{[compile]},
7089: because @code{[compile]} assumes that the compilation semantics of all
7090: @code{interpret/compile:} words are non-default. I.e., @code{[compile]
7091: opti-foobar} would compile compilation semantics, whereas
7092: @code{[compile] foobar} would compile interpretation semantics.
7093:
7094: @cindex state-smart words (are a bad idea)
7095: @anchor{state-smartness}
7096: Some people try to use @dfn{state-smart} words to emulate the feature provided
7097: by @code{interpret/compile:} (words are state-smart if they check
7098: @code{STATE} during execution). E.g., they would try to code
7099: @code{foobar} like this:
7100:
7101: @example
7102: : foobar
7103: STATE @@
7104: IF ( compilation state )
7105: POSTPONE foo POSTPONE bar
7106: ELSE
7107: foo bar
7108: ENDIF ; immediate
7109: @end example
7110:
7111: Although this works if @code{foobar} is only processed by the text
7112: interpreter, it does not work in other contexts (like @code{'} or
7113: @code{POSTPONE}). E.g., @code{' foobar} will produce an execution token
7114: for a state-smart word, not for the interpretation semantics of the
7115: original @code{foobar}; when you execute this execution token (directly
7116: with @code{EXECUTE} or indirectly through @code{COMPILE,}) in compile
7117: state, the result will not be what you expected (i.e., it will not
7118: perform @code{foo bar}). State-smart words are a bad idea. Simply don't
7119: write them@footnote{For a more detailed discussion of this topic, see
7120: M. Anton Ertl,
7121: @cite{@uref{http://www.complang.tuwien.ac.at/papers/ertl98.ps.gz,@code{State}-smartness---Why
7122: it is Evil and How to Exorcise it}}, EuroForth '98.}!
7123:
7124: @cindex defining words with arbitrary semantics combinations
7125: It is also possible to write defining words that define words with
7126: arbitrary combinations of interpretation and compilation semantics. In
7127: general, they look like this:
7128:
7129: @example
7130: : def-word
7131: create-interpret/compile
7132: @i{code1}
7133: interpretation>
7134: @i{code2}
7135: <interpretation
7136: compilation>
7137: @i{code3}
7138: <compilation ;
7139: @end example
7140:
7141: For a @i{word} defined with @code{def-word}, the interpretation
7142: semantics are to push the address of the body of @i{word} and perform
7143: @i{code2}, and the compilation semantics are to push the address of
7144: the body of @i{word} and perform @i{code3}. E.g., @code{constant}
7145: can also be defined like this (except that the defined constants don't
7146: behave correctly when @code{[compile]}d):
7147:
7148: @example
7149: : constant ( n "name" -- )
7150: create-interpret/compile
7151: ,
7152: interpretation> ( -- n )
7153: @@
7154: <interpretation
7155: compilation> ( compilation. -- ; run-time. -- n )
7156: @@ postpone literal
7157: <compilation ;
7158: @end example
7159:
7160:
7161: doc-create-interpret/compile
7162: doc-interpretation>
7163: doc-<interpretation
7164: doc-compilation>
7165: doc-<compilation
7166:
7167:
7168: Words defined with @code{interpret/compile:} and
7169: @code{create-interpret/compile} have an extended header structure that
7170: differs from other words; however, unless you try to access them with
7171: plain address arithmetic, you should not notice this. Words for
7172: accessing the header structure usually know how to deal with this; e.g.,
7173: @code{'} @i{word} @code{>body} also gives you the body of a word created
7174: with @code{create-interpret/compile}.
7175:
7176:
7177: @c -------------------------------------------------------------
7178: @node Tokens for Words, Compiling words, Interpretation and Compilation Semantics, Words
7179: @section Tokens for Words
7180: @cindex tokens for words
7181:
7182: This section describes the creation and use of tokens that represent
7183: words.
7184:
7185: @menu
7186: * Execution token:: represents execution/interpretation semantics
7187: * Compilation token:: represents compilation semantics
7188: * Name token:: represents named words
7189: @end menu
7190:
7191: @node Execution token, Compilation token, Tokens for Words, Tokens for Words
7192: @subsection Execution token
7193:
7194: @cindex xt
7195: @cindex execution token
7196: An @dfn{execution token} (@i{XT}) represents some behaviour of a word.
7197: You can use @code{execute} to invoke this behaviour.
7198:
7199: @cindex tick (')
7200: You can use @code{'} to get an execution token that represents the
7201: interpretation semantics of a named word:
7202:
7203: @example
7204: 5 ' . ( n xt )
7205: execute ( ) \ execute the xt (i.e., ".")
7206: @end example
7207:
7208: doc-'
7209:
7210: @code{'} parses at run-time; there is also a word @code{[']} that parses
7211: when it is compiled, and compiles the resulting XT:
7212:
7213: @example
7214: : foo ['] . execute ;
7215: 5 foo
7216: : bar ' execute ; \ by contrast,
7217: 5 bar . \ ' parses "." when bar executes
7218: @end example
7219:
7220: doc-[']
7221:
7222: If you want the execution token of @i{word}, write @code{['] @i{word}}
7223: in compiled code and @code{' @i{word}} in interpreted code. Gforth's
7224: @code{'} and @code{[']} behave somewhat unusually by complaining about
7225: compile-only words (because these words have no interpretation
7226: semantics). You might get what you want by using @code{COMP' @i{word}
7227: DROP} or @code{[COMP'] @i{word} DROP} (for details @pxref{Compilation
7228: token}).
7229:
7230: Another way to get an XT is @code{:noname} or @code{latestxt}
7231: (@pxref{Anonymous Definitions}). For anonymous words this gives an xt
7232: for the only behaviour the word has (the execution semantics). For
7233: named words, @code{latestxt} produces an XT for the same behaviour it
7234: would produce if the word was defined anonymously.
7235:
7236: @example
7237: :noname ." hello" ;
7238: execute
7239: @end example
7240:
7241: An XT occupies one cell and can be manipulated like any other cell.
7242:
7243: @cindex code field address
7244: @cindex CFA
7245: In ANS Forth the XT is just an abstract data type (i.e., defined by the
7246: operations that produce or consume it). For old hands: In Gforth, the
7247: XT is implemented as a code field address (CFA).
7248:
7249: doc-execute
7250: doc-perform
7251:
7252: @node Compilation token, Name token, Execution token, Tokens for Words
7253: @subsection Compilation token
7254:
7255: @cindex compilation token
7256: @cindex CT (compilation token)
7257: Gforth represents the compilation semantics of a named word by a
7258: @dfn{compilation token} consisting of two cells: @i{w xt}. The top cell
7259: @i{xt} is an execution token. The compilation semantics represented by
7260: the compilation token can be performed with @code{execute}, which
7261: consumes the whole compilation token, with an additional stack effect
7262: determined by the represented compilation semantics.
7263:
7264: At present, the @i{w} part of a compilation token is an execution token,
7265: and the @i{xt} part represents either @code{execute} or
7266: @code{compile,}@footnote{Depending upon the compilation semantics of the
7267: word. If the word has default compilation semantics, the @i{xt} will
7268: represent @code{compile,}. Otherwise (e.g., for immediate words), the
7269: @i{xt} will represent @code{execute}.}. However, don't rely on that
7270: knowledge, unless necessary; future versions of Gforth may introduce
7271: unusual compilation tokens (e.g., a compilation token that represents
7272: the compilation semantics of a literal).
7273:
7274: You can perform the compilation semantics represented by the compilation
7275: token with @code{execute}. You can compile the compilation semantics
7276: with @code{postpone,}. I.e., @code{COMP' @i{word} postpone,} is
7277: equivalent to @code{postpone @i{word}}.
7278:
7279: doc-[comp']
7280: doc-comp'
7281: doc-postpone,
7282:
7283: @node Name token, , Compilation token, Tokens for Words
7284: @subsection Name token
7285:
7286: @cindex name token
7287: Gforth represents named words by the @dfn{name token}, (@i{nt}). Name
7288: token is an abstract data type that occurs as argument or result of the
7289: words below.
7290:
7291: @c !! put this elswhere?
7292: @cindex name field address
7293: @cindex NFA
7294: The closest thing to the nt in older Forth systems is the name field
7295: address (NFA), but there are significant differences: in older Forth
7296: systems each word had a unique NFA, LFA, CFA and PFA (in this order, or
7297: LFA, NFA, CFA, PFA) and there were words for getting from one to the
7298: next. In contrast, in Gforth 0@dots{}n nts correspond to one xt; there
7299: is a link field in the structure identified by the name token, but
7300: searching usually uses a hash table external to these structures; the
7301: name in Gforth has a cell-wide count-and-flags field, and the nt is not
7302: implemented as the address of that count field.
7303:
7304: doc-find-name
7305: doc-latest
7306: doc->name
7307: doc-name>int
7308: doc-name?int
7309: doc-name>comp
7310: doc-name>string
7311: doc-id.
7312: doc-.name
7313: doc-.id
7314:
7315: @c ----------------------------------------------------------
7316: @node Compiling words, The Text Interpreter, Tokens for Words, Words
7317: @section Compiling words
7318: @cindex compiling words
7319: @cindex macros
7320:
7321: In contrast to most other languages, Forth has no strict boundary
7322: between compilation and run-time. E.g., you can run arbitrary code
7323: between defining words (or for computing data used by defining words
7324: like @code{constant}). Moreover, @code{Immediate} (@pxref{Interpretation
7325: and Compilation Semantics} and @code{[}...@code{]} (see below) allow
7326: running arbitrary code while compiling a colon definition (exception:
7327: you must not allot dictionary space).
7328:
7329: @menu
7330: * Literals:: Compiling data values
7331: * Macros:: Compiling words
7332: @end menu
7333:
7334: @node Literals, Macros, Compiling words, Compiling words
7335: @subsection Literals
7336: @cindex Literals
7337:
7338: The simplest and most frequent example is to compute a literal during
7339: compilation. E.g., the following definition prints an array of strings,
7340: one string per line:
7341:
7342: @example
7343: : .strings ( addr u -- ) \ gforth
7344: 2* cells bounds U+DO
7345: cr i 2@@ type
7346: 2 cells +LOOP ;
7347: @end example
7348:
7349: With a simple-minded compiler like Gforth's, this computes @code{2
7350: cells} on every loop iteration. You can compute this value once and for
7351: all at compile time and compile it into the definition like this:
7352:
7353: @example
7354: : .strings ( addr u -- ) \ gforth
7355: 2* cells bounds U+DO
7356: cr i 2@@ type
7357: [ 2 cells ] literal +LOOP ;
7358: @end example
7359:
7360: @code{[} switches the text interpreter to interpret state (you will get
7361: an @code{ok} prompt if you type this example interactively and insert a
7362: newline between @code{[} and @code{]}), so it performs the
7363: interpretation semantics of @code{2 cells}; this computes a number.
7364: @code{]} switches the text interpreter back into compile state. It then
7365: performs @code{Literal}'s compilation semantics, which are to compile
7366: this number into the current word. You can decompile the word with
7367: @code{see .strings} to see the effect on the compiled code.
7368:
7369: You can also optimize the @code{2* cells} into @code{[ 2 cells ] literal
7370: *} in this way.
7371:
7372: doc-[
7373: doc-]
7374: doc-literal
7375: doc-]L
7376:
7377: There are also words for compiling other data types than single cells as
7378: literals:
7379:
7380: doc-2literal
7381: doc-fliteral
7382: doc-sliteral
7383:
7384: @cindex colon-sys, passing data across @code{:}
7385: @cindex @code{:}, passing data across
7386: You might be tempted to pass data from outside a colon definition to the
7387: inside on the data stack. This does not work, because @code{:} puhes a
7388: colon-sys, making stuff below unaccessible. E.g., this does not work:
7389:
7390: @example
7391: 5 : foo literal ; \ error: "unstructured"
7392: @end example
7393:
7394: Instead, you have to pass the value in some other way, e.g., through a
7395: variable:
7396:
7397: @example
7398: variable temp
7399: 5 temp !
7400: : foo [ temp @@ ] literal ;
7401: @end example
7402:
7403:
7404: @node Macros, , Literals, Compiling words
7405: @subsection Macros
7406: @cindex Macros
7407: @cindex compiling compilation semantics
7408:
7409: @code{Literal} and friends compile data values into the current
7410: definition. You can also write words that compile other words into the
7411: current definition. E.g.,
7412:
7413: @example
7414: : compile-+ ( -- ) \ compiled code: ( n1 n2 -- n )
7415: POSTPONE + ;
7416:
7417: : foo ( n1 n2 -- n )
7418: [ compile-+ ] ;
7419: 1 2 foo .
7420: @end example
7421:
7422: This is equivalent to @code{: foo + ;} (@code{see foo} to check this).
7423: What happens in this example? @code{Postpone} compiles the compilation
7424: semantics of @code{+} into @code{compile-+}; later the text interpreter
7425: executes @code{compile-+} and thus the compilation semantics of +, which
7426: compile (the execution semantics of) @code{+} into
7427: @code{foo}.@footnote{A recent RFI answer requires that compiling words
7428: should only be executed in compile state, so this example is not
7429: guaranteed to work on all standard systems, but on any decent system it
7430: will work.}
7431:
7432: doc-postpone
7433:
7434: Compiling words like @code{compile-+} are usually immediate (or similar)
7435: so you do not have to switch to interpret state to execute them;
7436: modifying the last example accordingly produces:
7437:
7438: @example
7439: : [compile-+] ( compilation: --; interpretation: -- )
7440: \ compiled code: ( n1 n2 -- n )
7441: POSTPONE + ; immediate
7442:
7443: : foo ( n1 n2 -- n )
7444: [compile-+] ;
7445: 1 2 foo .
7446: @end example
7447:
7448: You will occassionally find the need to POSTPONE several words;
7449: putting POSTPONE before each such word is cumbersome, so Gforth
7450: provides a more convenient syntax: @code{]] ... [[}. This
7451: allows us to write @code{[compile-+]} as:
7452:
7453: @example
7454: : [compile-+] ( compilation: --; interpretation: -- )
7455: ]] + [[ ; immediate
7456: @end example
7457:
7458: doc-]]
7459: doc-[[
7460:
7461: The unusual direction of the brackets indicates their function:
7462: @code{]]} switches from compilation to postponing (i.e., compilation
7463: of compilation), just like @code{]} switches from immediate execution
7464: (interpretation) to compilation. Conversely, @code{[[} switches from
7465: postponing to compilation, ananlogous to @code{[} which switches from
7466: compilation to immediate execution.
7467:
7468: The real advantage of @code{]] }...@code{ [[} becomes apparent when
7469: there are many words to POSTPONE. E.g., the word
7470: @code{compile-map-array} (@pxref{Advanced macros Tutorial}) can be
7471: written much shorter as follows:
7472:
7473: @example
7474: : compile-map-array ( compilation: xt -- ; run-time: ... addr u -- ... )
7475: \ at run-time, execute xt ( ... x -- ... ) for each element of the
7476: \ array beginning at addr and containing u elements
7477: @{ xt @}
7478: ]] cells over + swap ?do
7479: i @@ [[ xt compile,
7480: 1 cells ]]L +loop [[ ;
7481: @end example
7482:
7483: This example also uses @code{]]L} as a shortcut for @code{]] literal}.
7484: There are also other shortcuts
7485:
7486: doc-]]L
7487: doc-]]2L
7488: doc-]]FL
7489: doc-]]SL
7490:
7491: Note that parsing words don't parse at postpone time; if you want to
7492: provide the parsed string right away, you have to switch back to
7493: compilation:
7494:
7495: @example
7496: ]] ... [[ s" some string" ]]2L ... [[
7497: ]] ... [[ ['] + ]]L ... [[
7498: @end example
7499:
7500: Definitions of @code{]]} and friends in ANS Forth are provided in
7501: @file{compat/macros.fs}.
7502:
7503: Immediate compiling words are similar to macros in other languages (in
7504: particular, Lisp). The important differences to macros in, e.g., C are:
7505:
7506: @itemize @bullet
7507:
7508: @item
7509: You use the same language for defining and processing macros, not a
7510: separate preprocessing language and processor.
7511:
7512: @item
7513: Consequently, the full power of Forth is available in macro definitions.
7514: E.g., you can perform arbitrarily complex computations, or generate
7515: different code conditionally or in a loop (e.g., @pxref{Advanced macros
7516: Tutorial}). This power is very useful when writing a parser generators
7517: or other code-generating software.
7518:
7519: @item
7520: Macros defined using @code{postpone} etc. deal with the language at a
7521: higher level than strings; name binding happens at macro definition
7522: time, so you can avoid the pitfalls of name collisions that can happen
7523: in C macros. Of course, Forth is a liberal language and also allows to
7524: shoot yourself in the foot with text-interpreted macros like
7525:
7526: @example
7527: : [compile-+] s" +" evaluate ; immediate
7528: @end example
7529:
7530: Apart from binding the name at macro use time, using @code{evaluate}
7531: also makes your definition @code{state}-smart (@pxref{state-smartness}).
7532: @end itemize
7533:
7534: You may want the macro to compile a number into a word. The word to do
7535: it is @code{literal}, but you have to @code{postpone} it, so its
7536: compilation semantics take effect when the macro is executed, not when
7537: it is compiled:
7538:
7539: @example
7540: : [compile-5] ( -- ) \ compiled code: ( -- n )
7541: 5 POSTPONE literal ; immediate
7542:
7543: : foo [compile-5] ;
7544: foo .
7545: @end example
7546:
7547: You may want to pass parameters to a macro, that the macro should
7548: compile into the current definition. If the parameter is a number, then
7549: you can use @code{postpone literal} (similar for other values).
7550:
7551: If you want to pass a word that is to be compiled, the usual way is to
7552: pass an execution token and @code{compile,} it:
7553:
7554: @example
7555: : twice1 ( xt -- ) \ compiled code: ... -- ...
7556: dup compile, compile, ;
7557:
7558: : 2+ ( n1 -- n2 )
7559: [ ' 1+ twice1 ] ;
7560: @end example
7561:
7562: doc-compile,
7563:
7564: An alternative available in Gforth, that allows you to pass compile-only
7565: words as parameters is to use the compilation token (@pxref{Compilation
7566: token}). The same example in this technique:
7567:
7568: @example
7569: : twice ( ... ct -- ... ) \ compiled code: ... -- ...
7570: 2dup 2>r execute 2r> execute ;
7571:
7572: : 2+ ( n1 -- n2 )
7573: [ comp' 1+ twice ] ;
7574: @end example
7575:
7576: In the example above @code{2>r} and @code{2r>} ensure that @code{twice}
7577: works even if the executed compilation semantics has an effect on the
7578: data stack.
7579:
7580: You can also define complete definitions with these words; this provides
7581: an alternative to using @code{does>} (@pxref{User-defined Defining
7582: Words}). E.g., instead of
7583:
7584: @example
7585: : curry+ ( n1 "name" -- )
7586: CREATE ,
7587: DOES> ( n2 -- n1+n2 )
7588: @@ + ;
7589: @end example
7590:
7591: you could define
7592:
7593: @example
7594: : curry+ ( n1 "name" -- )
7595: \ name execution: ( n2 -- n1+n2 )
7596: >r : r> POSTPONE literal POSTPONE + POSTPONE ; ;
7597:
7598: -3 curry+ 3-
7599: see 3-
7600: @end example
7601:
7602: The sequence @code{>r : r>} is necessary, because @code{:} puts a
7603: colon-sys on the data stack that makes everything below it unaccessible.
7604:
7605: This way of writing defining words is sometimes more, sometimes less
7606: convenient than using @code{does>} (@pxref{Advanced does> usage
7607: example}). One advantage of this method is that it can be optimized
7608: better, because the compiler knows that the value compiled with
7609: @code{literal} is fixed, whereas the data associated with a
7610: @code{create}d word can be changed.
7611:
7612: @c doc-[compile] !! not properly documented
7613:
7614: @c ----------------------------------------------------------
7615: @node The Text Interpreter, The Input Stream, Compiling words, Words
7616: @section The Text Interpreter
7617: @cindex interpreter - outer
7618: @cindex text interpreter
7619: @cindex outer interpreter
7620:
7621: @c Should we really describe all these ugly details? IMO the text
7622: @c interpreter should be much cleaner, but that may not be possible within
7623: @c ANS Forth. - anton
7624: @c nac-> I wanted to explain how it works to show how you can exploit
7625: @c it in your own programs. When I was writing a cross-compiler, figuring out
7626: @c some of these gory details was very helpful to me. None of the textbooks
7627: @c I've seen cover it, and the most modern Forth textbook -- Forth Inc's,
7628: @c seems to positively avoid going into too much detail for some of
7629: @c the internals.
7630:
7631: @c anton: ok. I wonder, though, if this is the right place; for some stuff
7632: @c it is; for the ugly details, I would prefer another place. I wonder
7633: @c whether we should have a chapter before "Words" that describes some
7634: @c basic concepts referred to in words, and a chapter after "Words" that
7635: @c describes implementation details.
7636:
7637: The text interpreter@footnote{This is an expanded version of the
7638: material in @ref{Introducing the Text Interpreter}.} is an endless loop
7639: that processes input from the current input device. It is also called
7640: the outer interpreter, in contrast to the inner interpreter
7641: (@pxref{Engine}) which executes the compiled Forth code on interpretive
7642: implementations.
7643:
7644: @cindex interpret state
7645: @cindex compile state
7646: The text interpreter operates in one of two states: @dfn{interpret
7647: state} and @dfn{compile state}. The current state is defined by the
7648: aptly-named variable @code{state}.
7649:
7650: This section starts by describing how the text interpreter behaves when
7651: it is in interpret state, processing input from the user input device --
7652: the keyboard. This is the mode that a Forth system is in after it starts
7653: up.
7654:
7655: @cindex input buffer
7656: @cindex terminal input buffer
7657: The text interpreter works from an area of memory called the @dfn{input
7658: buffer}@footnote{When the text interpreter is processing input from the
7659: keyboard, this area of memory is called the @dfn{terminal input buffer}
7660: (TIB) and is addressed by the (obsolescent) words @code{TIB} and
7661: @code{#TIB}.}, which stores your keyboard input when you press the
7662: @key{RET} key. Starting at the beginning of the input buffer, it skips
7663: leading spaces (called @dfn{delimiters}) then parses a string (a
7664: sequence of non-space characters) until it reaches either a space
7665: character or the end of the buffer. Having parsed a string, it makes two
7666: attempts to process it:
7667:
7668: @cindex dictionary
7669: @itemize @bullet
7670: @item
7671: It looks for the string in a @dfn{dictionary} of definitions. If the
7672: string is found, the string names a @dfn{definition} (also known as a
7673: @dfn{word}) and the dictionary search returns information that allows
7674: the text interpreter to perform the word's @dfn{interpretation
7675: semantics}. In most cases, this simply means that the word will be
7676: executed.
7677: @item
7678: If the string is not found in the dictionary, the text interpreter
7679: attempts to treat it as a number, using the rules described in
7680: @ref{Number Conversion}. If the string represents a legal number in the
7681: current radix, the number is pushed onto a parameter stack (the data
7682: stack for integers, the floating-point stack for floating-point
7683: numbers).
7684: @end itemize
7685:
7686: If both attempts fail, or if the word is found in the dictionary but has
7687: no interpretation semantics@footnote{This happens if the word was
7688: defined as @code{COMPILE-ONLY}.} the text interpreter discards the
7689: remainder of the input buffer, issues an error message and waits for
7690: more input. If one of the attempts succeeds, the text interpreter
7691: repeats the parsing process until the whole of the input buffer has been
7692: processed, at which point it prints the status message ``@code{ ok}''
7693: and waits for more input.
7694:
7695: @c anton: this should be in the input stream subsection (or below it)
7696:
7697: @cindex parse area
7698: The text interpreter keeps track of its position in the input buffer by
7699: updating a variable called @code{>IN} (pronounced ``to-in''). The value
7700: of @code{>IN} starts out as 0, indicating an offset of 0 from the start
7701: of the input buffer. The region from offset @code{>IN @@} to the end of
7702: the input buffer is called the @dfn{parse area}@footnote{In other words,
7703: the text interpreter processes the contents of the input buffer by
7704: parsing strings from the parse area until the parse area is empty.}.
7705: This example shows how @code{>IN} changes as the text interpreter parses
7706: the input buffer:
7707:
7708: @example
7709: : remaining >IN @@ SOURCE 2 PICK - -ROT + SWAP
7710: CR ." ->" TYPE ." <-" ; IMMEDIATE
7711:
7712: 1 2 3 remaining + remaining .
7713:
7714: : foo 1 2 3 remaining SWAP remaining ;
7715: @end example
7716:
7717: @noindent
7718: The result is:
7719:
7720: @example
7721: ->+ remaining .<-
7722: ->.<-5 ok
7723:
7724: ->SWAP remaining ;-<
7725: ->;<- ok
7726: @end example
7727:
7728: @cindex parsing words
7729: The value of @code{>IN} can also be modified by a word in the input
7730: buffer that is executed by the text interpreter. This means that a word
7731: can ``trick'' the text interpreter into either skipping a section of the
7732: input buffer@footnote{This is how parsing words work.} or into parsing a
7733: section twice. For example:
7734:
7735: @example
7736: : lat ." <<foo>>" ;
7737: : flat ." <<bar>>" >IN DUP @@ 3 - SWAP ! ;
7738: @end example
7739:
7740: @noindent
7741: When @code{flat} is executed, this output is produced@footnote{Exercise
7742: for the reader: what would happen if the @code{3} were replaced with
7743: @code{4}?}:
7744:
7745: @example
7746: <<bar>><<foo>>
7747: @end example
7748:
7749: This technique can be used to work around some of the interoperability
7750: problems of parsing words. Of course, it's better to avoid parsing
7751: words where possible.
7752:
7753: @noindent
7754: Two important notes about the behaviour of the text interpreter:
7755:
7756: @itemize @bullet
7757: @item
7758: It processes each input string to completion before parsing additional
7759: characters from the input buffer.
7760: @item
7761: It treats the input buffer as a read-only region (and so must your code).
7762: @end itemize
7763:
7764: @noindent
7765: When the text interpreter is in compile state, its behaviour changes in
7766: these ways:
7767:
7768: @itemize @bullet
7769: @item
7770: If a parsed string is found in the dictionary, the text interpreter will
7771: perform the word's @dfn{compilation semantics}. In most cases, this
7772: simply means that the execution semantics of the word will be appended
7773: to the current definition.
7774: @item
7775: When a number is encountered, it is compiled into the current definition
7776: (as a literal) rather than being pushed onto a parameter stack.
7777: @item
7778: If an error occurs, @code{state} is modified to put the text interpreter
7779: back into interpret state.
7780: @item
7781: Each time a line is entered from the keyboard, Gforth prints
7782: ``@code{ compiled}'' rather than `` @code{ok}''.
7783: @end itemize
7784:
7785: @cindex text interpreter - input sources
7786: When the text interpreter is using an input device other than the
7787: keyboard, its behaviour changes in these ways:
7788:
7789: @itemize @bullet
7790: @item
7791: When the parse area is empty, the text interpreter attempts to refill
7792: the input buffer from the input source. When the input source is
7793: exhausted, the input source is set back to the previous input source.
7794: @item
7795: It doesn't print out ``@code{ ok}'' or ``@code{ compiled}'' messages each
7796: time the parse area is emptied.
7797: @item
7798: If an error occurs, the input source is set back to the user input
7799: device.
7800: @end itemize
7801:
7802: You can read about this in more detail in @ref{Input Sources}.
7803:
7804: doc->in
7805: doc-source
7806:
7807: doc-tib
7808: doc-#tib
7809:
7810:
7811: @menu
7812: * Input Sources::
7813: * Number Conversion::
7814: * Interpret/Compile states::
7815: * Interpreter Directives::
7816: @end menu
7817:
7818: @node Input Sources, Number Conversion, The Text Interpreter, The Text Interpreter
7819: @subsection Input Sources
7820: @cindex input sources
7821: @cindex text interpreter - input sources
7822:
7823: By default, the text interpreter processes input from the user input
7824: device (the keyboard) when Forth starts up. The text interpreter can
7825: process input from any of these sources:
7826:
7827: @itemize @bullet
7828: @item
7829: The user input device -- the keyboard.
7830: @item
7831: A file, using the words described in @ref{Forth source files}.
7832: @item
7833: A block, using the words described in @ref{Blocks}.
7834: @item
7835: A text string, using @code{evaluate}.
7836: @end itemize
7837:
7838: A program can identify the current input device from the values of
7839: @code{source-id} and @code{blk}.
7840:
7841:
7842: doc-source-id
7843: doc-blk
7844:
7845: doc-save-input
7846: doc-restore-input
7847:
7848: doc-evaluate
7849: doc-query
7850:
7851:
7852:
7853: @node Number Conversion, Interpret/Compile states, Input Sources, The Text Interpreter
7854: @subsection Number Conversion
7855: @cindex number conversion
7856: @cindex double-cell numbers, input format
7857: @cindex input format for double-cell numbers
7858: @cindex single-cell numbers, input format
7859: @cindex input format for single-cell numbers
7860: @cindex floating-point numbers, input format
7861: @cindex input format for floating-point numbers
7862:
7863: This section describes the rules that the text interpreter uses when it
7864: tries to convert a string into a number.
7865:
7866: Let <digit> represent any character that is a legal digit in the current
7867: number base@footnote{For example, 0-9 when the number base is decimal or
7868: 0-9, A-F when the number base is hexadecimal.}.
7869:
7870: Let <decimal digit> represent any character in the range 0-9.
7871:
7872: Let @{@i{a b}@} represent the @i{optional} presence of any of the characters
7873: in the braces (@i{a} or @i{b} or neither).
7874:
7875: Let * represent any number of instances of the previous character
7876: (including none).
7877:
7878: Let any other character represent itself.
7879:
7880: @noindent
7881: Now, the conversion rules are:
7882:
7883: @itemize @bullet
7884: @item
7885: A string of the form <digit><digit>* is treated as a single-precision
7886: (cell-sized) positive integer. Examples are 0 123 6784532 32343212343456 42
7887: @item
7888: A string of the form -<digit><digit>* is treated as a single-precision
7889: (cell-sized) negative integer, and is represented using 2's-complement
7890: arithmetic. Examples are -45 -5681 -0
7891: @item
7892: A string of the form <digit><digit>*.<digit>* is treated as a double-precision
7893: (double-cell-sized) positive integer. Examples are 3465. 3.465 34.65
7894: (all three of these represent the same number).
7895: @item
7896: A string of the form -<digit><digit>*.<digit>* is treated as a
7897: double-precision (double-cell-sized) negative integer, and is
7898: represented using 2's-complement arithmetic. Examples are -3465. -3.465
7899: -34.65 (all three of these represent the same number).
7900: @item
7901: A string of the form @{+ -@}<decimal digit>@{.@}<decimal digit>*@{e
7902: E@}@{+ -@}<decimal digit><decimal digit>* is treated as a floating-point
7903: number. Examples are 1e 1e0 1.e 1.e0 +1e+0 (which all represent the same
7904: number) +12.E-4
7905: @end itemize
7906:
7907: By default, the number base used for integer number conversion is
7908: given by the contents of the variable @code{base}. Note that a lot of
7909: confusion can result from unexpected values of @code{base}. If you
7910: change @code{base} anywhere, make sure to save the old value and
7911: restore it afterwards; better yet, use @code{base-execute}, which does
7912: this for you. In general I recommend keeping @code{base} decimal, and
7913: using the prefixes described below for the popular non-decimal bases.
7914:
7915: doc-dpl
7916: doc-base-execute
7917: doc-base
7918: doc-hex
7919: doc-decimal
7920:
7921: @cindex '-prefix for character strings
7922: @cindex &-prefix for decimal numbers
7923: @cindex #-prefix for decimal numbers
7924: @cindex %-prefix for binary numbers
7925: @cindex $-prefix for hexadecimal numbers
7926: @cindex 0x-prefix for hexadecimal numbers
7927: Gforth allows you to override the value of @code{base} by using a
7928: prefix@footnote{Some Forth implementations provide a similar scheme by
7929: implementing @code{$} etc. as parsing words that process the subsequent
7930: number in the input stream and push it onto the stack. For example, see
7931: @cite{Number Conversion and Literals}, by Wil Baden; Forth Dimensions
7932: 20(3) pages 26--27. In such implementations, unlike in Gforth, a space
7933: is required between the prefix and the number.} before the first digit
7934: of an (integer) number. The following prefixes are supported:
7935:
7936: @itemize @bullet
7937: @item
7938: @code{&} -- decimal
7939: @item
7940: @code{#} -- decimal
7941: @item
7942: @code{%} -- binary
7943: @item
7944: @code{$} -- hexadecimal
7945: @item
7946: @code{0x} -- hexadecimal, if base<33.
7947: @item
7948: @code{'} -- numeric value (e.g., ASCII code) of next character; an
7949: optional @code{'} may be present after the character.
7950: @end itemize
7951:
7952: Here are some examples, with the equivalent decimal number shown after
7953: in braces:
7954:
7955: -$41 (-65), %1001101 (205), %1001.0001 (145 - a double-precision number),
7956: 'A (65),
7957: -'a' (-97),
7958: &905 (905), $abc (2478), $ABC (2478).
7959:
7960: @cindex number conversion - traps for the unwary
7961: @noindent
7962: Number conversion has a number of traps for the unwary:
7963:
7964: @itemize @bullet
7965: @item
7966: You cannot determine the current number base using the code sequence
7967: @code{base @@ .} -- the number base is always 10 in the current number
7968: base. Instead, use something like @code{base @@ dec.}
7969: @item
7970: If the number base is set to a value greater than 14 (for example,
7971: hexadecimal), the number 123E4 is ambiguous; the conversion rules allow
7972: it to be intepreted as either a single-precision integer or a
7973: floating-point number (Gforth treats it as an integer). The ambiguity
7974: can be resolved by explicitly stating the sign of the mantissa and/or
7975: exponent: 123E+4 or +123E4 -- if the number base is decimal, no
7976: ambiguity arises; either representation will be treated as a
7977: floating-point number.
7978: @item
7979: There is a word @code{bin} but it does @i{not} set the number base!
7980: It is used to specify file types.
7981: @item
7982: ANS Forth requires the @code{.} of a double-precision number to be the
7983: final character in the string. Gforth allows the @code{.} to be
7984: anywhere after the first digit.
7985: @item
7986: The number conversion process does not check for overflow.
7987: @item
7988: In an ANS Forth program @code{base} is required to be decimal when
7989: converting floating-point numbers. In Gforth, number conversion to
7990: floating-point numbers always uses base &10, irrespective of the value
7991: of @code{base}.
7992: @end itemize
7993:
7994: You can read numbers into your programs with the words described in
7995: @ref{Line input and conversion}.
7996:
7997: @node Interpret/Compile states, Interpreter Directives, Number Conversion, The Text Interpreter
7998: @subsection Interpret/Compile states
7999: @cindex Interpret/Compile states
8000:
8001: A standard program is not permitted to change @code{state}
8002: explicitly. However, it can change @code{state} implicitly, using the
8003: words @code{[} and @code{]}. When @code{[} is executed it switches
8004: @code{state} to interpret state, and therefore the text interpreter
8005: starts interpreting. When @code{]} is executed it switches @code{state}
8006: to compile state and therefore the text interpreter starts
8007: compiling. The most common usage for these words is for switching into
8008: interpret state and back from within a colon definition; this technique
8009: can be used to compile a literal (for an example, @pxref{Literals}) or
8010: for conditional compilation (for an example, @pxref{Interpreter
8011: Directives}).
8012:
8013:
8014: @c This is a bad example: It's non-standard, and it's not necessary.
8015: @c However, I can't think of a good example for switching into compile
8016: @c state when there is no current word (@code{state}-smart words are not a
8017: @c good reason). So maybe we should use an example for switching into
8018: @c interpret @code{state} in a colon def. - anton
8019: @c nac-> I agree. I started out by putting in the example, then realised
8020: @c that it was non-ANS, so wrote more words around it. I hope this
8021: @c re-written version is acceptable to you. I do want to keep the example
8022: @c as it is helpful for showing what is and what is not portable, particularly
8023: @c where it outlaws a style in common use.
8024:
8025: @c anton: it's more important to show what's portable. After we have done
8026: @c that, we can also show what's not. In any case, I have written a
8027: @c section Compiling Words which also deals with [ ].
8028:
8029: @c !! The following example does not work in Gforth 0.5.9 or later.
8030:
8031: @c @code{[} and @code{]} also give you the ability to switch into compile
8032: @c state and back, but we cannot think of any useful Standard application
8033: @c for this ability. Pre-ANS Forth textbooks have examples like this:
8034:
8035: @c @example
8036: @c : AA ." this is A" ;
8037: @c : BB ." this is B" ;
8038: @c : CC ." this is C" ;
8039:
8040: @c create table ] aa bb cc [
8041:
8042: @c : go ( n -- ) \ n is offset into table.. 0 for 1st entry
8043: @c cells table + @@ execute ;
8044: @c @end example
8045:
8046: @c This example builds a jump table; @code{0 go} will display ``@code{this
8047: @c is A}''. Using @code{[} and @code{]} in this example is equivalent to
8048: @c defining @code{table} like this:
8049:
8050: @c @example
8051: @c create table ' aa COMPILE, ' bb COMPILE, ' cc COMPILE,
8052: @c @end example
8053:
8054: @c The problem with this code is that the definition of @code{table} is not
8055: @c portable -- it @i{compile}s execution tokens into code space. Whilst it
8056: @c @i{may} work on systems where code space and data space co-incide, the
8057: @c Standard only allows data space to be assigned for a @code{CREATE}d
8058: @c word. In addition, the Standard only allows @code{@@} to access data
8059: @c space, whilst this example is using it to access code space. The only
8060: @c portable, Standard way to build this table is to build it in data space,
8061: @c like this:
8062:
8063: @c @example
8064: @c create table ' aa , ' bb , ' cc ,
8065: @c @end example
8066:
8067: @c doc-state
8068:
8069:
8070: @node Interpreter Directives, , Interpret/Compile states, The Text Interpreter
8071: @subsection Interpreter Directives
8072: @cindex interpreter directives
8073: @cindex conditional compilation
8074:
8075: These words are usually used in interpret state; typically to control
8076: which parts of a source file are processed by the text
8077: interpreter. There are only a few ANS Forth Standard words, but Gforth
8078: supplements these with a rich set of immediate control structure words
8079: to compensate for the fact that the non-immediate versions can only be
8080: used in compile state (@pxref{Control Structures}). Typical usages:
8081:
8082: @example
8083: FALSE Constant HAVE-ASSEMBLER
8084: .
8085: .
8086: HAVE-ASSEMBLER [IF]
8087: : ASSEMBLER-FEATURE
8088: ...
8089: ;
8090: [ENDIF]
8091: .
8092: .
8093: : SEE
8094: ... \ general-purpose SEE code
8095: [ HAVE-ASSEMBLER [IF] ]
8096: ... \ assembler-specific SEE code
8097: [ [ENDIF] ]
8098: ;
8099: @end example
8100:
8101:
8102: doc-[IF]
8103: doc-[ELSE]
8104: doc-[THEN]
8105: doc-[ENDIF]
8106:
8107: doc-[IFDEF]
8108: doc-[IFUNDEF]
8109:
8110: doc-[?DO]
8111: doc-[DO]
8112: doc-[FOR]
8113: doc-[LOOP]
8114: doc-[+LOOP]
8115: doc-[NEXT]
8116:
8117: doc-[BEGIN]
8118: doc-[UNTIL]
8119: doc-[AGAIN]
8120: doc-[WHILE]
8121: doc-[REPEAT]
8122:
8123:
8124: @c -------------------------------------------------------------
8125: @node The Input Stream, Word Lists, The Text Interpreter, Words
8126: @section The Input Stream
8127: @cindex input stream
8128:
8129: @c !! integrate this better with the "Text Interpreter" section
8130: The text interpreter reads from the input stream, which can come from
8131: several sources (@pxref{Input Sources}). Some words, in particular
8132: defining words, but also words like @code{'}, read parameters from the
8133: input stream instead of from the stack.
8134:
8135: Such words are called parsing words, because they parse the input
8136: stream. Parsing words are hard to use in other words, because it is
8137: hard to pass program-generated parameters through the input stream.
8138: They also usually have an unintuitive combination of interpretation and
8139: compilation semantics when implemented naively, leading to various
8140: approaches that try to produce a more intuitive behaviour
8141: (@pxref{Combined words}).
8142:
8143: It should be obvious by now that parsing words are a bad idea. If you
8144: want to implement a parsing word for convenience, also provide a factor
8145: of the word that does not parse, but takes the parameters on the stack.
8146: To implement the parsing word on top if it, you can use the following
8147: words:
8148:
8149: @c anton: these belong in the input stream section
8150: doc-parse
8151: doc-parse-name
8152: doc-parse-word
8153: doc-name
8154: doc-word
8155: doc-refill
8156:
8157: Conversely, if you have the bad luck (or lack of foresight) to have to
8158: deal with parsing words without having such factors, how do you pass a
8159: string that is not in the input stream to it?
8160:
8161: doc-execute-parsing
8162:
8163: A definition of this word in ANS Forth is provided in
8164: @file{compat/execute-parsing.fs}.
8165:
8166: If you want to run a parsing word on a file, the following word should
8167: help:
8168:
8169: doc-execute-parsing-file
8170:
8171: @c -------------------------------------------------------------
8172: @node Word Lists, Environmental Queries, The Input Stream, Words
8173: @section Word Lists
8174: @cindex word lists
8175: @cindex header space
8176:
8177: A wordlist is a list of named words; you can add new words and look up
8178: words by name (and you can remove words in a restricted way with
8179: markers). Every named (and @code{reveal}ed) word is in one wordlist.
8180:
8181: @cindex search order stack
8182: The text interpreter searches the wordlists present in the search order
8183: (a stack of wordlists), from the top to the bottom. Within each
8184: wordlist, the search starts conceptually at the newest word; i.e., if
8185: two words in a wordlist have the same name, the newer word is found.
8186:
8187: @cindex compilation word list
8188: New words are added to the @dfn{compilation wordlist} (aka current
8189: wordlist).
8190:
8191: @cindex wid
8192: A word list is identified by a cell-sized word list identifier (@i{wid})
8193: in much the same way as a file is identified by a file handle. The
8194: numerical value of the wid has no (portable) meaning, and might change
8195: from session to session.
8196:
8197: The ANS Forth ``Search order'' word set is intended to provide a set of
8198: low-level tools that allow various different schemes to be
8199: implemented. Gforth also provides @code{vocabulary}, a traditional Forth
8200: word. @file{compat/vocabulary.fs} provides an implementation in ANS
8201: Forth.
8202:
8203: @comment TODO: locals section refers to here, saying that every word list (aka
8204: @comment vocabulary) has its own methods for searching etc. Need to document that.
8205: @c anton: but better in a separate subsection on wordlist internals
8206:
8207: @comment TODO: document markers, reveal, tables, mappedwordlist
8208:
8209: @comment the gforthman- prefix is used to pick out the true definition of a
8210: @comment word from the source files, rather than some alias.
8211:
8212: doc-forth-wordlist
8213: doc-definitions
8214: doc-get-current
8215: doc-set-current
8216: doc-get-order
8217: doc-set-order
8218: doc-wordlist
8219: doc-table
8220: doc->order
8221: doc-previous
8222: doc-also
8223: doc-forth
8224: doc-only
8225: doc-order
8226:
8227: doc-find
8228: doc-search-wordlist
8229:
8230: doc-words
8231: doc-vlist
8232: @c doc-words-deferred
8233:
8234: @c doc-mappedwordlist @c map-structure undefined, implemantation-specific
8235: doc-root
8236: doc-vocabulary
8237: doc-seal
8238: doc-vocs
8239: doc-current
8240: doc-context
8241:
8242:
8243: @menu
8244: * Vocabularies::
8245: * Why use word lists?::
8246: * Word list example::
8247: @end menu
8248:
8249: @node Vocabularies, Why use word lists?, Word Lists, Word Lists
8250: @subsection Vocabularies
8251: @cindex Vocabularies, detailed explanation
8252:
8253: Here is an example of creating and using a new wordlist using ANS
8254: Forth words:
8255:
8256: @example
8257: wordlist constant my-new-words-wordlist
8258: : my-new-words get-order nip my-new-words-wordlist swap set-order ;
8259:
8260: \ add it to the search order
8261: also my-new-words
8262:
8263: \ alternatively, add it to the search order and make it
8264: \ the compilation word list
8265: also my-new-words definitions
8266: \ type "order" to see the problem
8267: @end example
8268:
8269: The problem with this example is that @code{order} has no way to
8270: associate the name @code{my-new-words} with the wid of the word list (in
8271: Gforth, @code{order} and @code{vocs} will display @code{???} for a wid
8272: that has no associated name). There is no Standard way of associating a
8273: name with a wid.
8274:
8275: In Gforth, this example can be re-coded using @code{vocabulary}, which
8276: associates a name with a wid:
8277:
8278: @example
8279: vocabulary my-new-words
8280:
8281: \ add it to the search order
8282: also my-new-words
8283:
8284: \ alternatively, add it to the search order and make it
8285: \ the compilation word list
8286: my-new-words definitions
8287: \ type "order" to see that the problem is solved
8288: @end example
8289:
8290:
8291: @node Why use word lists?, Word list example, Vocabularies, Word Lists
8292: @subsection Why use word lists?
8293: @cindex word lists - why use them?
8294:
8295: Here are some reasons why people use wordlists:
8296:
8297: @itemize @bullet
8298:
8299: @c anton: Gforth's hashing implementation makes the search speed
8300: @c independent from the number of words. But it is linear with the number
8301: @c of wordlists that have to be searched, so in effect using more wordlists
8302: @c actually slows down compilation.
8303:
8304: @c @item
8305: @c To improve compilation speed by reducing the number of header space
8306: @c entries that must be searched. This is achieved by creating a new
8307: @c word list that contains all of the definitions that are used in the
8308: @c definition of a Forth system but which would not usually be used by
8309: @c programs running on that system. That word list would be on the search
8310: @c list when the Forth system was compiled but would be removed from the
8311: @c search list for normal operation. This can be a useful technique for
8312: @c low-performance systems (for example, 8-bit processors in embedded
8313: @c systems) but is unlikely to be necessary in high-performance desktop
8314: @c systems.
8315:
8316: @item
8317: To prevent a set of words from being used outside the context in which
8318: they are valid. Two classic examples of this are an integrated editor
8319: (all of the edit commands are defined in a separate word list; the
8320: search order is set to the editor word list when the editor is invoked;
8321: the old search order is restored when the editor is terminated) and an
8322: integrated assembler (the op-codes for the machine are defined in a
8323: separate word list which is used when a @code{CODE} word is defined).
8324:
8325: @item
8326: To organize the words of an application or library into a user-visible
8327: set (in @code{forth-wordlist} or some other common wordlist) and a set
8328: of helper words used just for the implementation (hidden in a separate
8329: wordlist). This keeps @code{words}' output smaller, separates
8330: implementation and interface, and reduces the chance of name conflicts
8331: within the common wordlist.
8332:
8333: @item
8334: To prevent a name-space clash between multiple definitions with the same
8335: name. For example, when building a cross-compiler you might have a word
8336: @code{IF} that generates conditional code for your target system. By
8337: placing this definition in a different word list you can control whether
8338: the host system's @code{IF} or the target system's @code{IF} get used in
8339: any particular context by controlling the order of the word lists on the
8340: search order stack.
8341:
8342: @end itemize
8343:
8344: The downsides of using wordlists are:
8345:
8346: @itemize
8347:
8348: @item
8349: Debugging becomes more cumbersome.
8350:
8351: @item
8352: Name conflicts worked around with wordlists are still there, and you
8353: have to arrange the search order carefully to get the desired results;
8354: if you forget to do that, you get hard-to-find errors (as in any case
8355: where you read the code differently from the compiler; @code{see} can
8356: help seeing which of several possible words the name resolves to in such
8357: cases). @code{See} displays just the name of the words, not what
8358: wordlist they belong to, so it might be misleading. Using unique names
8359: is a better approach to avoid name conflicts.
8360:
8361: @item
8362: You have to explicitly undo any changes to the search order. In many
8363: cases it would be more convenient if this happened implicitly. Gforth
8364: currently does not provide such a feature, but it may do so in the
8365: future.
8366: @end itemize
8367:
8368:
8369: @node Word list example, , Why use word lists?, Word Lists
8370: @subsection Word list example
8371: @cindex word lists - example
8372:
8373: The following example is from the
8374: @uref{http://www.complang.tuwien.ac.at/forth/garbage-collection.zip,
8375: garbage collector} and uses wordlists to separate public words from
8376: helper words:
8377:
8378: @example
8379: get-current ( wid )
8380: vocabulary garbage-collector also garbage-collector definitions
8381: ... \ define helper words
8382: ( wid ) set-current \ restore original (i.e., public) compilation wordlist
8383: ... \ define the public (i.e., API) words
8384: \ they can refer to the helper words
8385: previous \ restore original search order (helper words become invisible)
8386: @end example
8387:
8388: @c -------------------------------------------------------------
8389: @node Environmental Queries, Files, Word Lists, Words
8390: @section Environmental Queries
8391: @cindex environmental queries
8392:
8393: ANS Forth introduced the idea of ``environmental queries'' as a way
8394: for a program running on a system to determine certain characteristics of the system.
8395: The Standard specifies a number of strings that might be recognised by a system.
8396:
8397: The Standard requires that the header space used for environmental queries
8398: be distinct from the header space used for definitions.
8399:
8400: Typically, environmental queries are supported by creating a set of
8401: definitions in a word list that is @i{only} used during environmental
8402: queries; that is what Gforth does. There is no Standard way of adding
8403: definitions to the set of recognised environmental queries, but any
8404: implementation that supports the loading of optional word sets must have
8405: some mechanism for doing this (after loading the word set, the
8406: associated environmental query string must return @code{true}). In
8407: Gforth, the word list used to honour environmental queries can be
8408: manipulated just like any other word list.
8409:
8410:
8411: doc-environment?
8412: doc-environment-wordlist
8413:
8414: doc-gforth
8415: doc-os-class
8416:
8417:
8418: Note that, whilst the documentation for (e.g.) @code{gforth} shows it
8419: returning two items on the stack, querying it using @code{environment?}
8420: will return an additional item; the @code{true} flag that shows that the
8421: string was recognised.
8422:
8423: @comment TODO Document the standard strings or note where they are documented herein
8424:
8425: Here are some examples of using environmental queries:
8426:
8427: @example
8428: s" address-unit-bits" environment? 0=
8429: [IF]
8430: cr .( environmental attribute address-units-bits unknown... ) cr
8431: [ELSE]
8432: drop \ ensure balanced stack effect
8433: [THEN]
8434:
8435: \ this might occur in the prelude of a standard program that uses THROW
8436: s" exception" environment? [IF]
8437: 0= [IF]
8438: : throw abort" exception thrown" ;
8439: [THEN]
8440: [ELSE] \ we don't know, so make sure
8441: : throw abort" exception thrown" ;
8442: [THEN]
8443:
8444: s" gforth" environment? [IF] .( Gforth version ) TYPE
8445: [ELSE] .( Not Gforth..) [THEN]
8446:
8447: \ a program using v*
8448: s" gforth" environment? [IF]
8449: s" 0.5.0" compare 0< [IF] \ v* is a primitive since 0.5.0
8450: : v* ( f_addr1 nstride1 f_addr2 nstride2 ucount -- r )
8451: >r swap 2swap swap 0e r> 0 ?DO
8452: dup f@@ over + 2swap dup f@@ f* f+ over + 2swap
8453: LOOP
8454: 2drop 2drop ;
8455: [THEN]
8456: [ELSE] \
8457: : v* ( f_addr1 nstride1 f_addr2 nstride2 ucount -- r )
8458: ...
8459: [THEN]
8460: @end example
8461:
8462: Here is an example of adding a definition to the environment word list:
8463:
8464: @example
8465: get-current environment-wordlist set-current
8466: true constant block
8467: true constant block-ext
8468: set-current
8469: @end example
8470:
8471: You can see what definitions are in the environment word list like this:
8472:
8473: @example
8474: environment-wordlist >order words previous
8475: @end example
8476:
8477:
8478: @c -------------------------------------------------------------
8479: @node Files, Blocks, Environmental Queries, Words
8480: @section Files
8481: @cindex files
8482: @cindex I/O - file-handling
8483:
8484: Gforth provides facilities for accessing files that are stored in the
8485: host operating system's file-system. Files that are processed by Gforth
8486: can be divided into two categories:
8487:
8488: @itemize @bullet
8489: @item
8490: Files that are processed by the Text Interpreter (@dfn{Forth source files}).
8491: @item
8492: Files that are processed by some other program (@dfn{general files}).
8493: @end itemize
8494:
8495: @menu
8496: * Forth source files::
8497: * General files::
8498: * Redirection::
8499: * Search Paths::
8500: @end menu
8501:
8502: @c -------------------------------------------------------------
8503: @node Forth source files, General files, Files, Files
8504: @subsection Forth source files
8505: @cindex including files
8506: @cindex Forth source files
8507:
8508: The simplest way to interpret the contents of a file is to use one of
8509: these two formats:
8510:
8511: @example
8512: include mysource.fs
8513: s" mysource.fs" included
8514: @end example
8515:
8516: You usually want to include a file only if it is not included already
8517: (by, say, another source file). In that case, you can use one of these
8518: three formats:
8519:
8520: @example
8521: require mysource.fs
8522: needs mysource.fs
8523: s" mysource.fs" required
8524: @end example
8525:
8526: @cindex stack effect of included files
8527: @cindex including files, stack effect
8528: It is good practice to write your source files such that interpreting them
8529: does not change the stack. Source files designed in this way can be used with
8530: @code{required} and friends without complications. For example:
8531:
8532: @example
8533: 1024 require foo.fs drop
8534: @end example
8535:
8536: Here you want to pass the argument 1024 (e.g., a buffer size) to
8537: @file{foo.fs}. Interpreting @file{foo.fs} has the stack effect ( n -- n
8538: ), which allows its use with @code{require}. Of course with such
8539: parameters to required files, you have to ensure that the first
8540: @code{require} fits for all uses (i.e., @code{require} it early in the
8541: master load file).
8542:
8543: doc-include-file
8544: doc-included
8545: doc-included?
8546: doc-include
8547: doc-required
8548: doc-require
8549: doc-needs
8550: @c doc-init-included-files @c internal
8551: doc-sourcefilename
8552: doc-sourceline#
8553:
8554: A definition in ANS Forth for @code{required} is provided in
8555: @file{compat/required.fs}.
8556:
8557: @c -------------------------------------------------------------
8558: @node General files, Redirection, Forth source files, Files
8559: @subsection General files
8560: @cindex general files
8561: @cindex file-handling
8562:
8563: Files are opened/created by name and type. The following file access
8564: methods (FAMs) are recognised:
8565:
8566: @cindex fam (file access method)
8567: doc-r/o
8568: doc-r/w
8569: doc-w/o
8570: doc-bin
8571:
8572:
8573: When a file is opened/created, it returns a file identifier,
8574: @i{wfileid} that is used for all other file commands. All file
8575: commands also return a status value, @i{wior}, that is 0 for a
8576: successful operation and an implementation-defined non-zero value in the
8577: case of an error.
8578:
8579:
8580: doc-open-file
8581: doc-create-file
8582:
8583: doc-close-file
8584: doc-delete-file
8585: doc-rename-file
8586: doc-read-file
8587: doc-read-line
8588: doc-key-file
8589: doc-key?-file
8590: doc-write-file
8591: doc-write-line
8592: doc-emit-file
8593: doc-flush-file
8594:
8595: doc-file-status
8596: doc-file-position
8597: doc-reposition-file
8598: doc-file-size
8599: doc-resize-file
8600:
8601: doc-slurp-file
8602: doc-slurp-fid
8603: doc-stdin
8604: doc-stdout
8605: doc-stderr
8606:
8607: @c ---------------------------------------------------------
8608: @node Redirection, Search Paths, General files, Files
8609: @subsection Redirection
8610: @cindex Redirection
8611: @cindex Input Redirection
8612: @cindex Output Redirection
8613:
8614: You can redirect the output of @code{type} and @code{emit} and all the
8615: words that use them (all output words that don't have an explicit
8616: target file) to an arbitrary file with the @code{outfile-execute},
8617: used like this:
8618:
8619: @example
8620: : some-warning ( n -- )
8621: cr ." warning# " . ;
8622:
8623: : print-some-warning ( n -- )
8624: ['] some-warning stderr outfile-execute ;
8625: @end example
8626:
8627: After @code{some-warning} is executed, the original output direction
8628: is restored; this construct is safe against exceptions. Similarly,
8629: there is @code{infile-execute} for redirecting the input of @code{key}
8630: and its users (any input word that does not take a file explicitly).
8631:
8632: doc-outfile-execute
8633: doc-infile-execute
8634:
8635: If you do not want to redirect the input or output to a file, you can
8636: also make use of the fact that @code{key}, @code{emit} and @code{type}
8637: are deferred words (@pxref{Deferred Words}). However, in that case
8638: you have to worry about the restoration and the protection against
8639: exceptions yourself; also, note that for redirecting the output in
8640: this way, you have to redirect both @code{emit} and @code{type}.
8641:
8642: @c ---------------------------------------------------------
8643: @node Search Paths, , Redirection, Files
8644: @subsection Search Paths
8645: @cindex path for @code{included}
8646: @cindex file search path
8647: @cindex @code{include} search path
8648: @cindex search path for files
8649:
8650: If you specify an absolute filename (i.e., a filename starting with
8651: @file{/} or @file{~}, or with @file{:} in the second position (as in
8652: @samp{C:...})) for @code{included} and friends, that file is included
8653: just as you would expect.
8654:
8655: If the filename starts with @file{./}, this refers to the directory that
8656: the present file was @code{included} from. This allows files to include
8657: other files relative to their own position (irrespective of the current
8658: working directory or the absolute position). This feature is essential
8659: for libraries consisting of several files, where a file may include
8660: other files from the library. It corresponds to @code{#include "..."}
8661: in C. If the current input source is not a file, @file{.} refers to the
8662: directory of the innermost file being included, or, if there is no file
8663: being included, to the current working directory.
8664:
8665: For relative filenames (not starting with @file{./}), Gforth uses a
8666: search path similar to Forth's search order (@pxref{Word Lists}). It
8667: tries to find the given filename in the directories present in the path,
8668: and includes the first one it finds. There are separate search paths for
8669: Forth source files and general files. If the search path contains the
8670: directory @file{.}, this refers to the directory of the current file, or
8671: the working directory, as if the file had been specified with @file{./}.
8672:
8673: Use @file{~+} to refer to the current working directory (as in the
8674: @code{bash}).
8675:
8676: @c anton: fold the following subsubsections into this subsection?
8677:
8678: @menu
8679: * Source Search Paths::
8680: * General Search Paths::
8681: @end menu
8682:
8683: @c ---------------------------------------------------------
8684: @node Source Search Paths, General Search Paths, Search Paths, Search Paths
8685: @subsubsection Source Search Paths
8686: @cindex search path control, source files
8687:
8688: The search path is initialized when you start Gforth (@pxref{Invoking
8689: Gforth}). You can display it and change it using @code{fpath} in
8690: combination with the general path handling words.
8691:
8692: doc-fpath
8693: @c the functionality of the following words is easily available through
8694: @c fpath and the general path words. The may go away.
8695: @c doc-.fpath
8696: @c doc-fpath+
8697: @c doc-fpath=
8698: @c doc-open-fpath-file
8699:
8700: @noindent
8701: Here is an example of using @code{fpath} and @code{require}:
8702:
8703: @example
8704: fpath path= /usr/lib/forth/|./
8705: require timer.fs
8706: @end example
8707:
8708:
8709: @c ---------------------------------------------------------
8710: @node General Search Paths, , Source Search Paths, Search Paths
8711: @subsubsection General Search Paths
8712: @cindex search path control, source files
8713:
8714: Your application may need to search files in several directories, like
8715: @code{included} does. To facilitate this, Gforth allows you to define
8716: and use your own search paths, by providing generic equivalents of the
8717: Forth search path words:
8718:
8719: doc-open-path-file
8720: doc-path-allot
8721: doc-clear-path
8722: doc-also-path
8723: doc-.path
8724: doc-path+
8725: doc-path=
8726:
8727: @c anton: better define a word for it, say "path-allot ( ucount -- path-addr )
8728:
8729: Here's an example of creating an empty search path:
8730: @c
8731: @example
8732: create mypath 500 path-allot \ maximum length 500 chars (is checked)
8733: @end example
8734:
8735: @c -------------------------------------------------------------
8736: @node Blocks, Other I/O, Files, Words
8737: @section Blocks
8738: @cindex I/O - blocks
8739: @cindex blocks
8740:
8741: When you run Gforth on a modern desk-top computer, it runs under the
8742: control of an operating system which provides certain services. One of
8743: these services is @var{file services}, which allows Forth source code
8744: and data to be stored in files and read into Gforth (@pxref{Files}).
8745:
8746: Traditionally, Forth has been an important programming language on
8747: systems where it has interfaced directly to the underlying hardware with
8748: no intervening operating system. Forth provides a mechanism, called
8749: @dfn{blocks}, for accessing mass storage on such systems.
8750:
8751: A block is a 1024-byte data area, which can be used to hold data or
8752: Forth source code. No structure is imposed on the contents of the
8753: block. A block is identified by its number; blocks are numbered
8754: contiguously from 1 to an implementation-defined maximum.
8755:
8756: A typical system that used blocks but no operating system might use a
8757: single floppy-disk drive for mass storage, with the disks formatted to
8758: provide 256-byte sectors. Blocks would be implemented by assigning the
8759: first four sectors of the disk to block 1, the second four sectors to
8760: block 2 and so on, up to the limit of the capacity of the disk. The disk
8761: would not contain any file system information, just the set of blocks.
8762:
8763: @cindex blocks file
8764: On systems that do provide file services, blocks are typically
8765: implemented by storing a sequence of blocks within a single @dfn{blocks
8766: file}. The size of the blocks file will be an exact multiple of 1024
8767: bytes, corresponding to the number of blocks it contains. This is the
8768: mechanism that Gforth uses.
8769:
8770: @cindex @file{blocks.fb}
8771: Only one blocks file can be open at a time. If you use block words without
8772: having specified a blocks file, Gforth defaults to the blocks file
8773: @file{blocks.fb}. Gforth uses the Forth search path when attempting to
8774: locate a blocks file (@pxref{Source Search Paths}).
8775:
8776: @cindex block buffers
8777: When you read and write blocks under program control, Gforth uses a
8778: number of @dfn{block buffers} as intermediate storage. These buffers are
8779: not used when you use @code{load} to interpret the contents of a block.
8780:
8781: The behaviour of the block buffers is analagous to that of a cache.
8782: Each block buffer has three states:
8783:
8784: @itemize @bullet
8785: @item
8786: Unassigned
8787: @item
8788: Assigned-clean
8789: @item
8790: Assigned-dirty
8791: @end itemize
8792:
8793: Initially, all block buffers are @i{unassigned}. In order to access a
8794: block, the block (specified by its block number) must be assigned to a
8795: block buffer.
8796:
8797: The assignment of a block to a block buffer is performed by @code{block}
8798: or @code{buffer}. Use @code{block} when you wish to modify the existing
8799: contents of a block. Use @code{buffer} when you don't care about the
8800: existing contents of the block@footnote{The ANS Forth definition of
8801: @code{buffer} is intended not to cause disk I/O; if the data associated
8802: with the particular block is already stored in a block buffer due to an
8803: earlier @code{block} command, @code{buffer} will return that block
8804: buffer and the existing contents of the block will be
8805: available. Otherwise, @code{buffer} will simply assign a new, empty
8806: block buffer for the block.}.
8807:
8808: Once a block has been assigned to a block buffer using @code{block} or
8809: @code{buffer}, that block buffer becomes the @i{current block
8810: buffer}. Data may only be manipulated (read or written) within the
8811: current block buffer.
8812:
8813: When the contents of the current block buffer has been modified it is
8814: necessary, @emph{before calling @code{block} or @code{buffer} again}, to
8815: either abandon the changes (by doing nothing) or mark the block as
8816: changed (assigned-dirty), using @code{update}. Using @code{update} does
8817: not change the blocks file; it simply changes a block buffer's state to
8818: @i{assigned-dirty}. The block will be written implicitly when it's
8819: buffer is needed for another block, or explicitly by @code{flush} or
8820: @code{save-buffers}.
8821:
8822: word @code{Flush} writes all @i{assigned-dirty} blocks back to the
8823: blocks file on disk. Leaving Gforth with @code{bye} also performs a
8824: @code{flush}.
8825:
8826: In Gforth, @code{block} and @code{buffer} use a @i{direct-mapped}
8827: algorithm to assign a block buffer to a block. That means that any
8828: particular block can only be assigned to one specific block buffer,
8829: called (for the particular operation) the @i{victim buffer}. If the
8830: victim buffer is @i{unassigned} or @i{assigned-clean} it is allocated to
8831: the new block immediately. If it is @i{assigned-dirty} its current
8832: contents are written back to the blocks file on disk before it is
8833: allocated to the new block.
8834:
8835: Although no structure is imposed on the contents of a block, it is
8836: traditional to display the contents as 16 lines each of 64 characters. A
8837: block provides a single, continuous stream of input (for example, it
8838: acts as a single parse area) -- there are no end-of-line characters
8839: within a block, and no end-of-file character at the end of a
8840: block. There are two consequences of this:
8841:
8842: @itemize @bullet
8843: @item
8844: The last character of one line wraps straight into the first character
8845: of the following line
8846: @item
8847: The word @code{\} -- comment to end of line -- requires special
8848: treatment; in the context of a block it causes all characters until the
8849: end of the current 64-character ``line'' to be ignored.
8850: @end itemize
8851:
8852: In Gforth, when you use @code{block} with a non-existent block number,
8853: the current blocks file will be extended to the appropriate size and the
8854: block buffer will be initialised with spaces.
8855:
8856: Gforth includes a simple block editor (type @code{use blocked.fb 0 list}
8857: for details) but doesn't encourage the use of blocks; the mechanism is
8858: only provided for backward compatibility -- ANS Forth requires blocks to
8859: be available when files are.
8860:
8861: Common techniques that are used when working with blocks include:
8862:
8863: @itemize @bullet
8864: @item
8865: A screen editor that allows you to edit blocks without leaving the Forth
8866: environment.
8867: @item
8868: Shadow screens; where every code block has an associated block
8869: containing comments (for example: code in odd block numbers, comments in
8870: even block numbers). Typically, the block editor provides a convenient
8871: mechanism to toggle between code and comments.
8872: @item
8873: Load blocks; a single block (typically block 1) contains a number of
8874: @code{thru} commands which @code{load} the whole of the application.
8875: @end itemize
8876:
8877: See Frank Sergeant's Pygmy Forth to see just how well blocks can be
8878: integrated into a Forth programming environment.
8879:
8880: @comment TODO what about errors on open-blocks?
8881:
8882: doc-open-blocks
8883: doc-use
8884: doc-block-offset
8885: doc-get-block-fid
8886: doc-block-position
8887:
8888: doc-list
8889: doc-scr
8890:
8891: doc-block
8892: doc-buffer
8893:
8894: doc-empty-buffers
8895: doc-empty-buffer
8896: doc-update
8897: doc-updated?
8898: doc-save-buffers
8899: doc-save-buffer
8900: doc-flush
8901:
8902: doc-load
8903: doc-thru
8904: doc-+load
8905: doc-+thru
8906: doc---gforthman--->
8907: doc-block-included
8908:
8909:
8910: @c -------------------------------------------------------------
8911: @node Other I/O, OS command line arguments, Blocks, Words
8912: @section Other I/O
8913: @cindex I/O - keyboard and display
8914:
8915: @menu
8916: * Simple numeric output:: Predefined formats
8917: * Formatted numeric output:: Formatted (pictured) output
8918: * String Formats:: How Forth stores strings in memory
8919: * Displaying characters and strings:: Other stuff
8920: * Terminal output:: Cursor positioning etc.
8921: * Single-key input::
8922: * Line input and conversion::
8923: * Pipes:: How to create your own pipes
8924: * Xchars and Unicode:: Non-ASCII characters
8925: @end menu
8926:
8927: @node Simple numeric output, Formatted numeric output, Other I/O, Other I/O
8928: @subsection Simple numeric output
8929: @cindex numeric output - simple/free-format
8930:
8931: The simplest output functions are those that display numbers from the
8932: data or floating-point stacks. Floating-point output is always displayed
8933: using base 10. Numbers displayed from the data stack use the value stored
8934: in @code{base}.
8935:
8936:
8937: doc-.
8938: doc-dec.
8939: doc-hex.
8940: doc-u.
8941: doc-.r
8942: doc-u.r
8943: doc-d.
8944: doc-ud.
8945: doc-d.r
8946: doc-ud.r
8947: doc-f.
8948: doc-fe.
8949: doc-fs.
8950: doc-f.rdp
8951:
8952: Examples of printing the number 1234.5678E23 in the different floating-point output
8953: formats are shown below:
8954:
8955: @example
8956: f. 123456779999999000000000000.
8957: fe. 123.456779999999E24
8958: fs. 1.23456779999999E26
8959: @end example
8960:
8961:
8962: @node Formatted numeric output, String Formats, Simple numeric output, Other I/O
8963: @subsection Formatted numeric output
8964: @cindex formatted numeric output
8965: @cindex pictured numeric output
8966: @cindex numeric output - formatted
8967:
8968: Forth traditionally uses a technique called @dfn{pictured numeric
8969: output} for formatted printing of integers. In this technique, digits
8970: are extracted from the number (using the current output radix defined by
8971: @code{base}), converted to ASCII codes and appended to a string that is
8972: built in a scratch-pad area of memory (@pxref{core-idef,
8973: Implementation-defined options, Implementation-defined
8974: options}). Arbitrary characters can be appended to the string during the
8975: extraction process. The completed string is specified by an address
8976: and length and can be manipulated (@code{TYPE}ed, copied, modified)
8977: under program control.
8978:
8979: All of the integer output words described in the previous section
8980: (@pxref{Simple numeric output}) are implemented in Gforth using pictured
8981: numeric output.
8982:
8983: Three important things to remember about pictured numeric output:
8984:
8985: @itemize @bullet
8986: @item
8987: It always operates on double-precision numbers; to display a
8988: single-precision number, convert it first (for ways of doing this
8989: @pxref{Double precision}).
8990: @item
8991: It always treats the double-precision number as though it were
8992: unsigned. The examples below show ways of printing signed numbers.
8993: @item
8994: The string is built up from right to left; least significant digit first.
8995: @end itemize
8996:
8997:
8998: doc-<#
8999: doc-<<#
9000: doc-#
9001: doc-#s
9002: doc-hold
9003: doc-sign
9004: doc-#>
9005: doc-#>>
9006:
9007: doc-represent
9008: doc-f>str-rdp
9009: doc-f>buf-rdp
9010:
9011:
9012: @noindent
9013: Here are some examples of using pictured numeric output:
9014:
9015: @example
9016: : my-u. ( u -- )
9017: \ Simplest use of pns.. behaves like Standard u.
9018: 0 \ convert to unsigned double
9019: <<# \ start conversion
9020: #s \ convert all digits
9021: #> \ complete conversion
9022: TYPE SPACE \ display, with trailing space
9023: #>> ; \ release hold area
9024:
9025: : cents-only ( u -- )
9026: 0 \ convert to unsigned double
9027: <<# \ start conversion
9028: # # \ convert two least-significant digits
9029: #> \ complete conversion, discard other digits
9030: TYPE SPACE \ display, with trailing space
9031: #>> ; \ release hold area
9032:
9033: : dollars-and-cents ( u -- )
9034: 0 \ convert to unsigned double
9035: <<# \ start conversion
9036: # # \ convert two least-significant digits
9037: [char] . hold \ insert decimal point
9038: #s \ convert remaining digits
9039: [char] $ hold \ append currency symbol
9040: #> \ complete conversion
9041: TYPE SPACE \ display, with trailing space
9042: #>> ; \ release hold area
9043:
9044: : my-. ( n -- )
9045: \ handling negatives.. behaves like Standard .
9046: s>d \ convert to signed double
9047: swap over dabs \ leave sign byte followed by unsigned double
9048: <<# \ start conversion
9049: #s \ convert all digits
9050: rot sign \ get at sign byte, append "-" if needed
9051: #> \ complete conversion
9052: TYPE SPACE \ display, with trailing space
9053: #>> ; \ release hold area
9054:
9055: : account. ( n -- )
9056: \ accountants don't like minus signs, they use parentheses
9057: \ for negative numbers
9058: s>d \ convert to signed double
9059: swap over dabs \ leave sign byte followed by unsigned double
9060: <<# \ start conversion
9061: 2 pick \ get copy of sign byte
9062: 0< IF [char] ) hold THEN \ right-most character of output
9063: #s \ convert all digits
9064: rot \ get at sign byte
9065: 0< IF [char] ( hold THEN
9066: #> \ complete conversion
9067: TYPE SPACE \ display, with trailing space
9068: #>> ; \ release hold area
9069:
9070: @end example
9071:
9072: Here are some examples of using these words:
9073:
9074: @example
9075: 1 my-u. 1
9076: hex -1 my-u. decimal FFFFFFFF
9077: 1 cents-only 01
9078: 1234 cents-only 34
9079: 2 dollars-and-cents $0.02
9080: 1234 dollars-and-cents $12.34
9081: 123 my-. 123
9082: -123 my. -123
9083: 123 account. 123
9084: -456 account. (456)
9085: @end example
9086:
9087:
9088: @node String Formats, Displaying characters and strings, Formatted numeric output, Other I/O
9089: @subsection String Formats
9090: @cindex strings - see character strings
9091: @cindex character strings - formats
9092: @cindex I/O - see character strings
9093: @cindex counted strings
9094:
9095: @c anton: this does not really belong here; maybe the memory section,
9096: @c or the principles chapter
9097:
9098: Forth commonly uses two different methods for representing character
9099: strings:
9100:
9101: @itemize @bullet
9102: @item
9103: @cindex address of counted string
9104: @cindex counted string
9105: As a @dfn{counted string}, represented by a @i{c-addr}. The char
9106: addressed by @i{c-addr} contains a character-count, @i{n}, of the
9107: string and the string occupies the subsequent @i{n} char addresses in
9108: memory.
9109: @item
9110: As cell pair on the stack; @i{c-addr u}, where @i{u} is the length
9111: of the string in characters, and @i{c-addr} is the address of the
9112: first byte of the string.
9113: @end itemize
9114:
9115: ANS Forth encourages the use of the second format when representing
9116: strings.
9117:
9118:
9119: doc-count
9120:
9121:
9122: For words that move, copy and search for strings see @ref{Memory
9123: Blocks}. For words that display characters and strings see
9124: @ref{Displaying characters and strings}.
9125:
9126: @node Displaying characters and strings, Terminal output, String Formats, Other I/O
9127: @subsection Displaying characters and strings
9128: @cindex characters - compiling and displaying
9129: @cindex character strings - compiling and displaying
9130:
9131: This section starts with a glossary of Forth words and ends with a set
9132: of examples.
9133:
9134: doc-bl
9135: doc-space
9136: doc-spaces
9137: doc-emit
9138: doc-toupper
9139: doc-."
9140: doc-.(
9141: doc-.\"
9142: doc-type
9143: doc-typewhite
9144: doc-cr
9145: @cindex cursor control
9146: doc-s"
9147: doc-s\"
9148: doc-c"
9149: doc-char
9150: doc-[char]
9151:
9152:
9153: @noindent
9154: As an example, consider the following text, stored in a file @file{test.fs}:
9155:
9156: @example
9157: .( text-1)
9158: : my-word
9159: ." text-2" cr
9160: .( text-3)
9161: ;
9162:
9163: ." text-4"
9164:
9165: : my-char
9166: [char] ALPHABET emit
9167: char emit
9168: ;
9169: @end example
9170:
9171: When you load this code into Gforth, the following output is generated:
9172:
9173: @example
9174: @kbd{include test.fs @key{RET}} text-1text-3text-4 ok
9175: @end example
9176:
9177: @itemize @bullet
9178: @item
9179: Messages @code{text-1} and @code{text-3} are displayed because @code{.(}
9180: is an immediate word; it behaves in the same way whether it is used inside
9181: or outside a colon definition.
9182: @item
9183: Message @code{text-4} is displayed because of Gforth's added interpretation
9184: semantics for @code{."}.
9185: @item
9186: Message @code{text-2} is @i{not} displayed, because the text interpreter
9187: performs the compilation semantics for @code{."} within the definition of
9188: @code{my-word}.
9189: @end itemize
9190:
9191: Here are some examples of executing @code{my-word} and @code{my-char}:
9192:
9193: @example
9194: @kbd{my-word @key{RET}} text-2
9195: ok
9196: @kbd{my-char fred @key{RET}} Af ok
9197: @kbd{my-char jim @key{RET}} Aj ok
9198: @end example
9199:
9200: @itemize @bullet
9201: @item
9202: Message @code{text-2} is displayed because of the run-time behaviour of
9203: @code{."}.
9204: @item
9205: @code{[char]} compiles the ``A'' from ``ALPHABET'' and puts its display code
9206: on the stack at run-time. @code{emit} always displays the character
9207: when @code{my-char} is executed.
9208: @item
9209: @code{char} parses a string at run-time and the second @code{emit} displays
9210: the first character of the string.
9211: @item
9212: If you type @code{see my-char} you can see that @code{[char]} discarded
9213: the text ``LPHABET'' and only compiled the display code for ``A'' into the
9214: definition of @code{my-char}.
9215: @end itemize
9216:
9217:
9218: @node Terminal output, Single-key input, Displaying characters and strings, Other I/O
9219: @subsection Terminal output
9220: @cindex output to terminal
9221: @cindex terminal output
9222:
9223: If you are outputting to a terminal, you may want to control the
9224: positioning of the cursor:
9225: @cindex cursor positioning
9226:
9227: doc-at-xy
9228:
9229: In order to know where to position the cursor, it is often helpful to
9230: know the size of the screen:
9231: @cindex terminal size
9232:
9233: doc-form
9234:
9235: And sometimes you want to use:
9236: @cindex clear screen
9237:
9238: doc-page
9239:
9240: Note that on non-terminals you should use @code{12 emit}, not
9241: @code{page}, to get a form feed.
9242:
9243:
9244: @node Single-key input, Line input and conversion, Terminal output, Other I/O
9245: @subsection Single-key input
9246: @cindex single-key input
9247: @cindex input, single-key
9248:
9249: If you want to get a single printable character, you can use
9250: @code{key}; to check whether a character is available for @code{key},
9251: you can use @code{key?}.
9252:
9253: doc-key
9254: doc-key?
9255:
9256: If you want to process a mix of printable and non-printable
9257: characters, you can do that with @code{ekey} and friends. @code{Ekey}
9258: produces a keyboard event that you have to convert into a character
9259: with @code{ekey>char} or into a key identifier with @code{ekey>fkey}.
9260:
9261: Typical code for using EKEY looks like this:
9262:
9263: @example
9264: ekey ekey>char if ( c )
9265: ... \ do something with the character
9266: else ekey>fkey if ( key-id )
9267: case
9268: k-up of ... endof
9269: k-f1 of ... endof
9270: k-left k-shift-mask or k-ctrl-mask or of ... endof
9271: ...
9272: endcase
9273: else ( keyboard-event )
9274: drop \ just ignore an unknown keyboard event type
9275: then then
9276: @end example
9277:
9278: doc-ekey
9279: doc-ekey>char
9280: doc-ekey>fkey
9281: doc-ekey?
9282:
9283: The key identifiers for cursor keys are:
9284:
9285: doc-k-left
9286: doc-k-right
9287: doc-k-up
9288: doc-k-down
9289: doc-k-home
9290: doc-k-end
9291: doc-k-prior
9292: doc-k-next
9293: doc-k-insert
9294: doc-k-delete
9295:
9296: The key identifiers for function keys (aka keypad keys) are:
9297:
9298: doc-k-f1
9299: doc-k-f2
9300: doc-k-f3
9301: doc-k-f4
9302: doc-k-f5
9303: doc-k-f6
9304: doc-k-f7
9305: doc-k-f8
9306: doc-k-f9
9307: doc-k-f10
9308: doc-k-f11
9309: doc-k-f12
9310:
9311: Note that @code{k-f11} and @code{k-f12} are not as widely available.
9312:
9313: You can combine these key identifiers with masks for various shift keys:
9314:
9315: doc-k-shift-mask
9316: doc-k-ctrl-mask
9317: doc-k-alt-mask
9318:
9319: Note that, even if a Forth system has @code{ekey>fkey} and the key
9320: identifier words, the keys are not necessarily available or it may not
9321: necessarily be able to report all the keys and all the possible
9322: combinations with shift masks. Therefore, write your programs in such
9323: a way that they are still useful even if the keys and key combinations
9324: cannot be pressed or are not recognized.
9325:
9326: Examples: Older keyboards often do not have an F11 and F12 key. If
9327: you run Gforth in an xterm, the xterm catches a number of combinations
9328: (e.g., @key{Shift-Up}), and never passes it to Gforth. Finally,
9329: Gforth currently does not recognize and report combinations with
9330: multiple shift keys (so the @key{shift-ctrl-left} case in the example
9331: above would never be entered).
9332:
9333: Gforth recognizes various keys available on ANSI terminals (in MS-DOS
9334: you need the ANSI.SYS driver to get that behaviour); it works by
9335: recognizing the escape sequences that ANSI terminals send when such a
9336: key is pressed. If you have a terminal that sends other escape
9337: sequences, you will not get useful results on Gforth. Other Forth
9338: systems may work in a different way.
9339:
9340: Gforth also provides a few words for outputting names of function
9341: keys:
9342:
9343: doc-fkey.
9344: doc-simple-fkey-string
9345:
9346:
9347: @node Line input and conversion, Pipes, Single-key input, Other I/O
9348: @subsection Line input and conversion
9349: @cindex line input from terminal
9350: @cindex input, linewise from terminal
9351: @cindex convertin strings to numbers
9352: @cindex I/O - see input
9353:
9354: For ways of storing character strings in memory see @ref{String Formats}.
9355:
9356: @comment TODO examples for >number >float accept key key? pad parse word refill
9357: @comment then index them
9358:
9359: Words for inputting one line from the keyboard:
9360:
9361: doc-accept
9362: doc-edit-line
9363:
9364: Conversion words:
9365:
9366: doc-s>number?
9367: doc-s>unumber?
9368: doc->number
9369: doc->float
9370:
9371:
9372: @comment obsolescent words..
9373: Obsolescent input and conversion words:
9374:
9375: doc-convert
9376: doc-expect
9377: doc-span
9378:
9379:
9380: @node Pipes, Xchars and Unicode, Line input and conversion, Other I/O
9381: @subsection Pipes
9382: @cindex pipes, creating your own
9383:
9384: In addition to using Gforth in pipes created by other processes
9385: (@pxref{Gforth in pipes}), you can create your own pipe with
9386: @code{open-pipe}, and read from or write to it.
9387:
9388: doc-open-pipe
9389: doc-close-pipe
9390:
9391: If you write to a pipe, Gforth can throw a @code{broken-pipe-error}; if
9392: you don't catch this exception, Gforth will catch it and exit, usually
9393: silently (@pxref{Gforth in pipes}). Since you probably do not want
9394: this, you should wrap a @code{catch} or @code{try} block around the code
9395: from @code{open-pipe} to @code{close-pipe}, so you can deal with the
9396: problem yourself, and then return to regular processing.
9397:
9398: doc-broken-pipe-error
9399:
9400: @node Xchars and Unicode, , Pipes, Other I/O
9401: @subsection Xchars and Unicode
9402:
9403: ASCII is only appropriate for the English language. Most western
9404: languages however fit somewhat into the Forth frame, since a byte is
9405: sufficient to encode the few special characters in each (though not
9406: always the same encoding can be used; latin-1 is most widely used,
9407: though). For other languages, different char-sets have to be used,
9408: several of them variable-width. Most prominent representant is
9409: UTF-8. Let's call these extended characters xchars. The primitive
9410: fixed-size characters stored as bytes are called pchars in this
9411: section.
9412:
9413: The xchar words add a few data types:
9414:
9415: @itemize
9416:
9417: @item
9418: @var{xc} is an extended char (xchar) on the stack. It occupies one cell,
9419: and is a subset of unsigned cell. Note: UTF-8 can not store more that
9420: 31 bits; on 16 bit systems, only the UCS16 subset of the UTF-8
9421: character set can be used.
9422:
9423: @item
9424: @var{xc-addr} is the address of an xchar in memory. Alignment
9425: requirements are the same as @var{c-addr}. The memory representation of an
9426: xchar differs from the stack representation, and depends on the
9427: encoding used. An xchar may use a variable number of pchars in memory.
9428:
9429: @item
9430: @var{xc-addr} @var{u} is a buffer of xchars in memory, starting at
9431: @var{xc-addr}, @var{u} pchars long.
9432:
9433: @end itemize
9434:
9435: doc-xc-size
9436: doc-x-size
9437: doc-xc@+
9438: doc-xc!+?
9439: doc-xchar+
9440: doc-xchar-
9441: doc-+x/string
9442: doc-x\string-
9443: doc--trailing-garbage
9444: doc-x-width
9445: doc-xkey
9446: doc-xemit
9447:
9448: There's a new environment query
9449:
9450: doc-xchar-encoding
9451:
9452: @node OS command line arguments, Locals, Other I/O, Words
9453: @section OS command line arguments
9454: @cindex OS command line arguments
9455: @cindex command line arguments, OS
9456: @cindex arguments, OS command line
9457:
9458: The usual way to pass arguments to Gforth programs on the command line
9459: is via the @option{-e} option, e.g.
9460:
9461: @example
9462: gforth -e "123 456" foo.fs -e bye
9463: @end example
9464:
9465: However, you may want to interpret the command-line arguments directly.
9466: In that case, you can access the (image-specific) command-line arguments
9467: through @code{next-arg}:
9468:
9469: doc-next-arg
9470:
9471: Here's an example program @file{echo.fs} for @code{next-arg}:
9472:
9473: @example
9474: : echo ( -- )
9475: begin
9476: next-arg 2dup 0 0 d<> while
9477: type space
9478: repeat
9479: 2drop ;
9480:
9481: echo cr bye
9482: @end example
9483:
9484: This can be invoked with
9485:
9486: @example
9487: gforth echo.fs hello world
9488: @end example
9489:
9490: and it will print
9491:
9492: @example
9493: hello world
9494: @end example
9495:
9496: The next lower level of dealing with the OS command line are the
9497: following words:
9498:
9499: doc-arg
9500: doc-shift-args
9501:
9502: Finally, at the lowest level Gforth provides the following words:
9503:
9504: doc-argc
9505: doc-argv
9506:
9507: @c -------------------------------------------------------------
9508: @node Locals, Structures, OS command line arguments, Words
9509: @section Locals
9510: @cindex locals
9511:
9512: Local variables can make Forth programming more enjoyable and Forth
9513: programs easier to read. Unfortunately, the locals of ANS Forth are
9514: laden with restrictions. Therefore, we provide not only the ANS Forth
9515: locals wordset, but also our own, more powerful locals wordset (we
9516: implemented the ANS Forth locals wordset through our locals wordset).
9517:
9518: The ideas in this section have also been published in M. Anton Ertl,
9519: @cite{@uref{http://www.complang.tuwien.ac.at/papers/ertl94l.ps.gz,
9520: Automatic Scoping of Local Variables}}, EuroForth '94.
9521:
9522: @menu
9523: * Gforth locals::
9524: * ANS Forth locals::
9525: @end menu
9526:
9527: @node Gforth locals, ANS Forth locals, Locals, Locals
9528: @subsection Gforth locals
9529: @cindex Gforth locals
9530: @cindex locals, Gforth style
9531:
9532: Locals can be defined with
9533:
9534: @example
9535: @{ local1 local2 ... -- comment @}
9536: @end example
9537: or
9538: @example
9539: @{ local1 local2 ... @}
9540: @end example
9541:
9542: E.g.,
9543: @example
9544: : max @{ n1 n2 -- n3 @}
9545: n1 n2 > if
9546: n1
9547: else
9548: n2
9549: endif ;
9550: @end example
9551:
9552: The similarity of locals definitions with stack comments is intended. A
9553: locals definition often replaces the stack comment of a word. The order
9554: of the locals corresponds to the order in a stack comment and everything
9555: after the @code{--} is really a comment.
9556:
9557: This similarity has one disadvantage: It is too easy to confuse locals
9558: declarations with stack comments, causing bugs and making them hard to
9559: find. However, this problem can be avoided by appropriate coding
9560: conventions: Do not use both notations in the same program. If you do,
9561: they should be distinguished using additional means, e.g. by position.
9562:
9563: @cindex types of locals
9564: @cindex locals types
9565: The name of the local may be preceded by a type specifier, e.g.,
9566: @code{F:} for a floating point value:
9567:
9568: @example
9569: : CX* @{ F: Ar F: Ai F: Br F: Bi -- Cr Ci @}
9570: \ complex multiplication
9571: Ar Br f* Ai Bi f* f-
9572: Ar Bi f* Ai Br f* f+ ;
9573: @end example
9574:
9575: @cindex flavours of locals
9576: @cindex locals flavours
9577: @cindex value-flavoured locals
9578: @cindex variable-flavoured locals
9579: Gforth currently supports cells (@code{W:}, @code{W^}), doubles
9580: (@code{D:}, @code{D^}), floats (@code{F:}, @code{F^}) and characters
9581: (@code{C:}, @code{C^}) in two flavours: a value-flavoured local (defined
9582: with @code{W:}, @code{D:} etc.) produces its value and can be changed
9583: with @code{TO}. A variable-flavoured local (defined with @code{W^} etc.)
9584: produces its address (which becomes invalid when the variable's scope is
9585: left). E.g., the standard word @code{emit} can be defined in terms of
9586: @code{type} like this:
9587:
9588: @example
9589: : emit @{ C^ char* -- @}
9590: char* 1 type ;
9591: @end example
9592:
9593: @cindex default type of locals
9594: @cindex locals, default type
9595: A local without type specifier is a @code{W:} local. Both flavours of
9596: locals are initialized with values from the data or FP stack.
9597:
9598: Currently there is no way to define locals with user-defined data
9599: structures, but we are working on it.
9600:
9601: Gforth allows defining locals everywhere in a colon definition. This
9602: poses the following questions:
9603:
9604: @menu
9605: * Where are locals visible by name?::
9606: * How long do locals live?::
9607: * Locals programming style::
9608: * Locals implementation::
9609: @end menu
9610:
9611: @node Where are locals visible by name?, How long do locals live?, Gforth locals, Gforth locals
9612: @subsubsection Where are locals visible by name?
9613: @cindex locals visibility
9614: @cindex visibility of locals
9615: @cindex scope of locals
9616:
9617: Basically, the answer is that locals are visible where you would expect
9618: it in block-structured languages, and sometimes a little longer. If you
9619: want to restrict the scope of a local, enclose its definition in
9620: @code{SCOPE}...@code{ENDSCOPE}.
9621:
9622:
9623: doc-scope
9624: doc-endscope
9625:
9626:
9627: These words behave like control structure words, so you can use them
9628: with @code{CS-PICK} and @code{CS-ROLL} to restrict the scope in
9629: arbitrary ways.
9630:
9631: If you want a more exact answer to the visibility question, here's the
9632: basic principle: A local is visible in all places that can only be
9633: reached through the definition of the local@footnote{In compiler
9634: construction terminology, all places dominated by the definition of the
9635: local.}. In other words, it is not visible in places that can be reached
9636: without going through the definition of the local. E.g., locals defined
9637: in @code{IF}...@code{ENDIF} are visible until the @code{ENDIF}, locals
9638: defined in @code{BEGIN}...@code{UNTIL} are visible after the
9639: @code{UNTIL} (until, e.g., a subsequent @code{ENDSCOPE}).
9640:
9641: The reasoning behind this solution is: We want to have the locals
9642: visible as long as it is meaningful. The user can always make the
9643: visibility shorter by using explicit scoping. In a place that can
9644: only be reached through the definition of a local, the meaning of a
9645: local name is clear. In other places it is not: How is the local
9646: initialized at the control flow path that does not contain the
9647: definition? Which local is meant, if the same name is defined twice in
9648: two independent control flow paths?
9649:
9650: This should be enough detail for nearly all users, so you can skip the
9651: rest of this section. If you really must know all the gory details and
9652: options, read on.
9653:
9654: In order to implement this rule, the compiler has to know which places
9655: are unreachable. It knows this automatically after @code{AHEAD},
9656: @code{AGAIN}, @code{EXIT} and @code{LEAVE}; in other cases (e.g., after
9657: most @code{THROW}s), you can use the word @code{UNREACHABLE} to tell the
9658: compiler that the control flow never reaches that place. If
9659: @code{UNREACHABLE} is not used where it could, the only consequence is
9660: that the visibility of some locals is more limited than the rule above
9661: says. If @code{UNREACHABLE} is used where it should not (i.e., if you
9662: lie to the compiler), buggy code will be produced.
9663:
9664:
9665: doc-unreachable
9666:
9667:
9668: Another problem with this rule is that at @code{BEGIN}, the compiler
9669: does not know which locals will be visible on the incoming
9670: back-edge. All problems discussed in the following are due to this
9671: ignorance of the compiler (we discuss the problems using @code{BEGIN}
9672: loops as examples; the discussion also applies to @code{?DO} and other
9673: loops). Perhaps the most insidious example is:
9674: @example
9675: AHEAD
9676: BEGIN
9677: x
9678: [ 1 CS-ROLL ] THEN
9679: @{ x @}
9680: ...
9681: UNTIL
9682: @end example
9683:
9684: This should be legal according to the visibility rule. The use of
9685: @code{x} can only be reached through the definition; but that appears
9686: textually below the use.
9687:
9688: From this example it is clear that the visibility rules cannot be fully
9689: implemented without major headaches. Our implementation treats common
9690: cases as advertised and the exceptions are treated in a safe way: The
9691: compiler makes a reasonable guess about the locals visible after a
9692: @code{BEGIN}; if it is too pessimistic, the
9693: user will get a spurious error about the local not being defined; if the
9694: compiler is too optimistic, it will notice this later and issue a
9695: warning. In the case above the compiler would complain about @code{x}
9696: being undefined at its use. You can see from the obscure examples in
9697: this section that it takes quite unusual control structures to get the
9698: compiler into trouble, and even then it will often do fine.
9699:
9700: If the @code{BEGIN} is reachable from above, the most optimistic guess
9701: is that all locals visible before the @code{BEGIN} will also be
9702: visible after the @code{BEGIN}. This guess is valid for all loops that
9703: are entered only through the @code{BEGIN}, in particular, for normal
9704: @code{BEGIN}...@code{WHILE}...@code{REPEAT} and
9705: @code{BEGIN}...@code{UNTIL} loops and it is implemented in our
9706: compiler. When the branch to the @code{BEGIN} is finally generated by
9707: @code{AGAIN} or @code{UNTIL}, the compiler checks the guess and
9708: warns the user if it was too optimistic:
9709: @example
9710: IF
9711: @{ x @}
9712: BEGIN
9713: \ x ?
9714: [ 1 cs-roll ] THEN
9715: ...
9716: UNTIL
9717: @end example
9718:
9719: Here, @code{x} lives only until the @code{BEGIN}, but the compiler
9720: optimistically assumes that it lives until the @code{THEN}. It notices
9721: this difference when it compiles the @code{UNTIL} and issues a
9722: warning. The user can avoid the warning, and make sure that @code{x}
9723: is not used in the wrong area by using explicit scoping:
9724: @example
9725: IF
9726: SCOPE
9727: @{ x @}
9728: ENDSCOPE
9729: BEGIN
9730: [ 1 cs-roll ] THEN
9731: ...
9732: UNTIL
9733: @end example
9734:
9735: Since the guess is optimistic, there will be no spurious error messages
9736: about undefined locals.
9737:
9738: If the @code{BEGIN} is not reachable from above (e.g., after
9739: @code{AHEAD} or @code{EXIT}), the compiler cannot even make an
9740: optimistic guess, as the locals visible after the @code{BEGIN} may be
9741: defined later. Therefore, the compiler assumes that no locals are
9742: visible after the @code{BEGIN}. However, the user can use
9743: @code{ASSUME-LIVE} to make the compiler assume that the same locals are
9744: visible at the BEGIN as at the point where the top control-flow stack
9745: item was created.
9746:
9747:
9748: doc-assume-live
9749:
9750:
9751: @noindent
9752: E.g.,
9753: @example
9754: @{ x @}
9755: AHEAD
9756: ASSUME-LIVE
9757: BEGIN
9758: x
9759: [ 1 CS-ROLL ] THEN
9760: ...
9761: UNTIL
9762: @end example
9763:
9764: Other cases where the locals are defined before the @code{BEGIN} can be
9765: handled by inserting an appropriate @code{CS-ROLL} before the
9766: @code{ASSUME-LIVE} (and changing the control-flow stack manipulation
9767: behind the @code{ASSUME-LIVE}).
9768:
9769: Cases where locals are defined after the @code{BEGIN} (but should be
9770: visible immediately after the @code{BEGIN}) can only be handled by
9771: rearranging the loop. E.g., the ``most insidious'' example above can be
9772: arranged into:
9773: @example
9774: BEGIN
9775: @{ x @}
9776: ... 0=
9777: WHILE
9778: x
9779: REPEAT
9780: @end example
9781:
9782: @node How long do locals live?, Locals programming style, Where are locals visible by name?, Gforth locals
9783: @subsubsection How long do locals live?
9784: @cindex locals lifetime
9785: @cindex lifetime of locals
9786:
9787: The right answer for the lifetime question would be: A local lives at
9788: least as long as it can be accessed. For a value-flavoured local this
9789: means: until the end of its visibility. However, a variable-flavoured
9790: local could be accessed through its address far beyond its visibility
9791: scope. Ultimately, this would mean that such locals would have to be
9792: garbage collected. Since this entails un-Forth-like implementation
9793: complexities, I adopted the same cowardly solution as some other
9794: languages (e.g., C): The local lives only as long as it is visible;
9795: afterwards its address is invalid (and programs that access it
9796: afterwards are erroneous).
9797:
9798: @node Locals programming style, Locals implementation, How long do locals live?, Gforth locals
9799: @subsubsection Locals programming style
9800: @cindex locals programming style
9801: @cindex programming style, locals
9802:
9803: The freedom to define locals anywhere has the potential to change
9804: programming styles dramatically. In particular, the need to use the
9805: return stack for intermediate storage vanishes. Moreover, all stack
9806: manipulations (except @code{PICK}s and @code{ROLL}s with run-time
9807: determined arguments) can be eliminated: If the stack items are in the
9808: wrong order, just write a locals definition for all of them; then
9809: write the items in the order you want.
9810:
9811: This seems a little far-fetched and eliminating stack manipulations is
9812: unlikely to become a conscious programming objective. Still, the number
9813: of stack manipulations will be reduced dramatically if local variables
9814: are used liberally (e.g., compare @code{max} (@pxref{Gforth locals}) with
9815: a traditional implementation of @code{max}).
9816:
9817: This shows one potential benefit of locals: making Forth programs more
9818: readable. Of course, this benefit will only be realized if the
9819: programmers continue to honour the principle of factoring instead of
9820: using the added latitude to make the words longer.
9821:
9822: @cindex single-assignment style for locals
9823: Using @code{TO} can and should be avoided. Without @code{TO},
9824: every value-flavoured local has only a single assignment and many
9825: advantages of functional languages apply to Forth. I.e., programs are
9826: easier to analyse, to optimize and to read: It is clear from the
9827: definition what the local stands for, it does not turn into something
9828: different later.
9829:
9830: E.g., a definition using @code{TO} might look like this:
9831: @example
9832: : strcmp @{ addr1 u1 addr2 u2 -- n @}
9833: u1 u2 min 0
9834: ?do
9835: addr1 c@@ addr2 c@@ -
9836: ?dup-if
9837: unloop exit
9838: then
9839: addr1 char+ TO addr1
9840: addr2 char+ TO addr2
9841: loop
9842: u1 u2 - ;
9843: @end example
9844: Here, @code{TO} is used to update @code{addr1} and @code{addr2} at
9845: every loop iteration. @code{strcmp} is a typical example of the
9846: readability problems of using @code{TO}. When you start reading
9847: @code{strcmp}, you think that @code{addr1} refers to the start of the
9848: string. Only near the end of the loop you realize that it is something
9849: else.
9850:
9851: This can be avoided by defining two locals at the start of the loop that
9852: are initialized with the right value for the current iteration.
9853: @example
9854: : strcmp @{ addr1 u1 addr2 u2 -- n @}
9855: addr1 addr2
9856: u1 u2 min 0
9857: ?do @{ s1 s2 @}
9858: s1 c@@ s2 c@@ -
9859: ?dup-if
9860: unloop exit
9861: then
9862: s1 char+ s2 char+
9863: loop
9864: 2drop
9865: u1 u2 - ;
9866: @end example
9867: Here it is clear from the start that @code{s1} has a different value
9868: in every loop iteration.
9869:
9870: @node Locals implementation, , Locals programming style, Gforth locals
9871: @subsubsection Locals implementation
9872: @cindex locals implementation
9873: @cindex implementation of locals
9874:
9875: @cindex locals stack
9876: Gforth uses an extra locals stack. The most compelling reason for
9877: this is that the return stack is not float-aligned; using an extra stack
9878: also eliminates the problems and restrictions of using the return stack
9879: as locals stack. Like the other stacks, the locals stack grows toward
9880: lower addresses. A few primitives allow an efficient implementation:
9881:
9882:
9883: doc-@local#
9884: doc-f@local#
9885: doc-laddr#
9886: doc-lp+!#
9887: doc-lp!
9888: doc->l
9889: doc-f>l
9890:
9891:
9892: In addition to these primitives, some specializations of these
9893: primitives for commonly occurring inline arguments are provided for
9894: efficiency reasons, e.g., @code{@@local0} as specialization of
9895: @code{@@local#} for the inline argument 0. The following compiling words
9896: compile the right specialized version, or the general version, as
9897: appropriate:
9898:
9899:
9900: @c doc-compile-@local
9901: @c doc-compile-f@local
9902: doc-compile-lp+!
9903:
9904:
9905: Combinations of conditional branches and @code{lp+!#} like
9906: @code{?branch-lp+!#} (the locals pointer is only changed if the branch
9907: is taken) are provided for efficiency and correctness in loops.
9908:
9909: A special area in the dictionary space is reserved for keeping the
9910: local variable names. @code{@{} switches the dictionary pointer to this
9911: area and @code{@}} switches it back and generates the locals
9912: initializing code. @code{W:} etc.@ are normal defining words. This
9913: special area is cleared at the start of every colon definition.
9914:
9915: @cindex word list for defining locals
9916: A special feature of Gforth's dictionary is used to implement the
9917: definition of locals without type specifiers: every word list (aka
9918: vocabulary) has its own methods for searching
9919: etc. (@pxref{Word Lists}). For the present purpose we defined a word list
9920: with a special search method: When it is searched for a word, it
9921: actually creates that word using @code{W:}. @code{@{} changes the search
9922: order to first search the word list containing @code{@}}, @code{W:} etc.,
9923: and then the word list for defining locals without type specifiers.
9924:
9925: The lifetime rules support a stack discipline within a colon
9926: definition: The lifetime of a local is either nested with other locals
9927: lifetimes or it does not overlap them.
9928:
9929: At @code{BEGIN}, @code{IF}, and @code{AHEAD} no code for locals stack
9930: pointer manipulation is generated. Between control structure words
9931: locals definitions can push locals onto the locals stack. @code{AGAIN}
9932: is the simplest of the other three control flow words. It has to
9933: restore the locals stack depth of the corresponding @code{BEGIN}
9934: before branching. The code looks like this:
9935: @format
9936: @code{lp+!#} current-locals-size @minus{} dest-locals-size
9937: @code{branch} <begin>
9938: @end format
9939:
9940: @code{UNTIL} is a little more complicated: If it branches back, it
9941: must adjust the stack just like @code{AGAIN}. But if it falls through,
9942: the locals stack must not be changed. The compiler generates the
9943: following code:
9944: @format
9945: @code{?branch-lp+!#} <begin> current-locals-size @minus{} dest-locals-size
9946: @end format
9947: The locals stack pointer is only adjusted if the branch is taken.
9948:
9949: @code{THEN} can produce somewhat inefficient code:
9950: @format
9951: @code{lp+!#} current-locals-size @minus{} orig-locals-size
9952: <orig target>:
9953: @code{lp+!#} orig-locals-size @minus{} new-locals-size
9954: @end format
9955: The second @code{lp+!#} adjusts the locals stack pointer from the
9956: level at the @i{orig} point to the level after the @code{THEN}. The
9957: first @code{lp+!#} adjusts the locals stack pointer from the current
9958: level to the level at the orig point, so the complete effect is an
9959: adjustment from the current level to the right level after the
9960: @code{THEN}.
9961:
9962: @cindex locals information on the control-flow stack
9963: @cindex control-flow stack items, locals information
9964: In a conventional Forth implementation a dest control-flow stack entry
9965: is just the target address and an orig entry is just the address to be
9966: patched. Our locals implementation adds a word list to every orig or dest
9967: item. It is the list of locals visible (or assumed visible) at the point
9968: described by the entry. Our implementation also adds a tag to identify
9969: the kind of entry, in particular to differentiate between live and dead
9970: (reachable and unreachable) orig entries.
9971:
9972: A few unusual operations have to be performed on locals word lists:
9973:
9974:
9975: doc-common-list
9976: doc-sub-list?
9977: doc-list-size
9978:
9979:
9980: Several features of our locals word list implementation make these
9981: operations easy to implement: The locals word lists are organised as
9982: linked lists; the tails of these lists are shared, if the lists
9983: contain some of the same locals; and the address of a name is greater
9984: than the address of the names behind it in the list.
9985:
9986: Another important implementation detail is the variable
9987: @code{dead-code}. It is used by @code{BEGIN} and @code{THEN} to
9988: determine if they can be reached directly or only through the branch
9989: that they resolve. @code{dead-code} is set by @code{UNREACHABLE},
9990: @code{AHEAD}, @code{EXIT} etc., and cleared at the start of a colon
9991: definition, by @code{BEGIN} and usually by @code{THEN}.
9992:
9993: Counted loops are similar to other loops in most respects, but
9994: @code{LEAVE} requires special attention: It performs basically the same
9995: service as @code{AHEAD}, but it does not create a control-flow stack
9996: entry. Therefore the information has to be stored elsewhere;
9997: traditionally, the information was stored in the target fields of the
9998: branches created by the @code{LEAVE}s, by organizing these fields into a
9999: linked list. Unfortunately, this clever trick does not provide enough
10000: space for storing our extended control flow information. Therefore, we
10001: introduce another stack, the leave stack. It contains the control-flow
10002: stack entries for all unresolved @code{LEAVE}s.
10003:
10004: Local names are kept until the end of the colon definition, even if
10005: they are no longer visible in any control-flow path. In a few cases
10006: this may lead to increased space needs for the locals name area, but
10007: usually less than reclaiming this space would cost in code size.
10008:
10009:
10010: @node ANS Forth locals, , Gforth locals, Locals
10011: @subsection ANS Forth locals
10012: @cindex locals, ANS Forth style
10013:
10014: The ANS Forth locals wordset does not define a syntax for locals, but
10015: words that make it possible to define various syntaxes. One of the
10016: possible syntaxes is a subset of the syntax we used in the Gforth locals
10017: wordset, i.e.:
10018:
10019: @example
10020: @{ local1 local2 ... -- comment @}
10021: @end example
10022: @noindent
10023: or
10024: @example
10025: @{ local1 local2 ... @}
10026: @end example
10027:
10028: The order of the locals corresponds to the order in a stack comment. The
10029: restrictions are:
10030:
10031: @itemize @bullet
10032: @item
10033: Locals can only be cell-sized values (no type specifiers are allowed).
10034: @item
10035: Locals can be defined only outside control structures.
10036: @item
10037: Locals can interfere with explicit usage of the return stack. For the
10038: exact (and long) rules, see the standard. If you don't use return stack
10039: accessing words in a definition using locals, you will be all right. The
10040: purpose of this rule is to make locals implementation on the return
10041: stack easier.
10042: @item
10043: The whole definition must be in one line.
10044: @end itemize
10045:
10046: Locals defined in ANS Forth behave like @code{VALUE}s
10047: (@pxref{Values}). I.e., they are initialized from the stack. Using their
10048: name produces their value. Their value can be changed using @code{TO}.
10049:
10050: Since the syntax above is supported by Gforth directly, you need not do
10051: anything to use it. If you want to port a program using this syntax to
10052: another ANS Forth system, use @file{compat/anslocal.fs} to implement the
10053: syntax on the other system.
10054:
10055: Note that a syntax shown in the standard, section A.13 looks
10056: similar, but is quite different in having the order of locals
10057: reversed. Beware!
10058:
10059: The ANS Forth locals wordset itself consists of one word:
10060:
10061: doc-(local)
10062:
10063: The ANS Forth locals extension wordset defines a syntax using
10064: @code{locals|}, but it is so awful that we strongly recommend not to use
10065: it. We have implemented this syntax to make porting to Gforth easy, but
10066: do not document it here. The problem with this syntax is that the locals
10067: are defined in an order reversed with respect to the standard stack
10068: comment notation, making programs harder to read, and easier to misread
10069: and miswrite. The only merit of this syntax is that it is easy to
10070: implement using the ANS Forth locals wordset.
10071:
10072:
10073: @c ----------------------------------------------------------
10074: @node Structures, Object-oriented Forth, Locals, Words
10075: @section Structures
10076: @cindex structures
10077: @cindex records
10078:
10079: This section presents the structure package that comes with Gforth. A
10080: version of the package implemented in ANS Forth is available in
10081: @file{compat/struct.fs}. This package was inspired by a posting on
10082: comp.lang.forth in 1989 (unfortunately I don't remember, by whom;
10083: possibly John Hayes). A version of this section has been published in
10084: M. Anton Ertl,
10085: @uref{http://www.complang.tuwien.ac.at/forth/objects/structs.html, Yet
10086: Another Forth Structures Package}, Forth Dimensions 19(3), pages
10087: 13--16. Marcel Hendrix provided helpful comments.
10088:
10089: @menu
10090: * Why explicit structure support?::
10091: * Structure Usage::
10092: * Structure Naming Convention::
10093: * Structure Implementation::
10094: * Structure Glossary::
10095: * Forth200x Structures::
10096: @end menu
10097:
10098: @node Why explicit structure support?, Structure Usage, Structures, Structures
10099: @subsection Why explicit structure support?
10100:
10101: @cindex address arithmetic for structures
10102: @cindex structures using address arithmetic
10103: If we want to use a structure containing several fields, we could simply
10104: reserve memory for it, and access the fields using address arithmetic
10105: (@pxref{Address arithmetic}). As an example, consider a structure with
10106: the following fields
10107:
10108: @table @code
10109: @item a
10110: is a float
10111: @item b
10112: is a cell
10113: @item c
10114: is a float
10115: @end table
10116:
10117: Given the (float-aligned) base address of the structure we get the
10118: address of the field
10119:
10120: @table @code
10121: @item a
10122: without doing anything further.
10123: @item b
10124: with @code{float+}
10125: @item c
10126: with @code{float+ cell+ faligned}
10127: @end table
10128:
10129: It is easy to see that this can become quite tiring.
10130:
10131: Moreover, it is not very readable, because seeing a
10132: @code{cell+} tells us neither which kind of structure is
10133: accessed nor what field is accessed; we have to somehow infer the kind
10134: of structure, and then look up in the documentation, which field of
10135: that structure corresponds to that offset.
10136:
10137: Finally, this kind of address arithmetic also causes maintenance
10138: troubles: If you add or delete a field somewhere in the middle of the
10139: structure, you have to find and change all computations for the fields
10140: afterwards.
10141:
10142: So, instead of using @code{cell+} and friends directly, how
10143: about storing the offsets in constants:
10144:
10145: @example
10146: 0 constant a-offset
10147: 0 float+ constant b-offset
10148: 0 float+ cell+ faligned c-offset
10149: @end example
10150:
10151: Now we can get the address of field @code{x} with @code{x-offset
10152: +}. This is much better in all respects. Of course, you still
10153: have to change all later offset definitions if you add a field. You can
10154: fix this by declaring the offsets in the following way:
10155:
10156: @example
10157: 0 constant a-offset
10158: a-offset float+ constant b-offset
10159: b-offset cell+ faligned constant c-offset
10160: @end example
10161:
10162: Since we always use the offsets with @code{+}, we could use a defining
10163: word @code{cfield} that includes the @code{+} in the action of the
10164: defined word:
10165:
10166: @example
10167: : cfield ( n "name" -- )
10168: create ,
10169: does> ( name execution: addr1 -- addr2 )
10170: @@ + ;
10171:
10172: 0 cfield a
10173: 0 a float+ cfield b
10174: 0 b cell+ faligned cfield c
10175: @end example
10176:
10177: Instead of @code{x-offset +}, we now simply write @code{x}.
10178:
10179: The structure field words now can be used quite nicely. However,
10180: their definition is still a bit cumbersome: We have to repeat the
10181: name, the information about size and alignment is distributed before
10182: and after the field definitions etc. The structure package presented
10183: here addresses these problems.
10184:
10185: @node Structure Usage, Structure Naming Convention, Why explicit structure support?, Structures
10186: @subsection Structure Usage
10187: @cindex structure usage
10188:
10189: @cindex @code{field} usage
10190: @cindex @code{struct} usage
10191: @cindex @code{end-struct} usage
10192: You can define a structure for a (data-less) linked list with:
10193: @example
10194: struct
10195: cell% field list-next
10196: end-struct list%
10197: @end example
10198:
10199: With the address of the list node on the stack, you can compute the
10200: address of the field that contains the address of the next node with
10201: @code{list-next}. E.g., you can determine the length of a list
10202: with:
10203:
10204: @example
10205: : list-length ( list -- n )
10206: \ "list" is a pointer to the first element of a linked list
10207: \ "n" is the length of the list
10208: 0 BEGIN ( list1 n1 )
10209: over
10210: WHILE ( list1 n1 )
10211: 1+ swap list-next @@ swap
10212: REPEAT
10213: nip ;
10214: @end example
10215:
10216: You can reserve memory for a list node in the dictionary with
10217: @code{list% %allot}, which leaves the address of the list node on the
10218: stack. For the equivalent allocation on the heap you can use @code{list%
10219: %alloc} (or, for an @code{allocate}-like stack effect (i.e., with ior),
10220: use @code{list% %allocate}). You can get the the size of a list
10221: node with @code{list% %size} and its alignment with @code{list%
10222: %alignment}.
10223:
10224: Note that in ANS Forth the body of a @code{create}d word is
10225: @code{aligned} but not necessarily @code{faligned};
10226: therefore, if you do a:
10227:
10228: @example
10229: create @emph{name} foo% %allot drop
10230: @end example
10231:
10232: @noindent
10233: then the memory alloted for @code{foo%} is guaranteed to start at the
10234: body of @code{@emph{name}} only if @code{foo%} contains only character,
10235: cell and double fields. Therefore, if your structure contains floats,
10236: better use
10237:
10238: @example
10239: foo% %allot constant @emph{name}
10240: @end example
10241:
10242: @cindex structures containing structures
10243: You can include a structure @code{foo%} as a field of
10244: another structure, like this:
10245: @example
10246: struct
10247: ...
10248: foo% field ...
10249: ...
10250: end-struct ...
10251: @end example
10252:
10253: @cindex structure extension
10254: @cindex extended records
10255: Instead of starting with an empty structure, you can extend an
10256: existing structure. E.g., a plain linked list without data, as defined
10257: above, is hardly useful; You can extend it to a linked list of integers,
10258: like this:@footnote{This feature is also known as @emph{extended
10259: records}. It is the main innovation in the Oberon language; in other
10260: words, adding this feature to Modula-2 led Wirth to create a new
10261: language, write a new compiler etc. Adding this feature to Forth just
10262: required a few lines of code.}
10263:
10264: @example
10265: list%
10266: cell% field intlist-int
10267: end-struct intlist%
10268: @end example
10269:
10270: @code{intlist%} is a structure with two fields:
10271: @code{list-next} and @code{intlist-int}.
10272:
10273: @cindex structures containing arrays
10274: You can specify an array type containing @emph{n} elements of
10275: type @code{foo%} like this:
10276:
10277: @example
10278: foo% @emph{n} *
10279: @end example
10280:
10281: You can use this array type in any place where you can use a normal
10282: type, e.g., when defining a @code{field}, or with
10283: @code{%allot}.
10284:
10285: @cindex first field optimization
10286: The first field is at the base address of a structure and the word for
10287: this field (e.g., @code{list-next}) actually does not change the address
10288: on the stack. You may be tempted to leave it away in the interest of
10289: run-time and space efficiency. This is not necessary, because the
10290: structure package optimizes this case: If you compile a first-field
10291: words, no code is generated. So, in the interest of readability and
10292: maintainability you should include the word for the field when accessing
10293: the field.
10294:
10295:
10296: @node Structure Naming Convention, Structure Implementation, Structure Usage, Structures
10297: @subsection Structure Naming Convention
10298: @cindex structure naming convention
10299:
10300: The field names that come to (my) mind are often quite generic, and,
10301: if used, would cause frequent name clashes. E.g., many structures
10302: probably contain a @code{counter} field. The structure names
10303: that come to (my) mind are often also the logical choice for the names
10304: of words that create such a structure.
10305:
10306: Therefore, I have adopted the following naming conventions:
10307:
10308: @itemize @bullet
10309: @cindex field naming convention
10310: @item
10311: The names of fields are of the form
10312: @code{@emph{struct}-@emph{field}}, where
10313: @code{@emph{struct}} is the basic name of the structure, and
10314: @code{@emph{field}} is the basic name of the field. You can
10315: think of field words as converting the (address of the)
10316: structure into the (address of the) field.
10317:
10318: @cindex structure naming convention
10319: @item
10320: The names of structures are of the form
10321: @code{@emph{struct}%}, where
10322: @code{@emph{struct}} is the basic name of the structure.
10323: @end itemize
10324:
10325: This naming convention does not work that well for fields of extended
10326: structures; e.g., the integer list structure has a field
10327: @code{intlist-int}, but has @code{list-next}, not
10328: @code{intlist-next}.
10329:
10330: @node Structure Implementation, Structure Glossary, Structure Naming Convention, Structures
10331: @subsection Structure Implementation
10332: @cindex structure implementation
10333: @cindex implementation of structures
10334:
10335: The central idea in the implementation is to pass the data about the
10336: structure being built on the stack, not in some global
10337: variable. Everything else falls into place naturally once this design
10338: decision is made.
10339:
10340: The type description on the stack is of the form @emph{align
10341: size}. Keeping the size on the top-of-stack makes dealing with arrays
10342: very simple.
10343:
10344: @code{field} is a defining word that uses @code{Create}
10345: and @code{DOES>}. The body of the field contains the offset
10346: of the field, and the normal @code{DOES>} action is simply:
10347:
10348: @example
10349: @@ +
10350: @end example
10351:
10352: @noindent
10353: i.e., add the offset to the address, giving the stack effect
10354: @i{addr1 -- addr2} for a field.
10355:
10356: @cindex first field optimization, implementation
10357: This simple structure is slightly complicated by the optimization
10358: for fields with offset 0, which requires a different
10359: @code{DOES>}-part (because we cannot rely on there being
10360: something on the stack if such a field is invoked during
10361: compilation). Therefore, we put the different @code{DOES>}-parts
10362: in separate words, and decide which one to invoke based on the
10363: offset. For a zero offset, the field is basically a noop; it is
10364: immediate, and therefore no code is generated when it is compiled.
10365:
10366: @node Structure Glossary, Forth200x Structures, Structure Implementation, Structures
10367: @subsection Structure Glossary
10368: @cindex structure glossary
10369:
10370:
10371: doc-%align
10372: doc-%alignment
10373: doc-%alloc
10374: doc-%allocate
10375: doc-%allot
10376: doc-cell%
10377: doc-char%
10378: doc-dfloat%
10379: doc-double%
10380: doc-end-struct
10381: doc-field
10382: doc-float%
10383: doc-naligned
10384: doc-sfloat%
10385: doc-%size
10386: doc-struct
10387:
10388:
10389: @node Forth200x Structures, , Structure Glossary, Structures
10390: @subsection Forth200x Structures
10391: @cindex Structures in Forth200x
10392:
10393: The Forth 200x standard defines a slightly less convenient form of
10394: structures. In general (when using @code{field+}, you have to perform
10395: the alignment yourself, but there are a number of convenience words
10396: (e.g., @code{field:} that perform the alignment for you.
10397:
10398: A typical usage example is:
10399:
10400: @example
10401: 0
10402: field: s-a
10403: faligned 2 floats +field s-b
10404: constant s-struct
10405: @end example
10406:
10407: An alternative way of writing this structure is:
10408:
10409: @example
10410: begin-structure s-struct
10411: field: s-a
10412: faligned 2 floats +field s-b
10413: end-structure
10414: @end example
10415:
10416: doc-begin-structure
10417: doc-end-structure
10418: doc-+field
10419: doc-cfield:
10420: doc-field:
10421: doc-2field:
10422: doc-ffield:
10423: doc-sffield:
10424: doc-dffield:
10425:
10426: @c -------------------------------------------------------------
10427: @node Object-oriented Forth, Programming Tools, Structures, Words
10428: @section Object-oriented Forth
10429:
10430: Gforth comes with three packages for object-oriented programming:
10431: @file{objects.fs}, @file{oof.fs}, and @file{mini-oof.fs}; none of them
10432: is preloaded, so you have to @code{include} them before use. The most
10433: important differences between these packages (and others) are discussed
10434: in @ref{Comparison with other object models}. All packages are written
10435: in ANS Forth and can be used with any other ANS Forth.
10436:
10437: @menu
10438: * Why object-oriented programming?::
10439: * Object-Oriented Terminology::
10440: * Objects::
10441: * OOF::
10442: * Mini-OOF::
10443: * Comparison with other object models::
10444: @end menu
10445:
10446: @c ----------------------------------------------------------------
10447: @node Why object-oriented programming?, Object-Oriented Terminology, Object-oriented Forth, Object-oriented Forth
10448: @subsection Why object-oriented programming?
10449: @cindex object-oriented programming motivation
10450: @cindex motivation for object-oriented programming
10451:
10452: Often we have to deal with several data structures (@emph{objects}),
10453: that have to be treated similarly in some respects, but differently in
10454: others. Graphical objects are the textbook example: circles, triangles,
10455: dinosaurs, icons, and others, and we may want to add more during program
10456: development. We want to apply some operations to any graphical object,
10457: e.g., @code{draw} for displaying it on the screen. However, @code{draw}
10458: has to do something different for every kind of object.
10459: @comment TODO add some other operations eg perimeter, area
10460: @comment and tie in to concrete examples later..
10461:
10462: We could implement @code{draw} as a big @code{CASE}
10463: control structure that executes the appropriate code depending on the
10464: kind of object to be drawn. This would be not be very elegant, and,
10465: moreover, we would have to change @code{draw} every time we add
10466: a new kind of graphical object (say, a spaceship).
10467:
10468: What we would rather do is: When defining spaceships, we would tell
10469: the system: ``Here's how you @code{draw} a spaceship; you figure
10470: out the rest''.
10471:
10472: This is the problem that all systems solve that (rightfully) call
10473: themselves object-oriented; the object-oriented packages presented here
10474: solve this problem (and not much else).
10475: @comment TODO ?list properties of oo systems.. oo vs o-based?
10476:
10477: @c ------------------------------------------------------------------------
10478: @node Object-Oriented Terminology, Objects, Why object-oriented programming?, Object-oriented Forth
10479: @subsection Object-Oriented Terminology
10480: @cindex object-oriented terminology
10481: @cindex terminology for object-oriented programming
10482:
10483: This section is mainly for reference, so you don't have to understand
10484: all of it right away. The terminology is mainly Smalltalk-inspired. In
10485: short:
10486:
10487: @table @emph
10488: @cindex class
10489: @item class
10490: a data structure definition with some extras.
10491:
10492: @cindex object
10493: @item object
10494: an instance of the data structure described by the class definition.
10495:
10496: @cindex instance variables
10497: @item instance variables
10498: fields of the data structure.
10499:
10500: @cindex selector
10501: @cindex method selector
10502: @cindex virtual function
10503: @item selector
10504: (or @emph{method selector}) a word (e.g.,
10505: @code{draw}) that performs an operation on a variety of data
10506: structures (classes). A selector describes @emph{what} operation to
10507: perform. In C++ terminology: a (pure) virtual function.
10508:
10509: @cindex method
10510: @item method
10511: the concrete definition that performs the operation
10512: described by the selector for a specific class. A method specifies
10513: @emph{how} the operation is performed for a specific class.
10514:
10515: @cindex selector invocation
10516: @cindex message send
10517: @cindex invoking a selector
10518: @item selector invocation
10519: a call of a selector. One argument of the call (the TOS (top-of-stack))
10520: is used for determining which method is used. In Smalltalk terminology:
10521: a message (consisting of the selector and the other arguments) is sent
10522: to the object.
10523:
10524: @cindex receiving object
10525: @item receiving object
10526: the object used for determining the method executed by a selector
10527: invocation. In the @file{objects.fs} model, it is the object that is on
10528: the TOS when the selector is invoked. (@emph{Receiving} comes from
10529: the Smalltalk @emph{message} terminology.)
10530:
10531: @cindex child class
10532: @cindex parent class
10533: @cindex inheritance
10534: @item child class
10535: a class that has (@emph{inherits}) all properties (instance variables,
10536: selectors, methods) from a @emph{parent class}. In Smalltalk
10537: terminology: The subclass inherits from the superclass. In C++
10538: terminology: The derived class inherits from the base class.
10539:
10540: @end table
10541:
10542: @c If you wonder about the message sending terminology, it comes from
10543: @c a time when each object had it's own task and objects communicated via
10544: @c message passing; eventually the Smalltalk developers realized that
10545: @c they can do most things through simple (indirect) calls. They kept the
10546: @c terminology.
10547:
10548: @c --------------------------------------------------------------
10549: @node Objects, OOF, Object-Oriented Terminology, Object-oriented Forth
10550: @subsection The @file{objects.fs} model
10551: @cindex objects
10552: @cindex object-oriented programming
10553:
10554: @cindex @file{objects.fs}
10555: @cindex @file{oof.fs}
10556:
10557: This section describes the @file{objects.fs} package. This material also
10558: has been published in M. Anton Ertl,
10559: @cite{@uref{http://www.complang.tuwien.ac.at/forth/objects/objects.html,
10560: Yet Another Forth Objects Package}}, Forth Dimensions 19(2), pages
10561: 37--43.
10562: @c McKewan's and Zsoter's packages
10563:
10564: This section assumes that you have read @ref{Structures}.
10565:
10566: The techniques on which this model is based have been used to implement
10567: the parser generator, Gray, and have also been used in Gforth for
10568: implementing the various flavours of word lists (hashed or not,
10569: case-sensitive or not, special-purpose word lists for locals etc.).
10570:
10571:
10572: @menu
10573: * Properties of the Objects model::
10574: * Basic Objects Usage::
10575: * The Objects base class::
10576: * Creating objects::
10577: * Object-Oriented Programming Style::
10578: * Class Binding::
10579: * Method conveniences::
10580: * Classes and Scoping::
10581: * Dividing classes::
10582: * Object Interfaces::
10583: * Objects Implementation::
10584: * Objects Glossary::
10585: @end menu
10586:
10587: Marcel Hendrix provided helpful comments on this section.
10588:
10589: @node Properties of the Objects model, Basic Objects Usage, Objects, Objects
10590: @subsubsection Properties of the @file{objects.fs} model
10591: @cindex @file{objects.fs} properties
10592:
10593: @itemize @bullet
10594: @item
10595: It is straightforward to pass objects on the stack. Passing
10596: selectors on the stack is a little less convenient, but possible.
10597:
10598: @item
10599: Objects are just data structures in memory, and are referenced by their
10600: address. You can create words for objects with normal defining words
10601: like @code{constant}. Likewise, there is no difference between instance
10602: variables that contain objects and those that contain other data.
10603:
10604: @item
10605: Late binding is efficient and easy to use.
10606:
10607: @item
10608: It avoids parsing, and thus avoids problems with state-smartness
10609: and reduced extensibility; for convenience there are a few parsing
10610: words, but they have non-parsing counterparts. There are also a few
10611: defining words that parse. This is hard to avoid, because all standard
10612: defining words parse (except @code{:noname}); however, such
10613: words are not as bad as many other parsing words, because they are not
10614: state-smart.
10615:
10616: @item
10617: It does not try to incorporate everything. It does a few things and does
10618: them well (IMO). In particular, this model was not designed to support
10619: information hiding (although it has features that may help); you can use
10620: a separate package for achieving this.
10621:
10622: @item
10623: It is layered; you don't have to learn and use all features to use this
10624: model. Only a few features are necessary (@pxref{Basic Objects Usage},
10625: @pxref{The Objects base class}, @pxref{Creating objects}.), the others
10626: are optional and independent of each other.
10627:
10628: @item
10629: An implementation in ANS Forth is available.
10630:
10631: @end itemize
10632:
10633:
10634: @node Basic Objects Usage, The Objects base class, Properties of the Objects model, Objects
10635: @subsubsection Basic @file{objects.fs} Usage
10636: @cindex basic objects usage
10637: @cindex objects, basic usage
10638:
10639: You can define a class for graphical objects like this:
10640:
10641: @cindex @code{class} usage
10642: @cindex @code{end-class} usage
10643: @cindex @code{selector} usage
10644: @example
10645: object class \ "object" is the parent class
10646: selector draw ( x y graphical -- )
10647: end-class graphical
10648: @end example
10649:
10650: This code defines a class @code{graphical} with an
10651: operation @code{draw}. We can perform the operation
10652: @code{draw} on any @code{graphical} object, e.g.:
10653:
10654: @example
10655: 100 100 t-rex draw
10656: @end example
10657:
10658: @noindent
10659: where @code{t-rex} is a word (say, a constant) that produces a
10660: graphical object.
10661:
10662: @comment TODO add a 2nd operation eg perimeter.. and use for
10663: @comment a concrete example
10664:
10665: @cindex abstract class
10666: How do we create a graphical object? With the present definitions,
10667: we cannot create a useful graphical object. The class
10668: @code{graphical} describes graphical objects in general, but not
10669: any concrete graphical object type (C++ users would call it an
10670: @emph{abstract class}); e.g., there is no method for the selector
10671: @code{draw} in the class @code{graphical}.
10672:
10673: For concrete graphical objects, we define child classes of the
10674: class @code{graphical}, e.g.:
10675:
10676: @cindex @code{overrides} usage
10677: @cindex @code{field} usage in class definition
10678: @example
10679: graphical class \ "graphical" is the parent class
10680: cell% field circle-radius
10681:
10682: :noname ( x y circle -- )
10683: circle-radius @@ draw-circle ;
10684: overrides draw
10685:
10686: :noname ( n-radius circle -- )
10687: circle-radius ! ;
10688: overrides construct
10689:
10690: end-class circle
10691: @end example
10692:
10693: Here we define a class @code{circle} as a child of @code{graphical},
10694: with field @code{circle-radius} (which behaves just like a field
10695: (@pxref{Structures}); it defines (using @code{overrides}) new methods
10696: for the selectors @code{draw} and @code{construct} (@code{construct} is
10697: defined in @code{object}, the parent class of @code{graphical}).
10698:
10699: Now we can create a circle on the heap (i.e.,
10700: @code{allocate}d memory) with:
10701:
10702: @cindex @code{heap-new} usage
10703: @example
10704: 50 circle heap-new constant my-circle
10705: @end example
10706:
10707: @noindent
10708: @code{heap-new} invokes @code{construct}, thus
10709: initializing the field @code{circle-radius} with 50. We can draw
10710: this new circle at (100,100) with:
10711:
10712: @example
10713: 100 100 my-circle draw
10714: @end example
10715:
10716: @cindex selector invocation, restrictions
10717: @cindex class definition, restrictions
10718: Note: You can only invoke a selector if the object on the TOS
10719: (the receiving object) belongs to the class where the selector was
10720: defined or one of its descendents; e.g., you can invoke
10721: @code{draw} only for objects belonging to @code{graphical}
10722: or its descendents (e.g., @code{circle}). Immediately before
10723: @code{end-class}, the search order has to be the same as
10724: immediately after @code{class}.
10725:
10726: @node The Objects base class, Creating objects, Basic Objects Usage, Objects
10727: @subsubsection The @file{object.fs} base class
10728: @cindex @code{object} class
10729:
10730: When you define a class, you have to specify a parent class. So how do
10731: you start defining classes? There is one class available from the start:
10732: @code{object}. It is ancestor for all classes and so is the
10733: only class that has no parent. It has two selectors: @code{construct}
10734: and @code{print}.
10735:
10736: @node Creating objects, Object-Oriented Programming Style, The Objects base class, Objects
10737: @subsubsection Creating objects
10738: @cindex creating objects
10739: @cindex object creation
10740: @cindex object allocation options
10741:
10742: @cindex @code{heap-new} discussion
10743: @cindex @code{dict-new} discussion
10744: @cindex @code{construct} discussion
10745: You can create and initialize an object of a class on the heap with
10746: @code{heap-new} ( ... class -- object ) and in the dictionary
10747: (allocation with @code{allot}) with @code{dict-new} (
10748: ... class -- object ). Both words invoke @code{construct}, which
10749: consumes the stack items indicated by "..." above.
10750:
10751: @cindex @code{init-object} discussion
10752: @cindex @code{class-inst-size} discussion
10753: If you want to allocate memory for an object yourself, you can get its
10754: alignment and size with @code{class-inst-size 2@@} ( class --
10755: align size ). Once you have memory for an object, you can initialize
10756: it with @code{init-object} ( ... class object -- );
10757: @code{construct} does only a part of the necessary work.
10758:
10759: @node Object-Oriented Programming Style, Class Binding, Creating objects, Objects
10760: @subsubsection Object-Oriented Programming Style
10761: @cindex object-oriented programming style
10762: @cindex programming style, object-oriented
10763:
10764: This section is not exhaustive.
10765:
10766: @cindex stack effects of selectors
10767: @cindex selectors and stack effects
10768: In general, it is a good idea to ensure that all methods for the
10769: same selector have the same stack effect: when you invoke a selector,
10770: you often have no idea which method will be invoked, so, unless all
10771: methods have the same stack effect, you will not know the stack effect
10772: of the selector invocation.
10773:
10774: One exception to this rule is methods for the selector
10775: @code{construct}. We know which method is invoked, because we
10776: specify the class to be constructed at the same place. Actually, I
10777: defined @code{construct} as a selector only to give the users a
10778: convenient way to specify initialization. The way it is used, a
10779: mechanism different from selector invocation would be more natural
10780: (but probably would take more code and more space to explain).
10781:
10782: @node Class Binding, Method conveniences, Object-Oriented Programming Style, Objects
10783: @subsubsection Class Binding
10784: @cindex class binding
10785: @cindex early binding
10786:
10787: @cindex late binding
10788: Normal selector invocations determine the method at run-time depending
10789: on the class of the receiving object. This run-time selection is called
10790: @i{late binding}.
10791:
10792: Sometimes it's preferable to invoke a different method. For example,
10793: you might want to use the simple method for @code{print}ing
10794: @code{object}s instead of the possibly long-winded @code{print} method
10795: of the receiver class. You can achieve this by replacing the invocation
10796: of @code{print} with:
10797:
10798: @cindex @code{[bind]} usage
10799: @example
10800: [bind] object print
10801: @end example
10802:
10803: @noindent
10804: in compiled code or:
10805:
10806: @cindex @code{bind} usage
10807: @example
10808: bind object print
10809: @end example
10810:
10811: @cindex class binding, alternative to
10812: @noindent
10813: in interpreted code. Alternatively, you can define the method with a
10814: name (e.g., @code{print-object}), and then invoke it through the
10815: name. Class binding is just a (often more convenient) way to achieve
10816: the same effect; it avoids name clutter and allows you to invoke
10817: methods directly without naming them first.
10818:
10819: @cindex superclass binding
10820: @cindex parent class binding
10821: A frequent use of class binding is this: When we define a method
10822: for a selector, we often want the method to do what the selector does
10823: in the parent class, and a little more. There is a special word for
10824: this purpose: @code{[parent]}; @code{[parent]
10825: @emph{selector}} is equivalent to @code{[bind] @emph{parent
10826: selector}}, where @code{@emph{parent}} is the parent
10827: class of the current class. E.g., a method definition might look like:
10828:
10829: @cindex @code{[parent]} usage
10830: @example
10831: :noname
10832: dup [parent] foo \ do parent's foo on the receiving object
10833: ... \ do some more
10834: ; overrides foo
10835: @end example
10836:
10837: @cindex class binding as optimization
10838: In @cite{Object-oriented programming in ANS Forth} (Forth Dimensions,
10839: March 1997), Andrew McKewan presents class binding as an optimization
10840: technique. I recommend not using it for this purpose unless you are in
10841: an emergency. Late binding is pretty fast with this model anyway, so the
10842: benefit of using class binding is small; the cost of using class binding
10843: where it is not appropriate is reduced maintainability.
10844:
10845: While we are at programming style questions: You should bind
10846: selectors only to ancestor classes of the receiving object. E.g., say,
10847: you know that the receiving object is of class @code{foo} or its
10848: descendents; then you should bind only to @code{foo} and its
10849: ancestors.
10850:
10851: @node Method conveniences, Classes and Scoping, Class Binding, Objects
10852: @subsubsection Method conveniences
10853: @cindex method conveniences
10854:
10855: In a method you usually access the receiving object pretty often. If
10856: you define the method as a plain colon definition (e.g., with
10857: @code{:noname}), you may have to do a lot of stack
10858: gymnastics. To avoid this, you can define the method with @code{m:
10859: ... ;m}. E.g., you could define the method for
10860: @code{draw}ing a @code{circle} with
10861:
10862: @cindex @code{this} usage
10863: @cindex @code{m:} usage
10864: @cindex @code{;m} usage
10865: @example
10866: m: ( x y circle -- )
10867: ( x y ) this circle-radius @@ draw-circle ;m
10868: @end example
10869:
10870: @cindex @code{exit} in @code{m: ... ;m}
10871: @cindex @code{exitm} discussion
10872: @cindex @code{catch} in @code{m: ... ;m}
10873: When this method is executed, the receiver object is removed from the
10874: stack; you can access it with @code{this} (admittedly, in this
10875: example the use of @code{m: ... ;m} offers no advantage). Note
10876: that I specify the stack effect for the whole method (i.e. including
10877: the receiver object), not just for the code between @code{m:}
10878: and @code{;m}. You cannot use @code{exit} in
10879: @code{m:...;m}; instead, use
10880: @code{exitm}.@footnote{Moreover, for any word that calls
10881: @code{catch} and was defined before loading
10882: @code{objects.fs}, you have to redefine it like I redefined
10883: @code{catch}: @code{: catch this >r catch r> to-this ;}}
10884:
10885: @cindex @code{inst-var} usage
10886: You will frequently use sequences of the form @code{this
10887: @emph{field}} (in the example above: @code{this
10888: circle-radius}). If you use the field only in this way, you can
10889: define it with @code{inst-var} and eliminate the
10890: @code{this} before the field name. E.g., the @code{circle}
10891: class above could also be defined with:
10892:
10893: @example
10894: graphical class
10895: cell% inst-var radius
10896:
10897: m: ( x y circle -- )
10898: radius @@ draw-circle ;m
10899: overrides draw
10900:
10901: m: ( n-radius circle -- )
10902: radius ! ;m
10903: overrides construct
10904:
10905: end-class circle
10906: @end example
10907:
10908: @code{radius} can only be used in @code{circle} and its
10909: descendent classes and inside @code{m:...;m}.
10910:
10911: @cindex @code{inst-value} usage
10912: You can also define fields with @code{inst-value}, which is
10913: to @code{inst-var} what @code{value} is to
10914: @code{variable}. You can change the value of such a field with
10915: @code{[to-inst]}. E.g., we could also define the class
10916: @code{circle} like this:
10917:
10918: @example
10919: graphical class
10920: inst-value radius
10921:
10922: m: ( x y circle -- )
10923: radius draw-circle ;m
10924: overrides draw
10925:
10926: m: ( n-radius circle -- )
10927: [to-inst] radius ;m
10928: overrides construct
10929:
10930: end-class circle
10931: @end example
10932:
10933: @c !! :m is easy to confuse with m:. Another name would be better.
10934:
10935: @c Finally, you can define named methods with @code{:m}. One use of this
10936: @c feature is the definition of words that occur only in one class and are
10937: @c not intended to be overridden, but which still need method context
10938: @c (e.g., for accessing @code{inst-var}s). Another use is for methods that
10939: @c would be bound frequently, if defined anonymously.
10940:
10941:
10942: @node Classes and Scoping, Dividing classes, Method conveniences, Objects
10943: @subsubsection Classes and Scoping
10944: @cindex classes and scoping
10945: @cindex scoping and classes
10946:
10947: Inheritance is frequent, unlike structure extension. This exacerbates
10948: the problem with the field name convention (@pxref{Structure Naming
10949: Convention}): One always has to remember in which class the field was
10950: originally defined; changing a part of the class structure would require
10951: changes for renaming in otherwise unaffected code.
10952:
10953: @cindex @code{inst-var} visibility
10954: @cindex @code{inst-value} visibility
10955: To solve this problem, I added a scoping mechanism (which was not in my
10956: original charter): A field defined with @code{inst-var} (or
10957: @code{inst-value}) is visible only in the class where it is defined and in
10958: the descendent classes of this class. Using such fields only makes
10959: sense in @code{m:}-defined methods in these classes anyway.
10960:
10961: This scoping mechanism allows us to use the unadorned field name,
10962: because name clashes with unrelated words become much less likely.
10963:
10964: @cindex @code{protected} discussion
10965: @cindex @code{private} discussion
10966: Once we have this mechanism, we can also use it for controlling the
10967: visibility of other words: All words defined after
10968: @code{protected} are visible only in the current class and its
10969: descendents. @code{public} restores the compilation
10970: (i.e. @code{current}) word list that was in effect before. If you
10971: have several @code{protected}s without an intervening
10972: @code{public} or @code{set-current}, @code{public}
10973: will restore the compilation word list in effect before the first of
10974: these @code{protected}s.
10975:
10976: @node Dividing classes, Object Interfaces, Classes and Scoping, Objects
10977: @subsubsection Dividing classes
10978: @cindex Dividing classes
10979: @cindex @code{methods}...@code{end-methods}
10980:
10981: You may want to do the definition of methods separate from the
10982: definition of the class, its selectors, fields, and instance variables,
10983: i.e., separate the implementation from the definition. You can do this
10984: in the following way:
10985:
10986: @example
10987: graphical class
10988: inst-value radius
10989: end-class circle
10990:
10991: ... \ do some other stuff
10992:
10993: circle methods \ now we are ready
10994:
10995: m: ( x y circle -- )
10996: radius draw-circle ;m
10997: overrides draw
10998:
10999: m: ( n-radius circle -- )
11000: [to-inst] radius ;m
11001: overrides construct
11002:
11003: end-methods
11004: @end example
11005:
11006: You can use several @code{methods}...@code{end-methods} sections. The
11007: only things you can do to the class in these sections are: defining
11008: methods, and overriding the class's selectors. You must not define new
11009: selectors or fields.
11010:
11011: Note that you often have to override a selector before using it. In
11012: particular, you usually have to override @code{construct} with a new
11013: method before you can invoke @code{heap-new} and friends. E.g., you
11014: must not create a circle before the @code{overrides construct} sequence
11015: in the example above.
11016:
11017: @node Object Interfaces, Objects Implementation, Dividing classes, Objects
11018: @subsubsection Object Interfaces
11019: @cindex object interfaces
11020: @cindex interfaces for objects
11021:
11022: In this model you can only call selectors defined in the class of the
11023: receiving objects or in one of its ancestors. If you call a selector
11024: with a receiving object that is not in one of these classes, the
11025: result is undefined; if you are lucky, the program crashes
11026: immediately.
11027:
11028: @cindex selectors common to hardly-related classes
11029: Now consider the case when you want to have a selector (or several)
11030: available in two classes: You would have to add the selector to a
11031: common ancestor class, in the worst case to @code{object}. You
11032: may not want to do this, e.g., because someone else is responsible for
11033: this ancestor class.
11034:
11035: The solution for this problem is interfaces. An interface is a
11036: collection of selectors. If a class implements an interface, the
11037: selectors become available to the class and its descendents. A class
11038: can implement an unlimited number of interfaces. For the problem
11039: discussed above, we would define an interface for the selector(s), and
11040: both classes would implement the interface.
11041:
11042: As an example, consider an interface @code{storage} for
11043: writing objects to disk and getting them back, and a class
11044: @code{foo} that implements it. The code would look like this:
11045:
11046: @cindex @code{interface} usage
11047: @cindex @code{end-interface} usage
11048: @cindex @code{implementation} usage
11049: @example
11050: interface
11051: selector write ( file object -- )
11052: selector read1 ( file object -- )
11053: end-interface storage
11054:
11055: bar class
11056: storage implementation
11057:
11058: ... overrides write
11059: ... overrides read1
11060: ...
11061: end-class foo
11062: @end example
11063:
11064: @noindent
11065: (I would add a word @code{read} @i{( file -- object )} that uses
11066: @code{read1} internally, but that's beyond the point illustrated
11067: here.)
11068:
11069: Note that you cannot use @code{protected} in an interface; and
11070: of course you cannot define fields.
11071:
11072: In the Neon model, all selectors are available for all classes;
11073: therefore it does not need interfaces. The price you pay in this model
11074: is slower late binding, and therefore, added complexity to avoid late
11075: binding.
11076:
11077: @node Objects Implementation, Objects Glossary, Object Interfaces, Objects
11078: @subsubsection @file{objects.fs} Implementation
11079: @cindex @file{objects.fs} implementation
11080:
11081: @cindex @code{object-map} discussion
11082: An object is a piece of memory, like one of the data structures
11083: described with @code{struct...end-struct}. It has a field
11084: @code{object-map} that points to the method map for the object's
11085: class.
11086:
11087: @cindex method map
11088: @cindex virtual function table
11089: The @emph{method map}@footnote{This is Self terminology; in C++
11090: terminology: virtual function table.} is an array that contains the
11091: execution tokens (@i{xt}s) of the methods for the object's class. Each
11092: selector contains an offset into a method map.
11093:
11094: @cindex @code{selector} implementation, class
11095: @code{selector} is a defining word that uses
11096: @code{CREATE} and @code{DOES>}. The body of the
11097: selector contains the offset; the @code{DOES>} action for a
11098: class selector is, basically:
11099:
11100: @example
11101: ( object addr ) @@ over object-map @@ + @@ execute
11102: @end example
11103:
11104: Since @code{object-map} is the first field of the object, it
11105: does not generate any code. As you can see, calling a selector has a
11106: small, constant cost.
11107:
11108: @cindex @code{current-interface} discussion
11109: @cindex class implementation and representation
11110: A class is basically a @code{struct} combined with a method
11111: map. During the class definition the alignment and size of the class
11112: are passed on the stack, just as with @code{struct}s, so
11113: @code{field} can also be used for defining class
11114: fields. However, passing more items on the stack would be
11115: inconvenient, so @code{class} builds a data structure in memory,
11116: which is accessed through the variable
11117: @code{current-interface}. After its definition is complete, the
11118: class is represented on the stack by a pointer (e.g., as parameter for
11119: a child class definition).
11120:
11121: A new class starts off with the alignment and size of its parent,
11122: and a copy of the parent's method map. Defining new fields extends the
11123: size and alignment; likewise, defining new selectors extends the
11124: method map. @code{overrides} just stores a new @i{xt} in the method
11125: map at the offset given by the selector.
11126:
11127: @cindex class binding, implementation
11128: Class binding just gets the @i{xt} at the offset given by the selector
11129: from the class's method map and @code{compile,}s (in the case of
11130: @code{[bind]}) it.
11131:
11132: @cindex @code{this} implementation
11133: @cindex @code{catch} and @code{this}
11134: @cindex @code{this} and @code{catch}
11135: I implemented @code{this} as a @code{value}. At the
11136: start of an @code{m:...;m} method the old @code{this} is
11137: stored to the return stack and restored at the end; and the object on
11138: the TOS is stored @code{TO this}. This technique has one
11139: disadvantage: If the user does not leave the method via
11140: @code{;m}, but via @code{throw} or @code{exit},
11141: @code{this} is not restored (and @code{exit} may
11142: crash). To deal with the @code{throw} problem, I have redefined
11143: @code{catch} to save and restore @code{this}; the same
11144: should be done with any word that can catch an exception. As for
11145: @code{exit}, I simply forbid it (as a replacement, there is
11146: @code{exitm}).
11147:
11148: @cindex @code{inst-var} implementation
11149: @code{inst-var} is just the same as @code{field}, with
11150: a different @code{DOES>} action:
11151: @example
11152: @@ this +
11153: @end example
11154: Similar for @code{inst-value}.
11155:
11156: @cindex class scoping implementation
11157: Each class also has a word list that contains the words defined with
11158: @code{inst-var} and @code{inst-value}, and its protected
11159: words. It also has a pointer to its parent. @code{class} pushes
11160: the word lists of the class and all its ancestors onto the search order stack,
11161: and @code{end-class} drops them.
11162:
11163: @cindex interface implementation
11164: An interface is like a class without fields, parent and protected
11165: words; i.e., it just has a method map. If a class implements an
11166: interface, its method map contains a pointer to the method map of the
11167: interface. The positive offsets in the map are reserved for class
11168: methods, therefore interface map pointers have negative
11169: offsets. Interfaces have offsets that are unique throughout the
11170: system, unlike class selectors, whose offsets are only unique for the
11171: classes where the selector is available (invokable).
11172:
11173: This structure means that interface selectors have to perform one
11174: indirection more than class selectors to find their method. Their body
11175: contains the interface map pointer offset in the class method map, and
11176: the method offset in the interface method map. The
11177: @code{does>} action for an interface selector is, basically:
11178:
11179: @example
11180: ( object selector-body )
11181: 2dup selector-interface @@ ( object selector-body object interface-offset )
11182: swap object-map @@ + @@ ( object selector-body map )
11183: swap selector-offset @@ + @@ execute
11184: @end example
11185:
11186: where @code{object-map} and @code{selector-offset} are
11187: first fields and generate no code.
11188:
11189: As a concrete example, consider the following code:
11190:
11191: @example
11192: interface
11193: selector if1sel1
11194: selector if1sel2
11195: end-interface if1
11196:
11197: object class
11198: if1 implementation
11199: selector cl1sel1
11200: cell% inst-var cl1iv1
11201:
11202: ' m1 overrides construct
11203: ' m2 overrides if1sel1
11204: ' m3 overrides if1sel2
11205: ' m4 overrides cl1sel2
11206: end-class cl1
11207:
11208: create obj1 object dict-new drop
11209: create obj2 cl1 dict-new drop
11210: @end example
11211:
11212: The data structure created by this code (including the data structure
11213: for @code{object}) is shown in the
11214: @uref{objects-implementation.eps,figure}, assuming a cell size of 4.
11215: @comment TODO add this diagram..
11216:
11217: @node Objects Glossary, , Objects Implementation, Objects
11218: @subsubsection @file{objects.fs} Glossary
11219: @cindex @file{objects.fs} Glossary
11220:
11221:
11222: doc---objects-bind
11223: doc---objects-<bind>
11224: doc---objects-bind'
11225: doc---objects-[bind]
11226: doc---objects-class
11227: doc---objects-class->map
11228: doc---objects-class-inst-size
11229: doc---objects-class-override!
11230: doc---objects-class-previous
11231: doc---objects-class>order
11232: doc---objects-construct
11233: doc---objects-current'
11234: doc---objects-[current]
11235: doc---objects-current-interface
11236: doc---objects-dict-new
11237: doc---objects-end-class
11238: doc---objects-end-class-noname
11239: doc---objects-end-interface
11240: doc---objects-end-interface-noname
11241: doc---objects-end-methods
11242: doc---objects-exitm
11243: doc---objects-heap-new
11244: doc---objects-implementation
11245: doc---objects-init-object
11246: doc---objects-inst-value
11247: doc---objects-inst-var
11248: doc---objects-interface
11249: doc---objects-m:
11250: doc---objects-:m
11251: doc---objects-;m
11252: doc---objects-method
11253: doc---objects-methods
11254: doc---objects-object
11255: doc---objects-overrides
11256: doc---objects-[parent]
11257: doc---objects-print
11258: doc---objects-protected
11259: doc---objects-public
11260: doc---objects-selector
11261: doc---objects-this
11262: doc---objects-<to-inst>
11263: doc---objects-[to-inst]
11264: doc---objects-to-this
11265: doc---objects-xt-new
11266:
11267:
11268: @c -------------------------------------------------------------
11269: @node OOF, Mini-OOF, Objects, Object-oriented Forth
11270: @subsection The @file{oof.fs} model
11271: @cindex oof
11272: @cindex object-oriented programming
11273:
11274: @cindex @file{objects.fs}
11275: @cindex @file{oof.fs}
11276:
11277: This section describes the @file{oof.fs} package.
11278:
11279: The package described in this section has been used in bigFORTH since 1991, and
11280: used for two large applications: a chromatographic system used to
11281: create new medicaments, and a graphic user interface library (MINOS).
11282:
11283: You can find a description (in German) of @file{oof.fs} in @cite{Object
11284: oriented bigFORTH} by Bernd Paysan, published in @cite{Vierte Dimension}
11285: 10(2), 1994.
11286:
11287: @menu
11288: * Properties of the OOF model::
11289: * Basic OOF Usage::
11290: * The OOF base class::
11291: * Class Declaration::
11292: * Class Implementation::
11293: @end menu
11294:
11295: @node Properties of the OOF model, Basic OOF Usage, OOF, OOF
11296: @subsubsection Properties of the @file{oof.fs} model
11297: @cindex @file{oof.fs} properties
11298:
11299: @itemize @bullet
11300: @item
11301: This model combines object oriented programming with information
11302: hiding. It helps you writing large application, where scoping is
11303: necessary, because it provides class-oriented scoping.
11304:
11305: @item
11306: Named objects, object pointers, and object arrays can be created,
11307: selector invocation uses the ``object selector'' syntax. Selector invocation
11308: to objects and/or selectors on the stack is a bit less convenient, but
11309: possible.
11310:
11311: @item
11312: Selector invocation and instance variable usage of the active object is
11313: straightforward, since both make use of the active object.
11314:
11315: @item
11316: Late binding is efficient and easy to use.
11317:
11318: @item
11319: State-smart objects parse selectors. However, extensibility is provided
11320: using a (parsing) selector @code{postpone} and a selector @code{'}.
11321:
11322: @item
11323: An implementation in ANS Forth is available.
11324:
11325: @end itemize
11326:
11327:
11328: @node Basic OOF Usage, The OOF base class, Properties of the OOF model, OOF
11329: @subsubsection Basic @file{oof.fs} Usage
11330: @cindex @file{oof.fs} usage
11331:
11332: This section uses the same example as for @code{objects} (@pxref{Basic Objects Usage}).
11333:
11334: You can define a class for graphical objects like this:
11335:
11336: @cindex @code{class} usage
11337: @cindex @code{class;} usage
11338: @cindex @code{method} usage
11339: @example
11340: object class graphical \ "object" is the parent class
11341: method draw ( x y -- )
11342: class;
11343: @end example
11344:
11345: This code defines a class @code{graphical} with an
11346: operation @code{draw}. We can perform the operation
11347: @code{draw} on any @code{graphical} object, e.g.:
11348:
11349: @example
11350: 100 100 t-rex draw
11351: @end example
11352:
11353: @noindent
11354: where @code{t-rex} is an object or object pointer, created with e.g.
11355: @code{graphical : t-rex}.
11356:
11357: @cindex abstract class
11358: How do we create a graphical object? With the present definitions,
11359: we cannot create a useful graphical object. The class
11360: @code{graphical} describes graphical objects in general, but not
11361: any concrete graphical object type (C++ users would call it an
11362: @emph{abstract class}); e.g., there is no method for the selector
11363: @code{draw} in the class @code{graphical}.
11364:
11365: For concrete graphical objects, we define child classes of the
11366: class @code{graphical}, e.g.:
11367:
11368: @example
11369: graphical class circle \ "graphical" is the parent class
11370: cell var circle-radius
11371: how:
11372: : draw ( x y -- )
11373: circle-radius @@ draw-circle ;
11374:
11375: : init ( n-radius -- )
11376: circle-radius ! ;
11377: class;
11378: @end example
11379:
11380: Here we define a class @code{circle} as a child of @code{graphical},
11381: with a field @code{circle-radius}; it defines new methods for the
11382: selectors @code{draw} and @code{init} (@code{init} is defined in
11383: @code{object}, the parent class of @code{graphical}).
11384:
11385: Now we can create a circle in the dictionary with:
11386:
11387: @example
11388: 50 circle : my-circle
11389: @end example
11390:
11391: @noindent
11392: @code{:} invokes @code{init}, thus initializing the field
11393: @code{circle-radius} with 50. We can draw this new circle at (100,100)
11394: with:
11395:
11396: @example
11397: 100 100 my-circle draw
11398: @end example
11399:
11400: @cindex selector invocation, restrictions
11401: @cindex class definition, restrictions
11402: Note: You can only invoke a selector if the receiving object belongs to
11403: the class where the selector was defined or one of its descendents;
11404: e.g., you can invoke @code{draw} only for objects belonging to
11405: @code{graphical} or its descendents (e.g., @code{circle}). The scoping
11406: mechanism will check if you try to invoke a selector that is not
11407: defined in this class hierarchy, so you'll get an error at compilation
11408: time.
11409:
11410:
11411: @node The OOF base class, Class Declaration, Basic OOF Usage, OOF
11412: @subsubsection The @file{oof.fs} base class
11413: @cindex @file{oof.fs} base class
11414:
11415: When you define a class, you have to specify a parent class. So how do
11416: you start defining classes? There is one class available from the start:
11417: @code{object}. You have to use it as ancestor for all classes. It is the
11418: only class that has no parent. Classes are also objects, except that
11419: they don't have instance variables; class manipulation such as
11420: inheritance or changing definitions of a class is handled through
11421: selectors of the class @code{object}.
11422:
11423: @code{object} provides a number of selectors:
11424:
11425: @itemize @bullet
11426: @item
11427: @code{class} for subclassing, @code{definitions} to add definitions
11428: later on, and @code{class?} to get type informations (is the class a
11429: subclass of the class passed on the stack?).
11430:
11431: doc---object-class
11432: doc---object-definitions
11433: doc---object-class?
11434:
11435:
11436: @item
11437: @code{init} and @code{dispose} as constructor and destructor of the
11438: object. @code{init} is invocated after the object's memory is allocated,
11439: while @code{dispose} also handles deallocation. Thus if you redefine
11440: @code{dispose}, you have to call the parent's dispose with @code{super
11441: dispose}, too.
11442:
11443: doc---object-init
11444: doc---object-dispose
11445:
11446:
11447: @item
11448: @code{new}, @code{new[]}, @code{:}, @code{ptr}, @code{asptr}, and
11449: @code{[]} to create named and unnamed objects and object arrays or
11450: object pointers.
11451:
11452: doc---object-new
11453: doc---object-new[]
11454: doc---object-:
11455: doc---object-ptr
11456: doc---object-asptr
11457: doc---object-[]
11458:
11459:
11460: @item
11461: @code{::} and @code{super} for explicit scoping. You should use explicit
11462: scoping only for super classes or classes with the same set of instance
11463: variables. Explicitly-scoped selectors use early binding.
11464:
11465: doc---object-::
11466: doc---object-super
11467:
11468:
11469: @item
11470: @code{self} to get the address of the object
11471:
11472: doc---object-self
11473:
11474:
11475: @item
11476: @code{bind}, @code{bound}, @code{link}, and @code{is} to assign object
11477: pointers and instance defers.
11478:
11479: doc---object-bind
11480: doc---object-bound
11481: doc---object-link
11482: doc---object-is
11483:
11484:
11485: @item
11486: @code{'} to obtain selector tokens, @code{send} to invocate selectors
11487: form the stack, and @code{postpone} to generate selector invocation code.
11488:
11489: doc---object-'
11490: doc---object-postpone
11491:
11492:
11493: @item
11494: @code{with} and @code{endwith} to select the active object from the
11495: stack, and enable its scope. Using @code{with} and @code{endwith}
11496: also allows you to create code using selector @code{postpone} without being
11497: trapped by the state-smart objects.
11498:
11499: doc---object-with
11500: doc---object-endwith
11501:
11502:
11503: @end itemize
11504:
11505: @node Class Declaration, Class Implementation, The OOF base class, OOF
11506: @subsubsection Class Declaration
11507: @cindex class declaration
11508:
11509: @itemize @bullet
11510: @item
11511: Instance variables
11512:
11513: doc---oof-var
11514:
11515:
11516: @item
11517: Object pointers
11518:
11519: doc---oof-ptr
11520: doc---oof-asptr
11521:
11522:
11523: @item
11524: Instance defers
11525:
11526: doc---oof-defer
11527:
11528:
11529: @item
11530: Method selectors
11531:
11532: doc---oof-early
11533: doc---oof-method
11534:
11535:
11536: @item
11537: Class-wide variables
11538:
11539: doc---oof-static
11540:
11541:
11542: @item
11543: End declaration
11544:
11545: doc---oof-how:
11546: doc---oof-class;
11547:
11548:
11549: @end itemize
11550:
11551: @c -------------------------------------------------------------
11552: @node Class Implementation, , Class Declaration, OOF
11553: @subsubsection Class Implementation
11554: @cindex class implementation
11555:
11556: @c -------------------------------------------------------------
11557: @node Mini-OOF, Comparison with other object models, OOF, Object-oriented Forth
11558: @subsection The @file{mini-oof.fs} model
11559: @cindex mini-oof
11560:
11561: Gforth's third object oriented Forth package is a 12-liner. It uses a
11562: mixture of the @file{objects.fs} and the @file{oof.fs} syntax,
11563: and reduces to the bare minimum of features. This is based on a posting
11564: of Bernd Paysan in comp.lang.forth.
11565:
11566: @menu
11567: * Basic Mini-OOF Usage::
11568: * Mini-OOF Example::
11569: * Mini-OOF Implementation::
11570: @end menu
11571:
11572: @c -------------------------------------------------------------
11573: @node Basic Mini-OOF Usage, Mini-OOF Example, Mini-OOF, Mini-OOF
11574: @subsubsection Basic @file{mini-oof.fs} Usage
11575: @cindex mini-oof usage
11576:
11577: There is a base class (@code{class}, which allocates one cell for the
11578: object pointer) plus seven other words: to define a method, a variable,
11579: a class; to end a class, to resolve binding, to allocate an object and
11580: to compile a class method.
11581: @comment TODO better description of the last one
11582:
11583:
11584: doc-object
11585: doc-method
11586: doc-var
11587: doc-class
11588: doc-end-class
11589: doc-defines
11590: doc-new
11591: doc-::
11592:
11593:
11594:
11595: @c -------------------------------------------------------------
11596: @node Mini-OOF Example, Mini-OOF Implementation, Basic Mini-OOF Usage, Mini-OOF
11597: @subsubsection Mini-OOF Example
11598: @cindex mini-oof example
11599:
11600: A short example shows how to use this package. This example, in slightly
11601: extended form, is supplied as @file{moof-exm.fs}
11602: @comment TODO could flesh this out with some comments from the Forthwrite article
11603:
11604: @example
11605: object class
11606: method init
11607: method draw
11608: end-class graphical
11609: @end example
11610:
11611: This code defines a class @code{graphical} with an
11612: operation @code{draw}. We can perform the operation
11613: @code{draw} on any @code{graphical} object, e.g.:
11614:
11615: @example
11616: 100 100 t-rex draw
11617: @end example
11618:
11619: where @code{t-rex} is an object or object pointer, created with e.g.
11620: @code{graphical new Constant t-rex}.
11621:
11622: For concrete graphical objects, we define child classes of the
11623: class @code{graphical}, e.g.:
11624:
11625: @example
11626: graphical class
11627: cell var circle-radius
11628: end-class circle \ "graphical" is the parent class
11629:
11630: :noname ( x y -- )
11631: circle-radius @@ draw-circle ; circle defines draw
11632: :noname ( r -- )
11633: circle-radius ! ; circle defines init
11634: @end example
11635:
11636: There is no implicit init method, so we have to define one. The creation
11637: code of the object now has to call init explicitely.
11638:
11639: @example
11640: circle new Constant my-circle
11641: 50 my-circle init
11642: @end example
11643:
11644: It is also possible to add a function to create named objects with
11645: automatic call of @code{init}, given that all objects have @code{init}
11646: on the same place:
11647:
11648: @example
11649: : new: ( .. o "name" -- )
11650: new dup Constant init ;
11651: 80 circle new: large-circle
11652: @end example
11653:
11654: We can draw this new circle at (100,100) with:
11655:
11656: @example
11657: 100 100 my-circle draw
11658: @end example
11659:
11660: @node Mini-OOF Implementation, , Mini-OOF Example, Mini-OOF
11661: @subsubsection @file{mini-oof.fs} Implementation
11662:
11663: Object-oriented systems with late binding typically use a
11664: ``vtable''-approach: the first variable in each object is a pointer to a
11665: table, which contains the methods as function pointers. The vtable
11666: may also contain other information.
11667:
11668: So first, let's declare selectors:
11669:
11670: @example
11671: : method ( m v "name" -- m' v ) Create over , swap cell+ swap
11672: DOES> ( ... o -- ... ) @@ over @@ + @@ execute ;
11673: @end example
11674:
11675: During selector declaration, the number of selectors and instance
11676: variables is on the stack (in address units). @code{method} creates one
11677: selector and increments the selector number. To execute a selector, it
11678: takes the object, fetches the vtable pointer, adds the offset, and
11679: executes the method @i{xt} stored there. Each selector takes the object
11680: it is invoked with as top of stack parameter; it passes the parameters
11681: (including the object) unchanged to the appropriate method which should
11682: consume that object.
11683:
11684: Now, we also have to declare instance variables
11685:
11686: @example
11687: : var ( m v size "name" -- m v' ) Create over , +
11688: DOES> ( o -- addr ) @@ + ;
11689: @end example
11690:
11691: As before, a word is created with the current offset. Instance
11692: variables can have different sizes (cells, floats, doubles, chars), so
11693: all we do is take the size and add it to the offset. If your machine
11694: has alignment restrictions, put the proper @code{aligned} or
11695: @code{faligned} before the variable, to adjust the variable
11696: offset. That's why it is on the top of stack.
11697:
11698: We need a starting point (the base object) and some syntactic sugar:
11699:
11700: @example
11701: Create object 1 cells , 2 cells ,
11702: : class ( class -- class selectors vars ) dup 2@@ ;
11703: @end example
11704:
11705: For inheritance, the vtable of the parent object has to be
11706: copied when a new, derived class is declared. This gives all the
11707: methods of the parent class, which can be overridden, though.
11708:
11709: @example
11710: : end-class ( class selectors vars "name" -- )
11711: Create here >r , dup , 2 cells ?DO ['] noop , 1 cells +LOOP
11712: cell+ dup cell+ r> rot @@ 2 cells /string move ;
11713: @end example
11714:
11715: The first line creates the vtable, initialized with
11716: @code{noop}s. The second line is the inheritance mechanism, it
11717: copies the xts from the parent vtable.
11718:
11719: We still have no way to define new methods, let's do that now:
11720:
11721: @example
11722: : defines ( xt class "name" -- ) ' >body @@ + ! ;
11723: @end example
11724:
11725: To allocate a new object, we need a word, too:
11726:
11727: @example
11728: : new ( class -- o ) here over @@ allot swap over ! ;
11729: @end example
11730:
11731: Sometimes derived classes want to access the method of the
11732: parent object. There are two ways to achieve this with Mini-OOF:
11733: first, you could use named words, and second, you could look up the
11734: vtable of the parent object.
11735:
11736: @example
11737: : :: ( class "name" -- ) ' >body @@ + @@ compile, ;
11738: @end example
11739:
11740:
11741: Nothing can be more confusing than a good example, so here is
11742: one. First let's declare a text object (called
11743: @code{button}), that stores text and position:
11744:
11745: @example
11746: object class
11747: cell var text
11748: cell var len
11749: cell var x
11750: cell var y
11751: method init
11752: method draw
11753: end-class button
11754: @end example
11755:
11756: @noindent
11757: Now, implement the two methods, @code{draw} and @code{init}:
11758:
11759: @example
11760: :noname ( o -- )
11761: >r r@@ x @@ r@@ y @@ at-xy r@@ text @@ r> len @@ type ;
11762: button defines draw
11763: :noname ( addr u o -- )
11764: >r 0 r@@ x ! 0 r@@ y ! r@@ len ! r> text ! ;
11765: button defines init
11766: @end example
11767:
11768: @noindent
11769: To demonstrate inheritance, we define a class @code{bold-button}, with no
11770: new data and no new selectors:
11771:
11772: @example
11773: button class
11774: end-class bold-button
11775:
11776: : bold 27 emit ." [1m" ;
11777: : normal 27 emit ." [0m" ;
11778: @end example
11779:
11780: @noindent
11781: The class @code{bold-button} has a different draw method to
11782: @code{button}, but the new method is defined in terms of the draw method
11783: for @code{button}:
11784:
11785: @example
11786: :noname bold [ button :: draw ] normal ; bold-button defines draw
11787: @end example
11788:
11789: @noindent
11790: Finally, create two objects and apply selectors:
11791:
11792: @example
11793: button new Constant foo
11794: s" thin foo" foo init
11795: page
11796: foo draw
11797: bold-button new Constant bar
11798: s" fat bar" bar init
11799: 1 bar y !
11800: bar draw
11801: @end example
11802:
11803:
11804: @node Comparison with other object models, , Mini-OOF, Object-oriented Forth
11805: @subsection Comparison with other object models
11806: @cindex comparison of object models
11807: @cindex object models, comparison
11808:
11809: Many object-oriented Forth extensions have been proposed (@cite{A survey
11810: of object-oriented Forths} (SIGPLAN Notices, April 1996) by Bradford
11811: J. Rodriguez and W. F. S. Poehlman lists 17). This section discusses the
11812: relation of the object models described here to two well-known and two
11813: closely-related (by the use of method maps) models. Andras Zsoter
11814: helped us with this section.
11815:
11816: @cindex Neon model
11817: The most popular model currently seems to be the Neon model (see
11818: @cite{Object-oriented programming in ANS Forth} (Forth Dimensions, March
11819: 1997) by Andrew McKewan) but this model has a number of limitations
11820: @footnote{A longer version of this critique can be
11821: found in @cite{On Standardizing Object-Oriented Forth Extensions} (Forth
11822: Dimensions, May 1997) by Anton Ertl.}:
11823:
11824: @itemize @bullet
11825: @item
11826: It uses a @code{@emph{selector object}} syntax, which makes it unnatural
11827: to pass objects on the stack.
11828:
11829: @item
11830: It requires that the selector parses the input stream (at
11831: compile time); this leads to reduced extensibility and to bugs that are
11832: hard to find.
11833:
11834: @item
11835: It allows using every selector on every object; this eliminates the
11836: need for interfaces, but makes it harder to create efficient
11837: implementations.
11838: @end itemize
11839:
11840: @cindex Pountain's object-oriented model
11841: Another well-known publication is @cite{Object-Oriented Forth} (Academic
11842: Press, London, 1987) by Dick Pountain. However, it is not really about
11843: object-oriented programming, because it hardly deals with late
11844: binding. Instead, it focuses on features like information hiding and
11845: overloading that are characteristic of modular languages like Ada (83).
11846:
11847: @cindex Zsoter's object-oriented model
11848: In @uref{http://www.forth.org/oopf.html, Does late binding have to be
11849: slow?} (Forth Dimensions 18(1) 1996, pages 31-35) Andras Zsoter
11850: describes a model that makes heavy use of an active object (like
11851: @code{this} in @file{objects.fs}): The active object is not only used
11852: for accessing all fields, but also specifies the receiving object of
11853: every selector invocation; you have to change the active object
11854: explicitly with @code{@{ ... @}}, whereas in @file{objects.fs} it
11855: changes more or less implicitly at @code{m: ... ;m}. Such a change at
11856: the method entry point is unnecessary with Zsoter's model, because the
11857: receiving object is the active object already. On the other hand, the
11858: explicit change is absolutely necessary in that model, because otherwise
11859: no one could ever change the active object. An ANS Forth implementation
11860: of this model is available through
11861: @uref{http://www.forth.org/oopf.html}.
11862:
11863: @cindex @file{oof.fs}, differences to other models
11864: The @file{oof.fs} model combines information hiding and overloading
11865: resolution (by keeping names in various word lists) with object-oriented
11866: programming. It sets the active object implicitly on method entry, but
11867: also allows explicit changing (with @code{>o...o>} or with
11868: @code{with...endwith}). It uses parsing and state-smart objects and
11869: classes for resolving overloading and for early binding: the object or
11870: class parses the selector and determines the method from this. If the
11871: selector is not parsed by an object or class, it performs a call to the
11872: selector for the active object (late binding), like Zsoter's model.
11873: Fields are always accessed through the active object. The big
11874: disadvantage of this model is the parsing and the state-smartness, which
11875: reduces extensibility and increases the opportunities for subtle bugs;
11876: essentially, you are only safe if you never tick or @code{postpone} an
11877: object or class (Bernd disagrees, but I (Anton) am not convinced).
11878:
11879: @cindex @file{mini-oof.fs}, differences to other models
11880: The @file{mini-oof.fs} model is quite similar to a very stripped-down
11881: version of the @file{objects.fs} model, but syntactically it is a
11882: mixture of the @file{objects.fs} and @file{oof.fs} models.
11883:
11884:
11885: @c -------------------------------------------------------------
11886: @node Programming Tools, C Interface, Object-oriented Forth, Words
11887: @section Programming Tools
11888: @cindex programming tools
11889:
11890: @c !! move this and assembler down below OO stuff.
11891:
11892: @menu
11893: * Examining:: Data and Code.
11894: * Forgetting words:: Usually before reloading.
11895: * Debugging:: Simple and quick.
11896: * Assertions:: Making your programs self-checking.
11897: * Singlestep Debugger:: Executing your program word by word.
11898: @end menu
11899:
11900: @node Examining, Forgetting words, Programming Tools, Programming Tools
11901: @subsection Examining data and code
11902: @cindex examining data and code
11903: @cindex data examination
11904: @cindex code examination
11905:
11906: The following words inspect the stack non-destructively:
11907:
11908: doc-.s
11909: doc-f.s
11910: doc-maxdepth-.s
11911:
11912: There is a word @code{.r} but it does @i{not} display the return stack!
11913: It is used for formatted numeric output (@pxref{Simple numeric output}).
11914:
11915: doc-depth
11916: doc-fdepth
11917: doc-clearstack
11918: doc-clearstacks
11919:
11920: The following words inspect memory.
11921:
11922: doc-?
11923: doc-dump
11924:
11925: And finally, @code{see} allows to inspect code:
11926:
11927: doc-see
11928: doc-xt-see
11929: doc-simple-see
11930: doc-simple-see-range
11931: doc-see-code
11932: doc-see-code-range
11933:
11934: @node Forgetting words, Debugging, Examining, Programming Tools
11935: @subsection Forgetting words
11936: @cindex words, forgetting
11937: @cindex forgeting words
11938:
11939: @c anton: other, maybe better places for this subsection: Defining Words;
11940: @c Dictionary allocation. At least a reference should be there.
11941:
11942: Forth allows you to forget words (and everything that was alloted in the
11943: dictonary after them) in a LIFO manner.
11944:
11945: doc-marker
11946:
11947: The most common use of this feature is during progam development: when
11948: you change a source file, forget all the words it defined and load it
11949: again (since you also forget everything defined after the source file
11950: was loaded, you have to reload that, too). Note that effects like
11951: storing to variables and destroyed system words are not undone when you
11952: forget words. With a system like Gforth, that is fast enough at
11953: starting up and compiling, I find it more convenient to exit and restart
11954: Gforth, as this gives me a clean slate.
11955:
11956: Here's an example of using @code{marker} at the start of a source file
11957: that you are debugging; it ensures that you only ever have one copy of
11958: the file's definitions compiled at any time:
11959:
11960: @example
11961: [IFDEF] my-code
11962: my-code
11963: [ENDIF]
11964:
11965: marker my-code
11966: init-included-files
11967:
11968: \ .. definitions start here
11969: \ .
11970: \ .
11971: \ end
11972: @end example
11973:
11974:
11975: @node Debugging, Assertions, Forgetting words, Programming Tools
11976: @subsection Debugging
11977: @cindex debugging
11978:
11979: Languages with a slow edit/compile/link/test development loop tend to
11980: require sophisticated tracing/stepping debuggers to facilate debugging.
11981:
11982: A much better (faster) way in fast-compiling languages is to add
11983: printing code at well-selected places, let the program run, look at
11984: the output, see where things went wrong, add more printing code, etc.,
11985: until the bug is found.
11986:
11987: The simple debugging aids provided in @file{debugs.fs}
11988: are meant to support this style of debugging.
11989:
11990: The word @code{~~} prints debugging information (by default the source
11991: location and the stack contents). It is easy to insert. If you use Emacs
11992: it is also easy to remove (@kbd{C-x ~} in the Emacs Forth mode to
11993: query-replace them with nothing). The deferred words
11994: @code{printdebugdata} and @code{.debugline} control the output of
11995: @code{~~}. The default source location output format works well with
11996: Emacs' compilation mode, so you can step through the program at the
11997: source level using @kbd{C-x `} (the advantage over a stepping debugger
11998: is that you can step in any direction and you know where the crash has
11999: happened or where the strange data has occurred).
12000:
12001: doc-~~
12002: doc-printdebugdata
12003: doc-.debugline
12004: doc-debug-fid
12005:
12006: @cindex filenames in @code{~~} output
12007: @code{~~} (and assertions) will usually print the wrong file name if a
12008: marker is executed in the same file after their occurance. They will
12009: print @samp{*somewhere*} as file name if a marker is executed in the
12010: same file before their occurance.
12011:
12012:
12013: @node Assertions, Singlestep Debugger, Debugging, Programming Tools
12014: @subsection Assertions
12015: @cindex assertions
12016:
12017: It is a good idea to make your programs self-checking, especially if you
12018: make an assumption that may become invalid during maintenance (for
12019: example, that a certain field of a data structure is never zero). Gforth
12020: supports @dfn{assertions} for this purpose. They are used like this:
12021:
12022: @example
12023: assert( @i{flag} )
12024: @end example
12025:
12026: The code between @code{assert(} and @code{)} should compute a flag, that
12027: should be true if everything is alright and false otherwise. It should
12028: not change anything else on the stack. The overall stack effect of the
12029: assertion is @code{( -- )}. E.g.
12030:
12031: @example
12032: assert( 1 1 + 2 = ) \ what we learn in school
12033: assert( dup 0<> ) \ assert that the top of stack is not zero
12034: assert( false ) \ this code should not be reached
12035: @end example
12036:
12037: The need for assertions is different at different times. During
12038: debugging, we want more checking, in production we sometimes care more
12039: for speed. Therefore, assertions can be turned off, i.e., the assertion
12040: becomes a comment. Depending on the importance of an assertion and the
12041: time it takes to check it, you may want to turn off some assertions and
12042: keep others turned on. Gforth provides several levels of assertions for
12043: this purpose:
12044:
12045:
12046: doc-assert0(
12047: doc-assert1(
12048: doc-assert2(
12049: doc-assert3(
12050: doc-assert(
12051: doc-)
12052:
12053:
12054: The variable @code{assert-level} specifies the highest assertions that
12055: are turned on. I.e., at the default @code{assert-level} of one,
12056: @code{assert0(} and @code{assert1(} assertions perform checking, while
12057: @code{assert2(} and @code{assert3(} assertions are treated as comments.
12058:
12059: The value of @code{assert-level} is evaluated at compile-time, not at
12060: run-time. Therefore you cannot turn assertions on or off at run-time;
12061: you have to set the @code{assert-level} appropriately before compiling a
12062: piece of code. You can compile different pieces of code at different
12063: @code{assert-level}s (e.g., a trusted library at level 1 and
12064: newly-written code at level 3).
12065:
12066:
12067: doc-assert-level
12068:
12069:
12070: If an assertion fails, a message compatible with Emacs' compilation mode
12071: is produced and the execution is aborted (currently with @code{ABORT"}.
12072: If there is interest, we will introduce a special throw code. But if you
12073: intend to @code{catch} a specific condition, using @code{throw} is
12074: probably more appropriate than an assertion).
12075:
12076: @cindex filenames in assertion output
12077: Assertions (and @code{~~}) will usually print the wrong file name if a
12078: marker is executed in the same file after their occurance. They will
12079: print @samp{*somewhere*} as file name if a marker is executed in the
12080: same file before their occurance.
12081:
12082: Definitions in ANS Forth for these assertion words are provided
12083: in @file{compat/assert.fs}.
12084:
12085:
12086: @node Singlestep Debugger, , Assertions, Programming Tools
12087: @subsection Singlestep Debugger
12088: @cindex singlestep Debugger
12089: @cindex debugging Singlestep
12090:
12091: The singlestep debugger works only with the engine @code{gforth-itc}.
12092:
12093: When you create a new word there's often the need to check whether it
12094: behaves correctly or not. You can do this by typing @code{dbg
12095: badword}. A debug session might look like this:
12096:
12097: @example
12098: : badword 0 DO i . LOOP ; ok
12099: 2 dbg badword
12100: : badword
12101: Scanning code...
12102:
12103: Nesting debugger ready!
12104:
12105: 400D4738 8049BC4 0 -> [ 2 ] 00002 00000
12106: 400D4740 8049F68 DO -> [ 0 ]
12107: 400D4744 804A0C8 i -> [ 1 ] 00000
12108: 400D4748 400C5E60 . -> 0 [ 0 ]
12109: 400D474C 8049D0C LOOP -> [ 0 ]
12110: 400D4744 804A0C8 i -> [ 1 ] 00001
12111: 400D4748 400C5E60 . -> 1 [ 0 ]
12112: 400D474C 8049D0C LOOP -> [ 0 ]
12113: 400D4758 804B384 ; -> ok
12114: @end example
12115:
12116: Each line displayed is one step. You always have to hit return to
12117: execute the next word that is displayed. If you don't want to execute
12118: the next word in a whole, you have to type @kbd{n} for @code{nest}. Here is
12119: an overview what keys are available:
12120:
12121: @table @i
12122:
12123: @item @key{RET}
12124: Next; Execute the next word.
12125:
12126: @item n
12127: Nest; Single step through next word.
12128:
12129: @item u
12130: Unnest; Stop debugging and execute rest of word. If we got to this word
12131: with nest, continue debugging with the calling word.
12132:
12133: @item d
12134: Done; Stop debugging and execute rest.
12135:
12136: @item s
12137: Stop; Abort immediately.
12138:
12139: @end table
12140:
12141: Debugging large application with this mechanism is very difficult, because
12142: you have to nest very deeply into the program before the interesting part
12143: begins. This takes a lot of time.
12144:
12145: To do it more directly put a @code{BREAK:} command into your source code.
12146: When program execution reaches @code{BREAK:} the single step debugger is
12147: invoked and you have all the features described above.
12148:
12149: If you have more than one part to debug it is useful to know where the
12150: program has stopped at the moment. You can do this by the
12151: @code{BREAK" string"} command. This behaves like @code{BREAK:} except that
12152: string is typed out when the ``breakpoint'' is reached.
12153:
12154:
12155: doc-dbg
12156: doc-break:
12157: doc-break"
12158:
12159: @c ------------------------------------------------------------
12160: @node C Interface, Assembler and Code Words, Programming Tools, Words
12161: @section C Interface
12162: @cindex C interface
12163: @cindex foreign language interface
12164: @cindex interface to C functions
12165:
12166: Note that the C interface is not yet complete; callbacks are missing,
12167: as well as a way of declaring structs, unions, and their fields.
12168:
12169: @menu
12170: * Calling C Functions::
12171: * Declaring C Functions::
12172: * Calling C function pointers::
12173: * Defining library interfaces::
12174: * Declaring OS-level libraries::
12175: * Callbacks::
12176: * C interface internals::
12177: * Low-Level C Interface Words::
12178: @end menu
12179:
12180: @node Calling C Functions, Declaring C Functions, C Interface, C Interface
12181: @subsection Calling C functions
12182: @cindex C functions, calls to
12183: @cindex calling C functions
12184:
12185: Once a C function is declared (see @pxref{Declaring C Functions}), you
12186: can call it as follows: You push the arguments on the stack(s), and
12187: then call the word for the C function. The arguments have to be
12188: pushed in the same order as the arguments appear in the C
12189: documentation (i.e., the first argument is deepest on the stack).
12190: Integer and pointer arguments have to be pushed on the data stack,
12191: floating-point arguments on the FP stack; these arguments are consumed
12192: by the called C function.
12193:
12194: On returning from the C function, the return value, if any, resides on
12195: the appropriate stack: an integer return value is pushed on the data
12196: stack, an FP return value on the FP stack, and a void return value
12197: results in not pushing anything. Note that most C functions have a
12198: return value, even if that is often not used in C; in Forth, you have
12199: to @code{drop} this return value explicitly if you do not use it.
12200:
12201: The C interface automatically converts between the C type and the
12202: Forth type as necessary, on a best-effort basis (in some cases, there
12203: may be some loss).
12204:
12205: As an example, consider the POSIX function @code{lseek()}:
12206:
12207: @example
12208: off_t lseek(int fd, off_t offset, int whence);
12209: @end example
12210:
12211: This function takes three integer arguments, and returns an integer
12212: argument, so a Forth call for setting the current file offset to the
12213: start of the file could look like this:
12214:
12215: @example
12216: fd @@ 0 SEEK_SET lseek -1 = if
12217: ... \ error handling
12218: then
12219: @end example
12220:
12221: You might be worried that an @code{off_t} does not fit into a cell, so
12222: you could not pass larger offsets to lseek, and might get only a part
12223: of the return values. In that case, in your declaration of the
12224: function (@pxref{Declaring C Functions}) you should declare it to use
12225: double-cells for the off_t argument and return value, and maybe give
12226: the resulting Forth word a different name, like @code{dlseek}; the
12227: result could be called like this:
12228:
12229: @example
12230: fd @@ 0. SEEK_SET dlseek -1. d= if
12231: ... \ error handling
12232: then
12233: @end example
12234:
12235: Passing and returning structs or unions is currently not supported by
12236: our interface@footnote{If you know the calling convention of your C
12237: compiler, you usually can call such functions in some way, but that
12238: way is usually not portable between platforms, and sometimes not even
12239: between C compilers.}.
12240:
12241: Calling functions with a variable number of arguments (@emph{variadic}
12242: functions, e.g., @code{printf()}) is only supported by having you
12243: declare one function-calling word for each argument pattern, and
12244: calling the appropriate word for the desired pattern.
12245:
12246:
12247:
12248: @node Declaring C Functions, Calling C function pointers, Calling C Functions, C Interface
12249: @subsection Declaring C Functions
12250: @cindex C functions, declarations
12251: @cindex declaring C functions
12252:
12253: Before you can call @code{lseek} or @code{dlseek}, you have to declare
12254: it. The declaration consists of two parts:
12255:
12256: @table @b
12257:
12258: @item The C part
12259: is the C declaration of the function, or more typically and portably,
12260: a C-style @code{#include} of a file that contains the declaration of
12261: the C function.
12262:
12263: @item The Forth part
12264: declares the Forth types of the parameters and the Forth word name
12265: corresponding to the C function.
12266:
12267: @end table
12268:
12269: For the words @code{lseek} and @code{dlseek} mentioned earlier, the
12270: declarations are:
12271:
12272: @example
12273: \c #define _FILE_OFFSET_BITS 64
12274: \c #include <sys/types.h>
12275: \c #include <unistd.h>
12276: c-function lseek lseek n n n -- n
12277: c-function dlseek lseek n d n -- d
12278: @end example
12279:
12280: The C part of the declarations is prefixed by @code{\c}, and the rest
12281: of the line is ordinary C code. You can use as many lines of C
12282: declarations as you like, and they are visible for all further
12283: function declarations.
12284:
12285: The Forth part declares each interface word with @code{c-function},
12286: followed by the Forth name of the word, the C name of the called
12287: function, and the stack effect of the word. The stack effect contains
12288: an arbitrary number of types of parameters, then @code{--}, and then
12289: exactly one type for the return value. The possible types are:
12290:
12291: @table @code
12292:
12293: @item n
12294: single-cell integer
12295:
12296: @item a
12297: address (single-cell)
12298:
12299: @item d
12300: double-cell integer
12301:
12302: @item r
12303: floating-point value
12304:
12305: @item func
12306: C function pointer
12307:
12308: @item void
12309: no value (used as return type for void functions)
12310:
12311: @end table
12312:
12313: @cindex variadic C functions
12314:
12315: To deal with variadic C functions, you can declare one Forth word for
12316: every pattern you want to use, e.g.:
12317:
12318: @example
12319: \c #include <stdio.h>
12320: c-function printf-nr printf a n r -- n
12321: c-function printf-rn printf a r n -- n
12322: @end example
12323:
12324: Note that with C functions declared as variadic (or if you don't
12325: provide a prototype), the C interface has no C type to convert to, so
12326: no automatic conversion happens, which may lead to portability
12327: problems in some cases. In such cases you can perform the conversion
12328: explicitly on the C level, e.g., as follows:
12329:
12330: @example
12331: \c #define printfll(s,ll) printf(s,(long long)ll)
12332: c-function printfll printfll a n -- n
12333: @end example
12334:
12335: Here, instead of calling @code{printf()} directly, we define a macro
12336: that casts (converts) the Forth single-cell integer into a
12337: C @code{long long} before calling @code{printf()}.
12338:
12339: doc-\c
12340: doc-c-function
12341: doc-c-value
12342: doc-c-variable
12343:
12344: In order to work, this C interface invokes GCC at run-time and uses
12345: dynamic linking. If these features are not available, there are
12346: other, less convenient and less portable C interfaces in @file{lib.fs}
12347: and @file{oldlib.fs}. These interfaces are mostly undocumented and
12348: mostly incompatible with each other and with the documented C
12349: interface; you can find some examples for the @file{lib.fs} interface
12350: in @file{lib.fs}.
12351:
12352:
12353: @node Calling C function pointers, Defining library interfaces, Declaring C Functions, C Interface
12354: @subsection Calling C function pointers from Forth
12355: @cindex C function pointers, calling from Forth
12356:
12357: If you come across a C function pointer (e.g., in some C-constructed
12358: structure) and want to call it from your Forth program, you can also
12359: use the features explained until now to achieve that, as follows:
12360:
12361: Let us assume that there is a C function pointer type @code{func1}
12362: defined in some header file @file{func1.h}, and you know that these
12363: functions take one integer argument and return an integer result; and
12364: you want to call functions through such pointers. Just define
12365:
12366: @example
12367: \c #include <func1.h>
12368: \c #define call_func1(par1,fptr) ((func1)fptr)(par1)
12369: c-function call-func1 call_func1 n func -- n
12370: @end example
12371:
12372: and then you can call a function pointed to by, say @code{func1a} as
12373: follows:
12374:
12375: @example
12376: -5 func1a call-func1 .
12377: @end example
12378:
12379: In the C part, @code{call_func} is defined as a macro to avoid having
12380: to declare the exact parameter and return types, so the C compiler
12381: knows them from the declaration of @code{func1}.
12382:
12383: The Forth word @code{call-func1} is similar to @code{execute}, except
12384: that it takes a C @code{func1} pointer instead of a Forth execution
12385: token, and it is specific to @code{func1} pointers. For each type of
12386: function pointer you want to call from Forth, you have to define
12387: a separate calling word.
12388:
12389:
12390: @node Defining library interfaces, Declaring OS-level libraries, Calling C function pointers, C Interface
12391: @subsection Defining library interfaces
12392: @cindex giving a name to a library interface
12393: @cindex library interface names
12394:
12395: You can give a name to a bunch of C function declarations (a library
12396: interface), as follows:
12397:
12398: @example
12399: c-library lseek-lib
12400: \c #define _FILE_OFFSET_BITS 64
12401: ...
12402: end-c-library
12403: @end example
12404:
12405: The effect of giving such a name to the interface is that the names of
12406: the generated files will contain that name, and when you use the
12407: interface a second time, it will use the existing files instead of
12408: generating and compiling them again, saving you time. Note that even
12409: if you change the declarations, the old (stale) files will be used,
12410: probably leading to errors. So, during development of the
12411: declarations we recommend not using @code{c-library}. Normally these
12412: files are cached in @file{$HOME/.gforth/libcc-named}, so by deleting
12413: that directory you can get rid of stale files.
12414:
12415: Note that you should use @code{c-library} before everything else
12416: having anything to do with that library, as it resets some setup
12417: stuff. The idea is that the typical use is to put each
12418: @code{c-library}...@code{end-library} unit in its own file, and to be
12419: able to include these files in any order.
12420:
12421: Note that the library name is not allocated in the dictionary and
12422: therefore does not shadow dictionary names. It is used in the file
12423: system, so you have to use naming conventions appropriate for file
12424: systems. Also, you must not call a function you declare after
12425: @code{c-library} before you perform @code{end-c-library}.
12426:
12427: A major benefit of these named library interfaces is that, once they
12428: are generated, the tools used to generated them (in particular, the C
12429: compiler and libtool) are no longer needed, so the interface can be
12430: used even on machines that do not have the tools installed.
12431:
12432: doc-c-library-name
12433: doc-c-library
12434: doc-end-c-library
12435:
12436:
12437: @node Declaring OS-level libraries, Callbacks, Defining library interfaces, C Interface
12438: @subsection Declaring OS-level libraries
12439: @cindex Shared libraries in C interface
12440: @cindex Dynamically linked libraries in C interface
12441: @cindex Libraries in C interface
12442:
12443: For calling some C functions, you need to link with a specific
12444: OS-level library that contains that function. E.g., the @code{sin}
12445: function requires linking a special library by using the command line
12446: switch @code{-lm}. In our C iterface you do the equivalent thing by
12447: calling @code{add-lib} as follows:
12448:
12449: @example
12450: clear-libs
12451: s" m" add-lib
12452: \c #include <math.h>
12453: c-function sin sin r -- r
12454: @end example
12455:
12456: First, you clear any libraries that may have been declared earlier
12457: (you don't need them for @code{sin}); then you add the @code{m}
12458: library (actually @code{libm.so} or somesuch) to the currently
12459: declared libraries; you can add as many as you need. Finally you
12460: declare the function as shown above. Typically you will use the same
12461: set of library declarations for many function declarations; you need
12462: to write only one set for that, right at the beginning.
12463:
12464: Note that you must not call @code{clear-libs} inside
12465: @code{c-library...end-c-library}; however, @code{c-library} performs
12466: the function of @code{clear-libs}, so @code{clear-libs} is not
12467: necessary, and you usually want to put @code{add-lib} calls inside
12468: @code{c-library...end-c-library}.
12469:
12470: doc-clear-libs
12471: doc-add-lib
12472:
12473:
12474: @node Callbacks, C interface internals, Declaring OS-level libraries, C Interface
12475: @subsection Callbacks
12476: @cindex Callback functions written in Forth
12477: @cindex C function pointers to Forth words
12478:
12479: Callbacks are not yet supported by the documented C interface. You
12480: can use the undocumented @file{lib.fs} interface for callbacks.
12481:
12482: In some cases you have to pass a function pointer to a C function,
12483: i.e., the library wants to call back to your application (and the
12484: pointed-to function is called a callback function). You can pass the
12485: address of an existing C function (that you get with @code{lib-sym},
12486: @pxref{Low-Level C Interface Words}), but if there is no appropriate C
12487: function, you probably want to define the function as a Forth word.
12488:
12489: @c I don't understand the existing callback interface from the example - anton
12490:
12491:
12492: @c > > Und dann gibt's noch die fptr-Deklaration, die einem
12493: @c > > C-Funktionspointer entspricht (Deklaration gleich wie bei
12494: @c > > Library-Funktionen, nur ohne den C-Namen, Aufruf mit der
12495: @c > > C-Funktionsadresse auf dem TOS).
12496: @c >
12497: @c > Ja, da bin ich dann ausgestiegen, weil ich aus dem Beispiel nicht
12498: @c > gesehen habe, wozu das gut ist.
12499: @c
12500: @c Irgendwie muss ich den Callback ja testen. Und es soll ja auch
12501: @c vorkommen, dass man von irgendwelchen kranken Interfaces einen
12502: @c Funktionspointer übergeben bekommt, den man dann bei Gelegenheit
12503: @c aufrufen muss. Also kann man den deklarieren, und das damit deklarierte
12504: @c Wort verhält sich dann wie ein EXECUTE für alle C-Funktionen mit
12505: @c demselben Prototyp.
12506:
12507:
12508: @node C interface internals, Low-Level C Interface Words, Callbacks, C Interface
12509: @subsection How the C interface works
12510:
12511: The documented C interface works by generating a C code out of the
12512: declarations.
12513:
12514: In particular, for every Forth word declared with @code{c-function},
12515: it generates a wrapper function in C that takes the Forth data from
12516: the Forth stacks, and calls the target C function with these data as
12517: arguments. The C compiler then performs an implicit conversion
12518: between the Forth type from the stack, and the C type for the
12519: parameter, which is given by the C function prototype. After the C
12520: function returns, the return value is likewise implicitly converted to
12521: a Forth type and written back on the stack.
12522:
12523: The @code{\c} lines are literally included in the C code (but without
12524: the @code{\c}), and provide the necessary declarations so that the C
12525: compiler knows the C types and has enough information to perform the
12526: conversion.
12527:
12528: These wrapper functions are eventually compiled and dynamically linked
12529: into Gforth, and then they can be called.
12530:
12531: The libraries added with @code{add-lib} are used in the compile
12532: command line to specify dependent libraries with @code{-l@var{lib}},
12533: causing these libraries to be dynamically linked when the wrapper
12534: function is linked.
12535:
12536:
12537: @node Low-Level C Interface Words, , C interface internals, C Interface
12538: @subsection Low-Level C Interface Words
12539:
12540: doc-open-lib
12541: doc-lib-sym
12542: doc-lib-error
12543: doc-call-c
12544:
12545: @c -------------------------------------------------------------
12546: @node Assembler and Code Words, Threading Words, C Interface, Words
12547: @section Assembler and Code Words
12548: @cindex assembler
12549: @cindex code words
12550:
12551: @menu
12552: * Code and ;code::
12553: * Common Assembler:: Assembler Syntax
12554: * Common Disassembler::
12555: * 386 Assembler:: Deviations and special cases
12556: * Alpha Assembler:: Deviations and special cases
12557: * MIPS assembler:: Deviations and special cases
12558: * PowerPC assembler:: Deviations and special cases
12559: * ARM Assembler:: Deviations and special cases
12560: * Other assemblers:: How to write them
12561: @end menu
12562:
12563: @node Code and ;code, Common Assembler, Assembler and Code Words, Assembler and Code Words
12564: @subsection @code{Code} and @code{;code}
12565:
12566: Gforth provides some words for defining primitives (words written in
12567: machine code), and for defining the machine-code equivalent of
12568: @code{DOES>}-based defining words. However, the machine-independent
12569: nature of Gforth poses a few problems: First of all, Gforth runs on
12570: several architectures, so it can provide no standard assembler. What's
12571: worse is that the register allocation not only depends on the processor,
12572: but also on the @code{gcc} version and options used.
12573:
12574: The words that Gforth offers encapsulate some system dependences (e.g.,
12575: the header structure), so a system-independent assembler may be used in
12576: Gforth. If you do not have an assembler, you can compile machine code
12577: directly with @code{,} and @code{c,}@footnote{This isn't portable,
12578: because these words emit stuff in @i{data} space; it works because
12579: Gforth has unified code/data spaces. Assembler isn't likely to be
12580: portable anyway.}.
12581:
12582:
12583: doc-assembler
12584: doc-init-asm
12585: doc-code
12586: doc-end-code
12587: doc-;code
12588: doc-flush-icache
12589:
12590:
12591: If @code{flush-icache} does not work correctly, @code{code} words
12592: etc. will not work (reliably), either.
12593:
12594: The typical usage of these @code{code} words can be shown most easily by
12595: analogy to the equivalent high-level defining words:
12596:
12597: @example
12598: : foo code foo
12599: <high-level Forth words> <assembler>
12600: ; end-code
12601:
12602: : bar : bar
12603: <high-level Forth words> <high-level Forth words>
12604: CREATE CREATE
12605: <high-level Forth words> <high-level Forth words>
12606: DOES> ;code
12607: <high-level Forth words> <assembler>
12608: ; end-code
12609: @end example
12610:
12611: @c anton: the following stuff is also in "Common Assembler", in less detail.
12612:
12613: @cindex registers of the inner interpreter
12614: In the assembly code you will want to refer to the inner interpreter's
12615: registers (e.g., the data stack pointer) and you may want to use other
12616: registers for temporary storage. Unfortunately, the register allocation
12617: is installation-dependent.
12618:
12619: In particular, @code{ip} (Forth instruction pointer) and @code{rp}
12620: (return stack pointer) may be in different places in @code{gforth} and
12621: @code{gforth-fast}, or different installations. This means that you
12622: cannot write a @code{NEXT} routine that works reliably on both versions
12623: or different installations; so for doing @code{NEXT}, I recommend
12624: jumping to @code{' noop >code-address}, which contains nothing but a
12625: @code{NEXT}.
12626:
12627: For general accesses to the inner interpreter's registers, the easiest
12628: solution is to use explicit register declarations (@pxref{Explicit Reg
12629: Vars, , Variables in Specified Registers, gcc.info, GNU C Manual}) for
12630: all of the inner interpreter's registers: You have to compile Gforth
12631: with @code{-DFORCE_REG} (configure option @code{--enable-force-reg}) and
12632: the appropriate declarations must be present in the @code{machine.h}
12633: file (see @code{mips.h} for an example; you can find a full list of all
12634: declarable register symbols with @code{grep register engine.c}). If you
12635: give explicit registers to all variables that are declared at the
12636: beginning of @code{engine()}, you should be able to use the other
12637: caller-saved registers for temporary storage. Alternatively, you can use
12638: the @code{gcc} option @code{-ffixed-REG} (@pxref{Code Gen Options, ,
12639: Options for Code Generation Conventions, gcc.info, GNU C Manual}) to
12640: reserve a register (however, this restriction on register allocation may
12641: slow Gforth significantly).
12642:
12643: If this solution is not viable (e.g., because @code{gcc} does not allow
12644: you to explicitly declare all the registers you need), you have to find
12645: out by looking at the code where the inner interpreter's registers
12646: reside and which registers can be used for temporary storage. You can
12647: get an assembly listing of the engine's code with @code{make engine.s}.
12648:
12649: In any case, it is good practice to abstract your assembly code from the
12650: actual register allocation. E.g., if the data stack pointer resides in
12651: register @code{$17}, create an alias for this register called @code{sp},
12652: and use that in your assembly code.
12653:
12654: @cindex code words, portable
12655: Another option for implementing normal and defining words efficiently
12656: is to add the desired functionality to the source of Gforth. For normal
12657: words you just have to edit @file{primitives} (@pxref{Automatic
12658: Generation}). Defining words (equivalent to @code{;CODE} words, for fast
12659: defined words) may require changes in @file{engine.c}, @file{kernel.fs},
12660: @file{prims2x.fs}, and possibly @file{cross.fs}.
12661:
12662: @node Common Assembler, Common Disassembler, Code and ;code, Assembler and Code Words
12663: @subsection Common Assembler
12664:
12665: The assemblers in Gforth generally use a postfix syntax, i.e., the
12666: instruction name follows the operands.
12667:
12668: The operands are passed in the usual order (the same that is used in the
12669: manual of the architecture). Since they all are Forth words, they have
12670: to be separated by spaces; you can also use Forth words to compute the
12671: operands.
12672:
12673: The instruction names usually end with a @code{,}. This makes it easier
12674: to visually separate instructions if you put several of them on one
12675: line; it also avoids shadowing other Forth words (e.g., @code{and}).
12676:
12677: Registers are usually specified by number; e.g., (decimal) @code{11}
12678: specifies registers R11 and F11 on the Alpha architecture (which one,
12679: depends on the instruction). The usual names are also available, e.g.,
12680: @code{s2} for R11 on Alpha.
12681:
12682: Control flow is specified similar to normal Forth code (@pxref{Arbitrary
12683: control structures}), with @code{if,}, @code{ahead,}, @code{then,},
12684: @code{begin,}, @code{until,}, @code{again,}, @code{cs-roll},
12685: @code{cs-pick}, @code{else,}, @code{while,}, and @code{repeat,}. The
12686: conditions are specified in a way specific to each assembler.
12687:
12688: Note that the register assignments of the Gforth engine can change
12689: between Gforth versions, or even between different compilations of the
12690: same Gforth version (e.g., if you use a different GCC version). So if
12691: you want to refer to Gforth's registers (e.g., the stack pointer or
12692: TOS), I recommend defining your own words for refering to these
12693: registers, and using them later on; then you can easily adapt to a
12694: changed register assignment. The stability of the register assignment
12695: is usually better if you build Gforth with @code{--enable-force-reg}.
12696:
12697: The most common use of these registers is to dispatch to the next word
12698: (the @code{next} routine). A portable way to do this is to jump to
12699: @code{' noop >code-address} (of course, this is less efficient than
12700: integrating the @code{next} code and scheduling it well).
12701:
12702: Another difference between Gforth version is that the top of stack is
12703: kept in memory in @code{gforth} and, on most platforms, in a register in
12704: @code{gforth-fast}.
12705:
12706: @node Common Disassembler, 386 Assembler, Common Assembler, Assembler and Code Words
12707: @subsection Common Disassembler
12708: @cindex disassembler, general
12709: @cindex gdb disassembler
12710:
12711: You can disassemble a @code{code} word with @code{see}
12712: (@pxref{Debugging}). You can disassemble a section of memory with
12713:
12714: doc-discode
12715:
12716: There are two kinds of disassembler for Gforth: The Forth disassembler
12717: (available on some CPUs) and the gdb disassembler (available on
12718: platforms with @command{gdb} and @command{mktemp}). If both are
12719: available, the Forth disassembler is used by default. If you prefer
12720: the gdb disassembler, say
12721:
12722: @example
12723: ' disasm-gdb is discode
12724: @end example
12725:
12726: If neither is available, @code{discode} performs @code{dump}.
12727:
12728: The Forth disassembler generally produces output that can be fed into the
12729: assembler (i.e., same syntax, etc.). It also includes additional
12730: information in comments. In particular, the address of the instruction
12731: is given in a comment before the instruction.
12732:
12733: The gdb disassembler produces output in the same format as the gdb
12734: @code{disassemble} command (@pxref{Machine Code,,Source and machine
12735: code,gdb,Debugging with GDB}), in the default flavour (AT&T syntax for
12736: the 386 and AMD64 architectures).
12737:
12738: @code{See} may display more or less than the actual code of the word,
12739: because the recognition of the end of the code is unreliable. You can
12740: use @code{discode} if it did not display enough. It may display more, if
12741: the code word is not immediately followed by a named word. If you have
12742: something else there, you can follow the word with @code{align latest ,}
12743: to ensure that the end is recognized.
12744:
12745: @node 386 Assembler, Alpha Assembler, Common Disassembler, Assembler and Code Words
12746: @subsection 386 Assembler
12747:
12748: The 386 assembler included in Gforth was written by Bernd Paysan, it's
12749: available under GPL, and originally part of bigFORTH.
12750:
12751: The 386 disassembler included in Gforth was written by Andrew McKewan
12752: and is in the public domain.
12753:
12754: The disassembler displays code in an Intel-like prefix syntax.
12755:
12756: The assembler uses a postfix syntax with reversed parameters.
12757:
12758: The assembler includes all instruction of the Athlon, i.e. 486 core
12759: instructions, Pentium and PPro extensions, floating point, MMX, 3Dnow!,
12760: but not ISSE. It's an integrated 16- and 32-bit assembler. Default is 32
12761: bit, you can switch to 16 bit with .86 and back to 32 bit with .386.
12762:
12763: There are several prefixes to switch between different operation sizes,
12764: @code{.b} for byte accesses, @code{.w} for word accesses, @code{.d} for
12765: double-word accesses. Addressing modes can be switched with @code{.wa}
12766: for 16 bit addresses, and @code{.da} for 32 bit addresses. You don't
12767: need a prefix for byte register names (@code{AL} et al).
12768:
12769: For floating point operations, the prefixes are @code{.fs} (IEEE
12770: single), @code{.fl} (IEEE double), @code{.fx} (extended), @code{.fw}
12771: (word), @code{.fd} (double-word), and @code{.fq} (quad-word).
12772:
12773: The MMX opcodes don't have size prefixes, they are spelled out like in
12774: the Intel assembler. Instead of move from and to memory, there are
12775: PLDQ/PLDD and PSTQ/PSTD.
12776:
12777: The registers lack the 'e' prefix; even in 32 bit mode, eax is called
12778: ax. Immediate values are indicated by postfixing them with @code{#},
12779: e.g., @code{3 #}. Here are some examples of addressing modes in various
12780: syntaxes:
12781:
12782: @example
12783: Gforth Intel (NASM) AT&T (gas) Name
12784: .w ax ax %ax register (16 bit)
12785: ax eax %eax register (32 bit)
12786: 3 # offset 3 $3 immediate
12787: 1000 #) byte ptr 1000 1000 displacement
12788: bx ) [ebx] (%ebx) base
12789: 100 di d) 100[edi] 100(%edi) base+displacement
12790: 20 ax *4 i#) 20[eax*4] 20(,%eax,4) (index*scale)+displacement
12791: di ax *4 i) [edi][eax*4] (%edi,%eax,4) base+(index*scale)
12792: 4 bx cx di) 4[ebx][ecx] 4(%ebx,%ecx) base+index+displacement
12793: 12 sp ax *2 di) 12[esp][eax*2] 12(%esp,%eax,2) base+(index*scale)+displacement
12794: @end example
12795:
12796: You can use @code{L)} and @code{LI)} instead of @code{D)} and
12797: @code{DI)} to enforce 32-bit displacement fields (useful for
12798: later patching).
12799:
12800: Some example of instructions are:
12801:
12802: @example
12803: ax bx mov \ move ebx,eax
12804: 3 # ax mov \ mov eax,3
12805: 100 di d) ax mov \ mov eax,100[edi]
12806: 4 bx cx di) ax mov \ mov eax,4[ebx][ecx]
12807: .w ax bx mov \ mov bx,ax
12808: @end example
12809:
12810: The following forms are supported for binary instructions:
12811:
12812: @example
12813: <reg> <reg> <inst>
12814: <n> # <reg> <inst>
12815: <mem> <reg> <inst>
12816: <reg> <mem> <inst>
12817: <n> # <mem> <inst>
12818: @end example
12819:
12820: The shift/rotate syntax is:
12821:
12822: @example
12823: <reg/mem> 1 # shl \ shortens to shift without immediate
12824: <reg/mem> 4 # shl
12825: <reg/mem> cl shl
12826: @end example
12827:
12828: Precede string instructions (@code{movs} etc.) with @code{.b} to get
12829: the byte version.
12830:
12831: The control structure words @code{IF} @code{UNTIL} etc. must be preceded
12832: by one of these conditions: @code{vs vc u< u>= 0= 0<> u<= u> 0< 0>= ps
12833: pc < >= <= >}. (Note that most of these words shadow some Forth words
12834: when @code{assembler} is in front of @code{forth} in the search path,
12835: e.g., in @code{code} words). Currently the control structure words use
12836: one stack item, so you have to use @code{roll} instead of @code{cs-roll}
12837: to shuffle them (you can also use @code{swap} etc.).
12838:
12839: Here is an example of a @code{code} word (assumes that the stack pointer
12840: is in esi and the TOS is in ebx):
12841:
12842: @example
12843: code my+ ( n1 n2 -- n )
12844: 4 si D) bx add
12845: 4 # si add
12846: Next
12847: end-code
12848: @end example
12849:
12850:
12851: @node Alpha Assembler, MIPS assembler, 386 Assembler, Assembler and Code Words
12852: @subsection Alpha Assembler
12853:
12854: The Alpha assembler and disassembler were originally written by Bernd
12855: Thallner.
12856:
12857: The register names @code{a0}--@code{a5} are not available to avoid
12858: shadowing hex numbers.
12859:
12860: Immediate forms of arithmetic instructions are distinguished by a
12861: @code{#} just before the @code{,}, e.g., @code{and#,} (note: @code{lda,}
12862: does not count as arithmetic instruction).
12863:
12864: You have to specify all operands to an instruction, even those that
12865: other assemblers consider optional, e.g., the destination register for
12866: @code{br,}, or the destination register and hint for @code{jmp,}.
12867:
12868: You can specify conditions for @code{if,} by removing the first @code{b}
12869: and the trailing @code{,} from a branch with a corresponding name; e.g.,
12870:
12871: @example
12872: 11 fgt if, \ if F11>0e
12873: ...
12874: endif,
12875: @end example
12876:
12877: @code{fbgt,} gives @code{fgt}.
12878:
12879: @node MIPS assembler, PowerPC assembler, Alpha Assembler, Assembler and Code Words
12880: @subsection MIPS assembler
12881:
12882: The MIPS assembler was originally written by Christian Pirker.
12883:
12884: Currently the assembler and disassembler only cover the MIPS-I
12885: architecture (R3000), and don't support FP instructions.
12886:
12887: The register names @code{$a0}--@code{$a3} are not available to avoid
12888: shadowing hex numbers.
12889:
12890: Because there is no way to distinguish registers from immediate values,
12891: you have to explicitly use the immediate forms of instructions, i.e.,
12892: @code{addiu,}, not just @code{addu,} (@command{as} does this
12893: implicitly).
12894:
12895: If the architecture manual specifies several formats for the instruction
12896: (e.g., for @code{jalr,}), you usually have to use the one with more
12897: arguments (i.e., two for @code{jalr,}). When in doubt, see
12898: @code{arch/mips/testasm.fs} for an example of correct use.
12899:
12900: Branches and jumps in the MIPS architecture have a delay slot. You have
12901: to fill it yourself (the simplest way is to use @code{nop,}), the
12902: assembler does not do it for you (unlike @command{as}). Even
12903: @code{if,}, @code{ahead,}, @code{until,}, @code{again,}, @code{while,},
12904: @code{else,} and @code{repeat,} need a delay slot. Since @code{begin,}
12905: and @code{then,} just specify branch targets, they are not affected.
12906:
12907: Note that you must not put branches, jumps, or @code{li,} into the delay
12908: slot: @code{li,} may expand to several instructions, and control flow
12909: instructions may not be put into the branch delay slot in any case.
12910:
12911: For branches the argument specifying the target is a relative address;
12912: You have to add the address of the delay slot to get the absolute
12913: address.
12914:
12915: The MIPS architecture also has load delay slots and restrictions on
12916: using @code{mfhi,} and @code{mflo,}; you have to order the instructions
12917: yourself to satisfy these restrictions, the assembler does not do it for
12918: you.
12919:
12920: You can specify the conditions for @code{if,} etc. by taking a
12921: conditional branch and leaving away the @code{b} at the start and the
12922: @code{,} at the end. E.g.,
12923:
12924: @example
12925: 4 5 eq if,
12926: ... \ do something if $4 equals $5
12927: then,
12928: @end example
12929:
12930:
12931: @node PowerPC assembler, ARM Assembler, MIPS assembler, Assembler and Code Words
12932: @subsection PowerPC assembler
12933:
12934: The PowerPC assembler and disassembler were contributed by Michal
12935: Revucky.
12936:
12937: This assembler does not follow the convention of ending mnemonic names
12938: with a ``,'', so some mnemonic names shadow regular Forth words (in
12939: particular: @code{and or xor fabs}); so if you want to use the Forth
12940: words, you have to make them visible first, e.g., with @code{also
12941: forth}.
12942:
12943: Registers are referred to by their number, e.g., @code{9} means the
12944: integer register 9 or the FP register 9 (depending on the
12945: instruction).
12946:
12947: Because there is no way to distinguish registers from immediate values,
12948: you have to explicitly use the immediate forms of instructions, i.e.,
12949: @code{addi,}, not just @code{add,}.
12950:
12951: The assembler and disassembler usually support the most general form
12952: of an instruction, but usually not the shorter forms (especially for
12953: branches).
12954:
12955:
12956: @node ARM Assembler, Other assemblers, PowerPC assembler, Assembler and Code Words
12957: @subsection ARM Assembler
12958:
12959: The ARM assembler included in Gforth was written from scratch by David
12960: Kuehling.
12961:
12962: The assembler includes all instruction of ARM architecture version 4,
12963: but does not (yet) have support for Thumb instructions. It also lacks
12964: support for any co-processors.
12965:
12966: The assembler uses a postfix syntax with the target operand specified
12967: last. For load/store instructions the last operand will be the
12968: register(s) to be loaded from/stored to.
12969:
12970: Registers are specified by their names @code{r0} through @code{r15},
12971: with the aliases @code{pc}, @code{lr}, @code{sp}, @code{ip} and
12972: @code{fp} provided for convenience. Note that @code{ip} means intra
12973: procedure call scratch register (@code{r12}) and does not refer to the
12974: instruction pointer.
12975:
12976: Condition codes can be specified anywhere in the instruction, but will
12977: be most readable if specified just in front of the mnemonic. The 'S'
12978: flag is not a separate word, but encoded into instruction mnemonics,
12979: ie. just use @code{adds,} instead of @code{add,} if you want the
12980: status register to be updated.
12981:
12982: The following table lists the syntax of operands for general
12983: instructions:
12984:
12985: @example
12986: Gforth normal assembler description
12987: 123 # #123 immediate
12988: r12 r12 register
12989: r12 4 #LSL r12, LSL #4 shift left by immediate
12990: r12 r1 #LSL r12, LSL r1 shift left by register
12991: r12 4 #LSR r12, LSR #4 shift right by immediate
12992: r12 r1 #LSR r12, LSR r1 shift right by register
12993: r12 4 #ASR r12, ASR #4 arithmetic shift right
12994: r12 r1 #ASR r12, ASR r1 ... by register
12995: r12 4 #ROR r12, ROR #4 rotate right by immediate
12996: r12 r1 #ROR r12, ROR r1 ... by register
12997: r12 RRX r12, RRX rotate right with extend by 1
12998: @end example
12999:
13000: Memory operand syntax is listed in this table:
13001:
13002: @example
13003: Gforth normal assembler description
13004: r4 ] [r4] register
13005: r4 4 #] [r4, #+4] register with immediate offset
13006: r4 -4 #] [r4, #-4] with negative offset
13007: r4 r1 +] [r4, +r1] register with register offset
13008: r4 r1 -] [r4, -r1] with negated register offset
13009: r4 r1 2 #LSL -] [r4, -r1, LSL #2] with negated and shifted offset
13010: r4 4 #]! [r4, #+4]! immediate preincrement
13011: r4 r1 +]! [r4, +r1]! register preincrement
13012: r4 r1 -]! [r4, +r1]! register predecrement
13013: r4 r1 2 #LSL +]! [r4, +r1, LSL #2]! shifted preincrement
13014: r4 -4 ]# [r4], #-4 immediate postdecrement
13015: r4 r1 ]+ [r4], r1 register postincrement
13016: r4 r1 ]- [r4], -r1 register postdecrement
13017: r4 r1 2 #LSL ]- [r4], -r1, LSL #2 shifted postdecrement
13018: ' xyz >body [#] xyz PC-relative addressing
13019: @end example
13020:
13021: Register lists for load/store multiple instructions are started and
13022: terminated by using the words @code{@{} and @code{@}}
13023: respectivly. Between braces, register names can be listed one by one,
13024: or register ranges can be formed by using the postfix operator
13025: @code{r-r}. The @code{^} flag is not encoded in the register list
13026: operand, but instead directly encoded into the instruction mnemonic,
13027: ie. use @code{^ldm,} and @code{^stm,}.
13028:
13029: Addressing modes for load/store multiple are not encoded as
13030: instruction suffixes, but instead specified after the register that
13031: supplies the address. Use one of @code{DA}, @code{IA}, @code{DB},
13032: @code{IB}, @code{DA!}, @code{IA!}, @code{DB!} or @code{IB!}.
13033:
13034: The following table gives some examples:
13035:
13036: @example
13037: Gforth normal assembler
13038: @{ r0 r7 r8 @} r4 ia stm, stmia @{r0,r7,r8@}, r4
13039: @{ r0 r7 r8 @} r4 db! ldm, ldmdb @{r0,r7,r8@}, r4!
13040: @{ r0 r15 r-r @} sp ia! ^ldm, ldmfd @{r0-r15@}^, sp!
13041: @end example
13042:
13043: Conditions for control structure words are specified in front of a
13044: word:
13045:
13046: @example
13047: r1 r2 cmp, \ compare r1 and r2
13048: eq if, \ equal?
13049: ... \ code executed if r1 == r2
13050: then,
13051: @end example
13052:
13053: Here is an example of a @code{code} word (assumes that the stack
13054: pointer is in @code{r9}, and that @code{r2} and @code{r3} can be
13055: clobbered):
13056:
13057: @example
13058: code my+ ( n1 n2 -- n3 )
13059: r9 IA! @{ r2 r3 @} ldm, \ pop r2 = n2, r3 = n1
13060: r2 r3 r3 add, \ r3 = n2+n1
13061: r9 -4 #]! r3 str, \ push r3
13062: next,
13063: end-code
13064: @end example
13065:
13066: Look at @file{arch/arm/asm-example.fs} for more examples.
13067:
13068: @node Other assemblers, , ARM Assembler, Assembler and Code Words
13069: @subsection Other assemblers
13070:
13071: If you want to contribute another assembler/disassembler, please contact
13072: us (@email{anton@@mips.complang.tuwien.ac.at}) to check if we have such
13073: an assembler already. If you are writing them from scratch, please use
13074: a similar syntax style as the one we use (i.e., postfix, commas at the
13075: end of the instruction names, @pxref{Common Assembler}); make the output
13076: of the disassembler be valid input for the assembler, and keep the style
13077: similar to the style we used.
13078:
13079: Hints on implementation: The most important part is to have a good test
13080: suite that contains all instructions. Once you have that, the rest is
13081: easy. For actual coding you can take a look at
13082: @file{arch/mips/disasm.fs} to get some ideas on how to use data for both
13083: the assembler and disassembler, avoiding redundancy and some potential
13084: bugs. You can also look at that file (and @pxref{Advanced does> usage
13085: example}) to get ideas how to factor a disassembler.
13086:
13087: Start with the disassembler, because it's easier to reuse data from the
13088: disassembler for the assembler than the other way round.
13089:
13090: For the assembler, take a look at @file{arch/alpha/asm.fs}, which shows
13091: how simple it can be.
13092:
13093:
13094:
13095:
13096: @c -------------------------------------------------------------
13097: @node Threading Words, Passing Commands to the OS, Assembler and Code Words, Words
13098: @section Threading Words
13099: @cindex threading words
13100:
13101: @cindex code address
13102: These words provide access to code addresses and other threading stuff
13103: in Gforth (and, possibly, other interpretive Forths). It more or less
13104: abstracts away the differences between direct and indirect threading
13105: (and, for direct threading, the machine dependences). However, at
13106: present this wordset is still incomplete. It is also pretty low-level;
13107: some day it will hopefully be made unnecessary by an internals wordset
13108: that abstracts implementation details away completely.
13109:
13110: The terminology used here stems from indirect threaded Forth systems; in
13111: such a system, the XT of a word is represented by the CFA (code field
13112: address) of a word; the CFA points to a cell that contains the code
13113: address. The code address is the address of some machine code that
13114: performs the run-time action of invoking the word (e.g., the
13115: @code{dovar:} routine pushes the address of the body of the word (a
13116: variable) on the stack
13117: ).
13118:
13119: @cindex code address
13120: @cindex code field address
13121: In an indirect threaded Forth, you can get the code address of @i{name}
13122: with @code{' @i{name} @@}; in Gforth you can get it with @code{' @i{name}
13123: >code-address}, independent of the threading method.
13124:
13125: doc-threading-method
13126: doc->code-address
13127: doc-code-address!
13128:
13129: @cindex @code{does>}-handler
13130: @cindex @code{does>}-code
13131: For a word defined with @code{DOES>}, the code address usually points to
13132: a jump instruction (the @dfn{does-handler}) that jumps to the dodoes
13133: routine (in Gforth on some platforms, it can also point to the dodoes
13134: routine itself). What you are typically interested in, though, is
13135: whether a word is a @code{DOES>}-defined word, and what Forth code it
13136: executes; @code{>does-code} tells you that.
13137:
13138: doc->does-code
13139:
13140: To create a @code{DOES>}-defined word with the following basic words,
13141: you have to set up a @code{DOES>}-handler with @code{does-handler!};
13142: @code{/does-handler} aus behind you have to place your executable Forth
13143: code. Finally you have to create a word and modify its behaviour with
13144: @code{does-handler!}.
13145:
13146: doc-does-code!
13147: doc-does-handler!
13148: doc-/does-handler
13149:
13150: The code addresses produced by various defining words are produced by
13151: the following words:
13152:
13153: doc-docol:
13154: doc-docon:
13155: doc-dovar:
13156: doc-douser:
13157: doc-dodefer:
13158: doc-dofield:
13159:
13160: @cindex definer
13161: The following two words generalize @code{>code-address},
13162: @code{>does-code}, @code{code-address!}, and @code{does-code!}:
13163:
13164: doc->definer
13165: doc-definer!
13166:
13167: @c -------------------------------------------------------------
13168: @node Passing Commands to the OS, Keeping track of Time, Threading Words, Words
13169: @section Passing Commands to the Operating System
13170: @cindex operating system - passing commands
13171: @cindex shell commands
13172:
13173: Gforth allows you to pass an arbitrary string to the host operating
13174: system shell (if such a thing exists) for execution.
13175:
13176: doc-sh
13177: doc-system
13178: doc-$?
13179: doc-getenv
13180:
13181: @c -------------------------------------------------------------
13182: @node Keeping track of Time, Miscellaneous Words, Passing Commands to the OS, Words
13183: @section Keeping track of Time
13184: @cindex time-related words
13185:
13186: doc-ms
13187: doc-time&date
13188: doc-utime
13189: doc-cputime
13190:
13191:
13192: @c -------------------------------------------------------------
13193: @node Miscellaneous Words, , Keeping track of Time, Words
13194: @section Miscellaneous Words
13195: @cindex miscellaneous words
13196:
13197: @comment TODO find homes for these
13198:
13199: These section lists the ANS Forth words that are not documented
13200: elsewhere in this manual. Ultimately, they all need proper homes.
13201:
13202: doc-quit
13203:
13204: The following ANS Forth words are not currently supported by Gforth
13205: (@pxref{ANS conformance}):
13206:
13207: @code{EDITOR}
13208: @code{EMIT?}
13209: @code{FORGET}
13210:
13211: @c ******************************************************************
13212: @node Error messages, Tools, Words, Top
13213: @chapter Error messages
13214: @cindex error messages
13215: @cindex backtrace
13216:
13217: A typical Gforth error message looks like this:
13218:
13219: @example
13220: in file included from \evaluated string/:-1
13221: in file included from ./yyy.fs:1
13222: ./xxx.fs:4: Invalid memory address
13223: >>>bar<<<
13224: Backtrace:
13225: $400E664C @@
13226: $400E6664 foo
13227: @end example
13228:
13229: The message identifying the error is @code{Invalid memory address}. The
13230: error happened when text-interpreting line 4 of the file
13231: @file{./xxx.fs}. This line is given (it contains @code{bar}), and the
13232: word on the line where the error happened, is pointed out (with
13233: @code{>>>} and @code{<<<}).
13234:
13235: The file containing the error was included in line 1 of @file{./yyy.fs},
13236: and @file{yyy.fs} was included from a non-file (in this case, by giving
13237: @file{yyy.fs} as command-line parameter to Gforth).
13238:
13239: At the end of the error message you find a return stack dump that can be
13240: interpreted as a backtrace (possibly empty). On top you find the top of
13241: the return stack when the @code{throw} happened, and at the bottom you
13242: find the return stack entry just above the return stack of the topmost
13243: text interpreter.
13244:
13245: To the right of most return stack entries you see a guess for the word
13246: that pushed that return stack entry as its return address. This gives a
13247: backtrace. In our case we see that @code{bar} called @code{foo}, and
13248: @code{foo} called @code{@@} (and @code{@@} had an @emph{Invalid memory
13249: address} exception).
13250:
13251: Note that the backtrace is not perfect: We don't know which return stack
13252: entries are return addresses (so we may get false positives); and in
13253: some cases (e.g., for @code{abort"}) we cannot determine from the return
13254: address the word that pushed the return address, so for some return
13255: addresses you see no names in the return stack dump.
13256:
13257: @cindex @code{catch} and backtraces
13258: The return stack dump represents the return stack at the time when a
13259: specific @code{throw} was executed. In programs that make use of
13260: @code{catch}, it is not necessarily clear which @code{throw} should be
13261: used for the return stack dump (e.g., consider one @code{throw} that
13262: indicates an error, which is caught, and during recovery another error
13263: happens; which @code{throw} should be used for the stack dump?).
13264: Gforth presents the return stack dump for the first @code{throw} after
13265: the last executed (not returned-to) @code{catch} or @code{nothrow};
13266: this works well in the usual case. To get the right backtrace, you
13267: usually want to insert @code{nothrow} or @code{['] false catch drop}
13268: after a @code{catch} if the error is not rethrown.
13269:
13270: @cindex @code{gforth-fast} and backtraces
13271: @cindex @code{gforth-fast}, difference from @code{gforth}
13272: @cindex backtraces with @code{gforth-fast}
13273: @cindex return stack dump with @code{gforth-fast}
13274: @code{Gforth} is able to do a return stack dump for throws generated
13275: from primitives (e.g., invalid memory address, stack empty etc.);
13276: @code{gforth-fast} is only able to do a return stack dump from a
13277: directly called @code{throw} (including @code{abort} etc.). Given an
13278: exception caused by a primitive in @code{gforth-fast}, you will
13279: typically see no return stack dump at all; however, if the exception is
13280: caught by @code{catch} (e.g., for restoring some state), and then
13281: @code{throw}n again, the return stack dump will be for the first such
13282: @code{throw}.
13283:
13284: @c ******************************************************************
13285: @node Tools, ANS conformance, Error messages, Top
13286: @chapter Tools
13287:
13288: @menu
13289: * ANS Report:: Report the words used, sorted by wordset.
13290: * Stack depth changes:: Where does this stack item come from?
13291: @end menu
13292:
13293: See also @ref{Emacs and Gforth}.
13294:
13295: @node ANS Report, Stack depth changes, Tools, Tools
13296: @section @file{ans-report.fs}: Report the words used, sorted by wordset
13297: @cindex @file{ans-report.fs}
13298: @cindex report the words used in your program
13299: @cindex words used in your program
13300:
13301: If you want to label a Forth program as ANS Forth Program, you must
13302: document which wordsets the program uses; for extension wordsets, it is
13303: helpful to list the words the program requires from these wordsets
13304: (because Forth systems are allowed to provide only some words of them).
13305:
13306: The @file{ans-report.fs} tool makes it easy for you to determine which
13307: words from which wordset and which non-ANS words your application
13308: uses. You simply have to include @file{ans-report.fs} before loading the
13309: program you want to check. After loading your program, you can get the
13310: report with @code{print-ans-report}. A typical use is to run this as
13311: batch job like this:
13312: @example
13313: gforth ans-report.fs myprog.fs -e "print-ans-report bye"
13314: @end example
13315:
13316: The output looks like this (for @file{compat/control.fs}):
13317: @example
13318: The program uses the following words
13319: from CORE :
13320: : POSTPONE THEN ; immediate ?dup IF 0=
13321: from BLOCK-EXT :
13322: \
13323: from FILE :
13324: (
13325: @end example
13326:
13327: @subsection Caveats
13328:
13329: Note that @file{ans-report.fs} just checks which words are used, not whether
13330: they are used in an ANS Forth conforming way!
13331:
13332: Some words are defined in several wordsets in the
13333: standard. @file{ans-report.fs} reports them for only one of the
13334: wordsets, and not necessarily the one you expect. It depends on usage
13335: which wordset is the right one to specify. E.g., if you only use the
13336: compilation semantics of @code{S"}, it is a Core word; if you also use
13337: its interpretation semantics, it is a File word.
13338:
13339:
13340: @node Stack depth changes, , ANS Report, Tools
13341: @section Stack depth changes during interpretation
13342: @cindex @file{depth-changes.fs}
13343: @cindex depth changes during interpretation
13344: @cindex stack depth changes during interpretation
13345: @cindex items on the stack after interpretation
13346:
13347: Sometimes you notice that, after loading a file, there are items left
13348: on the stack. The tool @file{depth-changes.fs} helps you find out
13349: quickly where in the file these stack items are coming from.
13350:
13351: The simplest way of using @file{depth-changes.fs} is to include it
13352: before the file(s) you want to check, e.g.:
13353:
13354: @example
13355: gforth depth-changes.fs my-file.fs
13356: @end example
13357:
13358: This will compare the stack depths of the data and FP stack at every
13359: empty line (in interpretation state) against these depths at the last
13360: empty line (in interpretation state). If the depths are not equal,
13361: the position in the file and the stack contents are printed with
13362: @code{~~} (@pxref{Debugging}). This indicates that a stack depth
13363: change has occured in the paragraph of non-empty lines before the
13364: indicated line. It is a good idea to leave an empty line at the end
13365: of the file, so the last paragraph is checked, too.
13366:
13367: Checking only at empty lines usually works well, but sometimes you
13368: have big blocks of non-empty lines (e.g., when building a big table),
13369: and you want to know where in this block the stack depth changed. You
13370: can check all interpreted lines with
13371:
13372: @example
13373: gforth depth-changes.fs -e "' all-lines is depth-changes-filter" my-file.fs
13374: @end example
13375:
13376: This checks the stack depth at every end-of-line. So the depth change
13377: occured in the line reported by the @code{~~} (not in the line
13378: before).
13379:
13380: Note that, while this offers better accuracy in indicating where the
13381: stack depth changes, it will often report many intentional stack depth
13382: changes (e.g., when an interpreted computation stretches across
13383: several lines). You can suppress the checking of some lines by
13384: putting backslashes at the end of these lines (not followed by white
13385: space), and using
13386:
13387: @example
13388: gforth depth-changes.fs -e "' most-lines is depth-changes-filter" my-file.fs
13389: @end example
13390:
13391: @c ******************************************************************
13392: @node ANS conformance, Standard vs Extensions, Tools, Top
13393: @chapter ANS conformance
13394: @cindex ANS conformance of Gforth
13395:
13396: To the best of our knowledge, Gforth is an
13397:
13398: ANS Forth System
13399: @itemize @bullet
13400: @item providing the Core Extensions word set
13401: @item providing the Block word set
13402: @item providing the Block Extensions word set
13403: @item providing the Double-Number word set
13404: @item providing the Double-Number Extensions word set
13405: @item providing the Exception word set
13406: @item providing the Exception Extensions word set
13407: @item providing the Facility word set
13408: @item providing @code{EKEY}, @code{EKEY>CHAR}, @code{EKEY?}, @code{MS} and @code{TIME&DATE} from the Facility Extensions word set
13409: @item providing the File Access word set
13410: @item providing the File Access Extensions word set
13411: @item providing the Floating-Point word set
13412: @item providing the Floating-Point Extensions word set
13413: @item providing the Locals word set
13414: @item providing the Locals Extensions word set
13415: @item providing the Memory-Allocation word set
13416: @item providing the Memory-Allocation Extensions word set (that one's easy)
13417: @item providing the Programming-Tools word set
13418: @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
13419: @item providing the Search-Order word set
13420: @item providing the Search-Order Extensions word set
13421: @item providing the String word set
13422: @item providing the String Extensions word set (another easy one)
13423: @end itemize
13424:
13425: Gforth has the following environmental restrictions:
13426:
13427: @cindex environmental restrictions
13428: @itemize @bullet
13429: @item
13430: While processing the OS command line, if an exception is not caught,
13431: Gforth exits with a non-zero exit code instyead of performing QUIT.
13432:
13433: @item
13434: When an @code{throw} is performed after a @code{query}, Gforth does not
13435: allways restore the input source specification in effect at the
13436: corresponding catch.
13437:
13438: @end itemize
13439:
13440:
13441: @cindex system documentation
13442: In addition, ANS Forth systems are required to document certain
13443: implementation choices. This chapter tries to meet these
13444: requirements. In many cases it gives a way to ask the system for the
13445: information instead of providing the information directly, in
13446: particular, if the information depends on the processor, the operating
13447: system or the installation options chosen, or if they are likely to
13448: change during the maintenance of Gforth.
13449:
13450: @comment The framework for the rest has been taken from pfe.
13451:
13452: @menu
13453: * The Core Words::
13454: * The optional Block word set::
13455: * The optional Double Number word set::
13456: * The optional Exception word set::
13457: * The optional Facility word set::
13458: * The optional File-Access word set::
13459: * The optional Floating-Point word set::
13460: * The optional Locals word set::
13461: * The optional Memory-Allocation word set::
13462: * The optional Programming-Tools word set::
13463: * The optional Search-Order word set::
13464: @end menu
13465:
13466:
13467: @c =====================================================================
13468: @node The Core Words, The optional Block word set, ANS conformance, ANS conformance
13469: @comment node-name, next, previous, up
13470: @section The Core Words
13471: @c =====================================================================
13472: @cindex core words, system documentation
13473: @cindex system documentation, core words
13474:
13475: @menu
13476: * core-idef:: Implementation Defined Options
13477: * core-ambcond:: Ambiguous Conditions
13478: * core-other:: Other System Documentation
13479: @end menu
13480:
13481: @c ---------------------------------------------------------------------
13482: @node core-idef, core-ambcond, The Core Words, The Core Words
13483: @subsection Implementation Defined Options
13484: @c ---------------------------------------------------------------------
13485: @cindex core words, implementation-defined options
13486: @cindex implementation-defined options, core words
13487:
13488:
13489: @table @i
13490: @item (Cell) aligned addresses:
13491: @cindex cell-aligned addresses
13492: @cindex aligned addresses
13493: processor-dependent. Gforth's alignment words perform natural alignment
13494: (e.g., an address aligned for a datum of size 8 is divisible by
13495: 8). Unaligned accesses usually result in a @code{-23 THROW}.
13496:
13497: @item @code{EMIT} and non-graphic characters:
13498: @cindex @code{EMIT} and non-graphic characters
13499: @cindex non-graphic characters and @code{EMIT}
13500: The character is output using the C library function (actually, macro)
13501: @code{putc}.
13502:
13503: @item character editing of @code{ACCEPT} and @code{EXPECT}:
13504: @cindex character editing of @code{ACCEPT} and @code{EXPECT}
13505: @cindex editing in @code{ACCEPT} and @code{EXPECT}
13506: @cindex @code{ACCEPT}, editing
13507: @cindex @code{EXPECT}, editing
13508: This is modeled on the GNU readline library (@pxref{Readline
13509: Interaction, , Command Line Editing, readline, The GNU Readline
13510: Library}) with Emacs-like key bindings. @kbd{Tab} deviates a little by
13511: producing a full word completion every time you type it (instead of
13512: producing the common prefix of all completions). @xref{Command-line editing}.
13513:
13514: @item character set:
13515: @cindex character set
13516: The character set of your computer and display device. Gforth is
13517: 8-bit-clean (but some other component in your system may make trouble).
13518:
13519: @item Character-aligned address requirements:
13520: @cindex character-aligned address requirements
13521: installation-dependent. Currently a character is represented by a C
13522: @code{unsigned char}; in the future we might switch to @code{wchar_t}
13523: (Comments on that requested).
13524:
13525: @item character-set extensions and matching of names:
13526: @cindex character-set extensions and matching of names
13527: @cindex case-sensitivity for name lookup
13528: @cindex name lookup, case-sensitivity
13529: @cindex locale and case-sensitivity
13530: Any character except the ASCII NUL character can be used in a
13531: name. Matching is case-insensitive (except in @code{TABLE}s). The
13532: matching is performed using the C library function @code{strncasecmp}, whose
13533: function is probably influenced by the locale. E.g., the @code{C} locale
13534: does not know about accents and umlauts, so they are matched
13535: case-sensitively in that locale. For portability reasons it is best to
13536: write programs such that they work in the @code{C} locale. Then one can
13537: use libraries written by a Polish programmer (who might use words
13538: containing ISO Latin-2 encoded characters) and by a French programmer
13539: (ISO Latin-1) in the same program (of course, @code{WORDS} will produce
13540: funny results for some of the words (which ones, depends on the font you
13541: are using)). Also, the locale you prefer may not be available in other
13542: operating systems. Hopefully, Unicode will solve these problems one day.
13543:
13544: @item conditions under which control characters match a space delimiter:
13545: @cindex space delimiters
13546: @cindex control characters as delimiters
13547: If @code{word} is called with the space character as a delimiter, all
13548: white-space characters (as identified by the C macro @code{isspace()})
13549: are delimiters. @code{Parse}, on the other hand, treats space like other
13550: delimiters. @code{Parse-name}, which is used by the outer
13551: interpreter (aka text interpreter) by default, treats all white-space
13552: characters as delimiters.
13553:
13554: @item format of the control-flow stack:
13555: @cindex control-flow stack, format
13556: The data stack is used as control-flow stack. The size of a control-flow
13557: stack item in cells is given by the constant @code{cs-item-size}. At the
13558: time of this writing, an item consists of a (pointer to a) locals list
13559: (third), an address in the code (second), and a tag for identifying the
13560: item (TOS). The following tags are used: @code{defstart},
13561: @code{live-orig}, @code{dead-orig}, @code{dest}, @code{do-dest},
13562: @code{scopestart}.
13563:
13564: @item conversion of digits > 35
13565: @cindex digits > 35
13566: The characters @code{[\]^_'} are the digits with the decimal value
13567: 36@minus{}41. There is no way to input many of the larger digits.
13568:
13569: @item display after input terminates in @code{ACCEPT} and @code{EXPECT}:
13570: @cindex @code{EXPECT}, display after end of input
13571: @cindex @code{ACCEPT}, display after end of input
13572: The cursor is moved to the end of the entered string. If the input is
13573: terminated using the @kbd{Return} key, a space is typed.
13574:
13575: @item exception abort sequence of @code{ABORT"}:
13576: @cindex exception abort sequence of @code{ABORT"}
13577: @cindex @code{ABORT"}, exception abort sequence
13578: The error string is stored into the variable @code{"error} and a
13579: @code{-2 throw} is performed.
13580:
13581: @item input line terminator:
13582: @cindex input line terminator
13583: @cindex line terminator on input
13584: @cindex newline character on input
13585: For interactive input, @kbd{C-m} (CR) and @kbd{C-j} (LF) terminate
13586: lines. One of these characters is typically produced when you type the
13587: @kbd{Enter} or @kbd{Return} key.
13588:
13589: @item maximum size of a counted string:
13590: @cindex maximum size of a counted string
13591: @cindex counted string, maximum size
13592: @code{s" /counted-string" environment? drop .}. Currently 255 characters
13593: on all platforms, but this may change.
13594:
13595: @item maximum size of a parsed string:
13596: @cindex maximum size of a parsed string
13597: @cindex parsed string, maximum size
13598: Given by the constant @code{/line}. Currently 255 characters.
13599:
13600: @item maximum size of a definition name, in characters:
13601: @cindex maximum size of a definition name, in characters
13602: @cindex name, maximum length
13603: MAXU/8
13604:
13605: @item maximum string length for @code{ENVIRONMENT?}, in characters:
13606: @cindex maximum string length for @code{ENVIRONMENT?}, in characters
13607: @cindex @code{ENVIRONMENT?} string length, maximum
13608: MAXU/8
13609:
13610: @item method of selecting the user input device:
13611: @cindex user input device, method of selecting
13612: The user input device is the standard input. There is currently no way to
13613: change it from within Gforth. However, the input can typically be
13614: redirected in the command line that starts Gforth.
13615:
13616: @item method of selecting the user output device:
13617: @cindex user output device, method of selecting
13618: @code{EMIT} and @code{TYPE} output to the file-id stored in the value
13619: @code{outfile-id} (@code{stdout} by default). Gforth uses unbuffered
13620: output when the user output device is a terminal, otherwise the output
13621: is buffered.
13622:
13623: @item methods of dictionary compilation:
13624: What are we expected to document here?
13625:
13626: @item number of bits in one address unit:
13627: @cindex number of bits in one address unit
13628: @cindex address unit, size in bits
13629: @code{s" address-units-bits" environment? drop .}. 8 in all current
13630: platforms.
13631:
13632: @item number representation and arithmetic:
13633: @cindex number representation and arithmetic
13634: Processor-dependent. Binary two's complement on all current platforms.
13635:
13636: @item ranges for integer types:
13637: @cindex ranges for integer types
13638: @cindex integer types, ranges
13639: Installation-dependent. Make environmental queries for @code{MAX-N},
13640: @code{MAX-U}, @code{MAX-D} and @code{MAX-UD}. The lower bounds for
13641: unsigned (and positive) types is 0. The lower bound for signed types on
13642: two's complement and one's complement machines machines can be computed
13643: by adding 1 to the upper bound.
13644:
13645: @item read-only data space regions:
13646: @cindex read-only data space regions
13647: @cindex data-space, read-only regions
13648: The whole Forth data space is writable.
13649:
13650: @item size of buffer at @code{WORD}:
13651: @cindex size of buffer at @code{WORD}
13652: @cindex @code{WORD} buffer size
13653: @code{PAD HERE - .}. 104 characters on 32-bit machines. The buffer is
13654: shared with the pictured numeric output string. If overwriting
13655: @code{PAD} is acceptable, it is as large as the remaining dictionary
13656: space, although only as much can be sensibly used as fits in a counted
13657: string.
13658:
13659: @item size of one cell in address units:
13660: @cindex cell size
13661: @code{1 cells .}.
13662:
13663: @item size of one character in address units:
13664: @cindex char size
13665: @code{1 chars .}. 1 on all current platforms.
13666:
13667: @item size of the keyboard terminal buffer:
13668: @cindex size of the keyboard terminal buffer
13669: @cindex terminal buffer, size
13670: Varies. You can determine the size at a specific time using @code{lp@@
13671: tib - .}. It is shared with the locals stack and TIBs of files that
13672: include the current file. You can change the amount of space for TIBs
13673: and locals stack at Gforth startup with the command line option
13674: @code{-l}.
13675:
13676: @item size of the pictured numeric output buffer:
13677: @cindex size of the pictured numeric output buffer
13678: @cindex pictured numeric output buffer, size
13679: @code{PAD HERE - .}. 104 characters on 32-bit machines. The buffer is
13680: shared with @code{WORD}.
13681:
13682: @item size of the scratch area returned by @code{PAD}:
13683: @cindex size of the scratch area returned by @code{PAD}
13684: @cindex @code{PAD} size
13685: The remainder of dictionary space. @code{unused pad here - - .}.
13686:
13687: @item system case-sensitivity characteristics:
13688: @cindex case-sensitivity characteristics
13689: Dictionary searches are case-insensitive (except in
13690: @code{TABLE}s). However, as explained above under @i{character-set
13691: extensions}, the matching for non-ASCII characters is determined by the
13692: locale you are using. In the default @code{C} locale all non-ASCII
13693: characters are matched case-sensitively.
13694:
13695: @item system prompt:
13696: @cindex system prompt
13697: @cindex prompt
13698: @code{ ok} in interpret state, @code{ compiled} in compile state.
13699:
13700: @item division rounding:
13701: @cindex division rounding
13702: The ordinary division words @code{/ mod /mod */ */mod} perform floored
13703: division (with the default installation of Gforth). You can check
13704: this with @code{s" floored" environment? drop .}. If you write
13705: programs that need a specific division rounding, best use
13706: @code{fm/mod} or @code{sm/rem} for portability.
13707:
13708: @item values of @code{STATE} when true:
13709: @cindex @code{STATE} values
13710: -1.
13711:
13712: @item values returned after arithmetic overflow:
13713: On two's complement machines, arithmetic is performed modulo
13714: 2**bits-per-cell for single arithmetic and 4**bits-per-cell for double
13715: arithmetic (with appropriate mapping for signed types). Division by
13716: zero typically results in a @code{-55 throw} (Floating-point
13717: unidentified fault) or @code{-10 throw} (divide by zero). Integer
13718: division overflow can result in these throws, or in @code{-11 throw};
13719: in @code{gforth-fast} division overflow and divide by zero may also
13720: result in returning bogus results without producing an exception.
13721:
13722: @item whether the current definition can be found after @t{DOES>}:
13723: @cindex @t{DOES>}, visibility of current definition
13724: No.
13725:
13726: @end table
13727:
13728: @c ---------------------------------------------------------------------
13729: @node core-ambcond, core-other, core-idef, The Core Words
13730: @subsection Ambiguous conditions
13731: @c ---------------------------------------------------------------------
13732: @cindex core words, ambiguous conditions
13733: @cindex ambiguous conditions, core words
13734:
13735: @table @i
13736:
13737: @item a name is neither a word nor a number:
13738: @cindex name not found
13739: @cindex undefined word
13740: @code{-13 throw} (Undefined word).
13741:
13742: @item a definition name exceeds the maximum length allowed:
13743: @cindex word name too long
13744: @code{-19 throw} (Word name too long)
13745:
13746: @item addressing a region not inside the various data spaces of the forth system:
13747: @cindex Invalid memory address
13748: The stacks, code space and header space are accessible. Machine code space is
13749: typically readable. Accessing other addresses gives results dependent on
13750: the operating system. On decent systems: @code{-9 throw} (Invalid memory
13751: address).
13752:
13753: @item argument type incompatible with parameter:
13754: @cindex argument type mismatch
13755: This is usually not caught. Some words perform checks, e.g., the control
13756: flow words, and issue a @code{ABORT"} or @code{-12 THROW} (Argument type
13757: mismatch).
13758:
13759: @item attempting to obtain the execution token of a word with undefined execution semantics:
13760: @cindex Interpreting a compile-only word, for @code{'} etc.
13761: @cindex execution token of words with undefined execution semantics
13762: @code{-14 throw} (Interpreting a compile-only word). In some cases, you
13763: get an execution token for @code{compile-only-error} (which performs a
13764: @code{-14 throw} when executed).
13765:
13766: @item dividing by zero:
13767: @cindex dividing by zero
13768: @cindex floating point unidentified fault, integer division
13769: On some platforms, this produces a @code{-10 throw} (Division by
13770: zero); on other systems, this typically results in a @code{-55 throw}
13771: (Floating-point unidentified fault).
13772:
13773: @item insufficient data stack or return stack space:
13774: @cindex insufficient data stack or return stack space
13775: @cindex stack overflow
13776: @cindex address alignment exception, stack overflow
13777: @cindex Invalid memory address, stack overflow
13778: Depending on the operating system, the installation, and the invocation
13779: of Gforth, this is either checked by the memory management hardware, or
13780: it is not checked. If it is checked, you typically get a @code{-3 throw}
13781: (Stack overflow), @code{-5 throw} (Return stack overflow), or @code{-9
13782: throw} (Invalid memory address) (depending on the platform and how you
13783: achieved the overflow) as soon as the overflow happens. If it is not
13784: checked, overflows typically result in mysterious illegal memory
13785: accesses, producing @code{-9 throw} (Invalid memory address) or
13786: @code{-23 throw} (Address alignment exception); they might also destroy
13787: the internal data structure of @code{ALLOCATE} and friends, resulting in
13788: various errors in these words.
13789:
13790: @item insufficient space for loop control parameters:
13791: @cindex insufficient space for loop control parameters
13792: Like other return stack overflows.
13793:
13794: @item insufficient space in the dictionary:
13795: @cindex insufficient space in the dictionary
13796: @cindex dictionary overflow
13797: If you try to allot (either directly with @code{allot}, or indirectly
13798: with @code{,}, @code{create} etc.) more memory than available in the
13799: dictionary, you get a @code{-8 throw} (Dictionary overflow). If you try
13800: to access memory beyond the end of the dictionary, the results are
13801: similar to stack overflows.
13802:
13803: @item interpreting a word with undefined interpretation semantics:
13804: @cindex interpreting a word with undefined interpretation semantics
13805: @cindex Interpreting a compile-only word
13806: For some words, we have defined interpretation semantics. For the
13807: others: @code{-14 throw} (Interpreting a compile-only word).
13808:
13809: @item modifying the contents of the input buffer or a string literal:
13810: @cindex modifying the contents of the input buffer or a string literal
13811: These are located in writable memory and can be modified.
13812:
13813: @item overflow of the pictured numeric output string:
13814: @cindex overflow of the pictured numeric output string
13815: @cindex pictured numeric output string, overflow
13816: @code{-17 throw} (Pictured numeric ouput string overflow).
13817:
13818: @item parsed string overflow:
13819: @cindex parsed string overflow
13820: @code{PARSE} cannot overflow. @code{WORD} does not check for overflow.
13821:
13822: @item producing a result out of range:
13823: @cindex result out of range
13824: On two's complement machines, arithmetic is performed modulo
13825: 2**bits-per-cell for single arithmetic and 4**bits-per-cell for double
13826: arithmetic (with appropriate mapping for signed types). Division by
13827: zero typically results in a @code{-10 throw} (divide by zero) or
13828: @code{-55 throw} (floating point unidentified fault). Overflow on
13829: division may result in these errors or in @code{-11 throw} (result out
13830: of range). @code{Gforth-fast} may silently produce bogus results on
13831: division overflow or division by zero. @code{Convert} and
13832: @code{>number} currently overflow silently.
13833:
13834: @item reading from an empty data or return stack:
13835: @cindex stack empty
13836: @cindex stack underflow
13837: @cindex return stack underflow
13838: The data stack is checked by the outer (aka text) interpreter after
13839: every word executed. If it has underflowed, a @code{-4 throw} (Stack
13840: underflow) is performed. Apart from that, stacks may be checked or not,
13841: depending on operating system, installation, and invocation. If they are
13842: caught by a check, they typically result in @code{-4 throw} (Stack
13843: underflow), @code{-6 throw} (Return stack underflow) or @code{-9 throw}
13844: (Invalid memory address), depending on the platform and which stack
13845: underflows and by how much. Note that even if the system uses checking
13846: (through the MMU), your program may have to underflow by a significant
13847: number of stack items to trigger the reaction (the reason for this is
13848: that the MMU, and therefore the checking, works with a page-size
13849: granularity). If there is no checking, the symptoms resulting from an
13850: underflow are similar to those from an overflow. Unbalanced return
13851: stack errors can result in a variety of symptoms, including @code{-9 throw}
13852: (Invalid memory address) and Illegal Instruction (typically @code{-260
13853: throw}).
13854:
13855: @item unexpected end of the input buffer, resulting in an attempt to use a zero-length string as a name:
13856: @cindex unexpected end of the input buffer
13857: @cindex zero-length string as a name
13858: @cindex Attempt to use zero-length string as a name
13859: @code{Create} and its descendants perform a @code{-16 throw} (Attempt to
13860: use zero-length string as a name). Words like @code{'} probably will not
13861: find what they search. Note that it is possible to create zero-length
13862: names with @code{nextname} (should it not?).
13863:
13864: @item @code{>IN} greater than input buffer:
13865: @cindex @code{>IN} greater than input buffer
13866: The next invocation of a parsing word returns a string with length 0.
13867:
13868: @item @code{RECURSE} appears after @code{DOES>}:
13869: @cindex @code{RECURSE} appears after @code{DOES>}
13870: Compiles a recursive call to the defining word, not to the defined word.
13871:
13872: @item argument input source different than current input source for @code{RESTORE-INPUT}:
13873: @cindex argument input source different than current input source for @code{RESTORE-INPUT}
13874: @cindex argument type mismatch, @code{RESTORE-INPUT}
13875: @cindex @code{RESTORE-INPUT}, Argument type mismatch
13876: @code{-12 THROW}. Note that, once an input file is closed (e.g., because
13877: the end of the file was reached), its source-id may be
13878: reused. Therefore, restoring an input source specification referencing a
13879: closed file may lead to unpredictable results instead of a @code{-12
13880: THROW}.
13881:
13882: In the future, Gforth may be able to restore input source specifications
13883: from other than the current input source.
13884:
13885: @item data space containing definitions gets de-allocated:
13886: @cindex data space containing definitions gets de-allocated
13887: Deallocation with @code{allot} is not checked. This typically results in
13888: memory access faults or execution of illegal instructions.
13889:
13890: @item data space read/write with incorrect alignment:
13891: @cindex data space read/write with incorrect alignment
13892: @cindex alignment faults
13893: @cindex address alignment exception
13894: Processor-dependent. Typically results in a @code{-23 throw} (Address
13895: alignment exception). Under Linux-Intel on a 486 or later processor with
13896: alignment turned on, incorrect alignment results in a @code{-9 throw}
13897: (Invalid memory address). There are reportedly some processors with
13898: alignment restrictions that do not report violations.
13899:
13900: @item data space pointer not properly aligned, @code{,}, @code{C,}:
13901: @cindex data space pointer not properly aligned, @code{,}, @code{C,}
13902: Like other alignment errors.
13903:
13904: @item less than u+2 stack items (@code{PICK} and @code{ROLL}):
13905: Like other stack underflows.
13906:
13907: @item loop control parameters not available:
13908: @cindex loop control parameters not available
13909: Not checked. The counted loop words simply assume that the top of return
13910: stack items are loop control parameters and behave accordingly.
13911:
13912: @item most recent definition does not have a name (@code{IMMEDIATE}):
13913: @cindex most recent definition does not have a name (@code{IMMEDIATE})
13914: @cindex last word was headerless
13915: @code{abort" last word was headerless"}.
13916:
13917: @item name not defined by @code{VALUE} used by @code{TO}:
13918: @cindex name not defined by @code{VALUE} used by @code{TO}
13919: @cindex @code{TO} on non-@code{VALUE}s
13920: @cindex Invalid name argument, @code{TO}
13921: @code{-32 throw} (Invalid name argument) (unless name is a local or was
13922: defined by @code{CONSTANT}; in the latter case it just changes the constant).
13923:
13924: @item name not found (@code{'}, @code{POSTPONE}, @code{[']}, @code{[COMPILE]}):
13925: @cindex name not found (@code{'}, @code{POSTPONE}, @code{[']}, @code{[COMPILE]})
13926: @cindex undefined word, @code{'}, @code{POSTPONE}, @code{[']}, @code{[COMPILE]}
13927: @code{-13 throw} (Undefined word)
13928:
13929: @item parameters are not of the same type (@code{DO}, @code{?DO}, @code{WITHIN}):
13930: @cindex parameters are not of the same type (@code{DO}, @code{?DO}, @code{WITHIN})
13931: Gforth behaves as if they were of the same type. I.e., you can predict
13932: the behaviour by interpreting all parameters as, e.g., signed.
13933:
13934: @item @code{POSTPONE} or @code{[COMPILE]} applied to @code{TO}:
13935: @cindex @code{POSTPONE} or @code{[COMPILE]} applied to @code{TO}
13936: Assume @code{: X POSTPONE TO ; IMMEDIATE}. @code{X} performs the
13937: compilation semantics of @code{TO}.
13938:
13939: @item String longer than a counted string returned by @code{WORD}:
13940: @cindex string longer than a counted string returned by @code{WORD}
13941: @cindex @code{WORD}, string overflow
13942: Not checked. The string will be ok, but the count will, of course,
13943: contain only the least significant bits of the length.
13944:
13945: @item u greater than or equal to the number of bits in a cell (@code{LSHIFT}, @code{RSHIFT}):
13946: @cindex @code{LSHIFT}, large shift counts
13947: @cindex @code{RSHIFT}, large shift counts
13948: Processor-dependent. Typical behaviours are returning 0 and using only
13949: the low bits of the shift count.
13950:
13951: @item word not defined via @code{CREATE}:
13952: @cindex @code{>BODY} of non-@code{CREATE}d words
13953: @code{>BODY} produces the PFA of the word no matter how it was defined.
13954:
13955: @cindex @code{DOES>} of non-@code{CREATE}d words
13956: @code{DOES>} changes the execution semantics of the last defined word no
13957: matter how it was defined. E.g., @code{CONSTANT DOES>} is equivalent to
13958: @code{CREATE , DOES>}.
13959:
13960: @item words improperly used outside @code{<#} and @code{#>}:
13961: Not checked. As usual, you can expect memory faults.
13962:
13963: @end table
13964:
13965:
13966: @c ---------------------------------------------------------------------
13967: @node core-other, , core-ambcond, The Core Words
13968: @subsection Other system documentation
13969: @c ---------------------------------------------------------------------
13970: @cindex other system documentation, core words
13971: @cindex core words, other system documentation
13972:
13973: @table @i
13974: @item nonstandard words using @code{PAD}:
13975: @cindex @code{PAD} use by nonstandard words
13976: None.
13977:
13978: @item operator's terminal facilities available:
13979: @cindex operator's terminal facilities available
13980: After processing the OS's command line, Gforth goes into interactive mode,
13981: and you can give commands to Gforth interactively. The actual facilities
13982: available depend on how you invoke Gforth.
13983:
13984: @item program data space available:
13985: @cindex program data space available
13986: @cindex data space available
13987: @code{UNUSED .} gives the remaining dictionary space. The total
13988: dictionary space can be specified with the @code{-m} switch
13989: (@pxref{Invoking Gforth}) when Gforth starts up.
13990:
13991: @item return stack space available:
13992: @cindex return stack space available
13993: You can compute the total return stack space in cells with
13994: @code{s" RETURN-STACK-CELLS" environment? drop .}. You can specify it at
13995: startup time with the @code{-r} switch (@pxref{Invoking Gforth}).
13996:
13997: @item stack space available:
13998: @cindex stack space available
13999: You can compute the total data stack space in cells with
14000: @code{s" STACK-CELLS" environment? drop .}. You can specify it at
14001: startup time with the @code{-d} switch (@pxref{Invoking Gforth}).
14002:
14003: @item system dictionary space required, in address units:
14004: @cindex system dictionary space required, in address units
14005: Type @code{here forthstart - .} after startup. At the time of this
14006: writing, this gives 80080 (bytes) on a 32-bit system.
14007: @end table
14008:
14009:
14010: @c =====================================================================
14011: @node The optional Block word set, The optional Double Number word set, The Core Words, ANS conformance
14012: @section The optional Block word set
14013: @c =====================================================================
14014: @cindex system documentation, block words
14015: @cindex block words, system documentation
14016:
14017: @menu
14018: * block-idef:: Implementation Defined Options
14019: * block-ambcond:: Ambiguous Conditions
14020: * block-other:: Other System Documentation
14021: @end menu
14022:
14023:
14024: @c ---------------------------------------------------------------------
14025: @node block-idef, block-ambcond, The optional Block word set, The optional Block word set
14026: @subsection Implementation Defined Options
14027: @c ---------------------------------------------------------------------
14028: @cindex implementation-defined options, block words
14029: @cindex block words, implementation-defined options
14030:
14031: @table @i
14032: @item the format for display by @code{LIST}:
14033: @cindex @code{LIST} display format
14034: First the screen number is displayed, then 16 lines of 64 characters,
14035: each line preceded by the line number.
14036:
14037: @item the length of a line affected by @code{\}:
14038: @cindex length of a line affected by @code{\}
14039: @cindex @code{\}, line length in blocks
14040: 64 characters.
14041: @end table
14042:
14043:
14044: @c ---------------------------------------------------------------------
14045: @node block-ambcond, block-other, block-idef, The optional Block word set
14046: @subsection Ambiguous conditions
14047: @c ---------------------------------------------------------------------
14048: @cindex block words, ambiguous conditions
14049: @cindex ambiguous conditions, block words
14050:
14051: @table @i
14052: @item correct block read was not possible:
14053: @cindex block read not possible
14054: Typically results in a @code{throw} of some OS-derived value (between
14055: -512 and -2048). If the blocks file was just not long enough, blanks are
14056: supplied for the missing portion.
14057:
14058: @item I/O exception in block transfer:
14059: @cindex I/O exception in block transfer
14060: @cindex block transfer, I/O exception
14061: Typically results in a @code{throw} of some OS-derived value (between
14062: -512 and -2048).
14063:
14064: @item invalid block number:
14065: @cindex invalid block number
14066: @cindex block number invalid
14067: @code{-35 throw} (Invalid block number)
14068:
14069: @item a program directly alters the contents of @code{BLK}:
14070: @cindex @code{BLK}, altering @code{BLK}
14071: The input stream is switched to that other block, at the same
14072: position. If the storing to @code{BLK} happens when interpreting
14073: non-block input, the system will get quite confused when the block ends.
14074:
14075: @item no current block buffer for @code{UPDATE}:
14076: @cindex @code{UPDATE}, no current block buffer
14077: @code{UPDATE} has no effect.
14078:
14079: @end table
14080:
14081: @c ---------------------------------------------------------------------
14082: @node block-other, , block-ambcond, The optional Block word set
14083: @subsection Other system documentation
14084: @c ---------------------------------------------------------------------
14085: @cindex other system documentation, block words
14086: @cindex block words, other system documentation
14087:
14088: @table @i
14089: @item any restrictions a multiprogramming system places on the use of buffer addresses:
14090: No restrictions (yet).
14091:
14092: @item the number of blocks available for source and data:
14093: depends on your disk space.
14094:
14095: @end table
14096:
14097:
14098: @c =====================================================================
14099: @node The optional Double Number word set, The optional Exception word set, The optional Block word set, ANS conformance
14100: @section The optional Double Number word set
14101: @c =====================================================================
14102: @cindex system documentation, double words
14103: @cindex double words, system documentation
14104:
14105: @menu
14106: * double-ambcond:: Ambiguous Conditions
14107: @end menu
14108:
14109:
14110: @c ---------------------------------------------------------------------
14111: @node double-ambcond, , The optional Double Number word set, The optional Double Number word set
14112: @subsection Ambiguous conditions
14113: @c ---------------------------------------------------------------------
14114: @cindex double words, ambiguous conditions
14115: @cindex ambiguous conditions, double words
14116:
14117: @table @i
14118: @item @i{d} outside of range of @i{n} in @code{D>S}:
14119: @cindex @code{D>S}, @i{d} out of range of @i{n}
14120: The least significant cell of @i{d} is produced.
14121:
14122: @end table
14123:
14124:
14125: @c =====================================================================
14126: @node The optional Exception word set, The optional Facility word set, The optional Double Number word set, ANS conformance
14127: @section The optional Exception word set
14128: @c =====================================================================
14129: @cindex system documentation, exception words
14130: @cindex exception words, system documentation
14131:
14132: @menu
14133: * exception-idef:: Implementation Defined Options
14134: @end menu
14135:
14136:
14137: @c ---------------------------------------------------------------------
14138: @node exception-idef, , The optional Exception word set, The optional Exception word set
14139: @subsection Implementation Defined Options
14140: @c ---------------------------------------------------------------------
14141: @cindex implementation-defined options, exception words
14142: @cindex exception words, implementation-defined options
14143:
14144: @table @i
14145: @item @code{THROW}-codes used in the system:
14146: @cindex @code{THROW}-codes used in the system
14147: The codes -256@minus{}-511 are used for reporting signals. The mapping
14148: from OS signal numbers to throw codes is -256@minus{}@i{signal}. The
14149: codes -512@minus{}-2047 are used for OS errors (for file and memory
14150: allocation operations). The mapping from OS error numbers to throw codes
14151: is -512@minus{}@code{errno}. One side effect of this mapping is that
14152: undefined OS errors produce a message with a strange number; e.g.,
14153: @code{-1000 THROW} results in @code{Unknown error 488} on my system.
14154: @end table
14155:
14156: @c =====================================================================
14157: @node The optional Facility word set, The optional File-Access word set, The optional Exception word set, ANS conformance
14158: @section The optional Facility word set
14159: @c =====================================================================
14160: @cindex system documentation, facility words
14161: @cindex facility words, system documentation
14162:
14163: @menu
14164: * facility-idef:: Implementation Defined Options
14165: * facility-ambcond:: Ambiguous Conditions
14166: @end menu
14167:
14168:
14169: @c ---------------------------------------------------------------------
14170: @node facility-idef, facility-ambcond, The optional Facility word set, The optional Facility word set
14171: @subsection Implementation Defined Options
14172: @c ---------------------------------------------------------------------
14173: @cindex implementation-defined options, facility words
14174: @cindex facility words, implementation-defined options
14175:
14176: @table @i
14177: @item encoding of keyboard events (@code{EKEY}):
14178: @cindex keyboard events, encoding in @code{EKEY}
14179: @cindex @code{EKEY}, encoding of keyboard events
14180: Keys corresponding to ASCII characters are encoded as ASCII characters.
14181: Other keys are encoded with the constants @code{k-left}, @code{k-right},
14182: @code{k-up}, @code{k-down}, @code{k-home}, @code{k-end}, @code{k1},
14183: @code{k2}, @code{k3}, @code{k4}, @code{k5}, @code{k6}, @code{k7},
14184: @code{k8}, @code{k9}, @code{k10}, @code{k11}, @code{k12}.
14185:
14186:
14187: @item duration of a system clock tick:
14188: @cindex duration of a system clock tick
14189: @cindex clock tick duration
14190: System dependent. With respect to @code{MS}, the time is specified in
14191: microseconds. How well the OS and the hardware implement this, is
14192: another question.
14193:
14194: @item repeatability to be expected from the execution of @code{MS}:
14195: @cindex repeatability to be expected from the execution of @code{MS}
14196: @cindex @code{MS}, repeatability to be expected
14197: System dependent. On Unix, a lot depends on load. If the system is
14198: lightly loaded, and the delay is short enough that Gforth does not get
14199: swapped out, the performance should be acceptable. Under MS-DOS and
14200: other single-tasking systems, it should be good.
14201:
14202: @end table
14203:
14204:
14205: @c ---------------------------------------------------------------------
14206: @node facility-ambcond, , facility-idef, The optional Facility word set
14207: @subsection Ambiguous conditions
14208: @c ---------------------------------------------------------------------
14209: @cindex facility words, ambiguous conditions
14210: @cindex ambiguous conditions, facility words
14211:
14212: @table @i
14213: @item @code{AT-XY} can't be performed on user output device:
14214: @cindex @code{AT-XY} can't be performed on user output device
14215: Largely terminal dependent. No range checks are done on the arguments.
14216: No errors are reported. You may see some garbage appearing, you may see
14217: simply nothing happen.
14218:
14219: @end table
14220:
14221:
14222: @c =====================================================================
14223: @node The optional File-Access word set, The optional Floating-Point word set, The optional Facility word set, ANS conformance
14224: @section The optional File-Access word set
14225: @c =====================================================================
14226: @cindex system documentation, file words
14227: @cindex file words, system documentation
14228:
14229: @menu
14230: * file-idef:: Implementation Defined Options
14231: * file-ambcond:: Ambiguous Conditions
14232: @end menu
14233:
14234: @c ---------------------------------------------------------------------
14235: @node file-idef, file-ambcond, The optional File-Access word set, The optional File-Access word set
14236: @subsection Implementation Defined Options
14237: @c ---------------------------------------------------------------------
14238: @cindex implementation-defined options, file words
14239: @cindex file words, implementation-defined options
14240:
14241: @table @i
14242: @item file access methods used:
14243: @cindex file access methods used
14244: @code{R/O}, @code{R/W} and @code{BIN} work as you would
14245: expect. @code{W/O} translates into the C file opening mode @code{w} (or
14246: @code{wb}): The file is cleared, if it exists, and created, if it does
14247: not (with both @code{open-file} and @code{create-file}). Under Unix
14248: @code{create-file} creates a file with 666 permissions modified by your
14249: umask.
14250:
14251: @item file exceptions:
14252: @cindex file exceptions
14253: The file words do not raise exceptions (except, perhaps, memory access
14254: faults when you pass illegal addresses or file-ids).
14255:
14256: @item file line terminator:
14257: @cindex file line terminator
14258: System-dependent. Gforth uses C's newline character as line
14259: terminator. What the actual character code(s) of this are is
14260: system-dependent.
14261:
14262: @item file name format:
14263: @cindex file name format
14264: System dependent. Gforth just uses the file name format of your OS.
14265:
14266: @item information returned by @code{FILE-STATUS}:
14267: @cindex @code{FILE-STATUS}, returned information
14268: @code{FILE-STATUS} returns the most powerful file access mode allowed
14269: for the file: Either @code{R/O}, @code{W/O} or @code{R/W}. If the file
14270: cannot be accessed, @code{R/O BIN} is returned. @code{BIN} is applicable
14271: along with the returned mode.
14272:
14273: @item input file state after an exception when including source:
14274: @cindex exception when including source
14275: All files that are left via the exception are closed.
14276:
14277: @item @i{ior} values and meaning:
14278: @cindex @i{ior} values and meaning
14279: @cindex @i{wior} values and meaning
14280: The @i{ior}s returned by the file and memory allocation words are
14281: intended as throw codes. They typically are in the range
14282: -512@minus{}-2047 of OS errors. The mapping from OS error numbers to
14283: @i{ior}s is -512@minus{}@i{errno}.
14284:
14285: @item maximum depth of file input nesting:
14286: @cindex maximum depth of file input nesting
14287: @cindex file input nesting, maximum depth
14288: limited by the amount of return stack, locals/TIB stack, and the number
14289: of open files available. This should not give you troubles.
14290:
14291: @item maximum size of input line:
14292: @cindex maximum size of input line
14293: @cindex input line size, maximum
14294: @code{/line}. Currently 255.
14295:
14296: @item methods of mapping block ranges to files:
14297: @cindex mapping block ranges to files
14298: @cindex files containing blocks
14299: @cindex blocks in files
14300: By default, blocks are accessed in the file @file{blocks.fb} in the
14301: current working directory. The file can be switched with @code{USE}.
14302:
14303: @item number of string buffers provided by @code{S"}:
14304: @cindex @code{S"}, number of string buffers
14305: 1
14306:
14307: @item size of string buffer used by @code{S"}:
14308: @cindex @code{S"}, size of string buffer
14309: @code{/line}. currently 255.
14310:
14311: @end table
14312:
14313: @c ---------------------------------------------------------------------
14314: @node file-ambcond, , file-idef, The optional File-Access word set
14315: @subsection Ambiguous conditions
14316: @c ---------------------------------------------------------------------
14317: @cindex file words, ambiguous conditions
14318: @cindex ambiguous conditions, file words
14319:
14320: @table @i
14321: @item attempting to position a file outside its boundaries:
14322: @cindex @code{REPOSITION-FILE}, outside the file's boundaries
14323: @code{REPOSITION-FILE} is performed as usual: Afterwards,
14324: @code{FILE-POSITION} returns the value given to @code{REPOSITION-FILE}.
14325:
14326: @item attempting to read from file positions not yet written:
14327: @cindex reading from file positions not yet written
14328: End-of-file, i.e., zero characters are read and no error is reported.
14329:
14330: @item @i{file-id} is invalid (@code{INCLUDE-FILE}):
14331: @cindex @code{INCLUDE-FILE}, @i{file-id} is invalid
14332: An appropriate exception may be thrown, but a memory fault or other
14333: problem is more probable.
14334:
14335: @item I/O exception reading or closing @i{file-id} (@code{INCLUDE-FILE}, @code{INCLUDED}):
14336: @cindex @code{INCLUDE-FILE}, I/O exception reading or closing @i{file-id}
14337: @cindex @code{INCLUDED}, I/O exception reading or closing @i{file-id}
14338: The @i{ior} produced by the operation, that discovered the problem, is
14339: thrown.
14340:
14341: @item named file cannot be opened (@code{INCLUDED}):
14342: @cindex @code{INCLUDED}, named file cannot be opened
14343: The @i{ior} produced by @code{open-file} is thrown.
14344:
14345: @item requesting an unmapped block number:
14346: @cindex unmapped block numbers
14347: There are no unmapped legal block numbers. On some operating systems,
14348: writing a block with a large number may overflow the file system and
14349: have an error message as consequence.
14350:
14351: @item using @code{source-id} when @code{blk} is non-zero:
14352: @cindex @code{SOURCE-ID}, behaviour when @code{BLK} is non-zero
14353: @code{source-id} performs its function. Typically it will give the id of
14354: the source which loaded the block. (Better ideas?)
14355:
14356: @end table
14357:
14358:
14359: @c =====================================================================
14360: @node The optional Floating-Point word set, The optional Locals word set, The optional File-Access word set, ANS conformance
14361: @section The optional Floating-Point word set
14362: @c =====================================================================
14363: @cindex system documentation, floating-point words
14364: @cindex floating-point words, system documentation
14365:
14366: @menu
14367: * floating-idef:: Implementation Defined Options
14368: * floating-ambcond:: Ambiguous Conditions
14369: @end menu
14370:
14371:
14372: @c ---------------------------------------------------------------------
14373: @node floating-idef, floating-ambcond, The optional Floating-Point word set, The optional Floating-Point word set
14374: @subsection Implementation Defined Options
14375: @c ---------------------------------------------------------------------
14376: @cindex implementation-defined options, floating-point words
14377: @cindex floating-point words, implementation-defined options
14378:
14379: @table @i
14380: @item format and range of floating point numbers:
14381: @cindex format and range of floating point numbers
14382: @cindex floating point numbers, format and range
14383: System-dependent; the @code{double} type of C.
14384:
14385: @item results of @code{REPRESENT} when @i{float} is out of range:
14386: @cindex @code{REPRESENT}, results when @i{float} is out of range
14387: System dependent; @code{REPRESENT} is implemented using the C library
14388: function @code{ecvt()} and inherits its behaviour in this respect.
14389:
14390: @item rounding or truncation of floating-point numbers:
14391: @cindex rounding of floating-point numbers
14392: @cindex truncation of floating-point numbers
14393: @cindex floating-point numbers, rounding or truncation
14394: System dependent; the rounding behaviour is inherited from the hosting C
14395: compiler. IEEE-FP-based (i.e., most) systems by default round to
14396: nearest, and break ties by rounding to even (i.e., such that the last
14397: bit of the mantissa is 0).
14398:
14399: @item size of floating-point stack:
14400: @cindex floating-point stack size
14401: @code{s" FLOATING-STACK" environment? drop .} gives the total size of
14402: the floating-point stack (in floats). You can specify this on startup
14403: with the command-line option @code{-f} (@pxref{Invoking Gforth}).
14404:
14405: @item width of floating-point stack:
14406: @cindex floating-point stack width
14407: @code{1 floats}.
14408:
14409: @end table
14410:
14411:
14412: @c ---------------------------------------------------------------------
14413: @node floating-ambcond, , floating-idef, The optional Floating-Point word set
14414: @subsection Ambiguous conditions
14415: @c ---------------------------------------------------------------------
14416: @cindex floating-point words, ambiguous conditions
14417: @cindex ambiguous conditions, floating-point words
14418:
14419: @table @i
14420: @item @code{df@@} or @code{df!} used with an address that is not double-float aligned:
14421: @cindex @code{df@@} or @code{df!} used with an address that is not double-float aligned
14422: System-dependent. Typically results in a @code{-23 THROW} like other
14423: alignment violations.
14424:
14425: @item @code{f@@} or @code{f!} used with an address that is not float aligned:
14426: @cindex @code{f@@} used with an address that is not float aligned
14427: @cindex @code{f!} used with an address that is not float aligned
14428: System-dependent. Typically results in a @code{-23 THROW} like other
14429: alignment violations.
14430:
14431: @item floating-point result out of range:
14432: @cindex floating-point result out of range
14433: System-dependent. Can result in a @code{-43 throw} (floating point
14434: overflow), @code{-54 throw} (floating point underflow), @code{-41 throw}
14435: (floating point inexact result), @code{-55 THROW} (Floating-point
14436: unidentified fault), or can produce a special value representing, e.g.,
14437: Infinity.
14438:
14439: @item @code{sf@@} or @code{sf!} used with an address that is not single-float aligned:
14440: @cindex @code{sf@@} or @code{sf!} used with an address that is not single-float aligned
14441: System-dependent. Typically results in an alignment fault like other
14442: alignment violations.
14443:
14444: @item @code{base} is not decimal (@code{REPRESENT}, @code{F.}, @code{FE.}, @code{FS.}):
14445: @cindex @code{base} is not decimal (@code{REPRESENT}, @code{F.}, @code{FE.}, @code{FS.})
14446: The floating-point number is converted into decimal nonetheless.
14447:
14448: @item Both arguments are equal to zero (@code{FATAN2}):
14449: @cindex @code{FATAN2}, both arguments are equal to zero
14450: System-dependent. @code{FATAN2} is implemented using the C library
14451: function @code{atan2()}.
14452:
14453: @item Using @code{FTAN} on an argument @i{r1} where cos(@i{r1}) is zero:
14454: @cindex @code{FTAN} on an argument @i{r1} where cos(@i{r1}) is zero
14455: System-dependent. Anyway, typically the cos of @i{r1} will not be zero
14456: because of small errors and the tan will be a very large (or very small)
14457: but finite number.
14458:
14459: @item @i{d} cannot be presented precisely as a float in @code{D>F}:
14460: @cindex @code{D>F}, @i{d} cannot be presented precisely as a float
14461: The result is rounded to the nearest float.
14462:
14463: @item dividing by zero:
14464: @cindex dividing by zero, floating-point
14465: @cindex floating-point dividing by zero
14466: @cindex floating-point unidentified fault, FP divide-by-zero
14467: Platform-dependent; can produce an Infinity, NaN, @code{-42 throw}
14468: (floating point divide by zero) or @code{-55 throw} (Floating-point
14469: unidentified fault).
14470:
14471: @item exponent too big for conversion (@code{DF!}, @code{DF@@}, @code{SF!}, @code{SF@@}):
14472: @cindex exponent too big for conversion (@code{DF!}, @code{DF@@}, @code{SF!}, @code{SF@@})
14473: System dependent. On IEEE-FP based systems the number is converted into
14474: an infinity.
14475:
14476: @item @i{float}<1 (@code{FACOSH}):
14477: @cindex @code{FACOSH}, @i{float}<1
14478: @cindex floating-point unidentified fault, @code{FACOSH}
14479: Platform-dependent; on IEEE-FP systems typically produces a NaN.
14480:
14481: @item @i{float}=<-1 (@code{FLNP1}):
14482: @cindex @code{FLNP1}, @i{float}=<-1
14483: @cindex floating-point unidentified fault, @code{FLNP1}
14484: Platform-dependent; on IEEE-FP systems typically produces a NaN (or a
14485: negative infinity for @i{float}=-1).
14486:
14487: @item @i{float}=<0 (@code{FLN}, @code{FLOG}):
14488: @cindex @code{FLN}, @i{float}=<0
14489: @cindex @code{FLOG}, @i{float}=<0
14490: @cindex floating-point unidentified fault, @code{FLN} or @code{FLOG}
14491: Platform-dependent; on IEEE-FP systems typically produces a NaN (or a
14492: negative infinity for @i{float}=0).
14493:
14494: @item @i{float}<0 (@code{FASINH}, @code{FSQRT}):
14495: @cindex @code{FASINH}, @i{float}<0
14496: @cindex @code{FSQRT}, @i{float}<0
14497: @cindex floating-point unidentified fault, @code{FASINH} or @code{FSQRT}
14498: Platform-dependent; for @code{fsqrt} this typically gives a NaN, for
14499: @code{fasinh} some platforms produce a NaN, others a number (bug in the
14500: C library?).
14501:
14502: @item |@i{float}|>1 (@code{FACOS}, @code{FASIN}, @code{FATANH}):
14503: @cindex @code{FACOS}, |@i{float}|>1
14504: @cindex @code{FASIN}, |@i{float}|>1
14505: @cindex @code{FATANH}, |@i{float}|>1
14506: @cindex floating-point unidentified fault, @code{FACOS}, @code{FASIN} or @code{FATANH}
14507: Platform-dependent; IEEE-FP systems typically produce a NaN.
14508:
14509: @item integer part of float cannot be represented by @i{d} in @code{F>D}:
14510: @cindex @code{F>D}, integer part of float cannot be represented by @i{d}
14511: @cindex floating-point unidentified fault, @code{F>D}
14512: Platform-dependent; typically, some double number is produced and no
14513: error is reported.
14514:
14515: @item string larger than pictured numeric output area (@code{f.}, @code{fe.}, @code{fs.}):
14516: @cindex string larger than pictured numeric output area (@code{f.}, @code{fe.}, @code{fs.})
14517: @code{Precision} characters of the numeric output area are used. If
14518: @code{precision} is too high, these words will smash the data or code
14519: close to @code{here}.
14520: @end table
14521:
14522: @c =====================================================================
14523: @node The optional Locals word set, The optional Memory-Allocation word set, The optional Floating-Point word set, ANS conformance
14524: @section The optional Locals word set
14525: @c =====================================================================
14526: @cindex system documentation, locals words
14527: @cindex locals words, system documentation
14528:
14529: @menu
14530: * locals-idef:: Implementation Defined Options
14531: * locals-ambcond:: Ambiguous Conditions
14532: @end menu
14533:
14534:
14535: @c ---------------------------------------------------------------------
14536: @node locals-idef, locals-ambcond, The optional Locals word set, The optional Locals word set
14537: @subsection Implementation Defined Options
14538: @c ---------------------------------------------------------------------
14539: @cindex implementation-defined options, locals words
14540: @cindex locals words, implementation-defined options
14541:
14542: @table @i
14543: @item maximum number of locals in a definition:
14544: @cindex maximum number of locals in a definition
14545: @cindex locals, maximum number in a definition
14546: @code{s" #locals" environment? drop .}. Currently 15. This is a lower
14547: bound, e.g., on a 32-bit machine there can be 41 locals of up to 8
14548: characters. The number of locals in a definition is bounded by the size
14549: of locals-buffer, which contains the names of the locals.
14550:
14551: @end table
14552:
14553:
14554: @c ---------------------------------------------------------------------
14555: @node locals-ambcond, , locals-idef, The optional Locals word set
14556: @subsection Ambiguous conditions
14557: @c ---------------------------------------------------------------------
14558: @cindex locals words, ambiguous conditions
14559: @cindex ambiguous conditions, locals words
14560:
14561: @table @i
14562: @item executing a named local in interpretation state:
14563: @cindex local in interpretation state
14564: @cindex Interpreting a compile-only word, for a local
14565: Locals have no interpretation semantics. If you try to perform the
14566: interpretation semantics, you will get a @code{-14 throw} somewhere
14567: (Interpreting a compile-only word). If you perform the compilation
14568: semantics, the locals access will be compiled (irrespective of state).
14569:
14570: @item @i{name} not defined by @code{VALUE} or @code{(LOCAL)} (@code{TO}):
14571: @cindex name not defined by @code{VALUE} or @code{(LOCAL)} used by @code{TO}
14572: @cindex @code{TO} on non-@code{VALUE}s and non-locals
14573: @cindex Invalid name argument, @code{TO}
14574: @code{-32 throw} (Invalid name argument)
14575:
14576: @end table
14577:
14578:
14579: @c =====================================================================
14580: @node The optional Memory-Allocation word set, The optional Programming-Tools word set, The optional Locals word set, ANS conformance
14581: @section The optional Memory-Allocation word set
14582: @c =====================================================================
14583: @cindex system documentation, memory-allocation words
14584: @cindex memory-allocation words, system documentation
14585:
14586: @menu
14587: * memory-idef:: Implementation Defined Options
14588: @end menu
14589:
14590:
14591: @c ---------------------------------------------------------------------
14592: @node memory-idef, , The optional Memory-Allocation word set, The optional Memory-Allocation word set
14593: @subsection Implementation Defined Options
14594: @c ---------------------------------------------------------------------
14595: @cindex implementation-defined options, memory-allocation words
14596: @cindex memory-allocation words, implementation-defined options
14597:
14598: @table @i
14599: @item values and meaning of @i{ior}:
14600: @cindex @i{ior} values and meaning
14601: The @i{ior}s returned by the file and memory allocation words are
14602: intended as throw codes. They typically are in the range
14603: -512@minus{}-2047 of OS errors. The mapping from OS error numbers to
14604: @i{ior}s is -512@minus{}@i{errno}.
14605:
14606: @end table
14607:
14608: @c =====================================================================
14609: @node The optional Programming-Tools word set, The optional Search-Order word set, The optional Memory-Allocation word set, ANS conformance
14610: @section The optional Programming-Tools word set
14611: @c =====================================================================
14612: @cindex system documentation, programming-tools words
14613: @cindex programming-tools words, system documentation
14614:
14615: @menu
14616: * programming-idef:: Implementation Defined Options
14617: * programming-ambcond:: Ambiguous Conditions
14618: @end menu
14619:
14620:
14621: @c ---------------------------------------------------------------------
14622: @node programming-idef, programming-ambcond, The optional Programming-Tools word set, The optional Programming-Tools word set
14623: @subsection Implementation Defined Options
14624: @c ---------------------------------------------------------------------
14625: @cindex implementation-defined options, programming-tools words
14626: @cindex programming-tools words, implementation-defined options
14627:
14628: @table @i
14629: @item ending sequence for input following @code{;CODE} and @code{CODE}:
14630: @cindex @code{;CODE} ending sequence
14631: @cindex @code{CODE} ending sequence
14632: @code{END-CODE}
14633:
14634: @item manner of processing input following @code{;CODE} and @code{CODE}:
14635: @cindex @code{;CODE}, processing input
14636: @cindex @code{CODE}, processing input
14637: The @code{ASSEMBLER} vocabulary is pushed on the search order stack, and
14638: the input is processed by the text interpreter, (starting) in interpret
14639: state.
14640:
14641: @item search order capability for @code{EDITOR} and @code{ASSEMBLER}:
14642: @cindex @code{ASSEMBLER}, search order capability
14643: The ANS Forth search order word set.
14644:
14645: @item source and format of display by @code{SEE}:
14646: @cindex @code{SEE}, source and format of output
14647: The source for @code{see} is the executable code used by the inner
14648: interpreter. The current @code{see} tries to output Forth source code
14649: (and on some platforms, assembly code for primitives) as well as
14650: possible.
14651:
14652: @end table
14653:
14654: @c ---------------------------------------------------------------------
14655: @node programming-ambcond, , programming-idef, The optional Programming-Tools word set
14656: @subsection Ambiguous conditions
14657: @c ---------------------------------------------------------------------
14658: @cindex programming-tools words, ambiguous conditions
14659: @cindex ambiguous conditions, programming-tools words
14660:
14661: @table @i
14662:
14663: @item deleting the compilation word list (@code{FORGET}):
14664: @cindex @code{FORGET}, deleting the compilation word list
14665: Not implemented (yet).
14666:
14667: @item fewer than @i{u}+1 items on the control-flow stack (@code{CS-PICK}, @code{CS-ROLL}):
14668: @cindex @code{CS-PICK}, fewer than @i{u}+1 items on the control flow-stack
14669: @cindex @code{CS-ROLL}, fewer than @i{u}+1 items on the control flow-stack
14670: @cindex control-flow stack underflow
14671: This typically results in an @code{abort"} with a descriptive error
14672: message (may change into a @code{-22 throw} (Control structure mismatch)
14673: in the future). You may also get a memory access error. If you are
14674: unlucky, this ambiguous condition is not caught.
14675:
14676: @item @i{name} can't be found (@code{FORGET}):
14677: @cindex @code{FORGET}, @i{name} can't be found
14678: Not implemented (yet).
14679:
14680: @item @i{name} not defined via @code{CREATE}:
14681: @cindex @code{;CODE}, @i{name} not defined via @code{CREATE}
14682: @code{;CODE} behaves like @code{DOES>} in this respect, i.e., it changes
14683: the execution semantics of the last defined word no matter how it was
14684: defined.
14685:
14686: @item @code{POSTPONE} applied to @code{[IF]}:
14687: @cindex @code{POSTPONE} applied to @code{[IF]}
14688: @cindex @code{[IF]} and @code{POSTPONE}
14689: After defining @code{: X POSTPONE [IF] ; IMMEDIATE}. @code{X} is
14690: equivalent to @code{[IF]}.
14691:
14692: @item reaching the end of the input source before matching @code{[ELSE]} or @code{[THEN]}:
14693: @cindex @code{[IF]}, end of the input source before matching @code{[ELSE]} or @code{[THEN]}
14694: Continue in the same state of conditional compilation in the next outer
14695: input source. Currently there is no warning to the user about this.
14696:
14697: @item removing a needed definition (@code{FORGET}):
14698: @cindex @code{FORGET}, removing a needed definition
14699: Not implemented (yet).
14700:
14701: @end table
14702:
14703:
14704: @c =====================================================================
14705: @node The optional Search-Order word set, , The optional Programming-Tools word set, ANS conformance
14706: @section The optional Search-Order word set
14707: @c =====================================================================
14708: @cindex system documentation, search-order words
14709: @cindex search-order words, system documentation
14710:
14711: @menu
14712: * search-idef:: Implementation Defined Options
14713: * search-ambcond:: Ambiguous Conditions
14714: @end menu
14715:
14716:
14717: @c ---------------------------------------------------------------------
14718: @node search-idef, search-ambcond, The optional Search-Order word set, The optional Search-Order word set
14719: @subsection Implementation Defined Options
14720: @c ---------------------------------------------------------------------
14721: @cindex implementation-defined options, search-order words
14722: @cindex search-order words, implementation-defined options
14723:
14724: @table @i
14725: @item maximum number of word lists in search order:
14726: @cindex maximum number of word lists in search order
14727: @cindex search order, maximum depth
14728: @code{s" wordlists" environment? drop .}. Currently 16.
14729:
14730: @item minimum search order:
14731: @cindex minimum search order
14732: @cindex search order, minimum
14733: @code{root root}.
14734:
14735: @end table
14736:
14737: @c ---------------------------------------------------------------------
14738: @node search-ambcond, , search-idef, The optional Search-Order word set
14739: @subsection Ambiguous conditions
14740: @c ---------------------------------------------------------------------
14741: @cindex search-order words, ambiguous conditions
14742: @cindex ambiguous conditions, search-order words
14743:
14744: @table @i
14745: @item changing the compilation word list (during compilation):
14746: @cindex changing the compilation word list (during compilation)
14747: @cindex compilation word list, change before definition ends
14748: The word is entered into the word list that was the compilation word list
14749: at the start of the definition. Any changes to the name field (e.g.,
14750: @code{immediate}) or the code field (e.g., when executing @code{DOES>})
14751: are applied to the latest defined word (as reported by @code{latest} or
14752: @code{latestxt}), if possible, irrespective of the compilation word list.
14753:
14754: @item search order empty (@code{previous}):
14755: @cindex @code{previous}, search order empty
14756: @cindex vocstack empty, @code{previous}
14757: @code{abort" Vocstack empty"}.
14758:
14759: @item too many word lists in search order (@code{also}):
14760: @cindex @code{also}, too many word lists in search order
14761: @cindex vocstack full, @code{also}
14762: @code{abort" Vocstack full"}.
14763:
14764: @end table
14765:
14766: @c ***************************************************************
14767: @node Standard vs Extensions, Model, ANS conformance, Top
14768: @chapter Should I use Gforth extensions?
14769: @cindex Gforth extensions
14770:
14771: As you read through the rest of this manual, you will see documentation
14772: for @i{Standard} words, and documentation for some appealing Gforth
14773: @i{extensions}. You might ask yourself the question: @i{``Should I
14774: restrict myself to the standard, or should I use the extensions?''}
14775:
14776: The answer depends on the goals you have for the program you are working
14777: on:
14778:
14779: @itemize @bullet
14780:
14781: @item Is it just for yourself or do you want to share it with others?
14782:
14783: @item
14784: If you want to share it, do the others all use Gforth?
14785:
14786: @item
14787: If it is just for yourself, do you want to restrict yourself to Gforth?
14788:
14789: @end itemize
14790:
14791: If restricting the program to Gforth is ok, then there is no reason not
14792: to use extensions. It is still a good idea to keep to the standard
14793: where it is easy, in case you want to reuse these parts in another
14794: program that you want to be portable.
14795:
14796: If you want to be able to port the program to other Forth systems, there
14797: are the following points to consider:
14798:
14799: @itemize @bullet
14800:
14801: @item
14802: Most Forth systems that are being maintained support the ANS Forth
14803: standard. So if your program complies with the standard, it will be
14804: portable among many systems.
14805:
14806: @item
14807: A number of the Gforth extensions can be implemented in ANS Forth using
14808: public-domain files provided in the @file{compat/} directory. These are
14809: mentioned in the text in passing. There is no reason not to use these
14810: extensions, your program will still be ANS Forth compliant; just include
14811: the appropriate compat files with your program.
14812:
14813: @item
14814: The tool @file{ans-report.fs} (@pxref{ANS Report}) makes it easy to
14815: analyse your program and determine what non-Standard words it relies
14816: upon. However, it does not check whether you use standard words in a
14817: non-standard way.
14818:
14819: @item
14820: Some techniques are not standardized by ANS Forth, and are hard or
14821: impossible to implement in a standard way, but can be implemented in
14822: most Forth systems easily, and usually in similar ways (e.g., accessing
14823: word headers). Forth has a rich historical precedent for programmers
14824: taking advantage of implementation-dependent features of their tools
14825: (for example, relying on a knowledge of the dictionary
14826: structure). Sometimes these techniques are necessary to extract every
14827: last bit of performance from the hardware, sometimes they are just a
14828: programming shorthand.
14829:
14830: @item
14831: Does using a Gforth extension save more work than the porting this part
14832: to other Forth systems (if any) will cost?
14833:
14834: @item
14835: Is the additional functionality worth the reduction in portability and
14836: the additional porting problems?
14837:
14838: @end itemize
14839:
14840: In order to perform these consideratios, you need to know what's
14841: standard and what's not. This manual generally states if something is
14842: non-standard, but the authoritative source is the
14843: @uref{http://www.taygeta.com/forth/dpans.html,standard document}.
14844: Appendix A of the Standard (@var{Rationale}) provides a valuable insight
14845: into the thought processes of the technical committee.
14846:
14847: Note also that portability between Forth systems is not the only
14848: portability issue; there is also the issue of portability between
14849: different platforms (processor/OS combinations).
14850:
14851: @c ***************************************************************
14852: @node Model, Integrating Gforth, Standard vs Extensions, Top
14853: @chapter Model
14854:
14855: This chapter has yet to be written. It will contain information, on
14856: which internal structures you can rely.
14857:
14858: @c ***************************************************************
14859: @node Integrating Gforth, Emacs and Gforth, Model, Top
14860: @chapter Integrating Gforth into C programs
14861:
14862: This is not yet implemented.
14863:
14864: Several people like to use Forth as scripting language for applications
14865: that are otherwise written in C, C++, or some other language.
14866:
14867: The Forth system ATLAST provides facilities for embedding it into
14868: applications; unfortunately it has several disadvantages: most
14869: importantly, it is not based on ANS Forth, and it is apparently dead
14870: (i.e., not developed further and not supported). The facilities
14871: provided by Gforth in this area are inspired by ATLAST's facilities, so
14872: making the switch should not be hard.
14873:
14874: We also tried to design the interface such that it can easily be
14875: implemented by other Forth systems, so that we may one day arrive at a
14876: standardized interface. Such a standard interface would allow you to
14877: replace the Forth system without having to rewrite C code.
14878:
14879: You embed the Gforth interpreter by linking with the library
14880: @code{libgforth.a} (give the compiler the option @code{-lgforth}). All
14881: global symbols in this library that belong to the interface, have the
14882: prefix @code{forth_}. (Global symbols that are used internally have the
14883: prefix @code{gforth_}).
14884:
14885: You can include the declarations of Forth types and the functions and
14886: variables of the interface with @code{#include <forth.h>}.
14887:
14888: Types.
14889:
14890: Variables.
14891:
14892: Data and FP Stack pointer. Area sizes.
14893:
14894: functions.
14895:
14896: forth_init(imagefile)
14897: forth_evaluate(string) exceptions?
14898: forth_goto(address) (or forth_execute(xt)?)
14899: forth_continue() (a corountining mechanism)
14900:
14901: Adding primitives.
14902:
14903: No checking.
14904:
14905: Signals?
14906:
14907: Accessing the Stacks
14908:
14909: @c ******************************************************************
14910: @node Emacs and Gforth, Image Files, Integrating Gforth, Top
14911: @chapter Emacs and Gforth
14912: @cindex Emacs and Gforth
14913:
14914: @cindex @file{gforth.el}
14915: @cindex @file{forth.el}
14916: @cindex Rydqvist, Goran
14917: @cindex Kuehling, David
14918: @cindex comment editing commands
14919: @cindex @code{\}, editing with Emacs
14920: @cindex debug tracer editing commands
14921: @cindex @code{~~}, removal with Emacs
14922: @cindex Forth mode in Emacs
14923:
14924: Gforth comes with @file{gforth.el}, an improved version of
14925: @file{forth.el} by Goran Rydqvist (included in the TILE package). The
14926: improvements are:
14927:
14928: @itemize @bullet
14929: @item
14930: A better handling of indentation.
14931: @item
14932: A custom hilighting engine for Forth-code.
14933: @item
14934: Comment paragraph filling (@kbd{M-q})
14935: @item
14936: Commenting (@kbd{C-x \}) and uncommenting (@kbd{C-u C-x \}) of regions
14937: @item
14938: Removal of debugging tracers (@kbd{C-x ~}, @pxref{Debugging}).
14939: @item
14940: Support of the @code{info-lookup} feature for looking up the
14941: documentation of a word.
14942: @item
14943: Support for reading and writing blocks files.
14944: @end itemize
14945:
14946: To get a basic description of these features, enter Forth mode and
14947: type @kbd{C-h m}.
14948:
14949: @cindex source location of error or debugging output in Emacs
14950: @cindex error output, finding the source location in Emacs
14951: @cindex debugging output, finding the source location in Emacs
14952: In addition, Gforth supports Emacs quite well: The source code locations
14953: given in error messages, debugging output (from @code{~~}) and failed
14954: assertion messages are in the right format for Emacs' compilation mode
14955: (@pxref{Compilation, , Running Compilations under Emacs, emacs, Emacs
14956: Manual}) so the source location corresponding to an error or other
14957: message is only a few keystrokes away (@kbd{C-x `} for the next error,
14958: @kbd{C-c C-c} for the error under the cursor).
14959:
14960: @cindex viewing the documentation of a word in Emacs
14961: @cindex context-sensitive help
14962: Moreover, for words documented in this manual, you can look up the
14963: glossary entry quickly by using @kbd{C-h TAB}
14964: (@code{info-lookup-symbol}, @pxref{Documentation, ,Documentation
14965: Commands, emacs, Emacs Manual}). This feature requires Emacs 20.3 or
14966: later and does not work for words containing @code{:}.
14967:
14968: @menu
14969: * Installing gforth.el:: Making Emacs aware of Forth.
14970: * Emacs Tags:: Viewing the source of a word in Emacs.
14971: * Hilighting:: Making Forth code look prettier.
14972: * Auto-Indentation:: Customizing auto-indentation.
14973: * Blocks Files:: Reading and writing blocks files.
14974: @end menu
14975:
14976: @c ----------------------------------
14977: @node Installing gforth.el, Emacs Tags, Emacs and Gforth, Emacs and Gforth
14978: @section Installing gforth.el
14979: @cindex @file{.emacs}
14980: @cindex @file{gforth.el}, installation
14981: To make the features from @file{gforth.el} available in Emacs, add
14982: the following lines to your @file{.emacs} file:
14983:
14984: @example
14985: (autoload 'forth-mode "gforth.el")
14986: (setq auto-mode-alist (cons '("\\.fs\\'" . forth-mode)
14987: auto-mode-alist))
14988: (autoload 'forth-block-mode "gforth.el")
14989: (setq auto-mode-alist (cons '("\\.fb\\'" . forth-block-mode)
14990: auto-mode-alist))
14991: (add-hook 'forth-mode-hook (function (lambda ()
14992: ;; customize variables here:
14993: (setq forth-indent-level 4)
14994: (setq forth-minor-indent-level 2)
14995: (setq forth-hilight-level 3)
14996: ;;; ...
14997: )))
14998: @end example
14999:
15000: @c ----------------------------------
15001: @node Emacs Tags, Hilighting, Installing gforth.el, Emacs and Gforth
15002: @section Emacs Tags
15003: @cindex @file{TAGS} file
15004: @cindex @file{etags.fs}
15005: @cindex viewing the source of a word in Emacs
15006: @cindex @code{require}, placement in files
15007: @cindex @code{include}, placement in files
15008: If you @code{require} @file{etags.fs}, a new @file{TAGS} file will be
15009: produced (@pxref{Tags, , Tags Tables, emacs, Emacs Manual}) that
15010: contains the definitions of all words defined afterwards. You can then
15011: find the source for a word using @kbd{M-.}. Note that Emacs can use
15012: several tags files at the same time (e.g., one for the Gforth sources
15013: and one for your program, @pxref{Select Tags Table,,Selecting a Tags
15014: Table,emacs, Emacs Manual}). The TAGS file for the preloaded words is
15015: @file{$(datadir)/gforth/$(VERSION)/TAGS} (e.g.,
15016: @file{/usr/local/share/gforth/0.2.0/TAGS}). To get the best behaviour
15017: with @file{etags.fs}, you should avoid putting definitions both before
15018: and after @code{require} etc., otherwise you will see the same file
15019: visited several times by commands like @code{tags-search}.
15020:
15021: @c ----------------------------------
15022: @node Hilighting, Auto-Indentation, Emacs Tags, Emacs and Gforth
15023: @section Hilighting
15024: @cindex hilighting Forth code in Emacs
15025: @cindex highlighting Forth code in Emacs
15026: @file{gforth.el} comes with a custom source hilighting engine. When
15027: you open a file in @code{forth-mode}, it will be completely parsed,
15028: assigning faces to keywords, comments, strings etc. While you edit
15029: the file, modified regions get parsed and updated on-the-fly.
15030:
15031: Use the variable `forth-hilight-level' to change the level of
15032: decoration from 0 (no hilighting at all) to 3 (the default). Even if
15033: you set the hilighting level to 0, the parser will still work in the
15034: background, collecting information about whether regions of text are
15035: ``compiled'' or ``interpreted''. Those information are required for
15036: auto-indentation to work properly. Set `forth-disable-parser' to
15037: non-nil if your computer is too slow to handle parsing. This will
15038: have an impact on the smartness of the auto-indentation engine,
15039: though.
15040:
15041: Sometimes Forth sources define new features that should be hilighted,
15042: new control structures, defining-words etc. You can use the variable
15043: `forth-custom-words' to make @code{forth-mode} hilight additional
15044: words and constructs. See the docstring of `forth-words' for details
15045: (in Emacs, type @kbd{C-h v forth-words}).
15046:
15047: `forth-custom-words' is meant to be customized in your
15048: @file{.emacs} file. To customize hilighing in a file-specific manner,
15049: set `forth-local-words' in a local-variables section at the end of
15050: your source file (@pxref{Local Variables in Files,, Variables, emacs, Emacs Manual}).
15051:
15052: Example:
15053: @example
15054: 0 [IF]
15055: Local Variables:
15056: forth-local-words:
15057: ((("t:") definition-starter (font-lock-keyword-face . 1)
15058: "[ \t\n]" t name (font-lock-function-name-face . 3))
15059: ((";t") definition-ender (font-lock-keyword-face . 1)))
15060: End:
15061: [THEN]
15062: @end example
15063:
15064: @c ----------------------------------
15065: @node Auto-Indentation, Blocks Files, Hilighting, Emacs and Gforth
15066: @section Auto-Indentation
15067: @cindex auto-indentation of Forth code in Emacs
15068: @cindex indentation of Forth code in Emacs
15069: @code{forth-mode} automatically tries to indent lines in a smart way,
15070: whenever you type @key{TAB} or break a line with @kbd{C-m}.
15071:
15072: Simple customization can be achieved by setting
15073: `forth-indent-level' and `forth-minor-indent-level' in your
15074: @file{.emacs} file. For historical reasons @file{gforth.el} indents
15075: per default by multiples of 4 columns. To use the more traditional
15076: 3-column indentation, add the following lines to your @file{.emacs}:
15077:
15078: @example
15079: (add-hook 'forth-mode-hook (function (lambda ()
15080: ;; customize variables here:
15081: (setq forth-indent-level 3)
15082: (setq forth-minor-indent-level 1)
15083: )))
15084: @end example
15085:
15086: If you want indentation to recognize non-default words, customize it
15087: by setting `forth-custom-indent-words' in your @file{.emacs}. See the
15088: docstring of `forth-indent-words' for details (in Emacs, type @kbd{C-h
15089: v forth-indent-words}).
15090:
15091: To customize indentation in a file-specific manner, set
15092: `forth-local-indent-words' in a local-variables section at the end of
15093: your source file (@pxref{Local Variables in Files, Variables,,emacs,
15094: Emacs Manual}).
15095:
15096: Example:
15097: @example
15098: 0 [IF]
15099: Local Variables:
15100: forth-local-indent-words:
15101: ((("t:") (0 . 2) (0 . 2))
15102: ((";t") (-2 . 0) (0 . -2)))
15103: End:
15104: [THEN]
15105: @end example
15106:
15107: @c ----------------------------------
15108: @node Blocks Files, , Auto-Indentation, Emacs and Gforth
15109: @section Blocks Files
15110: @cindex blocks files, use with Emacs
15111: @code{forth-mode} Autodetects blocks files by checking whether the
15112: length of the first line exceeds 1023 characters. It then tries to
15113: convert the file into normal text format. When you save the file, it
15114: will be written to disk as normal stream-source file.
15115:
15116: If you want to write blocks files, use @code{forth-blocks-mode}. It
15117: inherits all the features from @code{forth-mode}, plus some additions:
15118:
15119: @itemize @bullet
15120: @item
15121: Files are written to disk in blocks file format.
15122: @item
15123: Screen numbers are displayed in the mode line (enumerated beginning
15124: with the value of `forth-block-base')
15125: @item
15126: Warnings are displayed when lines exceed 64 characters.
15127: @item
15128: The beginning of the currently edited block is marked with an
15129: overlay-arrow.
15130: @end itemize
15131:
15132: There are some restrictions you should be aware of. When you open a
15133: blocks file that contains tabulator or newline characters, these
15134: characters will be translated into spaces when the file is written
15135: back to disk. If tabs or newlines are encountered during blocks file
15136: reading, an error is output to the echo area. So have a look at the
15137: `*Messages*' buffer, when Emacs' bell rings during reading.
15138:
15139: Please consult the docstring of @code{forth-blocks-mode} for more
15140: information by typing @kbd{C-h v forth-blocks-mode}).
15141:
15142: @c ******************************************************************
15143: @node Image Files, Engine, Emacs and Gforth, Top
15144: @chapter Image Files
15145: @cindex image file
15146: @cindex @file{.fi} files
15147: @cindex precompiled Forth code
15148: @cindex dictionary in persistent form
15149: @cindex persistent form of dictionary
15150:
15151: An image file is a file containing an image of the Forth dictionary,
15152: i.e., compiled Forth code and data residing in the dictionary. By
15153: convention, we use the extension @code{.fi} for image files.
15154:
15155: @menu
15156: * Image Licensing Issues:: Distribution terms for images.
15157: * Image File Background:: Why have image files?
15158: * Non-Relocatable Image Files:: don't always work.
15159: * Data-Relocatable Image Files:: are better.
15160: * Fully Relocatable Image Files:: better yet.
15161: * Stack and Dictionary Sizes:: Setting the default sizes for an image.
15162: * Running Image Files:: @code{gforth -i @i{file}} or @i{file}.
15163: * Modifying the Startup Sequence:: and turnkey applications.
15164: @end menu
15165:
15166: @node Image Licensing Issues, Image File Background, Image Files, Image Files
15167: @section Image Licensing Issues
15168: @cindex license for images
15169: @cindex image license
15170:
15171: An image created with @code{gforthmi} (@pxref{gforthmi}) or
15172: @code{savesystem} (@pxref{Non-Relocatable Image Files}) includes the
15173: original image; i.e., according to copyright law it is a derived work of
15174: the original image.
15175:
15176: Since Gforth is distributed under the GNU GPL, the newly created image
15177: falls under the GNU GPL, too. In particular, this means that if you
15178: distribute the image, you have to make all of the sources for the image
15179: available, including those you wrote. For details see @ref{Copying, ,
15180: GNU General Public License (Section 3)}.
15181:
15182: If you create an image with @code{cross} (@pxref{cross.fs}), the image
15183: contains only code compiled from the sources you gave it; if none of
15184: these sources is under the GPL, the terms discussed above do not apply
15185: to the image. However, if your image needs an engine (a gforth binary)
15186: that is under the GPL, you should make sure that you distribute both in
15187: a way that is at most a @emph{mere aggregation}, if you don't want the
15188: terms of the GPL to apply to the image.
15189:
15190: @node Image File Background, Non-Relocatable Image Files, Image Licensing Issues, Image Files
15191: @section Image File Background
15192: @cindex image file background
15193:
15194: Gforth consists not only of primitives (in the engine), but also of
15195: definitions written in Forth. Since the Forth compiler itself belongs to
15196: those definitions, it is not possible to start the system with the
15197: engine and the Forth source alone. Therefore we provide the Forth
15198: code as an image file in nearly executable form. When Gforth starts up,
15199: a C routine loads the image file into memory, optionally relocates the
15200: addresses, then sets up the memory (stacks etc.) according to
15201: information in the image file, and (finally) starts executing Forth
15202: code.
15203:
15204: The default image file is @file{gforth.fi} (in the @code{GFORTHPATH}).
15205: You can use a different image by using the @code{-i},
15206: @code{--image-file} or @code{--appl-image} options (@pxref{Invoking
15207: Gforth}), e.g.:
15208:
15209: @example
15210: gforth-fast -i myimage.fi
15211: @end example
15212:
15213: There are different variants of image files, and they represent
15214: different compromises between the goals of making it easy to generate
15215: image files and making them portable.
15216:
15217: @cindex relocation at run-time
15218: Win32Forth 3.4 and Mitch Bradley's @code{cforth} use relocation at
15219: run-time. This avoids many of the complications discussed below (image
15220: files are data relocatable without further ado), but costs performance
15221: (one addition per memory access) and makes it difficult to pass
15222: addresses between Forth and library calls or other programs.
15223:
15224: @cindex relocation at load-time
15225: By contrast, the Gforth loader performs relocation at image load time. The
15226: loader also has to replace tokens that represent primitive calls with the
15227: appropriate code-field addresses (or code addresses in the case of
15228: direct threading).
15229:
15230: There are three kinds of image files, with different degrees of
15231: relocatability: non-relocatable, data-relocatable, and fully relocatable
15232: image files.
15233:
15234: @cindex image file loader
15235: @cindex relocating loader
15236: @cindex loader for image files
15237: These image file variants have several restrictions in common; they are
15238: caused by the design of the image file loader:
15239:
15240: @itemize @bullet
15241: @item
15242: There is only one segment; in particular, this means, that an image file
15243: cannot represent @code{ALLOCATE}d memory chunks (and pointers to
15244: them). The contents of the stacks are not represented, either.
15245:
15246: @item
15247: The only kinds of relocation supported are: adding the same offset to
15248: all cells that represent data addresses; and replacing special tokens
15249: with code addresses or with pieces of machine code.
15250:
15251: If any complex computations involving addresses are performed, the
15252: results cannot be represented in the image file. Several applications that
15253: use such computations come to mind:
15254:
15255: @itemize @minus
15256: @item
15257: Hashing addresses (or data structures which contain addresses) for table
15258: lookup. If you use Gforth's @code{table}s or @code{wordlist}s for this
15259: purpose, you will have no problem, because the hash tables are
15260: recomputed automatically when the system is started. If you use your own
15261: hash tables, you will have to do something similar.
15262:
15263: @item
15264: There's a cute implementation of doubly-linked lists that uses
15265: @code{XOR}ed addresses. You could represent such lists as singly-linked
15266: in the image file, and restore the doubly-linked representation on
15267: startup.@footnote{In my opinion, though, you should think thrice before
15268: using a doubly-linked list (whatever implementation).}
15269:
15270: @item
15271: The code addresses of run-time routines like @code{docol:} cannot be
15272: represented in the image file (because their tokens would be replaced by
15273: machine code in direct threaded implementations). As a workaround,
15274: compute these addresses at run-time with @code{>code-address} from the
15275: executions tokens of appropriate words (see the definitions of
15276: @code{docol:} and friends in @file{kernel/getdoers.fs}).
15277:
15278: @item
15279: On many architectures addresses are represented in machine code in some
15280: shifted or mangled form. You cannot put @code{CODE} words that contain
15281: absolute addresses in this form in a relocatable image file. Workarounds
15282: are representing the address in some relative form (e.g., relative to
15283: the CFA, which is present in some register), or loading the address from
15284: a place where it is stored in a non-mangled form.
15285: @end itemize
15286: @end itemize
15287:
15288: @node Non-Relocatable Image Files, Data-Relocatable Image Files, Image File Background, Image Files
15289: @section Non-Relocatable Image Files
15290: @cindex non-relocatable image files
15291: @cindex image file, non-relocatable
15292:
15293: These files are simple memory dumps of the dictionary. They are
15294: specific to the executable (i.e., @file{gforth} file) they were
15295: created with. What's worse, they are specific to the place on which
15296: the dictionary resided when the image was created. Now, there is no
15297: guarantee that the dictionary will reside at the same place the next
15298: time you start Gforth, so there's no guarantee that a non-relocatable
15299: image will work the next time (Gforth will complain instead of
15300: crashing, though). Indeed, on OSs with (enabled) address-space
15301: randomization non-relocatable images are unlikely to work.
15302:
15303: You can create a non-relocatable image file with @code{savesystem}, e.g.:
15304:
15305: @example
15306: gforth app.fs -e "savesystem app.fi bye"
15307: @end example
15308:
15309: doc-savesystem
15310:
15311:
15312: @node Data-Relocatable Image Files, Fully Relocatable Image Files, Non-Relocatable Image Files, Image Files
15313: @section Data-Relocatable Image Files
15314: @cindex data-relocatable image files
15315: @cindex image file, data-relocatable
15316:
15317: These files contain relocatable data addresses, but fixed code
15318: addresses (instead of tokens). They are specific to the executable
15319: (i.e., @file{gforth} file) they were created with. Also, they disable
15320: dynamic native code generation (typically a factor of 2 in speed).
15321: You get a data-relocatable image, if you pass the engine you want to
15322: use through the @code{GFORTHD} environment variable to @file{gforthmi}
15323: (@pxref{gforthmi}), e.g.
15324:
15325: @example
15326: GFORTHD="/usr/bin/gforth-fast --no-dynamic" gforthmi myimage.fi source.fs
15327: @end example
15328:
15329: Note that the @code{--no-dynamic} is required here for the image to
15330: work (otherwise it will contain references to dynamically generated
15331: code that is not saved in the image).
15332:
15333:
15334: @node Fully Relocatable Image Files, Stack and Dictionary Sizes, Data-Relocatable Image Files, Image Files
15335: @section Fully Relocatable Image Files
15336: @cindex fully relocatable image files
15337: @cindex image file, fully relocatable
15338:
15339: @cindex @file{kern*.fi}, relocatability
15340: @cindex @file{gforth.fi}, relocatability
15341: These image files have relocatable data addresses, and tokens for code
15342: addresses. They can be used with different binaries (e.g., with and
15343: without debugging) on the same machine, and even across machines with
15344: the same data formats (byte order, cell size, floating point format),
15345: and they work with dynamic native code generation. However, they are
15346: usually specific to the version of Gforth they were created with. The
15347: files @file{gforth.fi} and @file{kernl*.fi} are fully relocatable.
15348:
15349: There are two ways to create a fully relocatable image file:
15350:
15351: @menu
15352: * gforthmi:: The normal way
15353: * cross.fs:: The hard way
15354: @end menu
15355:
15356: @node gforthmi, cross.fs, Fully Relocatable Image Files, Fully Relocatable Image Files
15357: @subsection @file{gforthmi}
15358: @cindex @file{comp-i.fs}
15359: @cindex @file{gforthmi}
15360:
15361: You will usually use @file{gforthmi}. If you want to create an
15362: image @i{file} that contains everything you would load by invoking
15363: Gforth with @code{gforth @i{options}}, you simply say:
15364: @example
15365: gforthmi @i{file} @i{options}
15366: @end example
15367:
15368: E.g., if you want to create an image @file{asm.fi} that has the file
15369: @file{asm.fs} loaded in addition to the usual stuff, you could do it
15370: like this:
15371:
15372: @example
15373: gforthmi asm.fi asm.fs
15374: @end example
15375:
15376: @file{gforthmi} is implemented as a sh script and works like this: It
15377: produces two non-relocatable images for different addresses and then
15378: compares them. Its output reflects this: first you see the output (if
15379: any) of the two Gforth invocations that produce the non-relocatable image
15380: files, then you see the output of the comparing program: It displays the
15381: offset used for data addresses and the offset used for code addresses;
15382: moreover, for each cell that cannot be represented correctly in the
15383: image files, it displays a line like this:
15384:
15385: @example
15386: 78DC BFFFFA50 BFFFFA40
15387: @end example
15388:
15389: This means that at offset $78dc from @code{forthstart}, one input image
15390: contains $bffffa50, and the other contains $bffffa40. Since these cells
15391: cannot be represented correctly in the output image, you should examine
15392: these places in the dictionary and verify that these cells are dead
15393: (i.e., not read before they are written).
15394:
15395: @cindex --application, @code{gforthmi} option
15396: If you insert the option @code{--application} in front of the image file
15397: name, you will get an image that uses the @code{--appl-image} option
15398: instead of the @code{--image-file} option (@pxref{Invoking
15399: Gforth}). When you execute such an image on Unix (by typing the image
15400: name as command), the Gforth engine will pass all options to the image
15401: instead of trying to interpret them as engine options.
15402:
15403: If you type @file{gforthmi} with no arguments, it prints some usage
15404: instructions.
15405:
15406: @cindex @code{savesystem} during @file{gforthmi}
15407: @cindex @code{bye} during @file{gforthmi}
15408: @cindex doubly indirect threaded code
15409: @cindex environment variables
15410: @cindex @code{GFORTHD} -- environment variable
15411: @cindex @code{GFORTH} -- environment variable
15412: @cindex @code{gforth-ditc}
15413: There are a few wrinkles: After processing the passed @i{options}, the
15414: words @code{savesystem} and @code{bye} must be visible. A special
15415: doubly indirect threaded version of the @file{gforth} executable is
15416: used for creating the non-relocatable images; you can pass the exact
15417: filename of this executable through the environment variable
15418: @code{GFORTHD} (default: @file{gforth-ditc}); if you pass a version
15419: that is not doubly indirect threaded, you will not get a fully
15420: relocatable image, but a data-relocatable image
15421: (@pxref{Data-Relocatable Image Files}), because there is no code
15422: address offset). The normal @file{gforth} executable is used for
15423: creating the relocatable image; you can pass the exact filename of
15424: this executable through the environment variable @code{GFORTH}.
15425:
15426: @node cross.fs, , gforthmi, Fully Relocatable Image Files
15427: @subsection @file{cross.fs}
15428: @cindex @file{cross.fs}
15429: @cindex cross-compiler
15430: @cindex metacompiler
15431: @cindex target compiler
15432:
15433: You can also use @code{cross}, a batch compiler that accepts a Forth-like
15434: programming language (@pxref{Cross Compiler}).
15435:
15436: @code{cross} allows you to create image files for machines with
15437: different data sizes and data formats than the one used for generating
15438: the image file. You can also use it to create an application image that
15439: does not contain a Forth compiler. These features are bought with
15440: restrictions and inconveniences in programming. E.g., addresses have to
15441: be stored in memory with special words (@code{A!}, @code{A,}, etc.) in
15442: order to make the code relocatable.
15443:
15444:
15445: @node Stack and Dictionary Sizes, Running Image Files, Fully Relocatable Image Files, Image Files
15446: @section Stack and Dictionary Sizes
15447: @cindex image file, stack and dictionary sizes
15448: @cindex dictionary size default
15449: @cindex stack size default
15450:
15451: If you invoke Gforth with a command line flag for the size
15452: (@pxref{Invoking Gforth}), the size you specify is stored in the
15453: dictionary. If you save the dictionary with @code{savesystem} or create
15454: an image with @file{gforthmi}, this size will become the default
15455: for the resulting image file. E.g., the following will create a
15456: fully relocatable version of @file{gforth.fi} with a 1MB dictionary:
15457:
15458: @example
15459: gforthmi gforth.fi -m 1M
15460: @end example
15461:
15462: In other words, if you want to set the default size for the dictionary
15463: and the stacks of an image, just invoke @file{gforthmi} with the
15464: appropriate options when creating the image.
15465:
15466: @cindex stack size, cache-friendly
15467: Note: For cache-friendly behaviour (i.e., good performance), you should
15468: make the sizes of the stacks modulo, say, 2K, somewhat different. E.g.,
15469: the default stack sizes are: data: 16k (mod 2k=0); fp: 15.5k (mod
15470: 2k=1.5k); return: 15k(mod 2k=1k); locals: 14.5k (mod 2k=0.5k).
15471:
15472: @node Running Image Files, Modifying the Startup Sequence, Stack and Dictionary Sizes, Image Files
15473: @section Running Image Files
15474: @cindex running image files
15475: @cindex invoking image files
15476: @cindex image file invocation
15477:
15478: @cindex -i, invoke image file
15479: @cindex --image file, invoke image file
15480: You can invoke Gforth with an image file @i{image} instead of the
15481: default @file{gforth.fi} with the @code{-i} flag (@pxref{Invoking Gforth}):
15482: @example
15483: gforth -i @i{image}
15484: @end example
15485:
15486: @cindex executable image file
15487: @cindex image file, executable
15488: If your operating system supports starting scripts with a line of the
15489: form @code{#! ...}, you just have to type the image file name to start
15490: Gforth with this image file (note that the file extension @code{.fi} is
15491: just a convention). I.e., to run Gforth with the image file @i{image},
15492: you can just type @i{image} instead of @code{gforth -i @i{image}}.
15493: This works because every @code{.fi} file starts with a line of this
15494: format:
15495:
15496: @example
15497: #! /usr/local/bin/gforth-0.4.0 -i
15498: @end example
15499:
15500: The file and pathname for the Gforth engine specified on this line is
15501: the specific Gforth executable that it was built against; i.e. the value
15502: of the environment variable @code{GFORTH} at the time that
15503: @file{gforthmi} was executed.
15504:
15505: You can make use of the same shell capability to make a Forth source
15506: file into an executable. For example, if you place this text in a file:
15507:
15508: @example
15509: #! /usr/local/bin/gforth
15510:
15511: ." Hello, world" CR
15512: bye
15513: @end example
15514:
15515: @noindent
15516: and then make the file executable (chmod +x in Unix), you can run it
15517: directly from the command line. The sequence @code{#!} is used in two
15518: ways; firstly, it is recognised as a ``magic sequence'' by the operating
15519: system@footnote{The Unix kernel actually recognises two types of files:
15520: executable files and files of data, where the data is processed by an
15521: interpreter that is specified on the ``interpreter line'' -- the first
15522: line of the file, starting with the sequence #!. There may be a small
15523: limit (e.g., 32) on the number of characters that may be specified on
15524: the interpreter line.} secondly it is treated as a comment character by
15525: Gforth. Because of the second usage, a space is required between
15526: @code{#!} and the path to the executable (moreover, some Unixes
15527: require the sequence @code{#! /}).
15528:
15529: The disadvantage of this latter technique, compared with using
15530: @file{gforthmi}, is that it is slightly slower; the Forth source code is
15531: compiled on-the-fly, each time the program is invoked.
15532:
15533: doc-#!
15534:
15535:
15536: @node Modifying the Startup Sequence, , Running Image Files, Image Files
15537: @section Modifying the Startup Sequence
15538: @cindex startup sequence for image file
15539: @cindex image file initialization sequence
15540: @cindex initialization sequence of image file
15541:
15542: You can add your own initialization to the startup sequence of an image
15543: through the deferred word @code{'cold}. @code{'cold} is invoked just
15544: before the image-specific command line processing (i.e., loading files
15545: and evaluating (@code{-e}) strings) starts.
15546:
15547: A sequence for adding your initialization usually looks like this:
15548:
15549: @example
15550: :noname
15551: Defers 'cold \ do other initialization stuff (e.g., rehashing wordlists)
15552: ... \ your stuff
15553: ; IS 'cold
15554: @end example
15555:
15556: After @code{'cold}, Gforth processes the image options
15557: (@pxref{Invoking Gforth}), and then it performs @code{bootmessage},
15558: another deferred word. This normally prints Gforth's startup message
15559: and does nothing else.
15560:
15561: @cindex turnkey image files
15562: @cindex image file, turnkey applications
15563: So, if you want to make a turnkey image (i.e., an image for an
15564: application instead of an extended Forth system), you can do this in
15565: two ways:
15566:
15567: @itemize @bullet
15568:
15569: @item
15570: If you want to do your interpretation of the OS command-line
15571: arguments, hook into @code{'cold}. In that case you probably also
15572: want to build the image with @code{gforthmi --application}
15573: (@pxref{gforthmi}) to keep the engine from processing OS command line
15574: options. You can then do your own command-line processing with
15575: @code{next-arg}
15576:
15577: @item
15578: If you want to have the normal Gforth processing of OS command-line
15579: arguments, hook into @code{bootmessage}.
15580:
15581: @end itemize
15582:
15583: In either case, you probably do not want the word that you execute in
15584: these hooks to exit normally, but use @code{bye} or @code{throw}.
15585: Otherwise the Gforth startup process would continue and eventually
15586: present the Forth command line to the user.
15587:
15588: doc-'cold
15589: doc-bootmessage
15590:
15591: @c ******************************************************************
15592: @node Engine, Cross Compiler, Image Files, Top
15593: @chapter Engine
15594: @cindex engine
15595: @cindex virtual machine
15596:
15597: Reading this chapter is not necessary for programming with Gforth. It
15598: may be helpful for finding your way in the Gforth sources.
15599:
15600: The ideas in this section have also been published in the following
15601: papers: Bernd Paysan, @cite{ANS fig/GNU/??? Forth} (in German),
15602: Forth-Tagung '93; M. Anton Ertl,
15603: @cite{@uref{http://www.complang.tuwien.ac.at/papers/ertl93.ps.Z, A
15604: Portable Forth Engine}}, EuroForth '93; M. Anton Ertl,
15605: @cite{@uref{http://www.complang.tuwien.ac.at/papers/ertl02.ps.gz,
15606: Threaded code variations and optimizations (extended version)}},
15607: Forth-Tagung '02.
15608:
15609: @menu
15610: * Portability::
15611: * Threading::
15612: * Primitives::
15613: * Performance::
15614: @end menu
15615:
15616: @node Portability, Threading, Engine, Engine
15617: @section Portability
15618: @cindex engine portability
15619:
15620: An important goal of the Gforth Project is availability across a wide
15621: range of personal machines. fig-Forth, and, to a lesser extent, F83,
15622: achieved this goal by manually coding the engine in assembly language
15623: for several then-popular processors. This approach is very
15624: labor-intensive and the results are short-lived due to progress in
15625: computer architecture.
15626:
15627: @cindex C, using C for the engine
15628: Others have avoided this problem by coding in C, e.g., Mitch Bradley
15629: (cforth), Mikael Patel (TILE) and Dirk Zoller (pfe). This approach is
15630: particularly popular for UNIX-based Forths due to the large variety of
15631: architectures of UNIX machines. Unfortunately an implementation in C
15632: does not mix well with the goals of efficiency and with using
15633: traditional techniques: Indirect or direct threading cannot be expressed
15634: in C, and switch threading, the fastest technique available in C, is
15635: significantly slower. Another problem with C is that it is very
15636: cumbersome to express double integer arithmetic.
15637:
15638: @cindex GNU C for the engine
15639: @cindex long long
15640: Fortunately, there is a portable language that does not have these
15641: limitations: GNU C, the version of C processed by the GNU C compiler
15642: (@pxref{C Extensions, , Extensions to the C Language Family, gcc.info,
15643: GNU C Manual}). Its labels as values feature (@pxref{Labels as Values, ,
15644: Labels as Values, gcc.info, GNU C Manual}) makes direct and indirect
15645: threading possible, its @code{long long} type (@pxref{Long Long, ,
15646: Double-Word Integers, gcc.info, GNU C Manual}) corresponds to Forth's
15647: double numbers on many systems. GNU C is freely available on all
15648: important (and many unimportant) UNIX machines, VMS, 80386s running
15649: MS-DOS, the Amiga, and the Atari ST, so a Forth written in GNU C can run
15650: on all these machines.
15651:
15652: Writing in a portable language has the reputation of producing code that
15653: is slower than assembly. For our Forth engine we repeatedly looked at
15654: the code produced by the compiler and eliminated most compiler-induced
15655: inefficiencies by appropriate changes in the source code.
15656:
15657: @cindex explicit register declarations
15658: @cindex --enable-force-reg, configuration flag
15659: @cindex -DFORCE_REG
15660: However, register allocation cannot be portably influenced by the
15661: programmer, leading to some inefficiencies on register-starved
15662: machines. We use explicit register declarations (@pxref{Explicit Reg
15663: Vars, , Variables in Specified Registers, gcc.info, GNU C Manual}) to
15664: improve the speed on some machines. They are turned on by using the
15665: configuration flag @code{--enable-force-reg} (@code{gcc} switch
15666: @code{-DFORCE_REG}). Unfortunately, this feature not only depends on the
15667: machine, but also on the compiler version: On some machines some
15668: compiler versions produce incorrect code when certain explicit register
15669: declarations are used. So by default @code{-DFORCE_REG} is not used.
15670:
15671: @node Threading, Primitives, Portability, Engine
15672: @section Threading
15673: @cindex inner interpreter implementation
15674: @cindex threaded code implementation
15675:
15676: @cindex labels as values
15677: GNU C's labels as values extension (available since @code{gcc-2.0},
15678: @pxref{Labels as Values, , Labels as Values, gcc.info, GNU C Manual})
15679: makes it possible to take the address of @i{label} by writing
15680: @code{&&@i{label}}. This address can then be used in a statement like
15681: @code{goto *@i{address}}. I.e., @code{goto *&&x} is the same as
15682: @code{goto x}.
15683:
15684: @cindex @code{NEXT}, indirect threaded
15685: @cindex indirect threaded inner interpreter
15686: @cindex inner interpreter, indirect threaded
15687: With this feature an indirect threaded @code{NEXT} looks like:
15688: @example
15689: cfa = *ip++;
15690: ca = *cfa;
15691: goto *ca;
15692: @end example
15693: @cindex instruction pointer
15694: For those unfamiliar with the names: @code{ip} is the Forth instruction
15695: pointer; the @code{cfa} (code-field address) corresponds to ANS Forths
15696: execution token and points to the code field of the next word to be
15697: executed; The @code{ca} (code address) fetched from there points to some
15698: executable code, e.g., a primitive or the colon definition handler
15699: @code{docol}.
15700:
15701: @cindex @code{NEXT}, direct threaded
15702: @cindex direct threaded inner interpreter
15703: @cindex inner interpreter, direct threaded
15704: Direct threading is even simpler:
15705: @example
15706: ca = *ip++;
15707: goto *ca;
15708: @end example
15709:
15710: Of course we have packaged the whole thing neatly in macros called
15711: @code{NEXT} and @code{NEXT1} (the part of @code{NEXT} after fetching the cfa).
15712:
15713: @menu
15714: * Scheduling::
15715: * Direct or Indirect Threaded?::
15716: * Dynamic Superinstructions::
15717: * DOES>::
15718: @end menu
15719:
15720: @node Scheduling, Direct or Indirect Threaded?, Threading, Threading
15721: @subsection Scheduling
15722: @cindex inner interpreter optimization
15723:
15724: There is a little complication: Pipelined and superscalar processors,
15725: i.e., RISC and some modern CISC machines can process independent
15726: instructions while waiting for the results of an instruction. The
15727: compiler usually reorders (schedules) the instructions in a way that
15728: achieves good usage of these delay slots. However, on our first tries
15729: the compiler did not do well on scheduling primitives. E.g., for
15730: @code{+} implemented as
15731: @example
15732: n=sp[0]+sp[1];
15733: sp++;
15734: sp[0]=n;
15735: NEXT;
15736: @end example
15737: the @code{NEXT} comes strictly after the other code, i.e., there is
15738: nearly no scheduling. After a little thought the problem becomes clear:
15739: The compiler cannot know that @code{sp} and @code{ip} point to different
15740: addresses (and the version of @code{gcc} we used would not know it even
15741: if it was possible), so it could not move the load of the cfa above the
15742: store to the TOS. Indeed the pointers could be the same, if code on or
15743: very near the top of stack were executed. In the interest of speed we
15744: chose to forbid this probably unused ``feature'' and helped the compiler
15745: in scheduling: @code{NEXT} is divided into several parts:
15746: @code{NEXT_P0}, @code{NEXT_P1} and @code{NEXT_P2}). @code{+} now looks
15747: like:
15748: @example
15749: NEXT_P0;
15750: n=sp[0]+sp[1];
15751: sp++;
15752: NEXT_P1;
15753: sp[0]=n;
15754: NEXT_P2;
15755: @end example
15756:
15757: There are various schemes that distribute the different operations of
15758: NEXT between these parts in several ways; in general, different schemes
15759: perform best on different processors. We use a scheme for most
15760: architectures that performs well for most processors of this
15761: architecture; in the future we may switch to benchmarking and chosing
15762: the scheme on installation time.
15763:
15764:
15765: @node Direct or Indirect Threaded?, Dynamic Superinstructions, Scheduling, Threading
15766: @subsection Direct or Indirect Threaded?
15767: @cindex threading, direct or indirect?
15768:
15769: Threaded forth code consists of references to primitives (simple machine
15770: code routines like @code{+}) and to non-primitives (e.g., colon
15771: definitions, variables, constants); for a specific class of
15772: non-primitives (e.g., variables) there is one code routine (e.g.,
15773: @code{dovar}), but each variable needs a separate reference to its data.
15774:
15775: Traditionally Forth has been implemented as indirect threaded code,
15776: because this allows to use only one cell to reference a non-primitive
15777: (basically you point to the data, and find the code address there).
15778:
15779: @cindex primitive-centric threaded code
15780: However, threaded code in Gforth (since 0.6.0) uses two cells for
15781: non-primitives, one for the code address, and one for the data address;
15782: the data pointer is an immediate argument for the virtual machine
15783: instruction represented by the code address. We call this
15784: @emph{primitive-centric} threaded code, because all code addresses point
15785: to simple primitives. E.g., for a variable, the code address is for
15786: @code{lit} (also used for integer literals like @code{99}).
15787:
15788: Primitive-centric threaded code allows us to use (faster) direct
15789: threading as dispatch method, completely portably (direct threaded code
15790: in Gforth before 0.6.0 required architecture-specific code). It also
15791: eliminates the performance problems related to I-cache consistency that
15792: 386 implementations have with direct threaded code, and allows
15793: additional optimizations.
15794:
15795: @cindex hybrid direct/indirect threaded code
15796: There is a catch, however: the @var{xt} parameter of @code{execute} can
15797: occupy only one cell, so how do we pass non-primitives with their code
15798: @emph{and} data addresses to them? Our answer is to use indirect
15799: threaded dispatch for @code{execute} and other words that use a
15800: single-cell xt. So, normal threaded code in colon definitions uses
15801: direct threading, and @code{execute} and similar words, which dispatch
15802: to xts on the data stack, use indirect threaded code. We call this
15803: @emph{hybrid direct/indirect} threaded code.
15804:
15805: @cindex engines, gforth vs. gforth-fast vs. gforth-itc
15806: @cindex gforth engine
15807: @cindex gforth-fast engine
15808: The engines @command{gforth} and @command{gforth-fast} use hybrid
15809: direct/indirect threaded code. This means that with these engines you
15810: cannot use @code{,} to compile an xt. Instead, you have to use
15811: @code{compile,}.
15812:
15813: @cindex gforth-itc engine
15814: If you want to compile xts with @code{,}, use @command{gforth-itc}.
15815: This engine uses plain old indirect threaded code. It still compiles in
15816: a primitive-centric style, so you cannot use @code{compile,} instead of
15817: @code{,} (e.g., for producing tables of xts with @code{] word1 word2
15818: ... [}). If you want to do that, you have to use @command{gforth-itc}
15819: and execute @code{' , is compile,}. Your program can check if it is
15820: running on a hybrid direct/indirect threaded engine or a pure indirect
15821: threaded engine with @code{threading-method} (@pxref{Threading Words}).
15822:
15823:
15824: @node Dynamic Superinstructions, DOES>, Direct or Indirect Threaded?, Threading
15825: @subsection Dynamic Superinstructions
15826: @cindex Dynamic superinstructions with replication
15827: @cindex Superinstructions
15828: @cindex Replication
15829:
15830: The engines @command{gforth} and @command{gforth-fast} use another
15831: optimization: Dynamic superinstructions with replication. As an
15832: example, consider the following colon definition:
15833:
15834: @example
15835: : squared ( n1 -- n2 )
15836: dup * ;
15837: @end example
15838:
15839: Gforth compiles this into the threaded code sequence
15840:
15841: @example
15842: dup
15843: *
15844: ;s
15845: @end example
15846:
15847: In normal direct threaded code there is a code address occupying one
15848: cell for each of these primitives. Each code address points to a
15849: machine code routine, and the interpreter jumps to this machine code in
15850: order to execute the primitive. The routines for these three
15851: primitives are (in @command{gforth-fast} on the 386):
15852:
15853: @example
15854: Code dup
15855: ( $804B950 ) add esi , # -4 \ $83 $C6 $FC
15856: ( $804B953 ) add ebx , # 4 \ $83 $C3 $4
15857: ( $804B956 ) mov dword ptr 4 [esi] , ecx \ $89 $4E $4
15858: ( $804B959 ) jmp dword ptr FC [ebx] \ $FF $63 $FC
15859: end-code
15860: Code *
15861: ( $804ACC4 ) mov eax , dword ptr 4 [esi] \ $8B $46 $4
15862: ( $804ACC7 ) add esi , # 4 \ $83 $C6 $4
15863: ( $804ACCA ) add ebx , # 4 \ $83 $C3 $4
15864: ( $804ACCD ) imul ecx , eax \ $F $AF $C8
15865: ( $804ACD0 ) jmp dword ptr FC [ebx] \ $FF $63 $FC
15866: end-code
15867: Code ;s
15868: ( $804A693 ) mov eax , dword ptr [edi] \ $8B $7
15869: ( $804A695 ) add edi , # 4 \ $83 $C7 $4
15870: ( $804A698 ) lea ebx , dword ptr 4 [eax] \ $8D $58 $4
15871: ( $804A69B ) jmp dword ptr FC [ebx] \ $FF $63 $FC
15872: end-code
15873: @end example
15874:
15875: With dynamic superinstructions and replication the compiler does not
15876: just lay down the threaded code, but also copies the machine code
15877: fragments, usually without the jump at the end.
15878:
15879: @example
15880: ( $4057D27D ) add esi , # -4 \ $83 $C6 $FC
15881: ( $4057D280 ) add ebx , # 4 \ $83 $C3 $4
15882: ( $4057D283 ) mov dword ptr 4 [esi] , ecx \ $89 $4E $4
15883: ( $4057D286 ) mov eax , dword ptr 4 [esi] \ $8B $46 $4
15884: ( $4057D289 ) add esi , # 4 \ $83 $C6 $4
15885: ( $4057D28C ) add ebx , # 4 \ $83 $C3 $4
15886: ( $4057D28F ) imul ecx , eax \ $F $AF $C8
15887: ( $4057D292 ) mov eax , dword ptr [edi] \ $8B $7
15888: ( $4057D294 ) add edi , # 4 \ $83 $C7 $4
15889: ( $4057D297 ) lea ebx , dword ptr 4 [eax] \ $8D $58 $4
15890: ( $4057D29A ) jmp dword ptr FC [ebx] \ $FF $63 $FC
15891: @end example
15892:
15893: Only when a threaded-code control-flow change happens (e.g., in
15894: @code{;s}), the jump is appended. This optimization eliminates many of
15895: these jumps and makes the rest much more predictable. The speedup
15896: depends on the processor and the application; on the Athlon and Pentium
15897: III this optimization typically produces a speedup by a factor of 2.
15898:
15899: The code addresses in the direct-threaded code are set to point to the
15900: appropriate points in the copied machine code, in this example like
15901: this:
15902:
15903: @example
15904: primitive code address
15905: dup $4057D27D
15906: * $4057D286
15907: ;s $4057D292
15908: @end example
15909:
15910: Thus there can be threaded-code jumps to any place in this piece of
15911: code. This also simplifies decompilation quite a bit.
15912:
15913: @cindex --no-dynamic command-line option
15914: @cindex --no-super command-line option
15915: You can disable this optimization with @option{--no-dynamic}. You can
15916: use the copying without eliminating the jumps (i.e., dynamic
15917: replication, but without superinstructions) with @option{--no-super};
15918: this gives the branch prediction benefit alone; the effect on
15919: performance depends on the CPU; on the Athlon and Pentium III the
15920: speedup is a little less than for dynamic superinstructions with
15921: replication.
15922:
15923: @cindex patching threaded code
15924: One use of these options is if you want to patch the threaded code.
15925: With superinstructions, many of the dispatch jumps are eliminated, so
15926: patching often has no effect. These options preserve all the dispatch
15927: jumps.
15928:
15929: @cindex --dynamic command-line option
15930: On some machines dynamic superinstructions are disabled by default,
15931: because it is unsafe on these machines. However, if you feel
15932: adventurous, you can enable it with @option{--dynamic}.
15933:
15934: @node DOES>, , Dynamic Superinstructions, Threading
15935: @subsection DOES>
15936: @cindex @code{DOES>} implementation
15937:
15938: @cindex @code{dodoes} routine
15939: @cindex @code{DOES>}-code
15940: One of the most complex parts of a Forth engine is @code{dodoes}, i.e.,
15941: the chunk of code executed by every word defined by a
15942: @code{CREATE}...@code{DOES>} pair; actually with primitive-centric code,
15943: this is only needed if the xt of the word is @code{execute}d. The main
15944: problem here is: How to find the Forth code to be executed, i.e. the
15945: code after the @code{DOES>} (the @code{DOES>}-code)? There are two
15946: solutions:
15947:
15948: In fig-Forth the code field points directly to the @code{dodoes} and the
15949: @code{DOES>}-code address is stored in the cell after the code address
15950: (i.e. at @code{@i{CFA} cell+}). It may seem that this solution is
15951: illegal in the Forth-79 and all later standards, because in fig-Forth
15952: this address lies in the body (which is illegal in these
15953: standards). However, by making the code field larger for all words this
15954: solution becomes legal again. We use this approach. Leaving a cell
15955: unused in most words is a bit wasteful, but on the machines we are
15956: targeting this is hardly a problem.
15957:
15958:
15959: @node Primitives, Performance, Threading, Engine
15960: @section Primitives
15961: @cindex primitives, implementation
15962: @cindex virtual machine instructions, implementation
15963:
15964: @menu
15965: * Automatic Generation::
15966: * TOS Optimization::
15967: * Produced code::
15968: @end menu
15969:
15970: @node Automatic Generation, TOS Optimization, Primitives, Primitives
15971: @subsection Automatic Generation
15972: @cindex primitives, automatic generation
15973:
15974: @cindex @file{prims2x.fs}
15975:
15976: Since the primitives are implemented in a portable language, there is no
15977: longer any need to minimize the number of primitives. On the contrary,
15978: having many primitives has an advantage: speed. In order to reduce the
15979: number of errors in primitives and to make programming them easier, we
15980: provide a tool, the primitive generator (@file{prims2x.fs} aka Vmgen,
15981: @pxref{Top, Vmgen, Introduction, vmgen, Vmgen}), that automatically
15982: generates most (and sometimes all) of the C code for a primitive from
15983: the stack effect notation. The source for a primitive has the following
15984: form:
15985:
15986: @cindex primitive source format
15987: @format
15988: @i{Forth-name} ( @i{stack-effect} ) @i{category} [@i{pronounc.}]
15989: [@code{""}@i{glossary entry}@code{""}]
15990: @i{C code}
15991: [@code{:}
15992: @i{Forth code}]
15993: @end format
15994:
15995: The items in brackets are optional. The category and glossary fields
15996: are there for generating the documentation, the Forth code is there
15997: for manual implementations on machines without GNU C. E.g., the source
15998: for the primitive @code{+} is:
15999: @example
16000: + ( n1 n2 -- n ) core plus
16001: n = n1+n2;
16002: @end example
16003:
16004: This looks like a specification, but in fact @code{n = n1+n2} is C
16005: code. Our primitive generation tool extracts a lot of information from
16006: the stack effect notations@footnote{We use a one-stack notation, even
16007: though we have separate data and floating-point stacks; The separate
16008: notation can be generated easily from the unified notation.}: The number
16009: of items popped from and pushed on the stack, their type, and by what
16010: name they are referred to in the C code. It then generates a C code
16011: prelude and postlude for each primitive. The final C code for @code{+}
16012: looks like this:
16013:
16014: @example
16015: I_plus: /* + ( n1 n2 -- n ) */ /* label, stack effect */
16016: /* */ /* documentation */
16017: NAME("+") /* debugging output (with -DDEBUG) */
16018: @{
16019: DEF_CA /* definition of variable ca (indirect threading) */
16020: Cell n1; /* definitions of variables */
16021: Cell n2;
16022: Cell n;
16023: NEXT_P0; /* NEXT part 0 */
16024: n1 = (Cell) sp[1]; /* input */
16025: n2 = (Cell) TOS;
16026: sp += 1; /* stack adjustment */
16027: @{
16028: n = n1+n2; /* C code taken from the source */
16029: @}
16030: NEXT_P1; /* NEXT part 1 */
16031: TOS = (Cell)n; /* output */
16032: NEXT_P2; /* NEXT part 2 */
16033: @}
16034: @end example
16035:
16036: This looks long and inefficient, but the GNU C compiler optimizes quite
16037: well and produces optimal code for @code{+} on, e.g., the R3000 and the
16038: HP RISC machines: Defining the @code{n}s does not produce any code, and
16039: using them as intermediate storage also adds no cost.
16040:
16041: There are also other optimizations that are not illustrated by this
16042: example: assignments between simple variables are usually for free (copy
16043: propagation). If one of the stack items is not used by the primitive
16044: (e.g. in @code{drop}), the compiler eliminates the load from the stack
16045: (dead code elimination). On the other hand, there are some things that
16046: the compiler does not do, therefore they are performed by
16047: @file{prims2x.fs}: The compiler does not optimize code away that stores
16048: a stack item to the place where it just came from (e.g., @code{over}).
16049:
16050: While programming a primitive is usually easy, there are a few cases
16051: where the programmer has to take the actions of the generator into
16052: account, most notably @code{?dup}, but also words that do not (always)
16053: fall through to @code{NEXT}.
16054:
16055: For more information
16056:
16057: @node TOS Optimization, Produced code, Automatic Generation, Primitives
16058: @subsection TOS Optimization
16059: @cindex TOS optimization for primitives
16060: @cindex primitives, keeping the TOS in a register
16061:
16062: An important optimization for stack machine emulators, e.g., Forth
16063: engines, is keeping one or more of the top stack items in
16064: registers. If a word has the stack effect @i{in1}...@i{inx} @code{--}
16065: @i{out1}...@i{outy}, keeping the top @i{n} items in registers
16066: @itemize @bullet
16067: @item
16068: is better than keeping @i{n-1} items, if @i{x>=n} and @i{y>=n},
16069: due to fewer loads from and stores to the stack.
16070: @item is slower than keeping @i{n-1} items, if @i{x<>y} and @i{x<n} and
16071: @i{y<n}, due to additional moves between registers.
16072: @end itemize
16073:
16074: @cindex -DUSE_TOS
16075: @cindex -DUSE_NO_TOS
16076: In particular, keeping one item in a register is never a disadvantage,
16077: if there are enough registers. Keeping two items in registers is a
16078: disadvantage for frequent words like @code{?branch}, constants,
16079: variables, literals and @code{i}. Therefore our generator only produces
16080: code that keeps zero or one items in registers. The generated C code
16081: covers both cases; the selection between these alternatives is made at
16082: C-compile time using the switch @code{-DUSE_TOS}. @code{TOS} in the C
16083: code for @code{+} is just a simple variable name in the one-item case,
16084: otherwise it is a macro that expands into @code{sp[0]}. Note that the
16085: GNU C compiler tries to keep simple variables like @code{TOS} in
16086: registers, and it usually succeeds, if there are enough registers.
16087:
16088: @cindex -DUSE_FTOS
16089: @cindex -DUSE_NO_FTOS
16090: The primitive generator performs the TOS optimization for the
16091: floating-point stack, too (@code{-DUSE_FTOS}). For floating-point
16092: operations the benefit of this optimization is even larger:
16093: floating-point operations take quite long on most processors, but can be
16094: performed in parallel with other operations as long as their results are
16095: not used. If the FP-TOS is kept in a register, this works. If
16096: it is kept on the stack, i.e., in memory, the store into memory has to
16097: wait for the result of the floating-point operation, lengthening the
16098: execution time of the primitive considerably.
16099:
16100: The TOS optimization makes the automatic generation of primitives a
16101: bit more complicated. Just replacing all occurrences of @code{sp[0]} by
16102: @code{TOS} is not sufficient. There are some special cases to
16103: consider:
16104: @itemize @bullet
16105: @item In the case of @code{dup ( w -- w w )} the generator must not
16106: eliminate the store to the original location of the item on the stack,
16107: if the TOS optimization is turned on.
16108: @item Primitives with stack effects of the form @code{--}
16109: @i{out1}...@i{outy} must store the TOS to the stack at the start.
16110: Likewise, primitives with the stack effect @i{in1}...@i{inx} @code{--}
16111: must load the TOS from the stack at the end. But for the null stack
16112: effect @code{--} no stores or loads should be generated.
16113: @end itemize
16114:
16115: @node Produced code, , TOS Optimization, Primitives
16116: @subsection Produced code
16117: @cindex primitives, assembly code listing
16118:
16119: @cindex @file{engine.s}
16120: To see what assembly code is produced for the primitives on your machine
16121: with your compiler and your flag settings, type @code{make engine.s} and
16122: look at the resulting file @file{engine.s}. Alternatively, you can also
16123: disassemble the code of primitives with @code{see} on some architectures.
16124:
16125: @node Performance, , Primitives, Engine
16126: @section Performance
16127: @cindex performance of some Forth interpreters
16128: @cindex engine performance
16129: @cindex benchmarking Forth systems
16130: @cindex Gforth performance
16131:
16132: On RISCs the Gforth engine is very close to optimal; i.e., it is usually
16133: impossible to write a significantly faster threaded-code engine.
16134:
16135: On register-starved machines like the 386 architecture processors
16136: improvements are possible, because @code{gcc} does not utilize the
16137: registers as well as a human, even with explicit register declarations;
16138: e.g., Bernd Beuster wrote a Forth system fragment in assembly language
16139: and hand-tuned it for the 486; this system is 1.19 times faster on the
16140: Sieve benchmark on a 486DX2/66 than Gforth compiled with
16141: @code{gcc-2.6.3} with @code{-DFORCE_REG}. The situation has improved
16142: with gcc-2.95 and gforth-0.4.9; now the most important virtual machine
16143: registers fit in real registers (and we can even afford to use the TOS
16144: optimization), resulting in a speedup of 1.14 on the sieve over the
16145: earlier results. And dynamic superinstructions provide another speedup
16146: (but only around a factor 1.2 on the 486).
16147:
16148: @cindex Win32Forth performance
16149: @cindex NT Forth performance
16150: @cindex eforth performance
16151: @cindex ThisForth performance
16152: @cindex PFE performance
16153: @cindex TILE performance
16154: The potential advantage of assembly language implementations is not
16155: necessarily realized in complete Forth systems: We compared Gforth-0.5.9
16156: (direct threaded, compiled with @code{gcc-2.95.1} and
16157: @code{-DFORCE_REG}) with Win32Forth 1.2093 (newer versions are
16158: reportedly much faster), LMI's NT Forth (Beta, May 1994) and Eforth
16159: (with and without peephole (aka pinhole) optimization of the threaded
16160: code); all these systems were written in assembly language. We also
16161: compared Gforth with three systems written in C: PFE-0.9.14 (compiled
16162: with @code{gcc-2.6.3} with the default configuration for Linux:
16163: @code{-O2 -fomit-frame-pointer -DUSE_REGS -DUNROLL_NEXT}), ThisForth
16164: Beta (compiled with @code{gcc-2.6.3 -O3 -fomit-frame-pointer}; ThisForth
16165: employs peephole optimization of the threaded code) and TILE (compiled
16166: with @code{make opt}). We benchmarked Gforth, PFE, ThisForth and TILE on
16167: a 486DX2/66 under Linux. Kenneth O'Heskin kindly provided the results
16168: for Win32Forth and NT Forth on a 486DX2/66 with similar memory
16169: performance under Windows NT. Marcel Hendrix ported Eforth to Linux,
16170: then extended it to run the benchmarks, added the peephole optimizer,
16171: ran the benchmarks and reported the results.
16172:
16173: We used four small benchmarks: the ubiquitous Sieve; bubble-sorting and
16174: matrix multiplication come from the Stanford integer benchmarks and have
16175: been translated into Forth by Martin Fraeman; we used the versions
16176: included in the TILE Forth package, but with bigger data set sizes; and
16177: a recursive Fibonacci number computation for benchmarking calling
16178: performance. The following table shows the time taken for the benchmarks
16179: scaled by the time taken by Gforth (in other words, it shows the speedup
16180: factor that Gforth achieved over the other systems).
16181:
16182: @example
16183: relative Win32- NT eforth This-
16184: time Gforth Forth Forth eforth +opt PFE Forth TILE
16185: sieve 1.00 2.16 1.78 2.16 1.32 2.46 4.96 13.37
16186: bubble 1.00 1.93 2.07 2.18 1.29 2.21 5.70
16187: matmul 1.00 1.92 1.76 1.90 0.96 2.06 5.32
16188: fib 1.00 2.32 2.03 1.86 1.31 2.64 4.55 6.54
16189: @end example
16190:
16191: You may be quite surprised by the good performance of Gforth when
16192: compared with systems written in assembly language. One important reason
16193: for the disappointing performance of these other systems is probably
16194: that they are not written optimally for the 486 (e.g., they use the
16195: @code{lods} instruction). In addition, Win32Forth uses a comfortable,
16196: but costly method for relocating the Forth image: like @code{cforth}, it
16197: computes the actual addresses at run time, resulting in two address
16198: computations per @code{NEXT} (@pxref{Image File Background}).
16199:
16200: The speedup of Gforth over PFE, ThisForth and TILE can be easily
16201: explained with the self-imposed restriction of the latter systems to
16202: standard C, which makes efficient threading impossible (however, the
16203: measured implementation of PFE uses a GNU C extension: @pxref{Global Reg
16204: Vars, , Defining Global Register Variables, gcc.info, GNU C Manual}).
16205: Moreover, current C compilers have a hard time optimizing other aspects
16206: of the ThisForth and the TILE source.
16207:
16208: The performance of Gforth on 386 architecture processors varies widely
16209: with the version of @code{gcc} used. E.g., @code{gcc-2.5.8} failed to
16210: allocate any of the virtual machine registers into real machine
16211: registers by itself and would not work correctly with explicit register
16212: declarations, giving a significantly slower engine (on a 486DX2/66
16213: running the Sieve) than the one measured above.
16214:
16215: Note that there have been several releases of Win32Forth since the
16216: release presented here, so the results presented above may have little
16217: predictive value for the performance of Win32Forth today (results for
16218: the current release on an i486DX2/66 are welcome).
16219:
16220: @cindex @file{Benchres}
16221: In
16222: @cite{@uref{http://www.complang.tuwien.ac.at/papers/ertl&maierhofer95.ps.gz,
16223: Translating Forth to Efficient C}} by M. Anton Ertl and Martin
16224: Maierhofer (presented at EuroForth '95), an indirect threaded version of
16225: Gforth is compared with Win32Forth, NT Forth, PFE, ThisForth, and
16226: several native code systems; that version of Gforth is slower on a 486
16227: than the version used here. You can find a newer version of these
16228: measurements at
16229: @uref{http://www.complang.tuwien.ac.at/forth/performance.html}. You can
16230: find numbers for Gforth on various machines in @file{Benchres}.
16231:
16232: @c ******************************************************************
16233: @c @node Binding to System Library, Cross Compiler, Engine, Top
16234: @c @chapter Binding to System Library
16235:
16236: @c ****************************************************************
16237: @node Cross Compiler, Bugs, Engine, Top
16238: @chapter Cross Compiler
16239: @cindex @file{cross.fs}
16240: @cindex cross-compiler
16241: @cindex metacompiler
16242: @cindex target compiler
16243:
16244: The cross compiler is used to bootstrap a Forth kernel. Since Gforth is
16245: mostly written in Forth, including crucial parts like the outer
16246: interpreter and compiler, it needs compiled Forth code to get
16247: started. The cross compiler allows to create new images for other
16248: architectures, even running under another Forth system.
16249:
16250: @menu
16251: * Using the Cross Compiler::
16252: * How the Cross Compiler Works::
16253: @end menu
16254:
16255: @node Using the Cross Compiler, How the Cross Compiler Works, Cross Compiler, Cross Compiler
16256: @section Using the Cross Compiler
16257:
16258: The cross compiler uses a language that resembles Forth, but isn't. The
16259: main difference is that you can execute Forth code after definition,
16260: while you usually can't execute the code compiled by cross, because the
16261: code you are compiling is typically for a different computer than the
16262: one you are compiling on.
16263:
16264: @c anton: This chapter is somewhat different from waht I would expect: I
16265: @c would expect an explanation of the cross language and how to create an
16266: @c application image with it. The section explains some aspects of
16267: @c creating a Gforth kernel.
16268:
16269: The Makefile is already set up to allow you to create kernels for new
16270: architectures with a simple make command. The generic kernels using the
16271: GCC compiled virtual machine are created in the normal build process
16272: with @code{make}. To create a embedded Gforth executable for e.g. the
16273: 8086 processor (running on a DOS machine), type
16274:
16275: @example
16276: make kernl-8086.fi
16277: @end example
16278:
16279: This will use the machine description from the @file{arch/8086}
16280: directory to create a new kernel. A machine file may look like that:
16281:
16282: @example
16283: \ Parameter for target systems 06oct92py
16284:
16285: 4 Constant cell \ cell size in bytes
16286: 2 Constant cell<< \ cell shift to bytes
16287: 5 Constant cell>bit \ cell shift to bits
16288: 8 Constant bits/char \ bits per character
16289: 8 Constant bits/byte \ bits per byte [default: 8]
16290: 8 Constant float \ bytes per float
16291: 8 Constant /maxalign \ maximum alignment in bytes
16292: false Constant bigendian \ byte order
16293: ( true=big, false=little )
16294:
16295: include machpc.fs \ feature list
16296: @end example
16297:
16298: This part is obligatory for the cross compiler itself, the feature list
16299: is used by the kernel to conditionally compile some features in and out,
16300: depending on whether the target supports these features.
16301:
16302: There are some optional features, if you define your own primitives,
16303: have an assembler, or need special, nonstandard preparation to make the
16304: boot process work. @code{asm-include} includes an assembler,
16305: @code{prims-include} includes primitives, and @code{>boot} prepares for
16306: booting.
16307:
16308: @example
16309: : asm-include ." Include assembler" cr
16310: s" arch/8086/asm.fs" included ;
16311:
16312: : prims-include ." Include primitives" cr
16313: s" arch/8086/prim.fs" included ;
16314:
16315: : >boot ." Prepare booting" cr
16316: s" ' boot >body into-forth 1+ !" evaluate ;
16317: @end example
16318:
16319: These words are used as sort of macro during the cross compilation in
16320: the file @file{kernel/main.fs}. Instead of using these macros, it would
16321: be possible --- but more complicated --- to write a new kernel project
16322: file, too.
16323:
16324: @file{kernel/main.fs} expects the machine description file name on the
16325: stack; the cross compiler itself (@file{cross.fs}) assumes that either
16326: @code{mach-file} leaves a counted string on the stack, or
16327: @code{machine-file} leaves an address, count pair of the filename on the
16328: stack.
16329:
16330: The feature list is typically controlled using @code{SetValue}, generic
16331: files that are used by several projects can use @code{DefaultValue}
16332: instead. Both functions work like @code{Value}, when the value isn't
16333: defined, but @code{SetValue} works like @code{to} if the value is
16334: defined, and @code{DefaultValue} doesn't set anything, if the value is
16335: defined.
16336:
16337: @example
16338: \ generic mach file for pc gforth 03sep97jaw
16339:
16340: true DefaultValue NIL \ relocating
16341:
16342: >ENVIRON
16343:
16344: true DefaultValue file \ controls the presence of the
16345: \ file access wordset
16346: true DefaultValue OS \ flag to indicate a operating system
16347:
16348: true DefaultValue prims \ true: primitives are c-code
16349:
16350: true DefaultValue floating \ floating point wordset is present
16351:
16352: true DefaultValue glocals \ gforth locals are present
16353: \ will be loaded
16354: true DefaultValue dcomps \ double number comparisons
16355:
16356: true DefaultValue hash \ hashing primitives are loaded/present
16357:
16358: true DefaultValue xconds \ used together with glocals,
16359: \ special conditionals supporting gforths'
16360: \ local variables
16361: true DefaultValue header \ save a header information
16362:
16363: true DefaultValue backtrace \ enables backtrace code
16364:
16365: false DefaultValue ec
16366: false DefaultValue crlf
16367:
16368: cell 2 = [IF] &32 [ELSE] &256 [THEN] KB DefaultValue kernel-size
16369:
16370: &16 KB DefaultValue stack-size
16371: &15 KB &512 + DefaultValue fstack-size
16372: &15 KB DefaultValue rstack-size
16373: &14 KB &512 + DefaultValue lstack-size
16374: @end example
16375:
16376: @node How the Cross Compiler Works, , Using the Cross Compiler, Cross Compiler
16377: @section How the Cross Compiler Works
16378:
16379: @node Bugs, Origin, Cross Compiler, Top
16380: @appendix Bugs
16381: @cindex bug reporting
16382:
16383: Known bugs are described in the file @file{BUGS} in the Gforth distribution.
16384:
16385: If you find a bug, please submit a bug report through
16386: @uref{https://savannah.gnu.org/bugs/?func=addbug&group=gforth}.
16387:
16388: @itemize @bullet
16389: @item
16390: A program (or a sequence of keyboard commands) that reproduces the bug.
16391: @item
16392: A description of what you think constitutes the buggy behaviour.
16393: @item
16394: The Gforth version used (it is announced at the start of an
16395: interactive Gforth session).
16396: @item
16397: The machine and operating system (on Unix
16398: systems @code{uname -a} will report this information).
16399: @item
16400: The installation options (you can find the configure options at the
16401: start of @file{config.status}) and configuration (@code{configure}
16402: output or @file{config.cache}).
16403: @item
16404: A complete list of changes (if any) you (or your installer) have made to the
16405: Gforth sources.
16406: @end itemize
16407:
16408: For a thorough guide on reporting bugs read @ref{Bug Reporting, , How
16409: to Report Bugs, gcc.info, GNU C Manual}.
16410:
16411:
16412: @node Origin, Forth-related information, Bugs, Top
16413: @appendix Authors and Ancestors of Gforth
16414:
16415: @section Authors and Contributors
16416: @cindex authors of Gforth
16417: @cindex contributors to Gforth
16418:
16419: The Gforth project was started in mid-1992 by Bernd Paysan and Anton
16420: Ertl. The third major author was Jens Wilke. Neal Crook contributed a
16421: lot to the manual. Assemblers and disassemblers were contributed by
16422: Andrew McKewan, Christian Pirker, Bernd Thallner, and Michal Revucky.
16423: Lennart Benschop (who was one of Gforth's first users, in mid-1993)
16424: and Stuart Ramsden inspired us with their continuous feedback. Lennart
16425: Benshop contributed @file{glosgen.fs}, while Stuart Ramsden has been
16426: working on automatic support for calling C libraries. Helpful comments
16427: also came from Paul Kleinrubatscher, Christian Pirker, Dirk Zoller,
16428: Marcel Hendrix, John Wavrik, Barrie Stott, Marc de Groot, Jorge
16429: Acerada, Bruce Hoyt, Robert Epprecht, Dennis Ruffer and David
16430: N. Williams. Since the release of Gforth-0.2.1 there were also helpful
16431: comments from many others; thank you all, sorry for not listing you
16432: here (but digging through my mailbox to extract your names is on my
16433: to-do list).
16434:
16435: Gforth also owes a lot to the authors of the tools we used (GCC, CVS,
16436: and autoconf, among others), and to the creators of the Internet: Gforth
16437: was developed across the Internet, and its authors did not meet
16438: physically for the first 4 years of development.
16439:
16440: @section Pedigree
16441: @cindex pedigree of Gforth
16442:
16443: Gforth descends from bigFORTH (1993) and fig-Forth. Of course, a
16444: significant part of the design of Gforth was prescribed by ANS Forth.
16445:
16446: Bernd Paysan wrote bigFORTH, a descendent from TurboForth, an unreleased
16447: 32 bit native code version of VolksForth for the Atari ST, written
16448: mostly by Dietrich Weineck.
16449:
16450: VolksForth was written by Klaus Schleisiek, Bernd Pennemann, Georg
16451: Rehfeld and Dietrich Weineck for the C64 (called UltraForth there) in
16452: the mid-80s and ported to the Atari ST in 1986. It descends from fig-Forth.
16453:
16454: @c Henry Laxen and Mike Perry wrote F83 as a model implementation of the
16455: @c Forth-83 standard. !! Pedigree? When?
16456:
16457: A team led by Bill Ragsdale implemented fig-Forth on many processors in
16458: 1979. Robert Selzer and Bill Ragsdale developed the original
16459: implementation of fig-Forth for the 6502 based on microForth.
16460:
16461: The principal architect of microForth was Dean Sanderson. microForth was
16462: FORTH, Inc.'s first off-the-shelf product. It was developed in 1976 for
16463: the 1802, and subsequently implemented on the 8080, the 6800 and the
16464: Z80.
16465:
16466: All earlier Forth systems were custom-made, usually by Charles Moore,
16467: who discovered (as he puts it) Forth during the late 60s. The first full
16468: Forth existed in 1971.
16469:
16470: A part of the information in this section comes from
16471: @cite{@uref{http://www.forth.com/Content/History/History1.htm,The
16472: Evolution of Forth}} by Elizabeth D. Rather, Donald R. Colburn and
16473: Charles H. Moore, presented at the HOPL-II conference and preprinted
16474: in SIGPLAN Notices 28(3), 1993. You can find more historical and
16475: genealogical information about Forth there. For a more general (and
16476: graphical) Forth family tree look see
16477: @cite{@uref{http://www.complang.tuwien.ac.at/forth/family-tree/},
16478: Forth Family Tree and Timeline}.
16479:
16480: @c ------------------------------------------------------------------
16481: @node Forth-related information, Licenses, Origin, Top
16482: @appendix Other Forth-related information
16483: @cindex Forth-related information
16484:
16485: @c anton: I threw most of this stuff out, because it can be found through
16486: @c the FAQ and the FAQ is more likely to be up-to-date.
16487:
16488: @cindex comp.lang.forth
16489: @cindex frequently asked questions
16490: There is an active news group (comp.lang.forth) discussing Forth
16491: (including Gforth) and Forth-related issues. Its
16492: @uref{http://www.complang.tuwien.ac.at/forth/faq/faq-general-2.html,FAQs}
16493: (frequently asked questions and their answers) contains a lot of
16494: information on Forth. You should read it before posting to
16495: comp.lang.forth.
16496:
16497: The ANS Forth standard is most usable in its
16498: @uref{http://www.taygeta.com/forth/dpans.html, HTML form}.
16499:
16500: @c ---------------------------------------------------
16501: @node Licenses, Word Index, Forth-related information, Top
16502: @appendix Licenses
16503:
16504: @menu
16505: * GNU Free Documentation License:: License for copying this manual.
16506: * Copying:: GPL (for copying this software).
16507: @end menu
16508:
16509: @node GNU Free Documentation License, Copying, Licenses, Licenses
16510: @appendixsec GNU Free Documentation License
16511: @include fdl.texi
16512:
16513: @node Copying, , GNU Free Documentation License, Licenses
16514: @appendixsec GNU GENERAL PUBLIC LICENSE
16515: @include gpl.texi
16516:
16517:
16518:
16519: @c ------------------------------------------------------------------
16520: @node Word Index, Concept Index, Licenses, Top
16521: @unnumbered Word Index
16522:
16523: This index is a list of Forth words that have ``glossary'' entries
16524: within this manual. Each word is listed with its stack effect and
16525: wordset.
16526:
16527: @printindex fn
16528:
16529: @c anton: the name index seems superfluous given the word and concept indices.
16530:
16531: @c @node Name Index, Concept Index, Word Index, Top
16532: @c @unnumbered Name Index
16533:
16534: @c This index is a list of Forth words that have ``glossary'' entries
16535: @c within this manual.
16536:
16537: @c @printindex ky
16538:
16539: @c -------------------------------------------------------
16540: @node Concept Index, , Word Index, Top
16541: @unnumbered Concept and Word Index
16542:
16543: Not all entries listed in this index are present verbatim in the
16544: text. This index also duplicates, in abbreviated form, all of the words
16545: listed in the Word Index (only the names are listed for the words here).
16546:
16547: @printindex cp
16548:
16549: @bye
16550:
16551:
16552:
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/cvsweb/icons/back.gif){kind=icon}
Return to gforth.ds CVS log ![[TXT]](/cvsweb/icons/text.gif){kind=icon}
![[DIR]](/cvsweb/icons/dir.gif){kind=icon}
Up to [gforth] / gforth / doc