Annotation of gforth/doc/gforth.ds, revision 1.62
1.1 anton 1: \input texinfo @c -*-texinfo-*-
2: @comment The source is gforth.ds, from which gforth.texi is generated
1.28 crook 3:
1.21 crook 4: @comment TODO: nac29jan99 - a list of things to add in the next edit:
1.28 crook 5: @comment 1. x-ref all ambiguous or implementation-defined features?
6: @comment 2. Describe the use of Auser Avariable AConstant A, etc.
7: @comment 3. words in miscellaneous section need a home.
8: @comment 4. search for TODO for other minor and major works required.
9: @comment 5. [rats] change all @var to @i in Forth source so that info
10: @comment file looks decent.
1.36 anton 11: @c Not an improvement IMO - anton
12: @c and anyway, this should be taken up
13: @c with Karl Berry (the texinfo guy) - anton
1.29 crook 14: @comment .. would be useful to have a word that identified all deferred words
15: @comment should semantics stuff in intro be moved to another section
16:
1.28 crook 17:
1.1 anton 18: @comment %**start of header (This is for running Texinfo on a region.)
19: @setfilename gforth.info
20: @settitle Gforth Manual
21: @dircategory GNU programming tools
22: @direntry
23: * Gforth: (gforth). A fast interpreter for the Forth language.
24: @end direntry
1.49 anton 25: @c The Texinfo manual also recommends doing this, but for Gforth it may
26: @c not make much sense
27: @c @dircategory Individual utilities
28: @c @direntry
29: @c * Gforth: (gforth)Invoking Gforth. gforth, gforth-fast, gforthmi
30: @c @end direntry
31:
1.1 anton 32: @comment @setchapternewpage odd
1.29 crook 33: @comment TODO this gets left in by HTML converter
1.12 anton 34: @macro progstyle {}
35: Programming style note:
1.3 anton 36: @end macro
1.48 anton 37:
38: @macro assignment {}
39: @table @i
40: @item Assignment:
41: @end macro
42: @macro endassignment {}
43: @end table
44: @end macro
45:
1.1 anton 46: @comment %**end of header (This is for running Texinfo on a region.)
47:
1.29 crook 48:
49: @comment ----------------------------------------------------------
50: @comment macros for beautifying glossary entries
51: @comment if these are used, need to strip them out for HTML converter
52: @comment else they get repeated verbatim in HTML output.
53: @comment .. not working yet.
54:
55: @macro GLOSS-START {}
56: @iftex
57: @ninerm
58: @end iftex
59: @end macro
60:
61: @macro GLOSS-END {}
62: @iftex
63: @rm
64: @end iftex
65: @end macro
66:
67: @comment ----------------------------------------------------------
68:
69:
1.10 anton 70: @include version.texi
71:
1.49 anton 72: @ifnottex
1.11 anton 73: This file documents Gforth @value{VERSION}
1.1 anton 74:
1.62 ! crook 75: Copyright @copyright{} 1995--2000 Free Software Foundation, Inc.
1.1 anton 76:
77: Permission is granted to make and distribute verbatim copies of
78: this manual provided the copyright notice and this permission notice
79: are preserved on all copies.
80:
81: @ignore
82: Permission is granted to process this file through TeX and print the
83: results, provided the printed document carries a copying permission
84: notice identical to this one except for the removal of this paragraph
85: (this paragraph not being relevant to the printed manual).
86:
87: @end ignore
88: Permission is granted to copy and distribute modified versions of this
89: manual under the conditions for verbatim copying, provided also that the
90: sections entitled "Distribution" and "General Public License" are
91: included exactly as in the original, and provided that the entire
92: resulting derived work is distributed under the terms of a permission
93: notice identical to this one.
94:
95: Permission is granted to copy and distribute translations of this manual
96: into another language, under the above conditions for modified versions,
97: except that the sections entitled "Distribution" and "General Public
98: License" may be included in a translation approved by the author instead
99: of in the original English.
1.49 anton 100: @end ifnottex
1.1 anton 101:
102: @finalout
103: @titlepage
104: @sp 10
105: @center @titlefont{Gforth Manual}
106: @sp 2
1.11 anton 107: @center for version @value{VERSION}
1.1 anton 108: @sp 2
1.34 anton 109: @center Neal Crook
1.1 anton 110: @center Anton Ertl
1.6 pazsan 111: @center Bernd Paysan
1.5 anton 112: @center Jens Wilke
1.1 anton 113: @sp 3
1.47 crook 114: @center This manual is permanently under construction and was last updated on 15-Mar-2000
1.1 anton 115:
116: @comment The following two commands start the copyright page.
117: @page
118: @vskip 0pt plus 1filll
1.62 ! crook 119: Copyright @copyright{} 1995--2000 Free Software Foundation, Inc.
1.1 anton 120:
121: @comment !! Published by ... or You can get a copy of this manual ...
122:
123: Permission is granted to make and distribute verbatim copies of
124: this manual provided the copyright notice and this permission notice
125: are preserved on all copies.
126:
127: Permission is granted to copy and distribute modified versions of this
128: manual under the conditions for verbatim copying, provided also that the
129: sections entitled "Distribution" and "General Public License" are
130: included exactly as in the original, and provided that the entire
131: resulting derived work is distributed under the terms of a permission
132: notice identical to this one.
133:
134: Permission is granted to copy and distribute translations of this manual
135: into another language, under the above conditions for modified versions,
136: except that the sections entitled "Distribution" and "General Public
137: License" may be included in a translation approved by the author instead
138: of in the original English.
139: @end titlepage
140:
141: @node Top, License, (dir), (dir)
1.49 anton 142: @ifnottex
1.1 anton 143: Gforth is a free implementation of ANS Forth available on many
1.11 anton 144: personal machines. This manual corresponds to version @value{VERSION}.
1.49 anton 145: @end ifnottex
1.1 anton 146:
147: @menu
1.21 crook 148: * License:: The GPL
1.26 crook 149: * Goals:: About the Gforth Project
1.29 crook 150: * Gforth Environment:: Starting (and exiting) Gforth
1.48 anton 151: * Tutorial:: Hands-on Forth Tutorial
1.21 crook 152: * Introduction:: An introduction to ANS Forth
1.1 anton 153: * Words:: Forth words available in Gforth
1.24 anton 154: * Error messages:: How to interpret them
1.1 anton 155: * Tools:: Programming tools
156: * ANS conformance:: Implementation-defined options etc.
157: * Model:: The abstract machine of Gforth
158: * Integrating Gforth:: Forth as scripting language for applications
159: * Emacs and Gforth:: The Gforth Mode
160: * Image Files:: @code{.fi} files contain compiled code
161: * Engine:: The inner interpreter and the primitives
1.24 anton 162: * Binding to System Library::
1.13 pazsan 163: * Cross Compiler:: The Cross Compiler
1.1 anton 164: * Bugs:: How to report them
165: * Origin:: Authors and ancestors of Gforth
1.21 crook 166: * Forth-related information:: Books and places to look on the WWW
1.1 anton 167: * Word Index:: An item for each Forth word
1.41 anton 168: * Name Index:: Forth words, only names listed
1.1 anton 169: * Concept Index:: A menu covering many topics
1.12 anton 170:
1.48 anton 171: @detailmenu --- The Detailed Node Listing ---
1.12 anton 172:
1.26 crook 173: Goals of Gforth
174:
175: * Gforth Extensions Sinful?::
176:
1.29 crook 177: Gforth Environment
178:
1.32 anton 179: * Invoking Gforth:: Getting in
180: * Leaving Gforth:: Getting out
181: * Command-line editing::
1.48 anton 182: * Upper and lower case::
183: * Environment variables:: that affect how Gforth starts up
1.32 anton 184: * Gforth Files:: What gets installed and where
1.48 anton 185: * Startup speed:: When 35ms is not fast enough ...
186:
187: Forth Tutorial
188:
189: * Starting Gforth Tutorial::
190: * Syntax Tutorial::
191: * Crash Course Tutorial::
192: * Stack Tutorial::
193: * Arithmetics Tutorial::
194: * Stack Manipulation Tutorial::
195: * Using files for Forth code Tutorial::
196: * Comments Tutorial::
197: * Colon Definitions Tutorial::
198: * Decompilation Tutorial::
199: * Stack-Effect Comments Tutorial::
200: * Types Tutorial::
201: * Factoring Tutorial::
202: * Designing the stack effect Tutorial::
203: * Local Variables Tutorial::
204: * Conditional execution Tutorial::
205: * Flags and Comparisons Tutorial::
206: * General Loops Tutorial::
207: * Counted loops Tutorial::
208: * Recursion Tutorial::
209: * Leaving definitions or loops Tutorial::
210: * Return Stack Tutorial::
211: * Memory Tutorial::
212: * Characters and Strings Tutorial::
213: * Alignment Tutorial::
214: * Interpretation and Compilation Semantics and Immediacy Tutorial::
215: * Execution Tokens Tutorial::
216: * Exceptions Tutorial::
217: * Defining Words Tutorial::
218: * Arrays and Records Tutorial::
219: * POSTPONE Tutorial::
220: * Literal Tutorial::
221: * Advanced macros Tutorial::
222: * Compilation Tokens Tutorial::
223: * Wordlists and Search Order Tutorial::
1.29 crook 224:
1.24 anton 225: An Introduction to ANS Forth
226:
227: * Introducing the Text Interpreter::
228: * Stacks and Postfix notation::
229: * Your first definition::
230: * How does that work?::
231: * Forth is written in Forth::
232: * Review - elements of a Forth system::
1.29 crook 233: * Where to go next::
1.24 anton 234: * Exercises::
235:
1.12 anton 236: Forth Words
237:
238: * Notation::
1.21 crook 239: * Comments::
240: * Boolean Flags::
1.12 anton 241: * Arithmetic::
242: * Stack Manipulation::
243: * Memory::
244: * Control Structures::
245: * Defining Words::
1.47 crook 246: * Interpretation and Compilation Semantics::
247: * Tokens for Words::
1.21 crook 248: * The Text Interpreter::
249: * Word Lists::
250: * Environmental Queries::
1.12 anton 251: * Files::
252: * Blocks::
253: * Other I/O::
254: * Programming Tools::
255: * Assembler and Code Words::
256: * Threading Words::
1.26 crook 257: * Locals::
258: * Structures::
259: * Object-oriented Forth::
1.21 crook 260: * Passing Commands to the OS::
1.47 crook 261: * Keeping track of Time::
1.21 crook 262: * Miscellaneous Words::
1.12 anton 263:
264: Arithmetic
265:
266: * Single precision::
267: * Bitwise operations::
1.21 crook 268: * Double precision:: Double-cell integer arithmetic
269: * Numeric comparison::
1.32 anton 270: * Mixed precision:: Operations with single and double-cell integers
1.12 anton 271: * Floating Point::
272:
273: Stack Manipulation
274:
275: * Data stack::
276: * Floating point stack::
277: * Return stack::
278: * Locals stack::
279: * Stack pointer manipulation::
280:
281: Memory
282:
1.32 anton 283: * Memory model::
284: * Dictionary allocation::
285: * Heap Allocation::
286: * Memory Access::
287: * Address arithmetic::
288: * Memory Blocks::
1.12 anton 289:
290: Control Structures
291:
1.41 anton 292: * Selection:: IF ... ELSE ... ENDIF
293: * Simple Loops:: BEGIN ...
1.32 anton 294: * Counted Loops:: DO
295: * Arbitrary control structures::
296: * Calls and returns::
1.12 anton 297: * Exception Handling::
298:
299: Defining Words
300:
1.45 crook 301: * CREATE::
1.44 crook 302: * Variables:: Variables and user variables
303: * Constants::
304: * Values:: Initialised variables
1.32 anton 305: * Colon Definitions::
1.44 crook 306: * Anonymous Definitions:: Definitions without names
1.32 anton 307: * User-defined Defining Words::
1.44 crook 308: * Deferred words:: Allow forward references
309: * Aliases::
1.32 anton 310: * Supplying names::
1.47 crook 311:
312: Interpretation and Compilation Semantics
313:
1.44 crook 314: * Combined words::
1.12 anton 315:
1.21 crook 316: The Text Interpreter
317:
1.29 crook 318: * Input Sources::
1.21 crook 319: * Number Conversion::
320: * Interpret/Compile states::
321: * Literals::
322: * Interpreter Directives::
323:
1.26 crook 324: Word Lists
325:
326: * Why use word lists?::
327: * Word list examples::
328:
329: Files
330:
1.48 anton 331: * Forth source files::
332: * General files::
333: * Search Paths::
334:
335: Search Paths
336:
337: * Forth Search Paths::
1.26 crook 338: * General Search Paths::
339:
340: Other I/O
341:
1.32 anton 342: * Simple numeric output:: Predefined formats
343: * Formatted numeric output:: Formatted (pictured) output
344: * String Formats:: How Forth stores strings in memory
345: * Displaying characters and strings:: Other stuff
346: * Input:: Input
1.26 crook 347:
348: Programming Tools
349:
350: * Debugging:: Simple and quick.
351: * Assertions:: Making your programs self-checking.
1.46 pazsan 352: * Singlestep Debugger:: Executing your program word by word.
1.26 crook 353:
354: Locals
355:
356: * Gforth locals::
357: * ANS Forth locals::
358:
359: Gforth locals
360:
361: * Where are locals visible by name?::
362: * How long do locals live?::
363: * Programming Style::
364: * Implementation::
365:
1.12 anton 366: Structures
367:
368: * Why explicit structure support?::
369: * Structure Usage::
370: * Structure Naming Convention::
371: * Structure Implementation::
372: * Structure Glossary::
373:
374: Object-oriented Forth
375:
1.48 anton 376: * Why object-oriented programming?::
377: * Object-Oriented Terminology::
378: * Objects::
379: * OOF::
380: * Mini-OOF::
1.23 crook 381: * Comparison with other object models::
1.12 anton 382:
1.24 anton 383: The @file{objects.fs} model
1.12 anton 384:
385: * Properties of the Objects model::
386: * Basic Objects Usage::
1.41 anton 387: * The Objects base class::
1.12 anton 388: * Creating objects::
389: * Object-Oriented Programming Style::
390: * Class Binding::
391: * Method conveniences::
392: * Classes and Scoping::
1.41 anton 393: * Dividing classes::
1.12 anton 394: * Object Interfaces::
395: * Objects Implementation::
396: * Objects Glossary::
397:
1.24 anton 398: The @file{oof.fs} model
1.12 anton 399:
400: * Properties of the OOF model::
401: * Basic OOF Usage::
1.23 crook 402: * The OOF base class::
1.12 anton 403: * Class Declaration::
404: * Class Implementation::
405:
1.24 anton 406: The @file{mini-oof.fs} model
1.23 crook 407:
1.48 anton 408: * Basic Mini-OOF Usage::
409: * Mini-OOF Example::
410: * Mini-OOF Implementation::
411: * Comparison with other object models::
1.23 crook 412:
1.12 anton 413: Tools
414:
415: * ANS Report:: Report the words used, sorted by wordset.
416:
417: ANS conformance
418:
419: * The Core Words::
420: * The optional Block word set::
421: * The optional Double Number word set::
422: * The optional Exception word set::
423: * The optional Facility word set::
424: * The optional File-Access word set::
425: * The optional Floating-Point word set::
426: * The optional Locals word set::
427: * The optional Memory-Allocation word set::
428: * The optional Programming-Tools word set::
429: * The optional Search-Order word set::
430:
431: The Core Words
432:
433: * core-idef:: Implementation Defined Options
434: * core-ambcond:: Ambiguous Conditions
435: * core-other:: Other System Documentation
436:
437: The optional Block word set
438:
439: * block-idef:: Implementation Defined Options
440: * block-ambcond:: Ambiguous Conditions
441: * block-other:: Other System Documentation
442:
443: The optional Double Number word set
444:
445: * double-ambcond:: Ambiguous Conditions
446:
447: The optional Exception word set
448:
449: * exception-idef:: Implementation Defined Options
450:
451: The optional Facility word set
452:
453: * facility-idef:: Implementation Defined Options
454: * facility-ambcond:: Ambiguous Conditions
455:
456: The optional File-Access word set
457:
458: * file-idef:: Implementation Defined Options
459: * file-ambcond:: Ambiguous Conditions
460:
461: The optional Floating-Point word set
462:
463: * floating-idef:: Implementation Defined Options
464: * floating-ambcond:: Ambiguous Conditions
465:
466: The optional Locals word set
467:
468: * locals-idef:: Implementation Defined Options
469: * locals-ambcond:: Ambiguous Conditions
470:
471: The optional Memory-Allocation word set
472:
473: * memory-idef:: Implementation Defined Options
474:
475: The optional Programming-Tools word set
476:
477: * programming-idef:: Implementation Defined Options
478: * programming-ambcond:: Ambiguous Conditions
479:
480: The optional Search-Order word set
481:
482: * search-idef:: Implementation Defined Options
483: * search-ambcond:: Ambiguous Conditions
484:
485: Image Files
486:
1.24 anton 487: * Image Licensing Issues:: Distribution terms for images.
488: * Image File Background:: Why have image files?
1.32 anton 489: * Non-Relocatable Image Files:: don't always work.
1.24 anton 490: * Data-Relocatable Image Files:: are better.
1.32 anton 491: * Fully Relocatable Image Files:: better yet.
1.24 anton 492: * Stack and Dictionary Sizes:: Setting the default sizes for an image.
1.32 anton 493: * Running Image Files:: @code{gforth -i @i{file}} or @i{file}.
1.24 anton 494: * Modifying the Startup Sequence:: and turnkey applications.
1.12 anton 495:
496: Fully Relocatable Image Files
497:
1.27 crook 498: * gforthmi:: The normal way
1.12 anton 499: * cross.fs:: The hard way
500:
501: Engine
502:
503: * Portability::
504: * Threading::
505: * Primitives::
506: * Performance::
507:
508: Threading
509:
510: * Scheduling::
511: * Direct or Indirect Threaded?::
512: * DOES>::
513:
514: Primitives
515:
516: * Automatic Generation::
517: * TOS Optimization::
518: * Produced code::
1.13 pazsan 519:
520: Cross Compiler
521:
522: * Using the Cross Compiler::
523: * How the Cross Compiler Works::
524:
1.24 anton 525: Other Forth-related information
1.21 crook 526:
527: * Internet resources::
528: * Books::
529: * The Forth Interest Group::
530: * Conferences::
531:
1.24 anton 532: @end detailmenu
1.1 anton 533: @end menu
534:
1.26 crook 535: @node License, Goals, Top, Top
1.1 anton 536: @unnumbered GNU GENERAL PUBLIC LICENSE
537: @center Version 2, June 1991
538:
539: @display
540: Copyright @copyright{} 1989, 1991 Free Software Foundation, Inc.
541: 675 Mass Ave, Cambridge, MA 02139, USA
542:
543: Everyone is permitted to copy and distribute verbatim copies
544: of this license document, but changing it is not allowed.
545: @end display
546:
547: @unnumberedsec Preamble
548:
549: The licenses for most software are designed to take away your
550: freedom to share and change it. By contrast, the GNU General Public
551: License is intended to guarantee your freedom to share and change free
552: software---to make sure the software is free for all its users. This
553: General Public License applies to most of the Free Software
554: Foundation's software and to any other program whose authors commit to
555: using it. (Some other Free Software Foundation software is covered by
556: the GNU Library General Public License instead.) You can apply it to
557: your programs, too.
558:
559: When we speak of free software, we are referring to freedom, not
560: price. Our General Public Licenses are designed to make sure that you
561: have the freedom to distribute copies of free software (and charge for
562: this service if you wish), that you receive source code or can get it
563: if you want it, that you can change the software or use pieces of it
564: in new free programs; and that you know you can do these things.
565:
566: To protect your rights, we need to make restrictions that forbid
567: anyone to deny you these rights or to ask you to surrender the rights.
568: These restrictions translate to certain responsibilities for you if you
569: distribute copies of the software, or if you modify it.
570:
571: For example, if you distribute copies of such a program, whether
572: gratis or for a fee, you must give the recipients all the rights that
573: you have. You must make sure that they, too, receive or can get the
574: source code. And you must show them these terms so they know their
575: rights.
576:
577: We protect your rights with two steps: (1) copyright the software, and
578: (2) offer you this license which gives you legal permission to copy,
579: distribute and/or modify the software.
580:
581: Also, for each author's protection and ours, we want to make certain
582: that everyone understands that there is no warranty for this free
583: software. If the software is modified by someone else and passed on, we
584: want its recipients to know that what they have is not the original, so
585: that any problems introduced by others will not reflect on the original
586: authors' reputations.
587:
588: Finally, any free program is threatened constantly by software
589: patents. We wish to avoid the danger that redistributors of a free
590: program will individually obtain patent licenses, in effect making the
591: program proprietary. To prevent this, we have made it clear that any
592: patent must be licensed for everyone's free use or not licensed at all.
593:
594: The precise terms and conditions for copying, distribution and
595: modification follow.
596:
597: @iftex
598: @unnumberedsec TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
599: @end iftex
1.49 anton 600: @ifnottex
1.1 anton 601: @center TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
1.49 anton 602: @end ifnottex
1.1 anton 603:
604: @enumerate 0
605: @item
606: This License applies to any program or other work which contains
607: a notice placed by the copyright holder saying it may be distributed
608: under the terms of this General Public License. The ``Program'', below,
609: refers to any such program or work, and a ``work based on the Program''
610: means either the Program or any derivative work under copyright law:
611: that is to say, a work containing the Program or a portion of it,
612: either verbatim or with modifications and/or translated into another
613: language. (Hereinafter, translation is included without limitation in
614: the term ``modification''.) Each licensee is addressed as ``you''.
615:
616: Activities other than copying, distribution and modification are not
617: covered by this License; they are outside its scope. The act of
618: running the Program is not restricted, and the output from the Program
619: is covered only if its contents constitute a work based on the
620: Program (independent of having been made by running the Program).
621: Whether that is true depends on what the Program does.
622:
623: @item
624: You may copy and distribute verbatim copies of the Program's
625: source code as you receive it, in any medium, provided that you
626: conspicuously and appropriately publish on each copy an appropriate
627: copyright notice and disclaimer of warranty; keep intact all the
628: notices that refer to this License and to the absence of any warranty;
629: and give any other recipients of the Program a copy of this License
630: along with the Program.
631:
632: You may charge a fee for the physical act of transferring a copy, and
633: you may at your option offer warranty protection in exchange for a fee.
634:
635: @item
636: You may modify your copy or copies of the Program or any portion
637: of it, thus forming a work based on the Program, and copy and
638: distribute such modifications or work under the terms of Section 1
639: above, provided that you also meet all of these conditions:
640:
641: @enumerate a
642: @item
643: You must cause the modified files to carry prominent notices
644: stating that you changed the files and the date of any change.
645:
646: @item
647: You must cause any work that you distribute or publish, that in
648: whole or in part contains or is derived from the Program or any
649: part thereof, to be licensed as a whole at no charge to all third
650: parties under the terms of this License.
651:
652: @item
653: If the modified program normally reads commands interactively
654: when run, you must cause it, when started running for such
655: interactive use in the most ordinary way, to print or display an
656: announcement including an appropriate copyright notice and a
657: notice that there is no warranty (or else, saying that you provide
658: a warranty) and that users may redistribute the program under
659: these conditions, and telling the user how to view a copy of this
660: License. (Exception: if the Program itself is interactive but
661: does not normally print such an announcement, your work based on
662: the Program is not required to print an announcement.)
663: @end enumerate
664:
665: These requirements apply to the modified work as a whole. If
666: identifiable sections of that work are not derived from the Program,
667: and can be reasonably considered independent and separate works in
668: themselves, then this License, and its terms, do not apply to those
669: sections when you distribute them as separate works. But when you
670: distribute the same sections as part of a whole which is a work based
671: on the Program, the distribution of the whole must be on the terms of
672: this License, whose permissions for other licensees extend to the
673: entire whole, and thus to each and every part regardless of who wrote it.
674:
675: Thus, it is not the intent of this section to claim rights or contest
676: your rights to work written entirely by you; rather, the intent is to
677: exercise the right to control the distribution of derivative or
678: collective works based on the Program.
679:
680: In addition, mere aggregation of another work not based on the Program
681: with the Program (or with a work based on the Program) on a volume of
682: a storage or distribution medium does not bring the other work under
683: the scope of this License.
684:
685: @item
686: You may copy and distribute the Program (or a work based on it,
687: under Section 2) in object code or executable form under the terms of
688: Sections 1 and 2 above provided that you also do one of the following:
689:
690: @enumerate a
691: @item
692: Accompany it with the complete corresponding machine-readable
693: source code, which must be distributed under the terms of Sections
694: 1 and 2 above on a medium customarily used for software interchange; or,
695:
696: @item
697: Accompany it with a written offer, valid for at least three
698: years, to give any third party, for a charge no more than your
699: cost of physically performing source distribution, a complete
700: machine-readable copy of the corresponding source code, to be
701: distributed under the terms of Sections 1 and 2 above on a medium
702: customarily used for software interchange; or,
703:
704: @item
705: Accompany it with the information you received as to the offer
706: to distribute corresponding source code. (This alternative is
707: allowed only for noncommercial distribution and only if you
708: received the program in object code or executable form with such
709: an offer, in accord with Subsection b above.)
710: @end enumerate
711:
712: The source code for a work means the preferred form of the work for
713: making modifications to it. For an executable work, complete source
714: code means all the source code for all modules it contains, plus any
715: associated interface definition files, plus the scripts used to
716: control compilation and installation of the executable. However, as a
717: special exception, the source code distributed need not include
718: anything that is normally distributed (in either source or binary
719: form) with the major components (compiler, kernel, and so on) of the
720: operating system on which the executable runs, unless that component
721: itself accompanies the executable.
722:
723: If distribution of executable or object code is made by offering
724: access to copy from a designated place, then offering equivalent
725: access to copy the source code from the same place counts as
726: distribution of the source code, even though third parties are not
727: compelled to copy the source along with the object code.
728:
729: @item
730: You may not copy, modify, sublicense, or distribute the Program
731: except as expressly provided under this License. Any attempt
732: otherwise to copy, modify, sublicense or distribute the Program is
733: void, and will automatically terminate your rights under this License.
734: However, parties who have received copies, or rights, from you under
735: this License will not have their licenses terminated so long as such
736: parties remain in full compliance.
737:
738: @item
739: You are not required to accept this License, since you have not
740: signed it. However, nothing else grants you permission to modify or
741: distribute the Program or its derivative works. These actions are
742: prohibited by law if you do not accept this License. Therefore, by
743: modifying or distributing the Program (or any work based on the
744: Program), you indicate your acceptance of this License to do so, and
745: all its terms and conditions for copying, distributing or modifying
746: the Program or works based on it.
747:
748: @item
749: Each time you redistribute the Program (or any work based on the
750: Program), the recipient automatically receives a license from the
751: original licensor to copy, distribute or modify the Program subject to
752: these terms and conditions. You may not impose any further
753: restrictions on the recipients' exercise of the rights granted herein.
754: You are not responsible for enforcing compliance by third parties to
755: this License.
756:
757: @item
758: If, as a consequence of a court judgment or allegation of patent
759: infringement or for any other reason (not limited to patent issues),
760: conditions are imposed on you (whether by court order, agreement or
761: otherwise) that contradict the conditions of this License, they do not
762: excuse you from the conditions of this License. If you cannot
763: distribute so as to satisfy simultaneously your obligations under this
764: License and any other pertinent obligations, then as a consequence you
765: may not distribute the Program at all. For example, if a patent
766: license would not permit royalty-free redistribution of the Program by
767: all those who receive copies directly or indirectly through you, then
768: the only way you could satisfy both it and this License would be to
769: refrain entirely from distribution of the Program.
770:
771: If any portion of this section is held invalid or unenforceable under
772: any particular circumstance, the balance of the section is intended to
773: apply and the section as a whole is intended to apply in other
774: circumstances.
775:
776: It is not the purpose of this section to induce you to infringe any
777: patents or other property right claims or to contest validity of any
778: such claims; this section has the sole purpose of protecting the
779: integrity of the free software distribution system, which is
780: implemented by public license practices. Many people have made
781: generous contributions to the wide range of software distributed
782: through that system in reliance on consistent application of that
783: system; it is up to the author/donor to decide if he or she is willing
784: to distribute software through any other system and a licensee cannot
785: impose that choice.
786:
787: This section is intended to make thoroughly clear what is believed to
788: be a consequence of the rest of this License.
789:
790: @item
791: If the distribution and/or use of the Program is restricted in
792: certain countries either by patents or by copyrighted interfaces, the
793: original copyright holder who places the Program under this License
794: may add an explicit geographical distribution limitation excluding
795: those countries, so that distribution is permitted only in or among
796: countries not thus excluded. In such case, this License incorporates
797: the limitation as if written in the body of this License.
798:
799: @item
800: The Free Software Foundation may publish revised and/or new versions
801: of the General Public License from time to time. Such new versions will
802: be similar in spirit to the present version, but may differ in detail to
803: address new problems or concerns.
804:
805: Each version is given a distinguishing version number. If the Program
806: specifies a version number of this License which applies to it and ``any
807: later version'', you have the option of following the terms and conditions
808: either of that version or of any later version published by the Free
809: Software Foundation. If the Program does not specify a version number of
810: this License, you may choose any version ever published by the Free Software
811: Foundation.
812:
813: @item
814: If you wish to incorporate parts of the Program into other free
815: programs whose distribution conditions are different, write to the author
816: to ask for permission. For software which is copyrighted by the Free
817: Software Foundation, write to the Free Software Foundation; we sometimes
818: make exceptions for this. Our decision will be guided by the two goals
819: of preserving the free status of all derivatives of our free software and
820: of promoting the sharing and reuse of software generally.
821:
822: @iftex
823: @heading NO WARRANTY
824: @end iftex
1.49 anton 825: @ifnottex
1.1 anton 826: @center NO WARRANTY
1.49 anton 827: @end ifnottex
1.1 anton 828:
829: @item
830: BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
831: FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN
832: OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
833: PROVIDE THE PROGRAM ``AS IS'' WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
834: OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
835: MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS
836: TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE
837: PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
838: REPAIR OR CORRECTION.
839:
840: @item
841: IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
842: WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
843: REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
844: INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
845: OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED
846: TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY
847: YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
848: PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
849: POSSIBILITY OF SUCH DAMAGES.
850: @end enumerate
851:
852: @iftex
853: @heading END OF TERMS AND CONDITIONS
854: @end iftex
1.49 anton 855: @ifnottex
1.1 anton 856: @center END OF TERMS AND CONDITIONS
1.49 anton 857: @end ifnottex
1.1 anton 858:
859: @page
860: @unnumberedsec How to Apply These Terms to Your New Programs
861:
862: If you develop a new program, and you want it to be of the greatest
863: possible use to the public, the best way to achieve this is to make it
864: free software which everyone can redistribute and change under these terms.
865:
866: To do so, attach the following notices to the program. It is safest
867: to attach them to the start of each source file to most effectively
868: convey the exclusion of warranty; and each file should have at least
869: the ``copyright'' line and a pointer to where the full notice is found.
870:
871: @smallexample
872: @var{one line to give the program's name and a brief idea of what it does.}
873: Copyright (C) 19@var{yy} @var{name of author}
874:
875: This program is free software; you can redistribute it and/or modify
876: it under the terms of the GNU General Public License as published by
877: the Free Software Foundation; either version 2 of the License, or
878: (at your option) any later version.
879:
880: This program is distributed in the hope that it will be useful,
881: but WITHOUT ANY WARRANTY; without even the implied warranty of
882: MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
883: GNU General Public License for more details.
884:
885: You should have received a copy of the GNU General Public License
886: along with this program; if not, write to the Free Software
887: Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
888: @end smallexample
889:
890: Also add information on how to contact you by electronic and paper mail.
891:
892: If the program is interactive, make it output a short notice like this
893: when it starts in an interactive mode:
894:
895: @smallexample
896: Gnomovision version 69, Copyright (C) 19@var{yy} @var{name of author}
897: Gnomovision comes with ABSOLUTELY NO WARRANTY; for details
898: type `show w'.
899: This is free software, and you are welcome to redistribute it
900: under certain conditions; type `show c' for details.
901: @end smallexample
902:
903: The hypothetical commands @samp{show w} and @samp{show c} should show
904: the appropriate parts of the General Public License. Of course, the
905: commands you use may be called something other than @samp{show w} and
906: @samp{show c}; they could even be mouse-clicks or menu items---whatever
907: suits your program.
908:
909: You should also get your employer (if you work as a programmer) or your
910: school, if any, to sign a ``copyright disclaimer'' for the program, if
911: necessary. Here is a sample; alter the names:
912:
913: @smallexample
914: Yoyodyne, Inc., hereby disclaims all copyright interest in the program
915: `Gnomovision' (which makes passes at compilers) written by James Hacker.
916:
917: @var{signature of Ty Coon}, 1 April 1989
918: Ty Coon, President of Vice
919: @end smallexample
920:
921: This General Public License does not permit incorporating your program into
922: proprietary programs. If your program is a subroutine library, you may
923: consider it more useful to permit linking proprietary applications with the
924: library. If this is what you want to do, use the GNU Library General
925: Public License instead of this License.
926:
927: @iftex
928: @unnumbered Preface
929: @cindex Preface
1.21 crook 930: This manual documents Gforth. Some introductory material is provided for
931: readers who are unfamiliar with Forth or who are migrating to Gforth
932: from other Forth compilers. However, this manual is primarily a
933: reference manual.
1.1 anton 934: @end iftex
935:
1.28 crook 936: @comment TODO much more blurb here.
1.26 crook 937:
938: @c ******************************************************************
1.29 crook 939: @node Goals, Gforth Environment, License, Top
1.26 crook 940: @comment node-name, next, previous, up
941: @chapter Goals of Gforth
942: @cindex goals of the Gforth project
943: The goal of the Gforth Project is to develop a standard model for
944: ANS Forth. This can be split into several subgoals:
945:
946: @itemize @bullet
947: @item
948: Gforth should conform to the ANS Forth Standard.
949: @item
950: It should be a model, i.e. it should define all the
951: implementation-dependent things.
952: @item
953: It should become standard, i.e. widely accepted and used. This goal
954: is the most difficult one.
955: @end itemize
956:
957: To achieve these goals Gforth should be
958: @itemize @bullet
959: @item
960: Similar to previous models (fig-Forth, F83)
961: @item
962: Powerful. It should provide for all the things that are considered
963: necessary today and even some that are not yet considered necessary.
964: @item
965: Efficient. It should not get the reputation of being exceptionally
966: slow.
967: @item
968: Free.
969: @item
970: Available on many machines/easy to port.
971: @end itemize
972:
973: Have we achieved these goals? Gforth conforms to the ANS Forth
974: standard. It may be considered a model, but we have not yet documented
975: which parts of the model are stable and which parts we are likely to
976: change. It certainly has not yet become a de facto standard, but it
977: appears to be quite popular. It has some similarities to and some
978: differences from previous models. It has some powerful features, but not
979: yet everything that we envisioned. We certainly have achieved our
980: execution speed goals (@pxref{Performance}). It is free and available
981: on many machines.
982:
983: @menu
984: * Gforth Extensions Sinful?::
985: @end menu
986:
1.48 anton 987: @node Gforth Extensions Sinful?, , Goals, Goals
1.26 crook 988: @comment node-name, next, previous, up
989: @section Is it a Sin to use Gforth Extensions?
990: @cindex Gforth extensions
991:
992: If you've been paying attention, you will have realised that there is an
993: ANS (American National Standard) for Forth. As you read through the rest
1.29 crook 994: of this manual, you will see documentation for @i{Standard} words, and
995: documentation for some appealing Gforth @i{extensions}. You might ask
996: yourself the question: @i{``Given that there is a standard, would I be
1.45 crook 997: committing a sin if I use (non-Standard) Gforth extensions?''}
1.26 crook 998:
999: The answer to that question is somewhat pragmatic and somewhat
1000: philosophical. Consider these points:
1001:
1002: @itemize @bullet
1003: @item
1004: A number of the Gforth extensions can be implemented in ANS Forth using
1005: files provided in the @file{compat/} directory. These are mentioned in
1006: the text in passing.
1007: @item
1008: Forth has a rich historical precedent for programmers taking advantage
1009: of implementation-dependent features of their tools (for example,
1010: relying on a knowledge of the dictionary structure). Sometimes these
1011: techniques are necessary to extract every last bit of performance from
1012: the hardware, sometimes they are just a programming shorthand.
1013: @item
1014: The best way to break the rules is to know what the rules are. To learn
1015: the rules, there is no substitute for studying the text of the Standard
1016: itself. In particular, Appendix A of the Standard (@var{Rationale})
1017: provides a valuable insight into the thought processes of the technical
1018: committee.
1019: @item
1020: The best reason to break a rule is because you have to; because it's
1021: more productive to do that, because it makes your code run fast enough
1022: or because you can see no Standard way to achieve what you want to
1023: achieve.
1024: @end itemize
1025:
1026: The tool @file{ans-report.fs} (@pxref{ANS Report}) makes it easy to
1027: analyse your program and determine what non-Standard definitions it
1028: relies upon.
1029:
1.29 crook 1030:
1.26 crook 1031: @c ******************************************************************
1.48 anton 1032: @node Gforth Environment, Tutorial, Goals, Top
1.29 crook 1033: @chapter Gforth Environment
1034: @cindex Gforth environment
1.21 crook 1035:
1.45 crook 1036: Note: ultimately, the Gforth man page will be auto-generated from the
1.29 crook 1037: material in this chapter.
1.21 crook 1038:
1039: @menu
1.29 crook 1040: * Invoking Gforth:: Getting in
1041: * Leaving Gforth:: Getting out
1042: * Command-line editing::
1.48 anton 1043: * Upper and lower case::
1044: * Environment variables:: that affect how Gforth starts up
1.29 crook 1045: * Gforth Files:: What gets installed and where
1.48 anton 1046: * Startup speed:: When 35ms is not fast enough ...
1.21 crook 1047: @end menu
1048:
1.49 anton 1049: For related information about the creation of images see @ref{Image Files}.
1.29 crook 1050:
1.21 crook 1051: @comment ----------------------------------------------
1.48 anton 1052: @node Invoking Gforth, Leaving Gforth, Gforth Environment, Gforth Environment
1.29 crook 1053: @section Invoking Gforth
1054: @cindex invoking Gforth
1055: @cindex running Gforth
1056: @cindex command-line options
1057: @cindex options on the command line
1058: @cindex flags on the command line
1.21 crook 1059:
1.30 anton 1060: Gforth is made up of two parts; an executable ``engine'' (named
1061: @file{gforth} or @file{gforth-fast}) and an image file. To start it, you
1062: will usually just say @code{gforth} -- this automatically loads the
1063: default image file @file{gforth.fi}. In many other cases the default
1064: Gforth image will be invoked like this:
1.21 crook 1065: @example
1.30 anton 1066: gforth [file | -e forth-code] ...
1.21 crook 1067: @end example
1.29 crook 1068: @noindent
1069: This interprets the contents of the files and the Forth code in the order they
1070: are given.
1.21 crook 1071:
1.30 anton 1072: In addition to the @file{gforth} engine, there is also an engine called
1073: @file{gforth-fast}, which is faster, but gives less informative error
1074: messages (@pxref{Error messages}).
1075:
1.29 crook 1076: In general, the command line looks like this:
1.21 crook 1077:
1078: @example
1.30 anton 1079: gforth[-fast] [engine options] [image options]
1.21 crook 1080: @end example
1081:
1.30 anton 1082: The engine options must come before the rest of the command
1.29 crook 1083: line. They are:
1.26 crook 1084:
1.29 crook 1085: @table @code
1086: @cindex -i, command-line option
1087: @cindex --image-file, command-line option
1088: @item --image-file @i{file}
1089: @itemx -i @i{file}
1090: Loads the Forth image @i{file} instead of the default
1091: @file{gforth.fi} (@pxref{Image Files}).
1.21 crook 1092:
1.39 anton 1093: @cindex --appl-image, command-line option
1094: @item --appl-image @i{file}
1095: Loads the image @i{file} and leaves all further command-line arguments
1096: to the image (instead of processing them as options). This is useful
1097: for building executable application images on Unix, built with
1098: @code{gforthmi --application ...}.
1099:
1.29 crook 1100: @cindex --path, command-line option
1101: @cindex -p, command-line option
1102: @item --path @i{path}
1103: @itemx -p @i{path}
1104: Uses @i{path} for searching the image file and Forth source code files
1105: instead of the default in the environment variable @code{GFORTHPATH} or
1106: the path specified at installation time (e.g.,
1107: @file{/usr/local/share/gforth/0.2.0:.}). A path is given as a list of
1108: directories, separated by @samp{:} (on Unix) or @samp{;} (on other OSs).
1.21 crook 1109:
1.29 crook 1110: @cindex --dictionary-size, command-line option
1111: @cindex -m, command-line option
1112: @cindex @i{size} parameters for command-line options
1113: @cindex size of the dictionary and the stacks
1114: @item --dictionary-size @i{size}
1115: @itemx -m @i{size}
1116: Allocate @i{size} space for the Forth dictionary space instead of
1117: using the default specified in the image (typically 256K). The
1118: @i{size} specification for this and subsequent options consists of
1119: an integer and a unit (e.g.,
1120: @code{4M}). The unit can be one of @code{b} (bytes), @code{e} (element
1121: size, in this case Cells), @code{k} (kilobytes), @code{M} (Megabytes),
1122: @code{G} (Gigabytes), and @code{T} (Terabytes). If no unit is specified,
1123: @code{e} is used.
1.21 crook 1124:
1.29 crook 1125: @cindex --data-stack-size, command-line option
1126: @cindex -d, command-line option
1127: @item --data-stack-size @i{size}
1128: @itemx -d @i{size}
1129: Allocate @i{size} space for the data stack instead of using the
1130: default specified in the image (typically 16K).
1.21 crook 1131:
1.29 crook 1132: @cindex --return-stack-size, command-line option
1133: @cindex -r, command-line option
1134: @item --return-stack-size @i{size}
1135: @itemx -r @i{size}
1136: Allocate @i{size} space for the return stack instead of using the
1137: default specified in the image (typically 15K).
1.21 crook 1138:
1.29 crook 1139: @cindex --fp-stack-size, command-line option
1140: @cindex -f, command-line option
1141: @item --fp-stack-size @i{size}
1142: @itemx -f @i{size}
1143: Allocate @i{size} space for the floating point stack instead of
1144: using the default specified in the image (typically 15.5K). In this case
1145: the unit specifier @code{e} refers to floating point numbers.
1.21 crook 1146:
1.48 anton 1147: @cindex --locals-stack-size, command-line option
1148: @cindex -l, command-line option
1149: @item --locals-stack-size @i{size}
1150: @itemx -l @i{size}
1151: Allocate @i{size} space for the locals stack instead of using the
1152: default specified in the image (typically 14.5K).
1153:
1154: @cindex -h, command-line option
1155: @cindex --help, command-line option
1156: @item --help
1157: @itemx -h
1158: Print a message about the command-line options
1159:
1160: @cindex -v, command-line option
1161: @cindex --version, command-line option
1162: @item --version
1163: @itemx -v
1164: Print version and exit
1165:
1166: @cindex --debug, command-line option
1167: @item --debug
1168: Print some information useful for debugging on startup.
1169:
1170: @cindex --offset-image, command-line option
1171: @item --offset-image
1172: Start the dictionary at a slightly different position than would be used
1173: otherwise (useful for creating data-relocatable images,
1174: @pxref{Data-Relocatable Image Files}).
1175:
1176: @cindex --no-offset-im, command-line option
1177: @item --no-offset-im
1178: Start the dictionary at the normal position.
1179:
1180: @cindex --clear-dictionary, command-line option
1181: @item --clear-dictionary
1182: Initialize all bytes in the dictionary to 0 before loading the image
1183: (@pxref{Data-Relocatable Image Files}).
1184:
1185: @cindex --die-on-signal, command-line-option
1186: @item --die-on-signal
1187: Normally Gforth handles most signals (e.g., the user interrupt SIGINT,
1188: or the segmentation violation SIGSEGV) by translating it into a Forth
1189: @code{THROW}. With this option, Gforth exits if it receives such a
1190: signal. This option is useful when the engine and/or the image might be
1191: severely broken (such that it causes another signal before recovering
1192: from the first); this option avoids endless loops in such cases.
1193: @end table
1194:
1195: @cindex loading files at startup
1196: @cindex executing code on startup
1197: @cindex batch processing with Gforth
1198: As explained above, the image-specific command-line arguments for the
1199: default image @file{gforth.fi} consist of a sequence of filenames and
1200: @code{-e @var{forth-code}} options that are interpreted in the sequence
1201: in which they are given. The @code{-e @var{forth-code}} or
1202: @code{--evaluate @var{forth-code}} option evaluates the Forth
1203: code. This option takes only one argument; if you want to evaluate more
1204: Forth words, you have to quote them or use @code{-e} several times. To exit
1205: after processing the command line (instead of entering interactive mode)
1206: append @code{-e bye} to the command line.
1207:
1208: @cindex versions, invoking other versions of Gforth
1209: If you have several versions of Gforth installed, @code{gforth} will
1210: invoke the version that was installed last. @code{gforth-@i{version}}
1211: invokes a specific version. If your environment contains the variable
1212: @code{GFORTHPATH}, you may want to override it by using the
1213: @code{--path} option.
1214:
1215: Not yet implemented:
1216: On startup the system first executes the system initialization file
1217: (unless the option @code{--no-init-file} is given; note that the system
1218: resulting from using this option may not be ANS Forth conformant). Then
1219: the user initialization file @file{.gforth.fs} is executed, unless the
1.62 ! crook 1220: option @code{--no-rc} is given; this file is searched for in @file{.},
1.48 anton 1221: then in @file{~}, then in the normal path (see above).
1222:
1223:
1224:
1225: @comment ----------------------------------------------
1226: @node Leaving Gforth, Command-line editing, Invoking Gforth, Gforth Environment
1227: @section Leaving Gforth
1228: @cindex Gforth - leaving
1229: @cindex leaving Gforth
1230:
1231: You can leave Gforth by typing @code{bye} or @kbd{Ctrl-d} (at the start
1232: of a line) or (if you invoked Gforth with the @code{--die-on-signal}
1233: option) @kbd{Ctrl-c}. When you leave Gforth, all of your definitions and
1.49 anton 1234: data are discarded. For ways of saving the state of the system before
1235: leaving Gforth see @ref{Image Files}.
1.48 anton 1236:
1237: doc-bye
1238:
1239:
1240: @comment ----------------------------------------------
1241: @node Command-line editing, Upper and lower case, Leaving Gforth, Gforth Environment
1242: @section Command-line editing
1243: @cindex command-line editing
1244:
1245: Gforth maintains a history file that records every line that you type to
1246: the text interpreter. This file is preserved between sessions, and is
1247: used to provide a command-line recall facility; if you type @kbd{Ctrl-P}
1248: repeatedly you can recall successively older commands from this (or
1249: previous) session(s). The full list of command-line editing facilities is:
1250:
1251: @itemize @bullet
1252: @item
1253: @kbd{Ctrl-p} (``previous'') (or up-arrow) to recall successively older
1254: commands from the history buffer.
1255: @item
1256: @kbd{Ctrl-n} (``next'') (or down-arrow) to recall successively newer commands
1257: from the history buffer.
1258: @item
1259: @kbd{Ctrl-f} (or right-arrow) to move the cursor right, non-destructively.
1260: @item
1261: @kbd{Ctrl-b} (or left-arrow) to move the cursor left, non-destructively.
1262: @item
1263: @kbd{Ctrl-h} (backspace) to delete the character to the left of the cursor,
1264: closing up the line.
1265: @item
1266: @kbd{Ctrl-k} to delete (``kill'') from the cursor to the end of the line.
1267: @item
1268: @kbd{Ctrl-a} to move the cursor to the start of the line.
1269: @item
1270: @kbd{Ctrl-e} to move the cursor to the end of the line.
1271: @item
1272: @key{RET} (@kbd{Ctrl-m}) or @key{LFD} (@kbd{Ctrl-j}) to submit the current
1273: line.
1274: @item
1275: @key{TAB} to step through all possible full-word completions of the word
1276: currently being typed.
1277: @item
1278: @kbd{Ctrl-d} at the start of the line to terminate Gforth (gracefully,
1279: using @code{bye}).
1280: @end itemize
1281:
1282: When editing, displayable characters are inserted to the left of the
1283: cursor position; the line is always in ``insert'' (as opposed to
1284: ``overstrike'') mode.
1285:
1286: @cindex history file
1287: @cindex @file{.gforth-history}
1288: On Unix systems, the history file is @file{~/.gforth-history} by
1289: default@footnote{i.e. it is stored in the user's home directory.}. You
1290: can find out the name and location of your history file using:
1291:
1292: @example
1293: history-file type \ Unix-class systems
1294:
1295: history-file type \ Other systems
1296: history-dir type
1297: @end example
1298:
1299: If you enter long definitions by hand, you can use a text editor to
1300: paste them out of the history file into a Forth source file for reuse at
1301: a later time.
1302:
1303: Gforth never trims the size of the history file, so you should do this
1304: periodically, if necessary.
1305:
1306: @comment this is all defined in history.fs
1307: @comment NAC TODO the ctrl-D behaviour can either do a bye or a beep.. how is that option
1308: @comment chosen?
1309:
1310:
1311:
1312: @comment ----------------------------------------------
1313: @node Upper and lower case, Environment variables, Command-line editing, Gforth Environment
1314: @section Upper and lower case
1315: @cindex case-sensitivity
1316: @cindex upper and lower case
1317:
1318: Gforth is case-insensitive; you can enter definitions and invoke
1319: Standard words using upper, lower or mixed case (however,
1320: @pxref{core-idef, Implementation-defined options, Implementation-defined
1321: options}).
1322:
1323: ANS Forth only @i{requires} implementations to recognise Standard words
1324: when they are typed entirely in upper case. Therefore, a Standard
1325: program must use upper case for all Standard words. You can use whatever
1.62 ! crook 1326: case you like for words that you define, but in a Standard program you
1.48 anton 1327: have to use the words in the same case that you defined them.
1328:
1329: Gforth supports case sensitivity through @code{table}s (case-sensitive
1330: wordlists, @pxref{Word Lists}).
1331:
1.62 ! crook 1332: Two people have asked how to convert Gforth to be case-sensitive; while
1.48 anton 1333: we think this is a bad idea, you can change all wordlists into tables
1334: like this:
1335:
1336: @example
1337: ' table-find forth-wordlist wordlist-map @ !
1338: @end example
1339:
1340: Note that you now have to type the predefined words in the same case
1341: that we defined them, which are varying. You may want to convert them
1342: to your favourite case before doing this operation (I won't explain how,
1.62 ! crook 1343: because if you are even contemplating doing this, you'd better have
1.48 anton 1344: enough knowledge of Forth systems to know this already).
1345:
1346: @comment ----------------------------------------------
1347: @node Environment variables, Gforth Files, Upper and lower case, Gforth Environment
1348: @section Environment variables
1349: @cindex environment variables
1350:
1351: Gforth uses these environment variables:
1352:
1353: @itemize @bullet
1354: @item
1355: @cindex @code{GFORTHHIST} -- environment variable
1356: @code{GFORTHHIST} -- (Unix systems only) specifies the directory in which to
1357: open/create the history file, @file{.gforth-history}. Default:
1358: @code{$HOME}.
1359:
1360: @item
1361: @cindex @code{GFORTHPATH} -- environment variable
1362: @code{GFORTHPATH} -- specifies the path used when searching for the gforth image file and
1363: for Forth source-code files.
1364:
1365: @item
1366: @cindex @code{GFORTH} -- environment variable
1.49 anton 1367: @code{GFORTH} -- used by @file{gforthmi}, @xref{gforthmi}.
1.48 anton 1368:
1369: @item
1370: @cindex @code{GFORTHD} -- environment variable
1.62 ! crook 1371: @code{GFORTHD} -- used by @file{gforthmi}, @xref{gforthmi}.
1.48 anton 1372:
1373: @item
1374: @cindex @code{TMP}, @code{TEMP} - environment variable
1375: @code{TMP}, @code{TEMP} - (non-Unix systems only) used as a potential
1376: location for the history file.
1377: @end itemize
1378:
1379: @comment also POSIXELY_CORRECT LINES COLUMNS HOME but no interest in
1380: @comment mentioning these.
1381:
1382: All the Gforth environment variables default to sensible values if they
1383: are not set.
1384:
1385:
1386: @comment ----------------------------------------------
1387: @node Gforth Files, Startup speed, Environment variables, Gforth Environment
1388: @section Gforth files
1389: @cindex Gforth files
1390:
1391: When you install Gforth on a Unix system, it installs files in these
1392: locations by default:
1393:
1394: @itemize @bullet
1395: @item
1396: @file{/usr/local/bin/gforth}
1397: @item
1398: @file{/usr/local/bin/gforthmi}
1399: @item
1400: @file{/usr/local/man/man1/gforth.1} - man page.
1401: @item
1402: @file{/usr/local/info} - the Info version of this manual.
1403: @item
1404: @file{/usr/local/lib/gforth/<version>/...} - Gforth @file{.fi} files.
1405: @item
1406: @file{/usr/local/share/gforth/<version>/TAGS} - Emacs TAGS file.
1407: @item
1408: @file{/usr/local/share/gforth/<version>/...} - Gforth source files.
1409: @item
1410: @file{.../emacs/site-lisp/gforth.el} - Emacs gforth mode.
1411: @end itemize
1412:
1413: You can select different places for installation by using
1414: @code{configure} options (listed with @code{configure --help}).
1415:
1416: @comment ----------------------------------------------
1417: @node Startup speed, , Gforth Files, Gforth Environment
1418: @section Startup speed
1419: @cindex Startup speed
1420: @cindex speed, startup
1421:
1422: If Gforth is used for CGI scripts or in shell scripts, its startup
1423: speed may become a problem. On a 300MHz 21064a under Linux-2.2.13 with
1424: glibc-2.0.7, @code{gforth -e bye} takes about 24.6ms user and 11.3ms
1425: system time.
1426:
1427: If startup speed is a problem, you may consider the following ways to
1428: improve it; or you may consider ways to reduce the number of startups
1.62 ! crook 1429: (for example, by using Fast-CGI).
1.48 anton 1430:
1431: The first step to improve startup speed is to statically link Gforth, by
1432: building it with @code{XLDFLAGS=-static}. This requires more memory for
1433: the code and will therefore slow down the first invocation, but
1434: subsequent invocations avoid the dynamic linking overhead. Another
1435: disadvantage is that Gforth won't profit from library upgrades. As a
1436: result, @code{gforth-static -e bye} takes about 17.1ms user and
1437: 8.2ms system time.
1438:
1439: The next step to improve startup speed is to use a non-relocatable image
1.62 ! crook 1440: (@xref{Non-Relocatable Image Files}). You can create this image with
1.48 anton 1441: @code{gforth -e "savesystem gforthnr.fi bye"} and later use it with
1442: @code{gforth -i gforthnr.fi ...}. This avoids the relocation overhead
1443: and a part of the copy-on-write overhead. The disadvantage is that the
1.62 ! crook 1444: non-relocatable image does not work if the OS gives Gforth a different
1.48 anton 1445: address for the dictionary, for whatever reason; so you better provide a
1446: fallback on a relocatable image. @code{gforth-static -i gforthnr.fi -e
1447: bye} takes about 15.3ms user and 7.5ms system time.
1448:
1449: The final step is to disable dictionary hashing in Gforth. Gforth
1450: builds the hash table on startup, which takes much of the startup
1451: overhead. You can do this by commenting out the @code{include hash.fs}
1452: in @file{startup.fs} and everything that requires @file{hash.fs} (at the
1453: moment @file{table.fs} and @file{ekey.fs}) and then doing @code{make}.
1454: The disadvantages are that functionality like @code{table} and
1455: @code{ekey} is missing and that text interpretation (e.g., compiling)
1456: now takes much longer. So, you should only use this method if there is
1457: no significant text interpretation to perform (the script should be
1.62 ! crook 1458: compiled into the image, amongst other things). @code{gforth-static -i
1.48 anton 1459: gforthnrnh.fi -e bye} takes about 2.1ms user and 6.1ms system time.
1460:
1461: @c ******************************************************************
1462: @node Tutorial, Introduction, Gforth Environment, Top
1463: @chapter Forth Tutorial
1464: @cindex Tutorial
1465: @cindex Forth Tutorial
1466:
1.62 ! crook 1467: This tutorial can be used with any ANS-compliant Forth; any
! 1468: Gforth-specific features are marked as such and you can skip them if you
! 1469: work with another Forth. This tutorial does not explain all features of
! 1470: Forth, just enough to get you started and give you some ideas about the
! 1471: facilities available in Forth. Read the rest of the manual and the
! 1472: standard when you are through this.
1.48 anton 1473:
1474: The intended way to use this tutorial is that you work through it while
1475: sitting in front of the console, take a look at the examples and predict
1476: what they will do, then try them out; if the outcome is not as expected,
1477: find out why (e.g., by trying out variations of the example), so you
1478: understand what's going on. There are also some assignments that you
1479: should solve.
1480:
1481: This tutorial assumes that you have programmed before and know what,
1482: e.g., a loop is.
1483:
1484: @c !! explain compat library
1485:
1486: @menu
1487: * Starting Gforth Tutorial::
1488: * Syntax Tutorial::
1489: * Crash Course Tutorial::
1490: * Stack Tutorial::
1491: * Arithmetics Tutorial::
1492: * Stack Manipulation Tutorial::
1493: * Using files for Forth code Tutorial::
1494: * Comments Tutorial::
1495: * Colon Definitions Tutorial::
1496: * Decompilation Tutorial::
1497: * Stack-Effect Comments Tutorial::
1498: * Types Tutorial::
1499: * Factoring Tutorial::
1500: * Designing the stack effect Tutorial::
1501: * Local Variables Tutorial::
1502: * Conditional execution Tutorial::
1503: * Flags and Comparisons Tutorial::
1504: * General Loops Tutorial::
1505: * Counted loops Tutorial::
1506: * Recursion Tutorial::
1507: * Leaving definitions or loops Tutorial::
1508: * Return Stack Tutorial::
1509: * Memory Tutorial::
1510: * Characters and Strings Tutorial::
1511: * Alignment Tutorial::
1512: * Interpretation and Compilation Semantics and Immediacy Tutorial::
1513: * Execution Tokens Tutorial::
1514: * Exceptions Tutorial::
1515: * Defining Words Tutorial::
1516: * Arrays and Records Tutorial::
1517: * POSTPONE Tutorial::
1518: * Literal Tutorial::
1519: * Advanced macros Tutorial::
1520: * Compilation Tokens Tutorial::
1521: * Wordlists and Search Order Tutorial::
1522: @end menu
1523:
1524: @node Starting Gforth Tutorial, Syntax Tutorial, Tutorial, Tutorial
1525: @section Starting Gforth
1526:
1527: You can start Gforth by typing its name:
1528:
1529: @example
1530: gforth
1531: @end example
1532:
1533: That puts you into interactive mode; you can leave Gforth by typing
1534: @code{bye}. While in Gforth, you can edit the command line and access
1535: the command line history with cursor keys, similar to bash.
1536:
1537:
1538: @node Syntax Tutorial, Crash Course Tutorial, Starting Gforth Tutorial, Tutorial
1539: @section Syntax
1540:
1541: A @dfn{word} is a sequence of arbitrary characters (expcept white
1542: space). Words are separated by white space. E.g., each of the
1543: following lines contains exactly one word:
1544:
1545: @example
1546: word
1547: !@@#$%^&*()
1548: 1234567890
1549: 5!a
1550: @end example
1551:
1552: A frequent beginner's error is to leave away necessary white space,
1553: resulting in an error like @samp{Undefined word}; so if you see such an
1554: error, check if you have put spaces wherever necessary.
1555:
1556: @example
1557: ." hello, world" \ correct
1558: ."hello, world" \ gives an "Undefined word" error
1559: @end example
1560:
1561: Gforth and most other Forth systems ignores differences in case (it is
1562: case-insensitive), i.e., @samp{word} is the same as @samp{Word}. If
1563: your system is case-sensitive, you may have to type all the examples
1564: given here in upper case.
1565:
1566:
1567: @node Crash Course Tutorial, Stack Tutorial, Syntax Tutorial, Tutorial
1568: @section Crash Course
1569:
1570: Type
1571:
1572: @example
1573: 0 0 !
1574: here execute
1575: ' catch >body 20 erase abort
1576: ' (quit) >body 20 erase
1577: @end example
1578:
1579: The last two examples are guaranteed to destroy parts of Gforth (and
1580: most other systems), so you better leave Gforth afterwards (if it has
1581: not finished by itself). On some systems you may have to kill gforth
1582: from outside (e.g., in Unix with @code{kill}).
1583:
1584: Now that you know how to produce crashes (and that there's not much to
1585: them), let's learn how to produce meaningful programs.
1586:
1587:
1588: @node Stack Tutorial, Arithmetics Tutorial, Crash Course Tutorial, Tutorial
1589: @section Stack
1590:
1591: The most obvious feature of Forth is the stack. When you type in a
1592: number, it is pushed on the stack. You can display the content of the
1593: stack with @code{.s}.
1594:
1595: @example
1596: 1 2 .s
1597: 3 .s
1598: @end example
1599:
1600: @code{.s} displays the top-of-stack to the right, i.e., the numbers
1601: appear in @code{.s} output as they appeared in the input.
1602:
1603: You can print the top of stack element with @code{.}.
1604:
1605: @example
1606: 1 2 3 . . .
1607: @end example
1608:
1609: In general, words consume their stack arguments (@code{.s} is an
1610: exception).
1611:
1612: @assignment
1613: What does the stack contain after @code{5 6 7 .}?
1614: @endassignment
1615:
1616:
1617: @node Arithmetics Tutorial, Stack Manipulation Tutorial, Stack Tutorial, Tutorial
1618: @section Arithmetics
1619:
1620: The words @code{+}, @code{-}, @code{*}, @code{/}, and @code{mod} always
1621: operate on the top two stack items:
1622:
1623: @example
1624: 2 2 + .
1625: 2 1 - .
1626: 7 3 mod .
1627: @end example
1628:
1629: The operands of @code{-}, @code{/}, and @code{mod} are in the same order
1630: as in the corresponding infix expression (this is generally the case in
1631: Forth).
1632:
1633: Parentheses are superfluous (and not available), because the order of
1634: the words unambiguously determines the order of evaluation and the
1635: operands:
1636:
1637: @example
1638: 3 4 + 5 * .
1639: 3 4 5 * + .
1640: @end example
1641:
1642: @assignment
1643: What are the infix expressions corresponding to the Forth code above?
1644: Write @code{6-7*8+9} in Forth notation@footnote{This notation is also
1645: known as Postfix or RPN (Reverse Polish Notation).}.
1646: @endassignment
1647:
1648: To change the sign, use @code{negate}:
1649:
1650: @example
1651: 2 negate .
1652: @end example
1653:
1654: @assignment
1655: Convert -(-3)*4-5 to Forth.
1656: @endassignment
1657:
1658: @code{/mod} performs both @code{/} and @code{mod}.
1659:
1660: @example
1661: 7 3 /mod . .
1662: @end example
1663:
1664: @node Stack Manipulation Tutorial, Using files for Forth code Tutorial, Arithmetics Tutorial, Tutorial
1665: @section Stack Manipulation
1666:
1667: Stack manipulation words rearrange the data on the stack.
1668:
1669: @example
1670: 1 .s drop .s
1671: 1 .s dup .s drop drop .s
1672: 1 2 .s over .s drop drop drop
1673: 1 2 .s swap .s drop drop
1674: 1 2 3 .s rot .s drop drop drop
1675: @end example
1676:
1677: These are the most important stack manipulation words. There are also
1678: variants that manipulate twice as many stack items:
1679:
1680: @example
1681: 1 2 3 4 .s 2swap .s 2drop 2drop
1682: @end example
1683:
1684: Two more stack manipulation words are:
1685:
1686: @example
1687: 1 2 .s nip .s drop
1688: 1 2 .s tuck .s 2drop drop
1689: @end example
1690:
1691: @assignment
1692: Replace @code{nip} and @code{tuck} with combinations of other stack
1693: manipulation words.
1694:
1695: @example
1696: Given: How do you get:
1697: 1 2 3 3 2 1
1698: 1 2 3 1 2 3 2
1699: 1 2 3 1 2 3 3
1700: 1 2 3 1 3 3
1701: 1 2 3 2 1 3
1702: 1 2 3 4 4 3 2 1
1703: 1 2 3 1 2 3 1 2 3
1704: 1 2 3 4 1 2 3 4 1 2
1705: 1 2 3
1706: 1 2 3 1 2 3 4
1707: 1 2 3 1 3
1708: @end example
1709: @endassignment
1710:
1711: @example
1712: 5 dup * .
1713: @end example
1714:
1715: @assignment
1716: Write 17^3 and 17^4 in Forth, without writing @code{17} more than once.
1717: Write a piece of Forth code that expects two numbers on the stack
1718: (@var{a} and @var{b}, with @var{b} on top) and computes
1719: @code{(a-b)(a+1)}.
1720: @endassignment
1721:
1722: @node Using files for Forth code Tutorial, Comments Tutorial, Stack Manipulation Tutorial, Tutorial
1723: @section Using files for Forth code
1724:
1725: While working at the Forth command line is convenient for one-line
1726: examples and short one-off code, you probably want to store your source
1727: code in files for convenient editing and persistence. You can use your
1728: favourite editor (Gforth includes Emacs support, @pxref{Emacs and
1729: Gforth}) to create @var{file} and use
1730:
1731: @example
1732: s" @var{file}" included
1733: @end example
1734:
1735: to load it into your Forth system. The file name extension I use for
1736: Forth files is @samp{.fs}.
1737:
1738: You can easily start Gforth with some files loaded like this:
1739:
1740: @example
1741: gforth @var{file1} @var{file2}
1742: @end example
1743:
1744: If an error occurs during loading these files, Gforth terminates,
1745: whereas an error during @code{INCLUDED} within Gforth usually gives you
1746: a Gforth command line. Starting the Forth system every time gives you a
1747: clean start every time, without interference from the results of earlier
1748: tries.
1749:
1750: I often put all the tests in a file, then load the code and run the
1751: tests with
1752:
1753: @example
1754: gforth @var{code} @var{tests} -e bye
1755: @end example
1756:
1757: (often by performing this command with @kbd{C-x C-e} in Emacs). The
1758: @code{-e bye} ensures that Gforth terminates afterwards so that I can
1759: restart this command without ado.
1760:
1761: The advantage of this approach is that the tests can be repeated easily
1762: every time the program ist changed, making it easy to catch bugs
1763: introduced by the change.
1764:
1765:
1766: @node Comments Tutorial, Colon Definitions Tutorial, Using files for Forth code Tutorial, Tutorial
1767: @section Comments
1768:
1769: @example
1770: \ That's a comment; it ends at the end of the line
1771: ( Another comment; it ends here: ) .s
1772: @end example
1773:
1774: @code{\} and @code{(} are ordinary Forth words and therefore have to be
1775: separated with white space from the following text.
1776:
1777: @example
1778: \This gives an "Undefined word" error
1779: @end example
1780:
1781: The first @code{)} ends a comment started with @code{(}, so you cannot
1782: nest @code{(}-comments; and you cannot comment out text containing a
1783: @code{)} with @code{( ... )}@footnote{therefore it's a good idea to
1784: avoid @code{)} in word names.}.
1785:
1786: I use @code{\}-comments for descriptive text and for commenting out code
1787: of one or more line; I use @code{(}-comments for describing the stack
1788: effect, the stack contents, or for commenting out sub-line pieces of
1789: code.
1790:
1791: The Emacs mode @file{gforth.el} (@pxref{Emacs and Gforth}) supports
1792: these uses by commenting out a region with @kbd{C-x \}, uncommenting a
1793: region with @kbd{C-u C-x \}, and filling a @code{\}-commented region
1794: with @kbd{M-q}.
1795:
1796:
1797: @node Colon Definitions Tutorial, Decompilation Tutorial, Comments Tutorial, Tutorial
1798: @section Colon Definitions
1799:
1800: are similar to procedures and functions in other programming languages.
1801:
1802: @example
1803: : squared ( n -- n^2 )
1804: dup * ;
1805: 5 squared .
1806: 7 squared .
1807: @end example
1808:
1809: @code{:} starts the colon definition; its name is @code{squared}. The
1810: following comment describes its stack effect. The words @code{dup *}
1811: are not executed, but compiled into the definition. @code{;} ends the
1812: colon definition.
1813:
1814: The newly-defined word can be used like any other word, including using
1815: it in other definitions:
1816:
1817: @example
1818: : cubed ( n -- n^3 )
1819: dup squared * ;
1820: -5 cubed .
1821: : fourth-power ( n -- n^4 )
1822: squared squared ;
1823: 3 fourth-power .
1824: @end example
1825:
1826: @assignment
1827: Write colon definitions for @code{nip}, @code{tuck}, @code{negate}, and
1828: @code{/mod} in terms of other Forth words, and check if they work (hint:
1829: test your tests on the originals first). Don't let the
1830: @samp{redefined}-Messages spook you, they are just warnings.
1831: @endassignment
1832:
1833:
1834: @node Decompilation Tutorial, Stack-Effect Comments Tutorial, Colon Definitions Tutorial, Tutorial
1835: @section Decompilation
1836:
1837: You can decompile colon definitions with @code{see}:
1838:
1839: @example
1840: see squared
1841: see cubed
1842: @end example
1843:
1844: In Gforth @code{see} shows you a reconstruction of the source code from
1845: the executable code. Informations that were present in the source, but
1846: not in the executable code, are lost (e.g., comments).
1847:
1848: @node Stack-Effect Comments Tutorial, Types Tutorial, Decompilation Tutorial, Tutorial
1849: @section Stack-Effect Comments
1850:
1851: By convention the comment after the name of a definition describes the
1852: stack effect: The part in from of the @samp{--} describes the state of
1853: the stack before the execution of the definition, i.e., the parameters
1854: that are passed into the colon definition; the part behind the @samp{--}
1855: is the state of the stack after the execution of the definition, i.e.,
1856: the results of the definition. The stack comment only shows the top
1857: stack items that the definition accesses and/or changes.
1858:
1859: You should put a correct stack effect on every definition, even if it is
1860: just @code{( -- )}. You should also add some descriptive comment to
1861: more complicated words (I usually do this in the lines following
1862: @code{:}). If you don't do this, your code becomes unreadable (because
1863: you have to work through every definition before you can undertsand
1864: any).
1865:
1866: @assignment
1867: The stack effect of @code{swap} can be written like this: @code{x1 x2 --
1868: x2 x1}. Describe the stack effect of @code{-}, @code{drop}, @code{dup},
1869: @code{over}, @code{rot}, @code{nip}, and @code{tuck}. Hint: When you
1870: are done, you can compare your stack effects to this in this manual
1871: (@pxref{Word Index}).
1872: @endassignment
1873:
1874: Sometimes programmers put comments at various places in colon
1875: definitions that describe the contents of the stack at that place (stack
1876: comments); i.e., they are like the first part of a stack-effect
1877: comment. E.g.,
1878:
1879: @example
1880: : cubed ( n -- n^3 )
1881: dup squared ( n n^2 ) * ;
1882: @end example
1883:
1884: In this case the stack comment is pretty superfluous, because the word
1885: is simple enough. If you think it would be a good idea to add such a
1886: comment to increase readability, you should also consider factoring the
1887: word into several simpler words (@pxref{Factoring Tutorial,,
1.60 anton 1888: Factoring}), which typically eliminates the need for the stack comment;
1.48 anton 1889: however, if you decide not to refactor it, then having such a comment is
1890: better than not having it.
1891:
1892: The names of the stack items in stack-effect and stack comments in the
1893: standard, in this manual, and in many programs specify the type through
1894: a type prefix, similar to Fortran and Hungarian notation. The most
1895: frequent prefixes are:
1896:
1897: @table @code
1898: @item n
1899: signed integer
1900: @item u
1901: unsigned integer
1902: @item c
1903: character
1904: @item f
1905: Boolean flags, i.e. @code{false} or @code{true}.
1906: @item a-addr,a-
1907: Cell-aligned address
1908: @item c-addr,c-
1909: Char-aligned address (note that a Char may have two bytes in Windows NT)
1910: @item xt
1911: Execution token, same size as Cell
1912: @item w,x
1913: Cell, can contain an integer or an address. It usually takes 32, 64 or
1914: 16 bits (depending on your platform and Forth system). A cell is more
1915: commonly known as machine word, but the term @emph{word} already means
1916: something different in Forth.
1917: @item d
1918: signed double-cell integer
1919: @item ud
1920: unsigned double-cell integer
1921: @item r
1922: Float (on the FP stack)
1923: @end table
1924:
1925: You can find a more complete list in @ref{Notation}.
1926:
1927: @assignment
1928: Write stack-effect comments for all definitions you have written up to
1929: now.
1930: @endassignment
1931:
1932:
1933: @node Types Tutorial, Factoring Tutorial, Stack-Effect Comments Tutorial, Tutorial
1934: @section Types
1935:
1936: In Forth the names of the operations are not overloaded; so similar
1937: operations on different types need different names; e.g., @code{+} adds
1938: integers, and you have to use @code{f+} to add floating-point numbers.
1939: The following prefixes are often used for related operations on
1940: different types:
1941:
1942: @table @code
1943: @item (none)
1944: signed integer
1945: @item u
1946: unsigned integer
1947: @item c
1948: character
1949: @item d
1950: signed double-cell integer
1951: @item ud, du
1952: unsigned double-cell integer
1953: @item 2
1954: two cells (not-necessarily double-cell numbers)
1955: @item m, um
1956: mixed single-cell and double-cell operations
1957: @item f
1958: floating-point (note that in stack comments @samp{f} represents flags,
1959: and @samp{r} represents FP number).
1960: @end table
1961:
1962: If there are no differences between the signed and the unsigned variant
1963: (e.g., for @code{+}), there is only the prefix-less variant.
1964:
1965: Forth does not perform type checking, neither at compile time, nor at
1966: run time. If you use the wrong oeration, the data are interpreted
1967: incorrectly:
1968:
1969: @example
1970: -1 u.
1971: @end example
1972:
1973: If you have only experience with type-checked languages until now, and
1974: have heard how important type-checking is, don't panic! In my
1975: experience (and that of other Forthers), type errors in Forth code are
1976: usually easy to find (once you get used to it), the increased vigilance
1977: of the programmer tends to catch some harder errors in addition to most
1978: type errors, and you never have to work around the type system, so in
1979: most situations the lack of type-checking seems to be a win (projects to
1980: add type checking to Forth have not caught on).
1981:
1982:
1983: @node Factoring Tutorial, Designing the stack effect Tutorial, Types Tutorial, Tutorial
1984: @section Factoring
1985:
1986: If you try to write longer definitions, you will soon find it hard to
1987: keep track of the stack contents. Therefore, good Forth programmers
1988: tend to write only short definitions (e.g., three lines). The art of
1989: finding meaningful short definitions is known as factoring (as in
1990: factoring polynomials).
1991:
1992: Well-factored programs offer additional advantages: smaller, more
1993: general words, are easier to test and debug and can be reused more and
1994: better than larger, specialized words.
1995:
1996: So, if you run into difficulties with stack management, when writing
1997: code, try to define meaningful factors for the word, and define the word
1998: in terms of those. Even if a factor contains only two words, it is
1999: often helpful.
2000:
2001: Good factoring is not easy, and even experienced Forth programmers often
2002: don't find the right solution right away, but only when rewriting the
2003: program. So, if you don't come up with a good solution immediately,
2004: keep trying, don't despair.
2005:
2006: @c example !!
2007:
2008:
2009: @node Designing the stack effect Tutorial, Local Variables Tutorial, Factoring Tutorial, Tutorial
2010: @section Designing the stack effect
2011:
2012: In other languages you can use an arbitrary order of parameters for a
2013: function; and since ther is only one result, you don't have to deal with
2014: the order of results, either.
2015:
2016: In Forth (and other stack-based languages, e.g., Postscript) the
2017: parameter and result order of a definition is important and should be
2018: designed well. The general guideline is to design the stack effect such
2019: that the word is simple to use in most cases, even if that complicates
2020: the implementation of the word. Some concrete rules are:
2021:
2022: @itemize @bullet
2023:
2024: @item
2025: Words consume all of their parameters (e.g., @code{.}).
2026:
2027: @item
2028: If there is a convention on the order of parameters (e.g., from
2029: mathematics or another programming language), stick with it (e.g.,
2030: @code{-}).
2031:
2032: @item
2033: If one parameter usually requires only a short computation (e.g., it is
2034: a constant), pass it on the top of the stack. Conversely, parameters
2035: that usually require a long sequence of code to compute should be passed
2036: as the bottom (i.e., first) parameter. This makes the code easier to
2037: read, because reader does not need to keep track of the bottom item
2038: through a long sequence of code (or, alternatively, through stack
1.49 anton 2039: manipulations). E.g., @code{!} (store, @pxref{Memory}) expects the
1.48 anton 2040: address on top of the stack because it is usually simpler to compute
2041: than the stored value (often the address is just a variable).
2042:
2043: @item
2044: Similarly, results that are usually consumed quickly should be returned
2045: on the top of stack, whereas a result that is often used in long
2046: computations should be passed as bottom result. E.g., the file words
2047: like @code{open-file} return the error code on the top of stack, because
2048: it is usually consumed quickly by @code{throw}; moreover, the error code
2049: has to be checked before doing anything with the other results.
2050:
2051: @end itemize
2052:
2053: These rules are just general guidelines, don't lose sight of the overall
2054: goal to make the words easy to use. E.g., if the convention rule
2055: conflicts with the computation-length rule, you might decide in favour
2056: of the convention if the word will be used rarely, and in favour of the
2057: computation-length rule if the word will be used frequently (because
2058: with frequent use the cost of breaking the computation-length rule would
2059: be quite high, and frequent use makes it easier to remember an
2060: unconventional order).
2061:
2062: @c example !! structure package
2063:
2064: @node Local Variables Tutorial, Conditional execution Tutorial, Designing the stack effect Tutorial, Tutorial
2065: @section Local Variables
2066:
2067: You can define local variables (@emph{locals}) in a colon definition:
2068:
2069: @example
2070: : swap @{ a b -- b a @}
2071: b a ;
2072: 1 2 swap .s 2drop
2073: @end example
2074:
2075: (If your Forth system does not support this syntax, include
2076: @file{compat/anslocals.fs} first).
2077:
2078: In this example @code{@{ a b -- b a @}} is the locals definition; it
2079: takes two cells from the stack, puts the top of stack in @code{b} and
2080: the next stack element in @code{a}. @code{--} starts a comment ending
2081: with @code{@}}. After the locals definition, using the name of the
2082: local will push its value on the stack. You can leave the comment
2083: part (@code{-- b a}) away:
2084:
2085: @example
2086: : swap ( x1 x2 -- x2 x1 )
2087: @{ a b @} b a ;
2088: @end example
2089:
2090: In Gforth you can have several locals definitions, anywhere in a colon
2091: definition; in contrast, in a standard program you can have only one
2092: locals definition per colon definition, and that locals definition must
2093: be outside any controll structure.
2094:
2095: With locals you can write slightly longer definitions without running
2096: into stack trouble. However, I recommend trying to write colon
2097: definitions without locals for exercise purposes to help you gain the
2098: essential factoring skills.
2099:
2100: @assignment
2101: Rewrite your definitions until now with locals
2102: @endassignment
2103:
2104:
2105: @node Conditional execution Tutorial, Flags and Comparisons Tutorial, Local Variables Tutorial, Tutorial
2106: @section Conditional execution
2107:
2108: In Forth you can use control structures only inside colon definitions.
2109: An @code{if}-structure looks like this:
2110:
2111: @example
2112: : abs ( n1 -- +n2 )
2113: dup 0 < if
2114: negate
2115: endif ;
2116: 5 abs .
2117: -5 abs .
2118: @end example
2119:
2120: @code{if} takes a flag from the stack. If the flag is non-zero (true),
2121: the following code is performed, otherwise execution continues after the
1.51 pazsan 2122: @code{endif} (or @code{else}). @code{<} compares the top two stack
1.48 anton 2123: elements and prioduces a flag:
2124:
2125: @example
2126: 1 2 < .
2127: 2 1 < .
2128: 1 1 < .
2129: @end example
2130:
2131: Actually the standard name for @code{endif} is @code{then}. This
2132: tutorial presents the examples using @code{endif}, because this is often
2133: less confusing for people familiar with other programming languages
2134: where @code{then} has a different meaning. If your system does not have
2135: @code{endif}, define it with
2136:
2137: @example
2138: : endif postpone then ; immediate
2139: @end example
2140:
2141: You can optionally use an @code{else}-part:
2142:
2143: @example
2144: : min ( n1 n2 -- n )
2145: 2dup < if
2146: drop
2147: else
2148: nip
2149: endif ;
2150: 2 3 min .
2151: 3 2 min .
2152: @end example
2153:
2154: @assignment
2155: Write @code{min} without @code{else}-part (hint: what's the definition
2156: of @code{nip}?).
2157: @endassignment
2158:
2159:
2160: @node Flags and Comparisons Tutorial, General Loops Tutorial, Conditional execution Tutorial, Tutorial
2161: @section Flags and Comparisons
2162:
2163: In a false-flag all bits are clear (0 when interpreted as integer). In
2164: a canonical true-flag all bits are set (-1 as a twos-complement signed
2165: integer); in many contexts (e.g., @code{if}) any non-zero value is
2166: treated as true flag.
2167:
2168: @example
2169: false .
2170: true .
2171: true hex u. decimal
2172: @end example
2173:
2174: Comparison words produce canonical flags:
2175:
2176: @example
2177: 1 1 = .
2178: 1 0= .
2179: 0 1 < .
2180: 0 0 < .
2181: -1 1 u< . \ type error, u< interprets -1 as large unsigned number
2182: -1 1 < .
2183: @end example
2184:
2185: Gforth supports all combinations of the prefixes @code{0 u d d0 du} (or
2186: none) and the comparisons @code{= <> < > <= >=}. Only a part of these
1.60 anton 2187: combinations are standard (for details see the standard or @ref{Word
1.61 anton 2188: Index}).
1.48 anton 2189:
2190: You can use @code{and or xor invert} can be used as operations on
2191: canonical flags. Actually they are bitwise operations:
2192:
2193: @example
2194: 1 2 and .
2195: 1 2 or .
2196: 1 3 xor .
2197: 1 invert .
2198: @end example
2199:
2200: You can convert a zero/non-zero flag into a canonical flag with
2201: @code{0<>} (and complement it on the way with @code{0=}).
2202:
2203: @example
2204: 1 0= .
2205: 1 0<> .
2206: @end example
2207:
2208: You can use the all-bits-set feature of canonicasl flags and the bitwise
2209: operation of the Boolean operations to avoid @code{if}s:
2210:
2211: @example
2212: : foo ( n1 -- n2 )
2213: 0= if
2214: 14
2215: else
2216: 0
2217: endif ;
2218: 0 foo .
2219: 1 foo .
2220:
2221: : foo ( n1 -- n2 )
2222: 0= 14 and ;
2223: 0 foo .
2224: 1 foo .
2225: @end example
2226:
2227: @assignment
2228: Write @code{min} without @code{if}.
2229: @endassignment
2230:
2231:
2232: @node General Loops Tutorial, Counted loops Tutorial, Flags and Comparisons Tutorial, Tutorial
2233: @section General Loops
2234:
2235: The endless loop is the most simple one:
2236:
2237: @example
2238: : endless ( -- )
2239: 0 begin
2240: dup . 1+
2241: again ;
2242: endless
2243: @end example
2244:
2245: Terminate this loop by pressing @kbd{Ctrl-C} (in Gforth). @code{begin}
2246: does nothing at run-time, @code{again} jumps back to @code{begin}.
2247:
2248: A loop with one exit at any place looks like this:
2249:
2250: @example
2251: : log2 ( +n1 -- n2 )
2252: \ logarithmus dualis of n1>0, rounded down to the next integer
2253: assert( dup 0> )
2254: 2/ 0 begin
2255: over 0> while
2256: 1+ swap 2/ swap
2257: repeat
2258: nip ;
2259: 7 log2 .
2260: 8 log2 .
2261: @end example
2262:
2263: At run-time @code{while} consumes a flag; if it is 0, execution
1.51 pazsan 2264: continues behind the @code{repeat}; if the flag is non-zero, execution
1.48 anton 2265: continues behind the @code{while}. @code{Repeat} jumps back to
2266: @code{begin}, just like @code{again}.
2267:
2268: In Forth there are many combinations/abbreviations, like @code{1+}.
2269: However, @code{2/} is not one of them; it shifts it's argument right by
2270: one bit (arithmetic shift right):
2271:
2272: @example
2273: -5 2 / .
2274: -5 2/ .
2275: @end example
2276:
2277: @code{assert(} is no standard word, but you can get it on systems other
2278: then Gforth by including @file{compat/assert.fs}. You can see what it
2279: does by trying
2280:
2281: @example
2282: 0 log2 .
2283: @end example
2284:
2285: Here's a loop with an exit at the end:
2286:
2287: @example
2288: : log2 ( +n1 -- n2 )
2289: \ logarithmus dualis of n1>0, rounded down to the next integer
2290: assert( dup 0 > )
2291: -1 begin
2292: 1+ swap 2/ swap
2293: over 0 <=
2294: until
2295: nip ;
2296: @end example
2297:
2298: @code{Until} consumes a flag; if it is non-zero, execution continues at
2299: the @code{begin}, otherwise after the @code{until}.
2300:
2301: @assignment
2302: Write a definition for computing the greatest common divisor.
2303: @endassignment
2304:
2305:
2306: @node Counted loops Tutorial, Recursion Tutorial, General Loops Tutorial, Tutorial
2307: @section Counted loops
2308:
2309: @example
2310: : ^ ( n1 u -- n )
2311: \ n = the uth power of u1
2312: 1 swap 0 u+do
2313: over *
2314: loop
2315: nip ;
2316: 3 2 ^ .
2317: 4 3 ^ .
2318: @end example
2319:
2320: @code{U+do} (from @file{compat/loops.fs}, if your Forth system doesn't
2321: have it) takes two numbers of the stack @code{( u3 u4 -- )}, and then
2322: performs the code between @code{u+do} and @code{loop} for @code{u3-u4}
2323: times (or not at all, if @code{u3-u4<0}).
2324:
2325: You can see the stack effect design rules at work in the stack effect of
2326: the loop start words: Since the start value of the loop is more
2327: frequently constant than the end value, the start value is passed on
2328: the top-of-stack.
2329:
2330: You can access the counter of a counted loop with @code{i}:
2331:
2332: @example
2333: : fac ( u -- u! )
2334: 1 swap 1+ 1 u+do
2335: i *
2336: loop ;
2337: 5 fac .
2338: 7 fac .
2339: @end example
2340:
2341: There is also @code{+do}, which expects signed numbers (important for
2342: deciding whether to enter the loop).
2343:
2344: @assignment
2345: Write a definition for computing the nth Fibonacci number.
2346: @endassignment
2347:
2348: !! +DO...+LOOP
2349: !! -DO...-LOOP
2350:
2351:
2352: @node Recursion Tutorial, Leaving definitions or loops Tutorial, Counted loops Tutorial, Tutorial
2353: @section Recursion
2354:
2355: Usually the name of a definition is not visible in the definition; but
2356: earlier definitions are usually visible:
2357:
2358: @example
2359: 1 0 / . \ "Floating-point unidentified fault" in Gforth on most platforms
2360: : / ( n1 n2 -- n )
2361: dup 0= if
2362: -10 throw \ report division by zero
2363: endif
2364: / \ old version
2365: ;
2366: 1 0 /
2367: @end example
2368:
2369: For recursive definitions you can use @code{recursive} (non-standard) or
2370: @code{recurse}:
2371:
2372: @example
2373: : fac1 ( n -- n! ) recursive
2374: dup 0> if
2375: dup 1- fac1 *
2376: else
2377: drop 1
2378: endif ;
2379: 7 fac1 .
2380:
2381: : fac2 ( n -- n! )
2382: dup 0> if
2383: dup 1- recurse *
2384: else
2385: drop 1
2386: endif ;
2387: 8 fac2 .
2388: @end example
2389:
2390: @assignment
2391: Write a recursive definition for computing the nth Fibonacci number.
2392: @endassignment
2393:
2394:
2395: @node Leaving definitions or loops Tutorial, Return Stack Tutorial, Recursion Tutorial, Tutorial
2396: @section Leaving definitions or loops
2397:
2398: @code{EXIT} exits the current definition right away. For every counted
2399: loop that is left in this way, an @code{UNLOOP} has to be performed
2400: before the @code{EXIT}:
2401:
2402: @c !! real examples
2403: @example
2404: : ...
2405: ... u+do
2406: ... if
2407: ... unloop exit
2408: endif
2409: ...
2410: loop
2411: ... ;
2412: @end example
2413:
2414: @code{LEAVE} leaves the innermost counted loop right away:
2415:
2416: @example
2417: : ...
2418: ... u+do
2419: ... if
2420: ... leave
2421: endif
2422: ...
2423: loop
2424: ... ;
2425: @end example
2426:
2427:
2428: @node Return Stack Tutorial, Memory Tutorial, Leaving definitions or loops Tutorial, Tutorial
2429: @section Return Stack
2430:
2431: In addition to the data stack Forth also has a second stack, the return
2432: stack; most Forth systems store the return addresses of procedure calls
2433: there (thus its name). Programmers can also use this stack:
2434:
2435: @example
2436: : foo ( n1 n2 -- )
2437: .s
2438: >r .s
1.50 anton 2439: r@@ .
1.48 anton 2440: >r .s
1.50 anton 2441: r@@ .
1.48 anton 2442: r> .
1.50 anton 2443: r@@ .
1.48 anton 2444: r> . ;
2445: 1 2 foo
2446: @end example
2447:
2448: @code{>r} takes an element from the data stack and pushes it onto the
2449: return stack; conversely, @code{r>} moves an elementm from the return to
2450: the data stack; @code{r@@} pushes a copy of the top of the return stack
2451: on the return stack.
2452:
2453: Forth programmers usually use the return stack for storing data
2454: temporarily, if using the data stack alone would be too complex, and
2455: factoring and locals are not an option:
2456:
2457: @example
2458: : 2swap ( x1 x2 x3 x4 -- x3 x4 x1 x2 )
2459: rot >r rot r> ;
2460: @end example
2461:
2462: The return address of the definition and the loop control parameters of
2463: counted loops usually reside on the return stack, so you have to take
2464: all items, that you have pushed on the return stack in a colon
2465: definition or counted loop, from the return stack before the definition
2466: or loop ends. You cannot access items that you pushed on the return
2467: stack outside some definition or loop within the definition of loop.
2468:
2469: If you miscount the return stack items, this usually ends in a crash:
2470:
2471: @example
2472: : crash ( n -- )
2473: >r ;
2474: 5 crash
2475: @end example
2476:
2477: You cannot mix using locals and using the return stack (according to the
2478: standard; Gforth has no problem). However, they solve the same
2479: problems, so this shouldn't be an issue.
2480:
2481: @assignment
2482: Can you rewrite any of the definitions you wrote until now in a better
2483: way using the return stack?
2484: @endassignment
2485:
2486:
2487: @node Memory Tutorial, Characters and Strings Tutorial, Return Stack Tutorial, Tutorial
2488: @section Memory
2489:
2490: You can create a global variable @code{v} with
2491:
2492: @example
2493: variable v ( -- addr )
2494: @end example
2495:
2496: @code{v} pushes the address of a cell in memory on the stack. This cell
2497: was reserved by @code{variable}. You can use @code{!} (store) to store
2498: values into this cell and @code{@@} (fetch) to load the value from the
2499: stack into memory:
2500:
2501: @example
2502: v .
2503: 5 v ! .s
1.50 anton 2504: v @@ .
1.48 anton 2505: @end example
2506:
2507: You can also reserve more memory:
2508:
2509: @example
2510: create v2 20 cells allot
2511: @end example
2512:
2513: creates a word @code{v2} and reserves 20 cells; the address pushed by
2514: @code{v2} points to the start of these 20 cells. You can use address
2515: arithmetic to access these cells:
2516:
2517: @example
2518: 3 v2 5 cells + !
2519: @end example
2520:
2521: You can reserve and initialize memory with @code{,}:
2522:
2523: @example
2524: create v3
2525: 5 , 4 , 3 , 2 , 1 ,
1.50 anton 2526: v3 @@ .
2527: v3 cell+ @@ .
2528: v3 2 cells + @@ .
1.48 anton 2529: @end example
2530:
2531: @assignment
2532: Write a definition @code{vsum ( addr u -- n )} that computes the sum of
2533: @code{u} cells, with the first of these cells at @code{addr}, the next
2534: one at @code{addr cell+} etc.
2535: @endassignment
2536:
2537: You can also reserve memory without creating a new word:
2538:
2539: @example
1.60 anton 2540: here 10 cells allot .
2541: here .
1.48 anton 2542: @end example
2543:
2544: @code{Here} pushes the start address of the memory area. You should
2545: store it somewhere, or you will have a hard time finding the memory area
2546: again.
2547:
2548: @code{Allot} manages dictionary memory. The dictionary memory contains
2549: the system's data structures for words etc. on Gforth and most other
2550: Forth systems. It is managed like a stack: You can free the memory that
2551: you have just @code{allot}ed with
2552:
2553: @example
2554: -10 cells allot
1.60 anton 2555: here .
1.48 anton 2556: @end example
2557:
2558: Note that you cannot do this if you have created a new word in the
2559: meantime (because then your @code{allot}ed memory is no longer on the
2560: top of the dictionary ``stack'').
2561:
2562: Alternatively, you can use @code{allocate} and @code{free} which allow
2563: freeing memory in any order:
2564:
2565: @example
2566: 10 cells allocate throw .s
2567: 20 cells allocate throw .s
2568: swap
2569: free throw
2570: free throw
2571: @end example
2572:
2573: The @code{throw}s deal with errors (e.g., out of memory).
2574:
1.60 anton 2575: And there is also a garbage collector
2576: @url{http://www.complang.tuwien.ac.at/forth/garbage-collection.zip},
2577: which eliminates the need to @code{free} memory explicitly.
1.48 anton 2578:
2579:
2580: @node Characters and Strings Tutorial, Alignment Tutorial, Memory Tutorial, Tutorial
2581: @section Characters and Strings
2582:
2583: On the stack characters take up a cell, like numbers. In memory they
2584: have their own size (one 8-bit byte on most systems), and therefore
2585: require their own words for memory access:
2586:
2587: @example
2588: create v4
2589: 104 c, 97 c, 108 c, 108 c, 111 c,
1.50 anton 2590: v4 4 chars + c@@ .
1.48 anton 2591: @end example
2592:
2593: The preferred representation of strings on the stack is @code{addr
2594: u-count}, where @code{addr} is the address of the first character and
2595: @code{u-count} is the number of characters in the string.
2596:
2597: @example
2598: v4 5 type
2599: @end example
2600:
2601: You get a string constant with
2602:
2603: @example
2604: s" hello, world" .s
2605: type
2606: @end example
2607:
2608: Make sure you have a space between @code{s"} and the string; @code{s"}
2609: is a normal Forth word and must be delimited with white space (try what
2610: happens when you remove the space).
2611:
2612: However, this interpretive use of @code{s"} is quite restricted: the
2613: string exists only until the next call of @code{s"} (some Forth systems
2614: keep more than one of these strings, but usually they still have a
1.62 ! crook 2615: limited lifetime).
1.48 anton 2616:
2617: @example
2618: s" hello," s" world" .s
2619: type
2620: type
2621: @end example
2622:
1.62 ! crook 2623: You can also use @code{s"} in a definition, and the resulting
! 2624: strings then live forever (well, for as long as the definition):
1.48 anton 2625:
2626: @example
2627: : foo s" hello," s" world" ;
2628: foo .s
2629: type
2630: type
2631: @end example
2632:
2633: @assignment
2634: @code{Emit ( c -- )} types @code{c} as character (not a number).
2635: Implement @code{type ( addr u -- )}.
2636: @endassignment
2637:
2638: @node Alignment Tutorial, Interpretation and Compilation Semantics and Immediacy Tutorial, Characters and Strings Tutorial, Tutorial
2639: @section Alignment
2640:
2641: On many processors cells have to be aligned in memory, if you want to
2642: access them with @code{@@} and @code{!} (and even if the processor does
1.62 ! crook 2643: not require alignment, access to aligned cells is faster).
1.48 anton 2644:
2645: @code{Create} aligns @code{here} (i.e., the place where the next
2646: allocation will occur, and that the @code{create}d word points to).
2647: Likewise, the memory produced by @code{allocate} starts at an aligned
2648: address. Adding a number of @code{cells} to an aligned address produces
2649: another aligned address.
2650:
2651: However, address arithmetic involving @code{char+} and @code{chars} can
2652: create an address that is not cell-aligned. @code{Aligned ( addr --
2653: a-addr )} produces the next aligned address:
2654:
2655: @example
1.50 anton 2656: v3 char+ aligned .s @@ .
2657: v3 char+ .s @@ .
1.48 anton 2658: @end example
2659:
2660: Similarly, @code{align} advances @code{here} to the next aligned
2661: address:
2662:
2663: @example
2664: create v5 97 c,
2665: here .
2666: align here .
2667: 1000 ,
2668: @end example
2669:
2670: Note that you should use aligned addresses even if your processor does
2671: not require them, if you want your program to be portable.
2672:
2673:
2674: @node Interpretation and Compilation Semantics and Immediacy Tutorial, Execution Tokens Tutorial, Alignment Tutorial, Tutorial
2675: @section Interpretation and Compilation Semantics and Immediacy
2676:
2677: When a word is compiled, it behaves differently from being interpreted.
2678: E.g., consider @code{+}:
2679:
2680: @example
2681: 1 2 + .
2682: : foo + ;
2683: @end example
2684:
2685: These two behaviours are known as compilation and interpretation
2686: semantics. For normal words (e.g., @code{+}), the compilation semantics
2687: is to append the interpretation semantics to the currently defined word
2688: (@code{foo} in the example above). I.e., when @code{foo} is executed
2689: later, the interpretation semantics of @code{+} (i.e., adding two
2690: numbers) will be performed.
2691:
2692: However, there are words with non-default compilation semantics, e.g.,
2693: the control-flow words like @code{if}. You can use @code{immediate} to
2694: change the compilation semantics of the last defined word to be equal to
2695: the interpretation semantics:
2696:
2697: @example
2698: : [FOO] ( -- )
2699: 5 . ; immediate
2700:
2701: [FOO]
2702: : bar ( -- )
2703: [FOO] ;
2704: bar
2705: see bar
2706: @end example
2707:
2708: Two conventions to mark words with non-default compilation semnatics are
2709: names with brackets (more frequently used) and to write them all in
2710: upper case (less frequently used).
2711:
2712: In Gforth (and many other systems) you can also remove the
2713: interpretation semantics with @code{compile-only} (the compilation
2714: semantics is derived from the original interpretation semantics):
2715:
2716: @example
2717: : flip ( -- )
2718: 6 . ; compile-only \ but not immediate
2719: flip
2720:
2721: : flop ( -- )
2722: flip ;
2723: flop
2724: @end example
2725:
2726: In this example the interpretation semantics of @code{flop} is equal to
2727: the original interpretation semantics of @code{flip}.
2728:
2729: The text interpreter has two states: in interpret state, it performs the
2730: interpretation semantics of words it encounters; in compile state, it
2731: performs the compilation semantics of these words.
2732:
2733: Among other things, @code{:} switches into compile state, and @code{;}
2734: switches back to interpret state. They contain the factors @code{]}
2735: (switch to compile state) and @code{[} (switch to interpret state), that
2736: do nothing but switch the state.
2737:
2738: @example
2739: : xxx ( -- )
2740: [ 5 . ]
2741: ;
2742:
2743: xxx
2744: see xxx
2745: @end example
2746:
2747: These brackets are also the source of the naming convention mentioned
2748: above.
2749:
2750:
2751: @node Execution Tokens Tutorial, Exceptions Tutorial, Interpretation and Compilation Semantics and Immediacy Tutorial, Tutorial
2752: @section Execution Tokens
2753:
2754: @code{' word} gives you the execution token (XT) of a word. The XT is a
2755: cell representing the interpretation semantics of a word. You can
2756: execute this semantics with @code{execute}:
2757:
2758: @example
2759: ' + .s
2760: 1 2 rot execute .
2761: @end example
2762:
2763: The XT is similar to a function pointer in C. However, parameter
2764: passing through the stack makes it a little more flexible:
2765:
2766: @example
2767: : map-array ( ... addr u xt -- ... )
1.50 anton 2768: \ executes xt ( ... x -- ... ) for every element of the array starting
2769: \ at addr and containing u elements
1.48 anton 2770: @{ xt @}
2771: cells over + swap ?do
1.50 anton 2772: i @@ xt execute
1.48 anton 2773: 1 cells +loop ;
2774:
2775: create a 3 , 4 , 2 , -1 , 4 ,
2776: a 5 ' . map-array .s
2777: 0 a 5 ' + map-array .
2778: s" max-n" environment? drop .s
2779: a 5 ' min map-array .
2780: @end example
2781:
2782: You can use map-array with the XTs of words that consume one element
2783: more than they produce. In theory you can also use it with other XTs,
2784: but the stack effect then depends on the size of the array, which is
2785: hard to understand.
2786:
1.51 pazsan 2787: Since XTs are cell-sized, you can store them in memory and manipulate
2788: them on the stack like other cells. You can also compile the XT into a
1.48 anton 2789: word with @code{compile,}:
2790:
2791: @example
2792: : foo1 ( n1 n2 -- n )
2793: [ ' + compile, ] ;
2794: see foo
2795: @end example
2796:
2797: This is non-standard, because @code{compile,} has no compilation
2798: semantics in the standard, but it works in good Forth systems. For the
2799: broken ones, use
2800:
2801: @example
2802: : [compile,] compile, ; immediate
2803:
2804: : foo1 ( n1 n2 -- n )
2805: [ ' + ] [compile,] ;
2806: see foo
2807: @end example
2808:
2809: @code{'} is a word with default compilation semantics; it parses the
2810: next word when its interpretation semantics are executed, not during
2811: compilation:
2812:
2813: @example
2814: : foo ( -- xt )
2815: ' ;
2816: see foo
2817: : bar ( ... "word" -- ... )
2818: ' execute ;
2819: see bar
1.60 anton 2820: 1 2 bar + .
1.48 anton 2821: @end example
2822:
2823: You often want to parse a word during compilation and compile its XT so
2824: it will be pushed on the stack at run-time. @code{[']} does this:
2825:
2826: @example
2827: : xt-+ ( -- xt )
2828: ['] + ;
2829: see xt-+
2830: 1 2 xt-+ execute .
2831: @end example
2832:
2833: Many programmers tend to see @code{'} and the word it parses as one
2834: unit, and expect it to behave like @code{[']} when compiled, and are
2835: confused by the actual behaviour. If you are, just remember that the
2836: Forth system just takes @code{'} as one unit and has no idea that it is
2837: a parsing word (attempts to convenience programmers in this issue have
2838: usually resulted in even worse pitfalls, see
2839: @uref{http://www.complang.tuwien.ac.at/papers/ertl98.ps.gz}).
2840:
2841: Note that the state of the interpreter does not come into play when
1.51 pazsan 2842: creating and executing XTs. I.e., even when you execute @code{'} in
1.48 anton 2843: compile state, it still gives you the interpretation semantics. And
2844: whatever that state is, @code{execute} performs the semantics
1.51 pazsan 2845: represented by the XT (i.e., the interpretation semantics).
1.48 anton 2846:
2847:
2848: @node Exceptions Tutorial, Defining Words Tutorial, Execution Tokens Tutorial, Tutorial
2849: @section Exceptions
2850:
2851: @code{throw ( n -- )} causes an exception unless n is zero.
2852:
2853: @example
2854: 100 throw .s
2855: 0 throw .s
2856: @end example
2857:
2858: @code{catch ( ... xt -- ... n )} behaves similar to @code{execute}, but
2859: it catches exceptions and pushes the number of the exception on the
2860: stack (or 0, if the xt executed without exception). If there was an
2861: exception, the stacks have the same depth as when entering @code{catch}:
2862:
2863: @example
2864: .s
2865: 3 0 ' / catch .s
2866: 3 2 ' / catch .s
2867: @end example
2868:
2869: @assignment
2870: Try the same with @code{execute} instead of @code{catch}.
2871: @endassignment
2872:
2873: @code{Throw} always jumps to the dynamically next enclosing
2874: @code{catch}, even if it has to leave several call levels to achieve
2875: this:
2876:
2877: @example
2878: : foo 100 throw ;
2879: : foo1 foo ." after foo" ;
1.51 pazsan 2880: : bar ['] foo1 catch ;
1.60 anton 2881: bar .
1.48 anton 2882: @end example
2883:
2884: It is often important to restore a value upon leaving a definition, even
2885: if the definition is left through an exception. You can ensure this
2886: like this:
2887:
2888: @example
2889: : ...
2890: save-x
1.51 pazsan 2891: ['] word-changing-x catch ( ... n )
1.48 anton 2892: restore-x
2893: ( ... n ) throw ;
2894: @end example
2895:
1.55 anton 2896: Gforth provides an alternative syntax in addition to @code{catch}:
1.48 anton 2897: @code{try ... recover ... endtry}. If the code between @code{try} and
2898: @code{recover} has an exception, the stack depths are restored, the
2899: exception number is pushed on the stack, and the code between
2900: @code{recover} and @code{endtry} is performed. E.g., the definition for
2901: @code{catch} is
2902:
2903: @example
2904: : catch ( x1 .. xn xt -- y1 .. ym 0 / z1 .. zn error ) \ exception
2905: try
2906: execute 0
2907: recover
2908: nip
2909: endtry ;
2910: @end example
2911:
2912: The equivalent to the restoration code above is
2913:
2914: @example
2915: : ...
2916: save-x
2917: try
2918: word-changing-x
2919: end-try
2920: restore-x
2921: throw ;
2922: @end example
2923:
2924: As you can see, the @code{recover} part is optional.
2925:
2926:
2927: @node Defining Words Tutorial, Arrays and Records Tutorial, Exceptions Tutorial, Tutorial
2928: @section Defining Words
2929:
2930: @code{:}, @code{create}, and @code{variable} are definition words: They
2931: define other words. @code{Constant} is another definition word:
2932:
2933: @example
2934: 5 constant foo
2935: foo .
2936: @end example
2937:
2938: You can also use the prefixes @code{2} (double-cell) and @code{f}
2939: (floating point) with @code{variable} and @code{constant}.
2940:
2941: You can also define your own defining words. E.g.:
2942:
2943: @example
2944: : variable ( "name" -- )
2945: create 0 , ;
2946: @end example
2947:
2948: You can also define defining words that create words that do something
2949: other than just producing their address:
2950:
2951: @example
2952: : constant ( n "name" -- )
2953: create ,
2954: does> ( -- n )
1.50 anton 2955: ( addr ) @@ ;
1.48 anton 2956:
2957: 5 constant foo
2958: foo .
2959: @end example
2960:
2961: The definition of @code{constant} above ends at the @code{does>}; i.e.,
2962: @code{does>} replaces @code{;}, but it also does something else: It
2963: changes the last defined word such that it pushes the address of the
2964: body of the word and then performs the code after the @code{does>}
2965: whenever it is called.
2966:
2967: In the example above, @code{constant} uses @code{,} to store 5 into the
2968: body of @code{foo}. When @code{foo} executes, it pushes the address of
2969: the body onto the stack, then (in the code after the @code{does>})
2970: fetches the 5 from there.
2971:
2972: The stack comment near the @code{does>} reflects the stack effect of the
2973: defined word, not the stack effect of the code after the @code{does>}
2974: (the difference is that the code expects the address of the body that
2975: the stack comment does not show).
2976:
2977: You can use these definition words to do factoring in cases that involve
2978: (other) definition words. E.g., a field offset is always added to an
2979: address. Instead of defining
2980:
2981: @example
2982: 2 cells constant offset-field1
2983: @end example
2984:
2985: and using this like
2986:
2987: @example
2988: ( addr ) offset-field1 +
2989: @end example
2990:
2991: you can define a definition word
2992:
2993: @example
2994: : simple-field ( n "name" -- )
2995: create ,
2996: does> ( n1 -- n1+n )
1.50 anton 2997: ( addr ) @@ + ;
1.48 anton 2998: @end example
1.21 crook 2999:
1.48 anton 3000: Definition and use of field offsets now look like this:
1.21 crook 3001:
1.48 anton 3002: @example
3003: 2 cells simple-field field1
1.60 anton 3004: create mystruct 4 cells allot
3005: mystruct .s field1 .s drop
1.48 anton 3006: @end example
1.21 crook 3007:
1.48 anton 3008: If you want to do something with the word without performing the code
3009: after the @code{does>}, you can access the body of a @code{create}d word
3010: with @code{>body ( xt -- addr )}:
1.21 crook 3011:
1.48 anton 3012: @example
3013: : value ( n "name" -- )
3014: create ,
3015: does> ( -- n1 )
1.50 anton 3016: @@ ;
1.48 anton 3017: : to ( n "name" -- )
3018: ' >body ! ;
1.21 crook 3019:
1.48 anton 3020: 5 value foo
3021: foo .
3022: 7 to foo
3023: foo .
3024: @end example
1.21 crook 3025:
1.48 anton 3026: @assignment
3027: Define @code{defer ( "name" -- )}, which creates a word that stores an
3028: XT (at the start the XT of @code{abort}), and upon execution
3029: @code{execute}s the XT. Define @code{is ( xt "name" -- )} that stores
3030: @code{xt} into @code{name}, a word defined with @code{defer}. Indirect
3031: recursion is one application of @code{defer}.
3032: @endassignment
1.29 crook 3033:
1.48 anton 3034: @node Arrays and Records Tutorial, POSTPONE Tutorial, Defining Words Tutorial, Tutorial
3035: @section Arrays and Records
1.29 crook 3036:
1.48 anton 3037: Forth has no standard words for defining data structures such as arrays
3038: and records (structs in C terminology), but you can build them yourself
3039: based on address arithmetic. You can also define words for defining
3040: arrays and records (@pxref{Defining Words Tutorial,, Defining Words}).
1.29 crook 3041:
1.48 anton 3042: One of the first projects a Forth newcomer sets out upon when learning
3043: about defining words is an array defining word (possibly for
3044: n-dimensional arrays). Go ahead and do it, I did it, too; you will
3045: learn something from it. However, don't be disappointed when you later
3046: learn that you have little use for these words (inappropriate use would
3047: be even worse). I have not yet found a set of useful array words yet;
3048: the needs are just too diverse, and named, global arrays (the result of
3049: naive use of defining words) are often not flexible enough (e.g.,
3050: consider how to pass them as parameters).
1.29 crook 3051:
1.48 anton 3052: On the other hand, there is a useful set of record words, and it has
3053: been defined in @file{compat/struct.fs}; these words are predefined in
3054: Gforth. They are explained in depth elsewhere in this manual (see
3055: @pxref{Structures}). The @code{simple-field} example above is
3056: simplified variant of fields in this package.
1.21 crook 3057:
3058:
1.48 anton 3059: @node POSTPONE Tutorial, Literal Tutorial, Arrays and Records Tutorial, Tutorial
3060: @section @code{POSTPONE}
1.21 crook 3061:
1.48 anton 3062: You can compile the compilation semantics (instead of compiling the
3063: interpretation semantics) of a word with @code{POSTPONE}:
1.21 crook 3064:
1.48 anton 3065: @example
3066: : MY-+ ( Compilation: -- ; Run-time of compiled code: n1 n2 -- n )
1.51 pazsan 3067: POSTPONE + ; immediate
1.48 anton 3068: : foo ( n1 n2 -- n )
3069: MY-+ ;
3070: 1 2 foo .
3071: see foo
3072: @end example
1.21 crook 3073:
1.48 anton 3074: During the definition of @code{foo} the text interpreter performs the
3075: compilation semantics of @code{MY-+}, which performs the compilation
3076: semantics of @code{+}, i.e., it compiles @code{+} into @code{foo}.
3077:
3078: This example also displays separate stack comments for the compilation
3079: semantics and for the stack effect of the compiled code. For words with
3080: default compilation semantics these stack effects are usually not
3081: displayed; the stack effect of the compilation semantics is always
3082: @code{( -- )} for these words, the stack effect for the compiled code is
3083: the stack effect of the interpretation semantics.
3084:
3085: Note that the state of the interpreter does not come into play when
3086: performing the compilation semantics in this way. You can also perform
3087: it interpretively, e.g.:
3088:
3089: @example
3090: : foo2 ( n1 n2 -- n )
3091: [ MY-+ ] ;
3092: 1 2 foo .
3093: see foo
3094: @end example
1.21 crook 3095:
1.48 anton 3096: However, there are some broken Forth systems where this does not always
1.62 ! crook 3097: work, and therefore this practice was been declared non-standard in
1.48 anton 3098: 1999.
3099: @c !! repair.fs
3100:
3101: Here is another example for using @code{POSTPONE}:
1.44 crook 3102:
1.48 anton 3103: @example
3104: : MY-- ( Compilation: -- ; Run-time of compiled code: n1 n2 -- n )
3105: POSTPONE negate POSTPONE + ; immediate compile-only
3106: : bar ( n1 n2 -- n )
3107: MY-- ;
3108: 2 1 bar .
3109: see bar
3110: @end example
1.21 crook 3111:
1.48 anton 3112: You can define @code{ENDIF} in this way:
1.21 crook 3113:
1.48 anton 3114: @example
3115: : ENDIF ( Compilation: orig -- )
3116: POSTPONE then ; immediate
3117: @end example
1.21 crook 3118:
1.48 anton 3119: @assignment
3120: Write @code{MY-2DUP} that has compilation semantics equivalent to
3121: @code{2dup}, but compiles @code{over over}.
3122: @endassignment
1.29 crook 3123:
1.48 anton 3124: @node Literal Tutorial, Advanced macros Tutorial, POSTPONE Tutorial, Tutorial
3125: @section @code{Literal}
1.29 crook 3126:
1.48 anton 3127: You cannot @code{POSTPONE} numbers:
1.21 crook 3128:
1.48 anton 3129: @example
3130: : [FOO] POSTPONE 500 ; immediate
1.21 crook 3131: @end example
3132:
1.48 anton 3133: Instead, you can use @code{LITERAL (compilation: n --; run-time: -- n )}:
1.29 crook 3134:
1.48 anton 3135: @example
3136: : [FOO] ( compilation: --; run-time: -- n )
3137: 500 POSTPONE literal ; immediate
1.29 crook 3138:
1.60 anton 3139: : flip [FOO] ;
1.48 anton 3140: flip .
3141: see flip
3142: @end example
1.29 crook 3143:
1.48 anton 3144: @code{LITERAL} consumes a number at compile-time (when it's compilation
3145: semantics are executed) and pushes it at run-time (when the code it
3146: compiled is executed). A frequent use of @code{LITERAL} is to compile a
3147: number computed at compile time into the current word:
1.29 crook 3148:
1.48 anton 3149: @example
3150: : bar ( -- n )
3151: [ 2 2 + ] literal ;
3152: see bar
3153: @end example
1.29 crook 3154:
1.48 anton 3155: @assignment
3156: Write @code{]L} which allows writing the example above as @code{: bar (
3157: -- n ) [ 2 2 + ]L ;}
3158: @endassignment
3159:
3160:
3161: @node Advanced macros Tutorial, Compilation Tokens Tutorial, Literal Tutorial, Tutorial
3162: @section Advanced macros
3163:
3164: Reconsider @code{map-array} from @ref{Execution Tokens
3165: Tutorial,, Execution Tokens}. It frequently performs @code{execute}, a
3166: relatively expensive operation in some implementations. You can use
3167: @code{compile,} and @code{POSTPONE} to eliminate these @code{execute}s
3168: and produce a word that contains the word to be performed directly:
3169:
3170: @c use ]] ... [[
3171: @example
3172: : compile-map-array ( compilation: xt -- ; run-time: ... addr u -- ... )
3173: \ at run-time, execute xt ( ... x -- ... ) for each element of the
3174: \ array beginning at addr and containing u elements
3175: @{ xt @}
3176: POSTPONE cells POSTPONE over POSTPONE + POSTPONE swap POSTPONE ?do
1.50 anton 3177: POSTPONE i POSTPONE @@ xt compile,
1.48 anton 3178: 1 cells POSTPONE literal POSTPONE +loop ;
3179:
3180: : sum-array ( addr u -- n )
3181: 0 rot rot [ ' + compile-map-array ] ;
3182: see sum-array
3183: a 5 sum-array .
3184: @end example
3185:
3186: You can use the full power of Forth for generating the code; here's an
3187: example where the code is generated in a loop:
3188:
3189: @example
3190: : compile-vmul-step ( compilation: n --; run-time: n1 addr1 -- n2 addr2 )
3191: \ n2=n1+(addr1)*n, addr2=addr1+cell
1.50 anton 3192: POSTPONE tuck POSTPONE @@
1.48 anton 3193: POSTPONE literal POSTPONE * POSTPONE +
3194: POSTPONE swap POSTPONE cell+ ;
3195:
3196: : compile-vmul ( compilation: addr1 u -- ; run-time: addr2 -- n )
1.51 pazsan 3197: \ n=v1*v2 (inner product), where the v_i are represented as addr_i u
1.48 anton 3198: 0 postpone literal postpone swap
3199: [ ' compile-vmul-step compile-map-array ]
3200: postpone drop ;
3201: see compile-vmul
3202:
3203: : a-vmul ( addr -- n )
1.51 pazsan 3204: \ n=a*v, where v is a vector that's as long as a and starts at addr
1.48 anton 3205: [ a 5 compile-vmul ] ;
3206: see a-vmul
3207: a a-vmul .
3208: @end example
3209:
3210: This example uses @code{compile-map-array} to show off, but you could
3211: also use @code{map-array} instead (try it now).
3212:
3213: You can use this technique for efficient multiplication of large
3214: matrices. In matrix multiplication, you multiply every line of one
3215: matrix with every column of the other matrix. You can generate the code
3216: for one line once, and use it for every column. The only downside of
3217: this technique is that it is cumbersome to recover the memory consumed
3218: by the generated code when you are done (and in more complicated cases
3219: it is not possible portably).
3220:
3221: @node Compilation Tokens Tutorial, Wordlists and Search Order Tutorial, Advanced macros Tutorial, Tutorial
3222: @section Compilation Tokens
3223:
3224: This section is Gforth-specific. You can skip it.
3225:
3226: @code{' word compile,} compiles the interpretation semantics. For words
3227: with default compilation semantics this is the same as performing the
3228: compilation semantics. To represent the compilation semantics of other
3229: words (e.g., words like @code{if} that have no interpretation
3230: semantics), Gforth has the concept of a compilation token (CT,
3231: consisting of two cells), and words @code{comp'} and @code{[comp']}.
3232: You can perform the compilation semantics represented by a CT with
3233: @code{execute}:
1.29 crook 3234:
1.48 anton 3235: @example
3236: : foo2 ( n1 n2 -- n )
3237: [ comp' + execute ] ;
3238: see foo
3239: @end example
1.29 crook 3240:
1.48 anton 3241: You can compile the compilation semantics represented by a CT with
3242: @code{postpone,}:
1.30 anton 3243:
1.48 anton 3244: @example
3245: : foo3 ( -- )
3246: [ comp' + postpone, ] ;
3247: see foo3
3248: @end example
1.30 anton 3249:
1.51 pazsan 3250: @code{[ comp' word postpone, ]} is equivalent to @code{POSTPONE word}.
1.48 anton 3251: @code{comp'} is particularly useful for words that have no
3252: interpretation semantics:
1.29 crook 3253:
1.30 anton 3254: @example
1.48 anton 3255: ' if
1.60 anton 3256: comp' if .s 2drop
1.30 anton 3257: @end example
3258:
1.29 crook 3259:
1.48 anton 3260: @node Wordlists and Search Order Tutorial, , Compilation Tokens Tutorial, Tutorial
3261: @section Wordlists and Search Order
3262:
3263: The dictionary is not just a memory area that allows you to allocate
3264: memory with @code{allot}, it also contains the Forth words, arranged in
3265: several wordlists. When searching for a word in a wordlist,
3266: conceptually you start searching at the youngest and proceed towards
3267: older words (in reality most systems nowadays use hash-tables); i.e., if
3268: you define a word with the same name as an older word, the new word
3269: shadows the older word.
3270:
3271: Which wordlists are searched in which order is determined by the search
3272: order. You can display the search order with @code{order}. It displays
3273: first the search order, starting with the wordlist searched first, then
3274: it displays the wordlist that will contain newly defined words.
1.21 crook 3275:
1.48 anton 3276: You can create a new, empty wordlist with @code{wordlist ( -- wid )}:
1.21 crook 3277:
1.48 anton 3278: @example
3279: wordlist constant mywords
3280: @end example
1.21 crook 3281:
1.48 anton 3282: @code{Set-current ( wid -- )} sets the wordlist that will contain newly
3283: defined words (the @emph{current} wordlist):
1.21 crook 3284:
1.48 anton 3285: @example
3286: mywords set-current
3287: order
3288: @end example
1.26 crook 3289:
1.48 anton 3290: Gforth does not display a name for the wordlist in @code{mywords}
3291: because this wordlist was created anonymously with @code{wordlist}.
1.21 crook 3292:
1.48 anton 3293: You can get the current wordlist with @code{get-current ( -- wid)}. If
3294: you want to put something into a specific wordlist without overall
3295: effect on the current wordlist, this typically looks like this:
1.21 crook 3296:
1.48 anton 3297: @example
3298: get-current mywords set-current ( wid )
3299: create someword
3300: ( wid ) set-current
3301: @end example
1.21 crook 3302:
1.48 anton 3303: You can write the search order with @code{set-order ( wid1 .. widn n --
3304: )} and read it with @code{get-order ( -- wid1 .. widn n )}. The first
3305: searched wordlist is topmost.
1.21 crook 3306:
1.48 anton 3307: @example
3308: get-order mywords swap 1+ set-order
3309: order
3310: @end example
1.21 crook 3311:
1.48 anton 3312: Yes, the order of wordlists in the output of @code{order} is reversed
3313: from stack comments and the output of @code{.s} and thus unintuitive.
1.21 crook 3314:
1.48 anton 3315: @assignment
3316: Define @code{>order ( wid -- )} with adds @code{wid} as first searched
3317: wordlist to the search order. Define @code{previous ( -- )}, which
3318: removes the first searched wordlist from the search order. Experiment
3319: with boundary conditions (you will see some crashes or situations that
3320: are hard or impossible to leave).
3321: @endassignment
1.21 crook 3322:
1.48 anton 3323: The search order is a powerful foundation for providing features similar
3324: to Modula-2 modules and C++ namespaces. However, trying to modularize
3325: programs in this way has disadvantages for debugging and reuse/factoring
3326: that overcome the advantages in my experience (I don't do huge projects,
1.55 anton 3327: though). These disadvantages are not so clear in other
1.48 anton 3328: languages/programming environments, because these langauges are not so
3329: strong in debugging and reuse.
1.21 crook 3330:
3331:
1.29 crook 3332: @c ******************************************************************
1.48 anton 3333: @node Introduction, Words, Tutorial, Top
1.29 crook 3334: @comment node-name, next, previous, up
3335: @chapter An Introduction to ANS Forth
3336: @cindex Forth - an introduction
1.21 crook 3337:
1.29 crook 3338: The primary purpose of this manual is to document Gforth. However, since
3339: Forth is not a widely-known language and there is a lack of up-to-date
3340: teaching material, it seems worthwhile to provide some introductory
1.49 anton 3341: material. For other sources of Forth-related
3342: information, see @ref{Forth-related information}.
1.21 crook 3343:
1.29 crook 3344: The examples in this section should work on any ANS Forth; the
3345: output shown was produced using Gforth. Each example attempts to
3346: reproduce the exact output that Gforth produces. If you try out the
3347: examples (and you should), what you should type is shown @kbd{like this}
3348: and Gforth's response is shown @code{like this}. The single exception is
1.30 anton 3349: that, where the example shows @key{RET} it means that you should
1.29 crook 3350: press the ``carriage return'' key. Unfortunately, some output formats for
3351: this manual cannot show the difference between @kbd{this} and
3352: @code{this} which will make trying out the examples harder (but not
3353: impossible).
1.21 crook 3354:
1.29 crook 3355: Forth is an unusual language. It provides an interactive development
3356: environment which includes both an interpreter and compiler. Forth
3357: programming style encourages you to break a problem down into many
3358: @cindex factoring
3359: small fragments (@dfn{factoring}), and then to develop and test each
3360: fragment interactively. Forth advocates assert that breaking the
3361: edit-compile-test cycle used by conventional programming languages can
3362: lead to great productivity improvements.
1.21 crook 3363:
1.29 crook 3364: @menu
3365: * Introducing the Text Interpreter::
3366: * Stacks and Postfix notation::
3367: * Your first definition::
3368: * How does that work?::
3369: * Forth is written in Forth::
3370: * Review - elements of a Forth system::
3371: * Where to go next::
3372: * Exercises::
3373: @end menu
1.21 crook 3374:
1.29 crook 3375: @comment ----------------------------------------------
3376: @node Introducing the Text Interpreter, Stacks and Postfix notation, Introduction, Introduction
3377: @section Introducing the Text Interpreter
3378: @cindex text interpreter
3379: @cindex outer interpreter
1.21 crook 3380:
1.30 anton 3381: @c IMO this is too detailed and the pace is too slow for
3382: @c an introduction. If you know German, take a look at
3383: @c http://www.complang.tuwien.ac.at/anton/lvas/skriptum-stack.html
3384: @c to see how I do it - anton
3385:
1.44 crook 3386: @c nac-> Where I have accepted your comments 100% and modified the text
3387: @c accordingly, I have deleted your comments. Elsewhere I have added a
3388: @c response like this to attempt to rationalise what I have done. Of
3389: @c course, this is a very clumsy mechanism for something that would be
3390: @c done far more efficiently over a beer. Please delete any dialogue
3391: @c you consider closed.
3392:
1.29 crook 3393: When you invoke the Forth image, you will see a startup banner printed
3394: and nothing else (if you have Gforth installed on your system, try
1.30 anton 3395: invoking it now, by typing @kbd{gforth@key{RET}}). Forth is now running
1.29 crook 3396: its command line interpreter, which is called the @dfn{Text Interpreter}
3397: (also known as the @dfn{Outer Interpreter}). (You will learn a lot
1.49 anton 3398: about the text interpreter as you read through this chapter, for more
3399: detail @pxref{The Text Interpreter}).
1.21 crook 3400:
1.29 crook 3401: Although it's not obvious, Forth is actually waiting for your
1.30 anton 3402: input. Type a number and press the @key{RET} key:
1.21 crook 3403:
1.26 crook 3404: @example
1.30 anton 3405: @kbd{45@key{RET}} ok
1.26 crook 3406: @end example
1.21 crook 3407:
1.29 crook 3408: Rather than give you a prompt to invite you to input something, the text
3409: interpreter prints a status message @i{after} it has processed a line
3410: of input. The status message in this case (``@code{ ok}'' followed by
3411: carriage-return) indicates that the text interpreter was able to process
3412: all of your input successfully. Now type something illegal:
3413:
3414: @example
1.30 anton 3415: @kbd{qwer341@key{RET}}
1.29 crook 3416: :1: Undefined word
3417: qwer341
3418: ^^^^^^^
3419: $400D2BA8 Bounce
3420: $400DBDA8 no.extensions
3421: @end example
1.23 crook 3422:
1.29 crook 3423: The exact text, other than the ``Undefined word'' may differ slightly on
3424: your system, but the effect is the same; when the text interpreter
3425: detects an error, it discards any remaining text on a line, resets
1.49 anton 3426: certain internal state and prints an error message. For a detailed description of error messages see @ref{Error
3427: messages}.
1.23 crook 3428:
1.29 crook 3429: The text interpreter waits for you to press carriage-return, and then
3430: processes your input line. Starting at the beginning of the line, it
3431: breaks the line into groups of characters separated by spaces. For each
3432: group of characters in turn, it makes two attempts to do something:
1.23 crook 3433:
1.29 crook 3434: @itemize @bullet
3435: @item
1.44 crook 3436: @cindex name dictionary
1.29 crook 3437: It tries to treat it as a command. It does this by searching a @dfn{name
3438: dictionary}. If the group of characters matches an entry in the name
3439: dictionary, the name dictionary provides the text interpreter with
3440: information that allows the text interpreter perform some actions. In
3441: Forth jargon, we say that the group
3442: @cindex word
3443: @cindex definition
3444: @cindex execution token
3445: @cindex xt
3446: of characters names a @dfn{word}, that the dictionary search returns an
3447: @dfn{execution token (xt)} corresponding to the @dfn{definition} of the
3448: word, and that the text interpreter executes the xt. Often, the terms
3449: @dfn{word} and @dfn{definition} are used interchangeably.
3450: @item
3451: If the text interpreter fails to find a match in the name dictionary, it
3452: tries to treat the group of characters as a number in the current number
3453: base (when you start up Forth, the current number base is base 10). If
3454: the group of characters legitimately represents a number, the text
3455: interpreter pushes the number onto a stack (we'll learn more about that
3456: in the next section).
3457: @end itemize
1.23 crook 3458:
1.29 crook 3459: If the text interpreter is unable to do either of these things with any
3460: group of characters, it discards the group of characters and the rest of
3461: the line, then prints an error message. If the text interpreter reaches
3462: the end of the line without error, it prints the status message ``@code{ ok}''
3463: followed by carriage-return.
1.21 crook 3464:
1.29 crook 3465: This is the simplest command we can give to the text interpreter:
1.23 crook 3466:
3467: @example
1.30 anton 3468: @key{RET} ok
1.23 crook 3469: @end example
1.21 crook 3470:
1.29 crook 3471: The text interpreter did everything we asked it to do (nothing) without
3472: an error, so it said that everything is ``@code{ ok}''. Try a slightly longer
3473: command:
1.21 crook 3474:
1.23 crook 3475: @example
1.30 anton 3476: @kbd{12 dup fred dup@key{RET}}
1.29 crook 3477: :1: Undefined word
3478: 12 dup fred dup
3479: ^^^^
3480: $400D2BA8 Bounce
3481: $400DBDA8 no.extensions
1.23 crook 3482: @end example
1.21 crook 3483:
1.29 crook 3484: When you press the carriage-return key, the text interpreter starts to
3485: work its way along the line:
1.21 crook 3486:
1.29 crook 3487: @itemize @bullet
3488: @item
3489: When it gets to the space after the @code{2}, it takes the group of
3490: characters @code{12} and looks them up in the name
3491: dictionary@footnote{We can't tell if it found them or not, but assume
3492: for now that it did not}. There is no match for this group of characters
3493: in the name dictionary, so it tries to treat them as a number. It is
3494: able to do this successfully, so it puts the number, 12, ``on the stack''
3495: (whatever that means).
3496: @item
3497: The text interpreter resumes scanning the line and gets the next group
3498: of characters, @code{dup}. It looks it up in the name dictionary and
3499: (you'll have to take my word for this) finds it, and executes the word
3500: @code{dup} (whatever that means).
3501: @item
3502: Once again, the text interpreter resumes scanning the line and gets the
3503: group of characters @code{fred}. It looks them up in the name
3504: dictionary, but can't find them. It tries to treat them as a number, but
3505: they don't represent any legal number.
3506: @end itemize
1.21 crook 3507:
1.29 crook 3508: At this point, the text interpreter gives up and prints an error
3509: message. The error message shows exactly how far the text interpreter
3510: got in processing the line. In particular, it shows that the text
3511: interpreter made no attempt to do anything with the final character
3512: group, @code{dup}, even though we have good reason to believe that the
3513: text interpreter would have no problem looking that word up and
3514: executing it a second time.
1.21 crook 3515:
3516:
1.29 crook 3517: @comment ----------------------------------------------
3518: @node Stacks and Postfix notation, Your first definition, Introducing the Text Interpreter, Introduction
3519: @section Stacks, postfix notation and parameter passing
3520: @cindex text interpreter
3521: @cindex outer interpreter
1.21 crook 3522:
1.29 crook 3523: In procedural programming languages (like C and Pascal), the
3524: building-block of programs is the @dfn{function} or @dfn{procedure}. These
3525: functions or procedures are called with @dfn{explicit parameters}. For
3526: example, in C we might write:
1.21 crook 3527:
1.23 crook 3528: @example
1.29 crook 3529: total = total + new_volume(length,height,depth);
1.23 crook 3530: @end example
1.21 crook 3531:
1.23 crook 3532: @noindent
1.29 crook 3533: where new_volume is a function-call to another piece of code, and total,
3534: length, height and depth are all variables. length, height and depth are
3535: parameters to the function-call.
1.21 crook 3536:
1.29 crook 3537: In Forth, the equivalent of the function or procedure is the
3538: @dfn{definition} and parameters are implicitly passed between
3539: definitions using a shared stack that is visible to the
3540: programmer. Although Forth does support variables, the existence of the
3541: stack means that they are used far less often than in most other
3542: programming languages. When the text interpreter encounters a number, it
3543: will place (@dfn{push}) it on the stack. There are several stacks (the
1.30 anton 3544: actual number is implementation-dependent ...) and the particular stack
1.29 crook 3545: used for any operation is implied unambiguously by the operation being
3546: performed. The stack used for all integer operations is called the @dfn{data
3547: stack} and, since this is the stack used most commonly, references to
3548: ``the data stack'' are often abbreviated to ``the stack''.
1.21 crook 3549:
1.29 crook 3550: The stacks have a last-in, first-out (LIFO) organisation. If you type:
1.21 crook 3551:
1.23 crook 3552: @example
1.30 anton 3553: @kbd{1 2 3@key{RET}} ok
1.23 crook 3554: @end example
1.21 crook 3555:
1.29 crook 3556: Then this instructs the text interpreter to placed three numbers on the
3557: (data) stack. An analogy for the behaviour of the stack is to take a
3558: pack of playing cards and deal out the ace (1), 2 and 3 into a pile on
3559: the table. The 3 was the last card onto the pile (``last-in'') and if
3560: you take a card off the pile then, unless you're prepared to fiddle a
3561: bit, the card that you take off will be the 3 (``first-out''). The
3562: number that will be first-out of the stack is called the @dfn{top of
3563: stack}, which
3564: @cindex TOS definition
3565: is often abbreviated to @dfn{TOS}.
1.21 crook 3566:
1.29 crook 3567: To understand how parameters are passed in Forth, consider the
3568: behaviour of the definition @code{+} (pronounced ``plus''). You will not
3569: be surprised to learn that this definition performs addition. More
3570: precisely, it adds two number together and produces a result. Where does
3571: it get the two numbers from? It takes the top two numbers off the
3572: stack. Where does it place the result? On the stack. You can act-out the
3573: behaviour of @code{+} with your playing cards like this:
1.21 crook 3574:
3575: @itemize @bullet
3576: @item
1.29 crook 3577: Pick up two cards from the stack on the table
1.21 crook 3578: @item
1.29 crook 3579: Stare at them intently and ask yourself ``what @i{is} the sum of these two
3580: numbers''
1.21 crook 3581: @item
1.29 crook 3582: Decide that the answer is 5
1.21 crook 3583: @item
1.29 crook 3584: Shuffle the two cards back into the pack and find a 5
1.21 crook 3585: @item
1.29 crook 3586: Put a 5 on the remaining ace that's on the table.
1.21 crook 3587: @end itemize
3588:
1.29 crook 3589: If you don't have a pack of cards handy but you do have Forth running,
3590: you can use the definition @code{.s} to show the current state of the stack,
3591: without affecting the stack. Type:
1.21 crook 3592:
3593: @example
1.30 anton 3594: @kbd{clearstack 1 2 3@key{RET}} ok
3595: @kbd{.s@key{RET}} <3> 1 2 3 ok
1.23 crook 3596: @end example
3597:
1.29 crook 3598: The text interpreter looks up the word @code{clearstack} and executes
3599: it; it tidies up the stack and removes any entries that may have been
3600: left on it by earlier examples. The text interpreter pushes each of the
3601: three numbers in turn onto the stack. Finally, the text interpreter
3602: looks up the word @code{.s} and executes it. The effect of executing
3603: @code{.s} is to print the ``<3>'' (the total number of items on the stack)
3604: followed by a list of all the items on the stack; the item on the far
3605: right-hand side is the TOS.
1.21 crook 3606:
1.29 crook 3607: You can now type:
1.21 crook 3608:
1.29 crook 3609: @example
1.30 anton 3610: @kbd{+ .s@key{RET}} <2> 1 5 ok
1.29 crook 3611: @end example
1.21 crook 3612:
1.29 crook 3613: @noindent
3614: which is correct; there are now 2 items on the stack and the result of
3615: the addition is 5.
1.23 crook 3616:
1.29 crook 3617: If you're playing with cards, try doing a second addition: pick up the
3618: two cards, work out that their sum is 6, shuffle them into the pack,
3619: look for a 6 and place that on the table. You now have just one item on
3620: the stack. What happens if you try to do a third addition? Pick up the
3621: first card, pick up the second card -- ah! There is no second card. This
3622: is called a @dfn{stack underflow} and consitutes an error. If you try to
3623: do the same thing with Forth it will report an error (probably a Stack
3624: Underflow or an Invalid Memory Address error).
1.23 crook 3625:
1.29 crook 3626: The opposite situation to a stack underflow is a @dfn{stack overflow},
3627: which simply accepts that there is a finite amount of storage space
3628: reserved for the stack. To stretch the playing card analogy, if you had
3629: enough packs of cards and you piled the cards up on the table, you would
3630: eventually be unable to add another card; you'd hit the ceiling. Gforth
3631: allows you to set the maximum size of the stacks. In general, the only
3632: time that you will get a stack overflow is because a definition has a
3633: bug in it and is generating data on the stack uncontrollably.
1.23 crook 3634:
1.29 crook 3635: There's one final use for the playing card analogy. If you model your
3636: stack using a pack of playing cards, the maximum number of items on
3637: your stack will be 52 (I assume you didn't use the Joker). The maximum
3638: @i{value} of any item on the stack is 13 (the King). In fact, the only
3639: possible numbers are positive integer numbers 1 through 13; you can't
3640: have (for example) 0 or 27 or 3.52 or -2. If you change the way you
3641: think about some of the cards, you can accommodate different
3642: numbers. For example, you could think of the Jack as representing 0,
3643: the Queen as representing -1 and the King as representing -2. Your
1.45 crook 3644: @i{range} remains unchanged (you can still only represent a total of 13
1.29 crook 3645: numbers) but the numbers that you can represent are -2 through 10.
1.28 crook 3646:
1.29 crook 3647: In that analogy, the limit was the amount of information that a single
3648: stack entry could hold, and Forth has a similar limit. In Forth, the
3649: size of a stack entry is called a @dfn{cell}. The actual size of a cell is
3650: implementation dependent and affects the maximum value that a stack
3651: entry can hold. A Standard Forth provides a cell size of at least
3652: 16-bits, and most desktop systems use a cell size of 32-bits.
1.21 crook 3653:
1.29 crook 3654: Forth does not do any type checking for you, so you are free to
3655: manipulate and combine stack items in any way you wish. A convenient way
3656: of treating stack items is as 2's complement signed integers, and that
3657: is what Standard words like @code{+} do. Therefore you can type:
1.21 crook 3658:
1.29 crook 3659: @example
1.30 anton 3660: @kbd{-5 12 + .s@key{RET}} <1> 7 ok
1.29 crook 3661: @end example
1.21 crook 3662:
1.29 crook 3663: If you use numbers and definitions like @code{+} in order to turn Forth
3664: into a great big pocket calculator, you will realise that it's rather
3665: different from a normal calculator. Rather than typing 2 + 3 = you had
3666: to type 2 3 + (ignore the fact that you had to use @code{.s} to see the
3667: result). The terminology used to describe this difference is to say that
3668: your calculator uses @dfn{Infix Notation} (parameters and operators are
3669: mixed) whilst Forth uses @dfn{Postfix Notation} (parameters and
3670: operators are separate), also called @dfn{Reverse Polish Notation}.
1.21 crook 3671:
1.29 crook 3672: Whilst postfix notation might look confusing to begin with, it has
3673: several important advantages:
1.21 crook 3674:
1.23 crook 3675: @itemize @bullet
3676: @item
1.29 crook 3677: it is unambiguous
1.23 crook 3678: @item
1.29 crook 3679: it is more concise
1.23 crook 3680: @item
1.29 crook 3681: it fits naturally with a stack-based system
1.23 crook 3682: @end itemize
1.21 crook 3683:
1.29 crook 3684: To examine these claims in more detail, consider these sums:
1.21 crook 3685:
1.29 crook 3686: @example
3687: 6 + 5 * 4 =
3688: 4 * 5 + 6 =
3689: @end example
1.21 crook 3690:
1.29 crook 3691: If you're just learning maths or your maths is very rusty, you will
3692: probably come up with the answer 44 for the first and 26 for the
3693: second. If you are a bit of a whizz at maths you will remember the
3694: @i{convention} that multiplication takes precendence over addition, and
3695: you'd come up with the answer 26 both times. To explain the answer 26
3696: to someone who got the answer 44, you'd probably rewrite the first sum
3697: like this:
1.21 crook 3698:
1.29 crook 3699: @example
3700: 6 + (5 * 4) =
3701: @end example
1.21 crook 3702:
1.29 crook 3703: If what you really wanted was to perform the addition before the
3704: multiplication, you would have to use parentheses to force it.
1.21 crook 3705:
1.29 crook 3706: If you did the first two sums on a pocket calculator you would probably
3707: get the right answers, unless you were very cautious and entered them using
3708: these keystroke sequences:
1.21 crook 3709:
1.29 crook 3710: 6 + 5 = * 4 =
3711: 4 * 5 = + 6 =
1.21 crook 3712:
1.29 crook 3713: Postfix notation is unambiguous because the order that the operators
3714: are applied is always explicit; that also means that parentheses are
3715: never required. The operators are @i{active} (the act of quoting the
3716: operator makes the operation occur) which removes the need for ``=''.
1.28 crook 3717:
1.29 crook 3718: The sum 6 + 5 * 4 can be written (in postfix notation) in two
3719: equivalent ways:
1.26 crook 3720:
3721: @example
1.29 crook 3722: 6 5 4 * + or:
3723: 5 4 * 6 +
1.26 crook 3724: @end example
1.23 crook 3725:
1.29 crook 3726: An important thing that you should notice about this notation is that
3727: the @i{order} of the numbers does not change; if you want to subtract
3728: 2 from 10 you type @code{10 2 -}.
1.1 anton 3729:
1.29 crook 3730: The reason that Forth uses postfix notation is very simple to explain: it
3731: makes the implementation extremely simple, and it follows naturally from
3732: using the stack as a mechanism for passing parameters. Another way of
3733: thinking about this is to realise that all Forth definitions are
3734: @i{active}; they execute as they are encountered by the text
3735: interpreter. The result of this is that the syntax of Forth is trivially
3736: simple.
1.1 anton 3737:
3738:
3739:
1.29 crook 3740: @comment ----------------------------------------------
3741: @node Your first definition, How does that work?, Stacks and Postfix notation, Introduction
3742: @section Your first Forth definition
3743: @cindex first definition
1.1 anton 3744:
1.29 crook 3745: Until now, the examples we've seen have been trivial; we've just been
3746: using Forth as a bigger-than-pocket calculator. Also, each calculation
3747: we've shown has been a ``one-off'' -- to repeat it we'd need to type it in
3748: again@footnote{That's not quite true. If you press the up-arrow key on
3749: your keyboard you should be able to scroll back to any earlier command,
3750: edit it and re-enter it.} In this section we'll see how to add new
3751: words to Forth's vocabulary.
1.1 anton 3752:
1.29 crook 3753: The easiest way to create a new word is to use a @dfn{colon
3754: definition}. We'll define a few and try them out before worrying too
3755: much about how they work. Try typing in these examples; be careful to
3756: copy the spaces accurately:
1.1 anton 3757:
1.29 crook 3758: @example
3759: : add-two 2 + . ;
3760: : greet ." Hello and welcome" ;
3761: : demo 5 add-two ;
3762: @end example
1.1 anton 3763:
1.29 crook 3764: @noindent
3765: Now try them out:
1.1 anton 3766:
1.29 crook 3767: @example
1.30 anton 3768: @kbd{greet@key{RET}} Hello and welcome ok
3769: @kbd{greet greet@key{RET}} Hello and welcomeHello and welcome ok
3770: @kbd{4 add-two@key{RET}} 6 ok
3771: @kbd{demo@key{RET}} 7 ok
3772: @kbd{9 greet demo add-two@key{RET}} Hello and welcome7 11 ok
1.29 crook 3773: @end example
1.1 anton 3774:
1.29 crook 3775: The first new thing that we've introduced here is the pair of words
3776: @code{:} and @code{;}. These are used to start and terminate a new
3777: definition, respectively. The first word after the @code{:} is the name
3778: for the new definition.
1.1 anton 3779:
1.29 crook 3780: As you can see from the examples, a definition is built up of words that
3781: have already been defined; Forth makes no distinction between
3782: definitions that existed when you started the system up, and those that
3783: you define yourself.
1.1 anton 3784:
1.29 crook 3785: The examples also introduce the words @code{.} (dot), @code{."}
3786: (dot-quote) and @code{dup} (dewp). Dot takes the value from the top of
3787: the stack and displays it. It's like @code{.s} except that it only
3788: displays the top item of the stack and it is destructive; after it has
3789: executed, the number is no longer on the stack. There is always one
3790: space printed after the number, and no spaces before it. Dot-quote
3791: defines a string (a sequence of characters) that will be printed when
3792: the word is executed. The string can contain any printable characters
3793: except @code{"}. A @code{"} has a special function; it is not a Forth
3794: word but it acts as a delimiter (the way that delimiters work is
3795: described in the next section). Finally, @code{dup} duplicates the value
3796: at the top of the stack. Try typing @code{5 dup .s} to see what it does.
1.1 anton 3797:
1.29 crook 3798: We already know that the text interpreter searches through the
3799: dictionary to locate names. If you've followed the examples earlier, you
3800: will already have a definition called @code{add-two}. Lets try modifying
3801: it by typing in a new definition:
1.1 anton 3802:
1.29 crook 3803: @example
1.30 anton 3804: @kbd{: add-two dup . ." + 2 =" 2 + . ;@key{RET}} redefined add-two ok
1.29 crook 3805: @end example
1.5 anton 3806:
1.29 crook 3807: Forth recognised that we were defining a word that already exists, and
3808: printed a message to warn us of that fact. Let's try out the new
3809: definition:
1.5 anton 3810:
1.29 crook 3811: @example
1.30 anton 3812: @kbd{9 add-two@key{RET}} 9 + 2 =11 ok
1.29 crook 3813: @end example
1.1 anton 3814:
1.29 crook 3815: @noindent
3816: All that we've actually done here, though, is to create a new
3817: definition, with a particular name. The fact that there was already a
3818: definition with the same name did not make any difference to the way
3819: that the new definition was created (except that Forth printed a warning
3820: message). The old definition of add-two still exists (try @code{demo}
3821: again to see that this is true). Any new definition will use the new
3822: definition of @code{add-two}, but old definitions continue to use the
3823: version that already existed at the time that they were @code{compiled}.
1.1 anton 3824:
1.29 crook 3825: Before you go on to the next section, try defining and redefining some
3826: words of your own.
1.1 anton 3827:
1.29 crook 3828: @comment ----------------------------------------------
3829: @node How does that work?, Forth is written in Forth, Your first definition, Introduction
3830: @section How does that work?
3831: @cindex parsing words
1.1 anton 3832:
1.30 anton 3833: @c That's pretty deep (IMO way too deep) for an introduction. - anton
3834:
3835: @c Is it a good idea to talk about the interpretation semantics of a
3836: @c number? We don't have an xt to go along with it. - anton
3837:
3838: @c Now that I have eliminated execution semantics, I wonder if it would not
3839: @c be better to keep them (or add run-time semantics), to make it easier to
3840: @c explain what compilation semantics usually does. - anton
3841:
1.44 crook 3842: @c nac-> I removed the term ``default compilation sematics'' from the
3843: @c introductory chapter. Removing ``execution semantics'' was making
3844: @c everything simpler to explain, then I think the use of this term made
3845: @c everything more complex again. I replaced it with ``default
3846: @c semantics'' (which is used elsewhere in the manual) by which I mean
3847: @c ``a definition that has neither the immediate nor the compile-only
3848: @c flag set''. I reworded big chunks of the ``how does that work''
3849: @c section (and, unusually for me, I think I even made it shorter!). See
3850: @c what you think -- I know I have not addressed your primary concern
3851: @c that it is too heavy-going for an introduction. From what I understood
3852: @c of your course notes it looks as though they might be a good framework.
3853: @c Things that I've tried to capture here are some things that came as a
3854: @c great revelation here when I first understood them. Also, I like the
3855: @c fact that a very simple code example shows up almost all of the issues
3856: @c that you need to understand to see how Forth works. That's unique and
3857: @c worthwhile to emphasise.
3858:
1.29 crook 3859: Now we're going to take another look at the definition of @code{add-two}
3860: from the previous section. From our knowledge of the way that the text
3861: interpreter works, we would have expected this result when we tried to
3862: define @code{add-two}:
1.21 crook 3863:
1.29 crook 3864: @example
1.44 crook 3865: @kbd{: add-two 2 + . ;@key{RET}}
1.29 crook 3866: ^^^^^^^
3867: Error: Undefined word
3868: @end example
1.28 crook 3869:
1.29 crook 3870: The reason that this didn't happen is bound up in the way that @code{:}
3871: works. The word @code{:} does two special things. The first special
3872: thing that it does prevents the text interpreter from ever seeing the
3873: characters @code{add-two}. The text interpreter uses a variable called
3874: @cindex modifying >IN
1.44 crook 3875: @code{>IN} (pronounced ``to-in'') to keep track of where it is in the
1.29 crook 3876: input line. When it encounters the word @code{:} it behaves in exactly
3877: the same way as it does for any other word; it looks it up in the name
3878: dictionary, finds its xt and executes it. When @code{:} executes, it
3879: looks at the input buffer, finds the word @code{add-two} and advances the
3880: value of @code{>IN} to point past it. It then does some other stuff
3881: associated with creating the new definition (including creating an entry
3882: for @code{add-two} in the name dictionary). When the execution of @code{:}
3883: completes, control returns to the text interpreter, which is oblivious
3884: to the fact that it has been tricked into ignoring part of the input
3885: line.
1.21 crook 3886:
1.29 crook 3887: @cindex parsing words
3888: Words like @code{:} -- words that advance the value of @code{>IN} and so
3889: prevent the text interpreter from acting on the whole of the input line
3890: -- are called @dfn{parsing words}.
1.21 crook 3891:
1.29 crook 3892: @cindex @code{state} - effect on the text interpreter
3893: @cindex text interpreter - effect of state
3894: The second special thing that @code{:} does is change the value of a
3895: variable called @code{state}, which affects the way that the text
3896: interpreter behaves. When Gforth starts up, @code{state} has the value
3897: 0, and the text interpreter is said to be @dfn{interpreting}. During a
3898: colon definition (started with @code{:}), @code{state} is set to -1 and
1.44 crook 3899: the text interpreter is said to be @dfn{compiling}.
3900:
3901: In this example, the text interpreter is compiling when it processes the
3902: string ``@code{2 + . ;}''. It still breaks the string down into
3903: character sequences in the same way. However, instead of pushing the
3904: number @code{2} onto the stack, it lays down (@dfn{compiles}) some magic
3905: into the definition of @code{add-two} that will make the number @code{2} get
3906: pushed onto the stack when @code{add-two} is @dfn{executed}. Similarly,
3907: the behaviours of @code{+} and @code{.} are also compiled into the
3908: definition.
3909:
3910: One category of words don't get compiled. These so-called @dfn{immediate
3911: words} get executed (performed @i{now}) regardless of whether the text
3912: interpreter is interpreting or compiling. The word @code{;} is an
3913: immediate word. Rather than being compiled into the definition, it
3914: executes. Its effect is to terminate the current definition, which
3915: includes changing the value of @code{state} back to 0.
3916:
3917: When you execute @code{add-two}, it has a @dfn{run-time effect} that is
3918: exactly the same as if you had typed @code{2 + . @key{RET}} outside of a
3919: definition.
1.28 crook 3920:
1.30 anton 3921: In Forth, every word or number can be described in terms of two
1.29 crook 3922: properties:
1.28 crook 3923:
3924: @itemize @bullet
3925: @item
1.29 crook 3926: @cindex interpretation semantics
1.44 crook 3927: Its @dfn{interpretation semantics} describe how it will behave when the
3928: text interpreter encounters it in @dfn{interpret} state. The
3929: interpretation semantics of a word are represented by an @dfn{execution
3930: token}.
1.28 crook 3931: @item
1.29 crook 3932: @cindex compilation semantics
1.44 crook 3933: Its @dfn{compilation semantics} describe how it will behave when the
3934: text interpreter encounters it in @dfn{compile} state. The compilation
3935: semantics of a word are represented in an implementation-dependent way;
3936: Gforth uses a @dfn{compilation token}.
1.29 crook 3937: @end itemize
3938:
3939: @noindent
3940: Numbers are always treated in a fixed way:
3941:
3942: @itemize @bullet
1.28 crook 3943: @item
1.44 crook 3944: When the number is @dfn{interpreted}, its behaviour is to push the
3945: number onto the stack.
1.28 crook 3946: @item
1.30 anton 3947: When the number is @dfn{compiled}, a piece of code is appended to the
3948: current definition that pushes the number when it runs. (In other words,
3949: the compilation semantics of a number are to postpone its interpretation
3950: semantics until the run-time of the definition that it is being compiled
3951: into.)
1.29 crook 3952: @end itemize
3953:
1.44 crook 3954: Words don't behave in such a regular way, but most have @i{default
3955: semantics} which means that they behave like this:
1.29 crook 3956:
3957: @itemize @bullet
1.28 crook 3958: @item
1.30 anton 3959: The @dfn{interpretation semantics} of the word are to do something useful.
3960: @item
1.29 crook 3961: The @dfn{compilation semantics} of the word are to append its
1.30 anton 3962: @dfn{interpretation semantics} to the current definition (so that its
3963: run-time behaviour is to do something useful).
1.28 crook 3964: @end itemize
3965:
1.30 anton 3966: @cindex immediate words
1.44 crook 3967: The actual behaviour of any particular word can be controlled by using
3968: the words @code{immediate} and @code{compile-only} when the word is
3969: defined. These words set flags in the name dictionary entry of the most
3970: recently defined word, and these flags are retrieved by the text
3971: interpreter when it finds the word in the name dictionary.
3972:
3973: A word that is marked as @dfn{immediate} has compilation semantics that
3974: are identical to its interpretation semantics. In other words, it
3975: behaves like this:
1.29 crook 3976:
3977: @itemize @bullet
3978: @item
1.30 anton 3979: The @dfn{interpretation semantics} of the word are to do something useful.
1.29 crook 3980: @item
1.30 anton 3981: The @dfn{compilation semantics} of the word are to do something useful
3982: (and actually the same thing); i.e., it is executed during compilation.
1.29 crook 3983: @end itemize
1.28 crook 3984:
1.44 crook 3985: Marking a word as @dfn{compile-only} prohibits the text interpreter from
3986: performing the interpretation semantics of the word directly; an attempt
3987: to do so will generate an error. It is never necessary to use
3988: @code{compile-only} (and it is not even part of ANS Forth, though it is
3989: provided by many implementations) but it is good etiquette to apply it
3990: to a word that will not behave correctly (and might have unexpected
3991: side-effects) in interpret state. For example, it is only legal to use
3992: the conditional word @code{IF} within a definition. If you forget this
3993: and try to use it elsewhere, the fact that (in Gforth) it is marked as
3994: @code{compile-only} allows the text interpreter to generate a helpful
3995: error message rather than subjecting you to the consequences of your
3996: folly.
3997:
1.29 crook 3998: This example shows the difference between an immediate and a
3999: non-immediate word:
1.28 crook 4000:
1.29 crook 4001: @example
4002: : show-state state @@ . ;
4003: : show-state-now show-state ; immediate
4004: : word1 show-state ;
4005: : word2 show-state-now ;
1.28 crook 4006: @end example
1.23 crook 4007:
1.29 crook 4008: The word @code{immediate} after the definition of @code{show-state-now}
4009: makes that word an immediate word. These definitions introduce a new
4010: word: @code{@@} (pronounced ``fetch''). This word fetches the value of a
4011: variable, and leaves it on the stack. Therefore, the behaviour of
4012: @code{show-state} is to print a number that represents the current value
4013: of @code{state}.
1.28 crook 4014:
1.29 crook 4015: When you execute @code{word1}, it prints the number 0, indicating that
4016: the system is interpreting. When the text interpreter compiled the
4017: definition of @code{word1}, it encountered @code{show-state} whose
1.30 anton 4018: compilation semantics are to append its interpretation semantics to the
1.29 crook 4019: current definition. When you execute @code{word1}, it performs the
1.30 anton 4020: interpretation semantics of @code{show-state}. At the time that @code{word1}
1.29 crook 4021: (and therefore @code{show-state}) are executed, the system is
4022: interpreting.
1.28 crook 4023:
1.30 anton 4024: When you pressed @key{RET} after entering the definition of @code{word2},
1.29 crook 4025: you should have seen the number -1 printed, followed by ``@code{
4026: ok}''. When the text interpreter compiled the definition of
4027: @code{word2}, it encountered @code{show-state-now}, an immediate word,
1.30 anton 4028: whose compilation semantics are therefore to perform its interpretation
1.29 crook 4029: semantics. It is executed straight away (even before the text
4030: interpreter has moved on to process another group of characters; the
4031: @code{;} in this example). The effect of executing it are to display the
4032: value of @code{state} @i{at the time that the definition of}
4033: @code{word2} @i{is being defined}. Printing -1 demonstrates that the
4034: system is compiling at this time. If you execute @code{word2} it does
4035: nothing at all.
1.28 crook 4036:
1.29 crook 4037: @cindex @code{."}, how it works
4038: Before leaving the subject of immediate words, consider the behaviour of
4039: @code{."} in the definition of @code{greet}, in the previous
4040: section. This word is both a parsing word and an immediate word. Notice
4041: that there is a space between @code{."} and the start of the text
4042: @code{Hello and welcome}, but that there is no space between the last
4043: letter of @code{welcome} and the @code{"} character. The reason for this
4044: is that @code{."} is a Forth word; it must have a space after it so that
4045: the text interpreter can identify it. The @code{"} is not a Forth word;
4046: it is a @dfn{delimiter}. The examples earlier show that, when the string
4047: is displayed, there is neither a space before the @code{H} nor after the
4048: @code{e}. Since @code{."} is an immediate word, it executes at the time
4049: that @code{greet} is defined. When it executes, its behaviour is to
4050: search forward in the input line looking for the delimiter. When it
4051: finds the delimiter, it updates @code{>IN} to point past the
4052: delimiter. It also compiles some magic code into the definition of
4053: @code{greet}; the xt of a run-time routine that prints a text string. It
4054: compiles the string @code{Hello and welcome} into memory so that it is
4055: available to be printed later. When the text interpreter gains control,
4056: the next word it finds in the input stream is @code{;} and so it
4057: terminates the definition of @code{greet}.
1.28 crook 4058:
4059:
4060: @comment ----------------------------------------------
1.29 crook 4061: @node Forth is written in Forth, Review - elements of a Forth system, How does that work?, Introduction
4062: @section Forth is written in Forth
4063: @cindex structure of Forth programs
4064:
4065: When you start up a Forth compiler, a large number of definitions
4066: already exist. In Forth, you develop a new application using bottom-up
4067: programming techniques to create new definitions that are defined in
4068: terms of existing definitions. As you create each definition you can
4069: test and debug it interactively.
4070:
4071: If you have tried out the examples in this section, you will probably
4072: have typed them in by hand; when you leave Gforth, your definitions will
4073: be lost. You can avoid this by using a text editor to enter Forth source
4074: code into a file, and then loading code from the file using
1.49 anton 4075: @code{include} (@pxref{Forth source files}). A Forth source file is
1.29 crook 4076: processed by the text interpreter, just as though you had typed it in by
4077: hand@footnote{Actually, there are some subtle differences -- see
4078: @ref{The Text Interpreter}.}.
4079:
4080: Gforth also supports the traditional Forth alternative to using text
1.49 anton 4081: files for program entry (@pxref{Blocks}).
1.28 crook 4082:
1.29 crook 4083: In common with many, if not most, Forth compilers, most of Gforth is
4084: actually written in Forth. All of the @file{.fs} files in the
4085: installation directory@footnote{For example,
1.30 anton 4086: @file{/usr/local/share/gforth...}} are Forth source files, which you can
1.29 crook 4087: study to see examples of Forth programming.
1.28 crook 4088:
1.29 crook 4089: Gforth maintains a history file that records every line that you type to
4090: the text interpreter. This file is preserved between sessions, and is
4091: used to provide a command-line recall facility. If you enter long
4092: definitions by hand, you can use a text editor to paste them out of the
4093: history file into a Forth source file for reuse at a later time
1.49 anton 4094: (for more information @pxref{Command-line editing}).
1.28 crook 4095:
4096:
4097: @comment ----------------------------------------------
1.29 crook 4098: @node Review - elements of a Forth system, Where to go next, Forth is written in Forth, Introduction
4099: @section Review - elements of a Forth system
4100: @cindex elements of a Forth system
1.28 crook 4101:
1.29 crook 4102: To summarise this chapter:
1.28 crook 4103:
4104: @itemize @bullet
4105: @item
1.29 crook 4106: Forth programs use @dfn{factoring} to break a problem down into small
4107: fragments called @dfn{words} or @dfn{definitions}.
4108: @item
4109: Forth program development is an interactive process.
4110: @item
4111: The main command loop that accepts input, and controls both
4112: interpretation and compilation, is called the @dfn{text interpreter}
4113: (also known as the @dfn{outer interpreter}).
4114: @item
4115: Forth has a very simple syntax, consisting of words and numbers
4116: separated by spaces or carriage-return characters. Any additional syntax
4117: is imposed by @dfn{parsing words}.
4118: @item
4119: Forth uses a stack to pass parameters between words. As a result, it
4120: uses postfix notation.
4121: @item
4122: To use a word that has previously been defined, the text interpreter
4123: searches for the word in the @dfn{name dictionary}.
4124: @item
1.30 anton 4125: Words have @dfn{interpretation semantics} and @dfn{compilation semantics}.
1.28 crook 4126: @item
1.29 crook 4127: The text interpreter uses the value of @code{state} to select between
4128: the use of the @dfn{interpretation semantics} and the @dfn{compilation
4129: semantics} of a word that it encounters.
1.28 crook 4130: @item
1.30 anton 4131: The relationship between the @dfn{interpretation semantics} and
4132: @dfn{compilation semantics} for a word
1.29 crook 4133: depend upon the way in which the word was defined (for example, whether
4134: it is an @dfn{immediate} word).
1.28 crook 4135: @item
1.29 crook 4136: Forth definitions can be implemented in Forth (called @dfn{high-level
4137: definitions}) or in some other way (usually a lower-level language and
4138: as a result often called @dfn{low-level definitions}, @dfn{code
4139: definitions} or @dfn{primitives}).
1.28 crook 4140: @item
1.29 crook 4141: Many Forth systems are implemented mainly in Forth.
1.28 crook 4142: @end itemize
4143:
4144:
1.29 crook 4145: @comment ----------------------------------------------
1.48 anton 4146: @node Where to go next, Exercises, Review - elements of a Forth system, Introduction
1.29 crook 4147: @section Where To Go Next
4148: @cindex where to go next
1.28 crook 4149:
1.29 crook 4150: Amazing as it may seem, if you have read (and understood) this far, you
4151: know almost all the fundamentals about the inner workings of a Forth
4152: system. You certainly know enough to be able to read and understand the
4153: rest of this manual and the ANS Forth document, to learn more about the
4154: facilities that Forth in general and Gforth in particular provide. Even
4155: scarier, you know almost enough to implement your own Forth system.
1.30 anton 4156: However, that's not a good idea just yet... better to try writing some
1.29 crook 4157: programs in Gforth.
1.28 crook 4158:
1.29 crook 4159: Forth has such a rich vocabulary that it can be hard to know where to
4160: start in learning it. This section suggests a few sets of words that are
4161: enough to write small but useful programs. Use the word index in this
4162: document to learn more about each word, then try it out and try to write
4163: small definitions using it. Start by experimenting with these words:
1.28 crook 4164:
4165: @itemize @bullet
4166: @item
1.29 crook 4167: Arithmetic: @code{+ - * / /MOD */ ABS INVERT}
4168: @item
4169: Comparison: @code{MIN MAX =}
4170: @item
4171: Logic: @code{AND OR XOR NOT}
4172: @item
4173: Stack manipulation: @code{DUP DROP SWAP OVER}
1.28 crook 4174: @item
1.29 crook 4175: Loops and decisions: @code{IF ELSE ENDIF ?DO I LOOP}
1.28 crook 4176: @item
1.29 crook 4177: Input/Output: @code{. ." EMIT CR KEY}
1.28 crook 4178: @item
1.29 crook 4179: Defining words: @code{: ; CREATE}
1.28 crook 4180: @item
1.29 crook 4181: Memory allocation words: @code{ALLOT ,}
1.28 crook 4182: @item
1.29 crook 4183: Tools: @code{SEE WORDS .S MARKER}
4184: @end itemize
4185:
4186: When you have mastered those, go on to:
4187:
4188: @itemize @bullet
1.28 crook 4189: @item
1.29 crook 4190: More defining words: @code{VARIABLE CONSTANT VALUE TO CREATE DOES>}
1.28 crook 4191: @item
1.29 crook 4192: Memory access: @code{@@ !}
1.28 crook 4193: @end itemize
1.23 crook 4194:
1.29 crook 4195: When you have mastered these, there's nothing for it but to read through
4196: the whole of this manual and find out what you've missed.
4197:
4198: @comment ----------------------------------------------
1.48 anton 4199: @node Exercises, , Where to go next, Introduction
1.29 crook 4200: @section Exercises
4201: @cindex exercises
4202:
4203: TODO: provide a set of programming excercises linked into the stuff done
4204: already and into other sections of the manual. Provide solutions to all
4205: the exercises in a .fs file in the distribution.
4206:
4207: @c Get some inspiration from Starting Forth and Kelly&Spies.
4208:
4209: @c excercises:
4210: @c 1. take inches and convert to feet and inches.
4211: @c 2. take temperature and convert from fahrenheight to celcius;
4212: @c may need to care about symmetric vs floored??
4213: @c 3. take input line and do character substitution
4214: @c to encipher or decipher
4215: @c 4. as above but work on a file for in and out
4216: @c 5. take input line and convert to pig-latin
4217: @c
4218: @c thing of sets of things to exercise then come up with
4219: @c problems that need those things.
4220:
4221:
1.26 crook 4222: @c ******************************************************************
1.29 crook 4223: @node Words, Error messages, Introduction, Top
1.1 anton 4224: @chapter Forth Words
1.26 crook 4225: @cindex words
1.1 anton 4226:
4227: @menu
4228: * Notation::
1.21 crook 4229: * Comments::
4230: * Boolean Flags::
1.1 anton 4231: * Arithmetic::
4232: * Stack Manipulation::
1.5 anton 4233: * Memory::
1.1 anton 4234: * Control Structures::
4235: * Defining Words::
1.47 crook 4236: * Interpretation and Compilation Semantics::
4237: * Tokens for Words::
1.21 crook 4238: * The Text Interpreter::
4239: * Word Lists::
4240: * Environmental Queries::
1.12 anton 4241: * Files::
4242: * Blocks::
4243: * Other I/O::
4244: * Programming Tools::
4245: * Assembler and Code Words::
4246: * Threading Words::
1.26 crook 4247: * Locals::
4248: * Structures::
4249: * Object-oriented Forth::
1.21 crook 4250: * Passing Commands to the OS::
1.47 crook 4251: * Keeping track of Time::
1.21 crook 4252: * Miscellaneous Words::
1.1 anton 4253: @end menu
4254:
1.21 crook 4255: @node Notation, Comments, Words, Words
1.1 anton 4256: @section Notation
4257: @cindex notation of glossary entries
4258: @cindex format of glossary entries
4259: @cindex glossary notation format
4260: @cindex word glossary entry format
4261:
4262: The Forth words are described in this section in the glossary notation
4263: that has become a de-facto standard for Forth texts, i.e.,
4264:
4265: @format
1.29 crook 4266: @i{word} @i{Stack effect} @i{wordset} @i{pronunciation}
1.1 anton 4267: @end format
1.29 crook 4268: @i{Description}
1.1 anton 4269:
4270: @table @var
4271: @item word
1.28 crook 4272: The name of the word.
1.1 anton 4273:
4274: @item Stack effect
4275: @cindex stack effect
1.29 crook 4276: The stack effect is written in the notation @code{@i{before} --
4277: @i{after}}, where @i{before} and @i{after} describe the top of
1.1 anton 4278: stack entries before and after the execution of the word. The rest of
4279: the stack is not touched by the word. The top of stack is rightmost,
4280: i.e., a stack sequence is written as it is typed in. Note that Gforth
4281: uses a separate floating point stack, but a unified stack
1.29 crook 4282: notation. Also, return stack effects are not shown in @i{stack
4283: effect}, but in @i{Description}. The name of a stack item describes
1.1 anton 4284: the type and/or the function of the item. See below for a discussion of
4285: the types.
4286:
4287: All words have two stack effects: A compile-time stack effect and a
4288: run-time stack effect. The compile-time stack-effect of most words is
1.29 crook 4289: @i{ -- }. If the compile-time stack-effect of a word deviates from
1.1 anton 4290: this standard behaviour, or the word does other unusual things at
4291: compile time, both stack effects are shown; otherwise only the run-time
4292: stack effect is shown.
4293:
4294: @cindex pronounciation of words
4295: @item pronunciation
4296: How the word is pronounced.
4297:
4298: @cindex wordset
4299: @item wordset
1.21 crook 4300: The ANS Forth standard is divided into several word sets. A standard
4301: system need not support all of them. Therefore, in theory, the fewer
4302: word sets your program uses the more portable it will be. However, we
4303: suspect that most ANS Forth systems on personal machines will feature
1.26 crook 4304: all word sets. Words that are not defined in ANS Forth have
1.21 crook 4305: @code{gforth} or @code{gforth-internal} as word set. @code{gforth}
1.1 anton 4306: describes words that will work in future releases of Gforth;
4307: @code{gforth-internal} words are more volatile. Environmental query
4308: strings are also displayed like words; you can recognize them by the
1.21 crook 4309: @code{environment} in the word set field.
1.1 anton 4310:
4311: @item Description
4312: A description of the behaviour of the word.
4313: @end table
4314:
4315: @cindex types of stack items
4316: @cindex stack item types
4317: The type of a stack item is specified by the character(s) the name
4318: starts with:
4319:
4320: @table @code
4321: @item f
4322: @cindex @code{f}, stack item type
4323: Boolean flags, i.e. @code{false} or @code{true}.
4324: @item c
4325: @cindex @code{c}, stack item type
4326: Char
4327: @item w
4328: @cindex @code{w}, stack item type
4329: Cell, can contain an integer or an address
4330: @item n
4331: @cindex @code{n}, stack item type
4332: signed integer
4333: @item u
4334: @cindex @code{u}, stack item type
4335: unsigned integer
4336: @item d
4337: @cindex @code{d}, stack item type
4338: double sized signed integer
4339: @item ud
4340: @cindex @code{ud}, stack item type
4341: double sized unsigned integer
4342: @item r
4343: @cindex @code{r}, stack item type
4344: Float (on the FP stack)
1.21 crook 4345: @item a-
1.1 anton 4346: @cindex @code{a_}, stack item type
4347: Cell-aligned address
1.21 crook 4348: @item c-
1.1 anton 4349: @cindex @code{c_}, stack item type
4350: Char-aligned address (note that a Char may have two bytes in Windows NT)
1.21 crook 4351: @item f-
1.1 anton 4352: @cindex @code{f_}, stack item type
4353: Float-aligned address
1.21 crook 4354: @item df-
1.1 anton 4355: @cindex @code{df_}, stack item type
4356: Address aligned for IEEE double precision float
1.21 crook 4357: @item sf-
1.1 anton 4358: @cindex @code{sf_}, stack item type
4359: Address aligned for IEEE single precision float
4360: @item xt
4361: @cindex @code{xt}, stack item type
4362: Execution token, same size as Cell
4363: @item wid
4364: @cindex @code{wid}, stack item type
1.21 crook 4365: Word list ID, same size as Cell
1.1 anton 4366: @item f83name
4367: @cindex @code{f83name}, stack item type
4368: Pointer to a name structure
4369: @item "
4370: @cindex @code{"}, stack item type
1.12 anton 4371: string in the input stream (not on the stack). The terminating character
4372: is a blank by default. If it is not a blank, it is shown in @code{<>}
1.1 anton 4373: quotes.
4374: @end table
4375:
1.21 crook 4376: @node Comments, Boolean Flags, Notation, Words
4377: @section Comments
1.26 crook 4378: @cindex comments
1.21 crook 4379:
1.29 crook 4380: Forth supports two styles of comment; the traditional @i{in-line} comment,
4381: @code{(} and its modern cousin, the @i{comment to end of line}; @code{\}.
1.21 crook 4382:
1.44 crook 4383:
1.23 crook 4384: doc-(
1.21 crook 4385: doc-\
1.23 crook 4386: doc-\G
1.21 crook 4387:
1.44 crook 4388:
1.21 crook 4389: @node Boolean Flags, Arithmetic, Comments, Words
4390: @section Boolean Flags
1.26 crook 4391: @cindex Boolean flags
1.21 crook 4392:
4393: A Boolean flag is cell-sized. A cell with all bits clear represents the
4394: flag @code{false} and a flag with all bits set represents the flag
1.26 crook 4395: @code{true}. Words that check a flag (for example, @code{IF}) will treat
1.29 crook 4396: a cell that has @i{any} bit set as @code{true}.
1.21 crook 4397:
1.44 crook 4398:
1.21 crook 4399: doc-true
4400: doc-false
1.29 crook 4401: doc-on
4402: doc-off
1.21 crook 4403:
1.44 crook 4404:
1.21 crook 4405: @node Arithmetic, Stack Manipulation, Boolean Flags, Words
1.1 anton 4406: @section Arithmetic
4407: @cindex arithmetic words
4408:
4409: @cindex division with potentially negative operands
4410: Forth arithmetic is not checked, i.e., you will not hear about integer
4411: overflow on addition or multiplication, you may hear about division by
4412: zero if you are lucky. The operator is written after the operands, but
4413: the operands are still in the original order. I.e., the infix @code{2-1}
4414: corresponds to @code{2 1 -}. Forth offers a variety of division
4415: operators. If you perform division with potentially negative operands,
4416: you do not want to use @code{/} or @code{/mod} with its undefined
4417: behaviour, but rather @code{fm/mod} or @code{sm/mod} (probably the
4418: former, @pxref{Mixed precision}).
1.26 crook 4419: @comment TODO discuss the different division forms and the std approach
1.1 anton 4420:
4421: @menu
4422: * Single precision::
4423: * Bitwise operations::
1.21 crook 4424: * Double precision:: Double-cell integer arithmetic
4425: * Numeric comparison::
1.29 crook 4426: * Mixed precision:: Operations with single and double-cell integers
1.1 anton 4427: * Floating Point::
4428: @end menu
4429:
4430: @node Single precision, Bitwise operations, Arithmetic, Arithmetic
4431: @subsection Single precision
4432: @cindex single precision arithmetic words
4433:
1.21 crook 4434: By default, numbers in Forth are single-precision integers that are 1
1.26 crook 4435: cell in size. They can be signed or unsigned, depending upon how you
1.49 anton 4436: treat them. For the rules used by the text interpreter for recognising
4437: single-precision integers see @ref{Number Conversion}.
1.21 crook 4438:
1.44 crook 4439:
1.1 anton 4440: doc-+
1.21 crook 4441: doc-1+
1.1 anton 4442: doc--
1.21 crook 4443: doc-1-
1.1 anton 4444: doc-*
4445: doc-/
4446: doc-mod
4447: doc-/mod
4448: doc-negate
4449: doc-abs
4450: doc-min
4451: doc-max
1.21 crook 4452: doc-d>s
1.27 crook 4453: doc-floored
1.1 anton 4454:
1.44 crook 4455:
1.21 crook 4456: @node Bitwise operations, Double precision, Single precision, Arithmetic
1.1 anton 4457: @subsection Bitwise operations
4458: @cindex bitwise operation words
4459:
1.44 crook 4460:
1.1 anton 4461: doc-and
4462: doc-or
4463: doc-xor
4464: doc-invert
1.21 crook 4465: doc-lshift
4466: doc-rshift
1.1 anton 4467: doc-2*
1.21 crook 4468: doc-d2*
1.1 anton 4469: doc-2/
1.21 crook 4470: doc-d2/
4471:
1.44 crook 4472:
1.21 crook 4473: @node Double precision, Numeric comparison, Bitwise operations, Arithmetic
4474: @subsection Double precision
4475: @cindex double precision arithmetic words
4476:
1.49 anton 4477: For the rules used by the text interpreter for
4478: recognising double-precision integers, see @ref{Number Conversion}.
1.21 crook 4479:
4480: A double precision number is represented by a cell pair, with the most
1.31 anton 4481: significant cell at the TOS. It is trivial to convert an unsigned
1.26 crook 4482: single to an (unsigned) double; simply push a @code{0} onto the
4483: TOS. Since numbers are represented by Gforth using 2's complement
4484: arithmetic, converting a signed single to a (signed) double requires
1.31 anton 4485: sign-extension across the most significant cell. This can be achieved
1.26 crook 4486: using @code{s>d}. The moral of the story is that you cannot convert a
4487: number without knowing whether it represents an unsigned or a
4488: signed number.
1.21 crook 4489:
1.44 crook 4490:
1.21 crook 4491: doc-s>d
4492: doc-d+
4493: doc-d-
4494: doc-dnegate
4495: doc-dabs
4496: doc-dmin
4497: doc-dmax
4498:
1.44 crook 4499:
1.21 crook 4500: @node Numeric comparison, Mixed precision, Double precision, Arithmetic
4501: @subsection Numeric comparison
4502: @cindex numeric comparison words
4503:
1.44 crook 4504:
1.28 crook 4505: doc-<
4506: doc-<=
4507: doc-<>
4508: doc-=
4509: doc->
4510: doc->=
4511:
1.21 crook 4512: doc-0<
1.23 crook 4513: doc-0<=
1.21 crook 4514: doc-0<>
4515: doc-0=
1.23 crook 4516: doc-0>
4517: doc-0>=
1.28 crook 4518:
4519: doc-u<
4520: doc-u<=
1.44 crook 4521: @c u<> and u= exist but are the same as <> and =
1.31 anton 4522: @c doc-u<>
4523: @c doc-u=
1.28 crook 4524: doc-u>
4525: doc-u>=
4526:
4527: doc-within
4528:
4529: doc-d<
4530: doc-d<=
4531: doc-d<>
4532: doc-d=
4533: doc-d>
4534: doc-d>=
1.23 crook 4535:
1.21 crook 4536: doc-d0<
1.23 crook 4537: doc-d0<=
4538: doc-d0<>
1.21 crook 4539: doc-d0=
1.23 crook 4540: doc-d0>
4541: doc-d0>=
4542:
1.21 crook 4543: doc-du<
1.28 crook 4544: doc-du<=
1.44 crook 4545: @c du<> and du= exist but are the same as d<> and d=
1.31 anton 4546: @c doc-du<>
4547: @c doc-du=
1.28 crook 4548: doc-du>
4549: doc-du>=
1.1 anton 4550:
1.44 crook 4551:
1.21 crook 4552: @node Mixed precision, Floating Point, Numeric comparison, Arithmetic
1.1 anton 4553: @subsection Mixed precision
4554: @cindex mixed precision arithmetic words
4555:
1.44 crook 4556:
1.1 anton 4557: doc-m+
4558: doc-*/
4559: doc-*/mod
4560: doc-m*
4561: doc-um*
4562: doc-m*/
4563: doc-um/mod
4564: doc-fm/mod
4565: doc-sm/rem
4566:
1.44 crook 4567:
1.21 crook 4568: @node Floating Point, , Mixed precision, Arithmetic
1.1 anton 4569: @subsection Floating Point
4570: @cindex floating point arithmetic words
4571:
1.49 anton 4572: For the rules used by the text interpreter for
4573: recognising floating-point numbers see @ref{Number Conversion}.
1.1 anton 4574:
1.32 anton 4575: Gforth has a separate floating point
1.26 crook 4576: stack, but the documentation uses the unified notation.
1.1 anton 4577:
4578: @cindex floating-point arithmetic, pitfalls
4579: Floating point numbers have a number of unpleasant surprises for the
4580: unwary (e.g., floating point addition is not associative) and even a few
4581: for the wary. You should not use them unless you know what you are doing
4582: or you don't care that the results you get are totally bogus. If you
4583: want to learn about the problems of floating point numbers (and how to
4584: avoid them), you might start with @cite{David Goldberg, What Every
4585: Computer Scientist Should Know About Floating-Point Arithmetic, ACM
1.17 anton 4586: Computing Surveys 23(1):5@minus{}48, March 1991}
1.47 crook 4587: (@uref{http://www.validgh.com/goldberg/paper.ps}).
1.1 anton 4588:
1.44 crook 4589:
1.21 crook 4590: doc-d>f
4591: doc-f>d
1.1 anton 4592: doc-f+
4593: doc-f-
4594: doc-f*
4595: doc-f/
4596: doc-fnegate
4597: doc-fabs
4598: doc-fmax
4599: doc-fmin
4600: doc-floor
4601: doc-fround
4602: doc-f**
4603: doc-fsqrt
4604: doc-fexp
4605: doc-fexpm1
4606: doc-fln
4607: doc-flnp1
4608: doc-flog
4609: doc-falog
1.32 anton 4610: doc-f2*
4611: doc-f2/
4612: doc-1/f
4613: doc-precision
4614: doc-set-precision
4615:
4616: @cindex angles in trigonometric operations
4617: @cindex trigonometric operations
4618: Angles in floating point operations are given in radians (a full circle
4619: has 2 pi radians).
4620:
1.1 anton 4621: doc-fsin
4622: doc-fcos
4623: doc-fsincos
4624: doc-ftan
4625: doc-fasin
4626: doc-facos
4627: doc-fatan
4628: doc-fatan2
4629: doc-fsinh
4630: doc-fcosh
4631: doc-ftanh
4632: doc-fasinh
4633: doc-facosh
4634: doc-fatanh
1.21 crook 4635: doc-pi
1.28 crook 4636:
1.32 anton 4637: @cindex equality of floats
4638: @cindex floating-point comparisons
1.31 anton 4639: One particular problem with floating-point arithmetic is that comparison
4640: for equality often fails when you would expect it to succeed. For this
4641: reason approximate equality is often preferred (but you still have to
4642: know what you are doing). The comparison words are:
4643:
4644: doc-f~rel
4645: doc-f~abs
4646: doc-f=
4647: doc-f~
4648: doc-f<>
4649:
4650: doc-f<
4651: doc-f<=
4652: doc-f>
4653: doc-f>=
4654:
1.21 crook 4655: doc-f0<
1.28 crook 4656: doc-f0<=
4657: doc-f0<>
1.21 crook 4658: doc-f0=
1.28 crook 4659: doc-f0>
4660: doc-f0>=
4661:
1.1 anton 4662:
4663: @node Stack Manipulation, Memory, Arithmetic, Words
4664: @section Stack Manipulation
4665: @cindex stack manipulation words
4666:
4667: @cindex floating-point stack in the standard
1.21 crook 4668: Gforth maintains a number of separate stacks:
4669:
1.29 crook 4670: @cindex data stack
4671: @cindex parameter stack
1.21 crook 4672: @itemize @bullet
4673: @item
1.29 crook 4674: A data stack (also known as the @dfn{parameter stack}) -- for
4675: characters, cells, addresses, and double cells.
1.21 crook 4676:
1.29 crook 4677: @cindex floating-point stack
1.21 crook 4678: @item
1.44 crook 4679: A floating point stack -- for holding floating point (FP) numbers.
1.21 crook 4680:
1.29 crook 4681: @cindex return stack
1.21 crook 4682: @item
1.44 crook 4683: A return stack -- for holding the return addresses of colon
1.32 anton 4684: definitions and other (non-FP) data.
1.21 crook 4685:
1.29 crook 4686: @cindex locals stack
1.21 crook 4687: @item
1.44 crook 4688: A locals stack -- for holding local variables.
1.21 crook 4689: @end itemize
4690:
1.1 anton 4691: @menu
4692: * Data stack::
4693: * Floating point stack::
4694: * Return stack::
4695: * Locals stack::
4696: * Stack pointer manipulation::
4697: @end menu
4698:
4699: @node Data stack, Floating point stack, Stack Manipulation, Stack Manipulation
4700: @subsection Data stack
4701: @cindex data stack manipulation words
4702: @cindex stack manipulations words, data stack
4703:
1.44 crook 4704:
1.1 anton 4705: doc-drop
4706: doc-nip
4707: doc-dup
4708: doc-over
4709: doc-tuck
4710: doc-swap
1.21 crook 4711: doc-pick
1.1 anton 4712: doc-rot
4713: doc--rot
4714: doc-?dup
4715: doc-roll
4716: doc-2drop
4717: doc-2nip
4718: doc-2dup
4719: doc-2over
4720: doc-2tuck
4721: doc-2swap
4722: doc-2rot
4723:
1.44 crook 4724:
1.1 anton 4725: @node Floating point stack, Return stack, Data stack, Stack Manipulation
4726: @subsection Floating point stack
4727: @cindex floating-point stack manipulation words
4728: @cindex stack manipulation words, floating-point stack
4729:
1.32 anton 4730: Whilst every sane Forth has a separate floating-point stack, it is not
4731: strictly required; an ANS Forth system could theoretically keep
4732: floating-point numbers on the data stack. As an additional difficulty,
4733: you don't know how many cells a floating-point number takes. It is
4734: reportedly possible to write words in a way that they work also for a
4735: unified stack model, but we do not recommend trying it. Instead, just
4736: say that your program has an environmental dependency on a separate
4737: floating-point stack.
4738:
4739: doc-floating-stack
4740:
1.1 anton 4741: doc-fdrop
4742: doc-fnip
4743: doc-fdup
4744: doc-fover
4745: doc-ftuck
4746: doc-fswap
1.21 crook 4747: doc-fpick
1.1 anton 4748: doc-frot
4749:
1.44 crook 4750:
1.1 anton 4751: @node Return stack, Locals stack, Floating point stack, Stack Manipulation
4752: @subsection Return stack
4753: @cindex return stack manipulation words
4754: @cindex stack manipulation words, return stack
4755:
1.32 anton 4756: @cindex return stack and locals
4757: @cindex locals and return stack
4758: A Forth system is allowed to keep local variables on the
4759: return stack. This is reasonable, as local variables usually eliminate
4760: the need to use the return stack explicitly. So, if you want to produce
4761: a standard compliant program and you are using local variables in a
4762: word, forget about return stack manipulations in that word (refer to the
4763: standard document for the exact rules).
4764:
1.1 anton 4765: doc->r
4766: doc-r>
4767: doc-r@
4768: doc-rdrop
4769: doc-2>r
4770: doc-2r>
4771: doc-2r@
4772: doc-2rdrop
4773:
1.44 crook 4774:
1.1 anton 4775: @node Locals stack, Stack pointer manipulation, Return stack, Stack Manipulation
4776: @subsection Locals stack
4777:
1.47 crook 4778: Gforth uses an extra locals stack. It is described, along with the
4779: reasons for its existence, in @ref{Implementation,Implementation of locals}.
1.21 crook 4780:
1.1 anton 4781: @node Stack pointer manipulation, , Locals stack, Stack Manipulation
4782: @subsection Stack pointer manipulation
4783: @cindex stack pointer manipulation words
4784:
1.44 crook 4785: @c removed s0 r0 l0 -- they are obsolete aliases for sp0 rp0 lp0
1.21 crook 4786: doc-sp0
1.1 anton 4787: doc-sp@
4788: doc-sp!
1.21 crook 4789: doc-fp0
1.1 anton 4790: doc-fp@
4791: doc-fp!
1.21 crook 4792: doc-rp0
1.1 anton 4793: doc-rp@
4794: doc-rp!
1.21 crook 4795: doc-lp0
1.1 anton 4796: doc-lp@
4797: doc-lp!
4798:
1.44 crook 4799:
1.1 anton 4800: @node Memory, Control Structures, Stack Manipulation, Words
4801: @section Memory
1.26 crook 4802: @cindex memory words
1.1 anton 4803:
1.32 anton 4804: @menu
4805: * Memory model::
4806: * Dictionary allocation::
4807: * Heap Allocation::
4808: * Memory Access::
4809: * Address arithmetic::
4810: * Memory Blocks::
4811: @end menu
4812:
4813: @node Memory model, Dictionary allocation, Memory, Memory
4814: @subsection ANS Forth and Gforth memory models
4815:
4816: @c The ANS Forth description is a mess (e.g., is the heap part of
4817: @c the dictionary?), so let's not stick to closely with it.
4818:
4819: ANS Forth considers a Forth system as consisting of several memories, of
4820: which only @dfn{data space} is managed and accessible with the memory
4821: words. Memory not necessarily in data space includes the stacks, the
4822: code (called code space) and the headers (called name space). In Gforth
4823: everything is in data space, but the code for the primitives is usually
4824: read-only.
4825:
4826: Data space is divided into a number of areas: The (data space portion of
4827: the) dictionary@footnote{Sometimes, the term @dfn{dictionary} is used to
4828: refer to the search data structure embodied in word lists and headers,
4829: because it is used for looking up names, just as you would in a
4830: conventional dictionary.}, the heap, and a number of system-allocated
4831: buffers.
4832:
4833: In ANS Forth data space is also divided into contiguous regions. You
4834: can only use address arithmetic within a contiguous region, not between
4835: them. Usually each allocation gives you one contiguous region, but the
1.33 anton 4836: dictionary allocation words have additional rules (@pxref{Dictionary
1.32 anton 4837: allocation}).
4838:
4839: Gforth provides one big address space, and address arithmetic can be
4840: performed between any addresses. However, in the dictionary headers or
4841: code are interleaved with data, so almost the only contiguous data space
4842: regions there are those described by ANS Forth as contiguous; but you
4843: can be sure that the dictionary is allocated towards increasing
4844: addresses even between contiguous regions. The memory order of
4845: allocations in the heap is platform-dependent (and possibly different
4846: from one run to the next).
4847:
4848: @subsubsection ANS Forth dictionary details
4849:
4850: This section is just informative, you can skip it if you are in a hurry.
1.27 crook 4851:
1.29 crook 4852: When you create a colon definition, the text interpreter compiles the
1.32 anton 4853: code for the definition into the code space and compiles the name
4854: of the definition into the header space, together with other
1.27 crook 4855: information about the definition (such as its execution token).
4856:
1.44 crook 4857: When you create a variable, the execution of @code{Variable} will
1.32 anton 4858: compile some code, assign one cell in data space, and compile the name
4859: of the variable into the header space.
1.27 crook 4860:
4861: @cindex memory regions - relationship between them
4862: ANS Forth does not specify the relationship between the three memory
4863: regions, and specifies that a Standard program must not access code or
4864: data space directly -- it may only access data space directly. In
4865: addition, the Standard defines what relationships you may and may not
4866: rely on when allocating regions in data space. These constraints are
4867: simply a reflection of the many diverse techniques that are used to
4868: implement Forth systems; understanding and following the requirements of
4869: the Standard allows you to write portable programs -- programs that run
4870: in the same way on any of these diverse systems. Another way of looking
4871: at this is to say that ANS Forth was designed to permit compliant Forth
4872: systems to be implemented in many diverse ways.
4873:
4874: @cindex memory regions - how they are assigned
1.29 crook 4875: Here are some examples of ways in which name, code and data spaces
4876: might be assigned in different Forth implementations:
1.27 crook 4877:
4878: @itemize @bullet
4879: @item
4880: For a Forth system that runs from RAM under a general-purpose operating
4881: system, it can be convenient to interleave name, code and data spaces in
4882: a single contiguous memory region. This organisation can be
4883: memory-efficient (for example, because the relationship between the name
1.32 anton 4884: dictionary entry and the associated code space entry can be
1.27 crook 4885: implicit, rather than requiring an explicit memory pointer to reference
1.32 anton 4886: from the header space and the code space). This is the
1.27 crook 4887: organisation used by Gforth, as this example@footnote{The addresses
4888: in the example have been truncated to fit it onto the page, and the
4889: addresses and data shown will not match the output from your system} shows:
4890: @example
4891: hex
4892: variable fred 123456 fred !
4893: variable jim abcd jim !
4894: : foo + / - ;
4895: ' fred 10 - 50 dump
4896: ..80: 5C 46 0E 40 84 66 72 65 - 64 20 20 20 20 20 20 20 \F.@.fred
1.50 anton 4897: ..90: D0 9B 04 08 00 00 00 00 - 56 34 12 00 80 46 0E 40 ........V4...F.@@
1.27 crook 4898: ..A0: 83 6A 69 6D 20 20 20 20 - D0 9B 04 08 00 00 00 00 .jim ........
4899: ..B0: CD AB 00 00 9C 46 0E 40 - 83 66 6F 6F 20 20 20 20 .....F.@.foo
4900: ..C0: 80 9B 04 08 00 00 00 00 - E4 2E 05 08 0C 2F 05 08 ............./..
4901: @end example
4902:
4903: @item
4904: For a high-performance system running on a modern RISC processor with a
4905: modified Harvard architecture (one that has a unified main memory but
4906: separate instruction and data caches), it is desirable to separate
4907: processor instructions from processor data. This encourages a high cache
1.32 anton 4908: density and therefore a high cache hit rate. The Forth code space
1.27 crook 4909: is not necessarily made up entirely of processor instructions; its
4910: nature is dependent upon the Forth implementation.
4911:
4912: @item
4913: A Forth compiler that runs on a segmented 8086 processor could be
4914: designed to interleave the name, code and data spaces within a single
4915: 64Kbyte segment. A more common implementation choice is to use a
4916: separate 64Kbyte segment for each region, which provides more memory
4917: overall but provides an address map in which only the data space is
4918: accessible.
4919:
4920: @item
4921: Microprocessors exist that run Forth (or many of the primitives required
4922: to implement the Forth virtual machine efficiently) directly. On these
4923: processors, the relationship between name, code and data spaces may be
1.32 anton 4924: imposed as a side-effect of the architecture of the processor.
1.27 crook 4925:
4926: @item
4927: A Forth compiler that executes from ROM on an embedded system needs its
4928: data space separated from the name and code spaces so that the data
4929: space can be mapped to a RAM area.
4930:
4931: @item
4932: A Forth compiler that runs on an embedded system may have a requirement
4933: for a small memory footprint. On such a system it can be useful to
1.32 anton 4934: separate the header space from the data and code spaces; once the
4935: application has been compiled, the header space is no longer
1.27 crook 4936: required@footnote{more strictly speaking, most applications can be
1.32 anton 4937: designed so that this is the case}. The header space can be deleted
1.29 crook 4938: entirely, or could be stored in memory on a remote @i{host} system for
1.27 crook 4939: debug and development purposes. In the latter case, the compiler running
1.29 crook 4940: on the @i{target} system could implement a protocol across a
1.32 anton 4941: communication link that would allow it to interrogate the header space.
1.27 crook 4942: @end itemize
4943:
1.32 anton 4944: @node Dictionary allocation, Heap Allocation, Memory model, Memory
4945: @subsection Dictionary allocation
1.27 crook 4946: @cindex reserving data space
4947: @cindex data space - reserving some
4948:
1.32 anton 4949: Dictionary allocation is a stack-oriented allocation scheme, i.e., if
4950: you want to deallocate X, you also deallocate everything
4951: allocated after X.
4952:
4953: The allocations using the words below are contiguous and grow the region
4954: towards increasing addresses. Other words that allocate dictionary
4955: memory of any kind (i.e., defining words including @code{:noname}) end
4956: the contiguous region and start a new one.
4957:
4958: In ANS Forth only @code{create}d words are guaranteed to produce an
4959: address that is the start of the following contiguous region. In
4960: particular, the cell allocated by @code{variable} is not guaranteed to
4961: be contiguous with following @code{allot}ed memory.
4962:
4963: You can deallocate memory by using @code{allot} with a negative argument
4964: (with some restrictions, see @code{allot}). For larger deallocations use
4965: @code{marker}.
1.27 crook 4966:
1.29 crook 4967:
1.27 crook 4968: doc-here
4969: doc-unused
4970: doc-allot
4971: doc-c,
1.29 crook 4972: doc-f,
1.27 crook 4973: doc-,
4974: doc-2,
1.29 crook 4975: @cindex user space
4976: doc-udp
4977: doc-uallot
1.27 crook 4978:
1.32 anton 4979: Memory accesses have to be aligned (@pxref{Address arithmetic}). So of
4980: course you should allocate memory in an aligned way, too. I.e., before
4981: allocating allocating a cell, @code{here} must be cell-aligned, etc.
4982: The words below align @code{here} if it is not already. Basically it is
4983: only already aligned for a type, if the last allocation was a multiple
4984: of the size of this type and if @code{here} was aligned for this type
4985: before.
4986:
4987: After freshly @code{create}ing a word, @code{here} is @code{align}ed in
4988: ANS Forth (@code{maxalign}ed in Gforth).
4989:
4990: doc-align
4991: doc-falign
4992: doc-sfalign
4993: doc-dfalign
4994: doc-maxalign
4995: doc-cfalign
4996:
4997:
4998: @node Heap Allocation, Memory Access, Dictionary allocation, Memory
4999: @subsection Heap allocation
5000: @cindex heap allocation
5001: @cindex dynamic allocation of memory
5002: @cindex memory-allocation word set
5003:
5004: Heap allocation supports deallocation of allocated memory in any
5005: order. Dictionary allocation is not affected by it (i.e., it does not
5006: end a contiguous region). In Gforth, these words are implemented using
5007: the standard C library calls malloc(), free() and resize().
5008:
5009: doc-allocate
5010: doc-free
5011: doc-resize
5012:
1.27 crook 5013:
1.32 anton 5014: @node Memory Access, Address arithmetic, Heap Allocation, Memory
1.1 anton 5015: @subsection Memory Access
5016: @cindex memory access words
5017:
1.44 crook 5018:
1.1 anton 5019: doc-@
5020: doc-!
5021: doc-+!
5022: doc-c@
5023: doc-c!
5024: doc-2@
5025: doc-2!
5026: doc-f@
5027: doc-f!
5028: doc-sf@
5029: doc-sf!
5030: doc-df@
5031: doc-df!
5032:
1.32 anton 5033: @node Address arithmetic, Memory Blocks, Memory Access, Memory
5034: @subsection Address arithmetic
1.1 anton 5035: @cindex address arithmetic words
5036:
1.32 anton 5037: Address arithmetic is the foundation on which data structures like
5038: arrays, records (@pxref{Structures}) and objects (@pxref{Object-oriented
5039: Forth}) are built.
5040:
1.1 anton 5041: ANS Forth does not specify the sizes of the data types. Instead, it
5042: offers a number of words for computing sizes and doing address
1.29 crook 5043: arithmetic. Address arithmetic is performed in terms of address units
5044: (aus); on most systems the address unit is one byte. Note that a
5045: character may have more than one au, so @code{chars} is no noop (on
5046: systems where it is a noop, it compiles to nothing).
1.1 anton 5047:
5048: @cindex alignment of addresses for types
5049: ANS Forth also defines words for aligning addresses for specific
5050: types. Many computers require that accesses to specific data types
5051: must only occur at specific addresses; e.g., that cells may only be
5052: accessed at addresses divisible by 4. Even if a machine allows unaligned
5053: accesses, it can usually perform aligned accesses faster.
5054:
5055: For the performance-conscious: alignment operations are usually only
5056: necessary during the definition of a data structure, not during the
5057: (more frequent) accesses to it.
5058:
5059: ANS Forth defines no words for character-aligning addresses. This is not
5060: an oversight, but reflects the fact that addresses that are not
5061: char-aligned have no use in the standard and therefore will not be
5062: created.
5063:
5064: @cindex @code{CREATE} and alignment
1.29 crook 5065: ANS Forth guarantees that addresses returned by @code{CREATE}d words
1.1 anton 5066: are cell-aligned; in addition, Gforth guarantees that these addresses
5067: are aligned for all purposes.
5068:
1.26 crook 5069: Note that the ANS Forth word @code{char} has nothing to do with address
5070: arithmetic.
1.1 anton 5071:
1.44 crook 5072:
1.1 anton 5073: doc-chars
5074: doc-char+
5075: doc-cells
5076: doc-cell+
5077: doc-cell
5078: doc-aligned
5079: doc-floats
5080: doc-float+
5081: doc-float
5082: doc-faligned
5083: doc-sfloats
5084: doc-sfloat+
5085: doc-sfaligned
5086: doc-dfloats
5087: doc-dfloat+
5088: doc-dfaligned
5089: doc-maxaligned
5090: doc-cfaligned
5091: doc-address-unit-bits
5092:
1.44 crook 5093:
1.32 anton 5094: @node Memory Blocks, , Address arithmetic, Memory
1.1 anton 5095: @subsection Memory Blocks
5096: @cindex memory block words
1.27 crook 5097: @cindex character strings - moving and copying
5098:
1.49 anton 5099: Memory blocks often represent character strings; For ways of storing
5100: character strings in memory see @ref{String Formats}. For other
5101: string-processing words see @ref{Displaying characters and strings}.
1.1 anton 5102:
1.32 anton 5103: Some of these words work on address units. Others work on character
5104: units (increments of @code{CHAR}), and expect a @code{CHAR}-aligned
5105: address. Choose the correct operation depending upon your data type.
1.21 crook 5106:
5107: When copying characters between overlapping memory regions, choose
5108: carefully between @code{cmove} and @code{cmove>}.
5109:
1.29 crook 5110: You can only use any of these words @i{portably} to access data space.
1.21 crook 5111:
1.27 crook 5112: @comment TODO - think the naming of the arguments is wrong for move
1.29 crook 5113: @comment well, really it seems to be the Standard that's wrong; it
5114: @comment describes MOVE as a word that requires a CELL-aligned source
5115: @comment and destination address but a xtranfer count that need not
5116: @comment be a multiple of CELL.
1.44 crook 5117:
1.1 anton 5118: doc-move
5119: doc-erase
5120: doc-cmove
5121: doc-cmove>
5122: doc-fill
5123: doc-blank
1.21 crook 5124: doc-compare
5125: doc-search
1.27 crook 5126: doc--trailing
5127: doc-/string
5128:
1.44 crook 5129:
1.27 crook 5130: @comment TODO examples
5131:
1.1 anton 5132:
1.26 crook 5133: @node Control Structures, Defining Words, Memory, Words
1.1 anton 5134: @section Control Structures
5135: @cindex control structures
5136:
1.33 anton 5137: Control structures in Forth cannot be used interpretively, only in a
5138: colon definition@footnote{To be precise, they have no interpretation
5139: semantics (@pxref{Interpretation and Compilation Semantics}).}. We do
5140: not like this limitation, but have not seen a satisfying way around it
5141: yet, although many schemes have been proposed.
1.1 anton 5142:
5143: @menu
1.33 anton 5144: * Selection:: IF ... ELSE ... ENDIF
5145: * Simple Loops:: BEGIN ...
1.29 crook 5146: * Counted Loops:: DO
5147: * Arbitrary control structures::
5148: * Calls and returns::
1.1 anton 5149: * Exception Handling::
5150: @end menu
5151:
5152: @node Selection, Simple Loops, Control Structures, Control Structures
5153: @subsection Selection
5154: @cindex selection control structures
5155: @cindex control structures for selection
5156:
1.33 anton 5157: @c what's the purpose of all these @i? Maybe we should define a macro
5158: @c so we can produce logical markup. - anton
5159:
1.44 crook 5160: @c nac-> When I started working on the manual, a mixture of @i and @var
5161: @c were used inconsistently in code examples and \Glossary entries. These
5162: @c two behave differently in info format so I decided to standardize on @i.
5163: @c Logical markup would be better but texi isn't really upto it, and
5164: @c texi2html just ignores macros.
1.47 crook 5165: @c nac02dec1999-> update: the latest texinfo release can spit out html
5166: @c and it handles macros, so we could do some logical markup. Unfortunately
5167: @c texinfo will not split html output, which would be a big pain if you
5168: @c wanted to put the document on the web, which would be nice.
1.44 crook 5169:
1.1 anton 5170: @cindex @code{IF} control structure
5171: @example
1.29 crook 5172: @i{flag}
1.1 anton 5173: IF
1.29 crook 5174: @i{code}
1.1 anton 5175: ENDIF
5176: @end example
1.21 crook 5177: @noindent
1.33 anton 5178:
1.44 crook 5179: If @i{flag} is non-zero (as far as @code{IF} etc. are concerned, a cell
5180: with any bit set represents truth) @i{code} is executed.
1.33 anton 5181:
1.1 anton 5182: @example
1.29 crook 5183: @i{flag}
1.1 anton 5184: IF
1.29 crook 5185: @i{code1}
1.1 anton 5186: ELSE
1.29 crook 5187: @i{code2}
1.1 anton 5188: ENDIF
5189: @end example
5190:
1.44 crook 5191: If @var{flag} is true, @i{code1} is executed, otherwise @i{code2} is
5192: executed.
1.33 anton 5193:
1.1 anton 5194: You can use @code{THEN} instead of @code{ENDIF}. Indeed, @code{THEN} is
5195: standard, and @code{ENDIF} is not, although it is quite popular. We
5196: recommend using @code{ENDIF}, because it is less confusing for people
5197: who also know other languages (and is not prone to reinforcing negative
5198: prejudices against Forth in these people). Adding @code{ENDIF} to a
5199: system that only supplies @code{THEN} is simple:
5200: @example
1.21 crook 5201: : ENDIF POSTPONE THEN ; immediate
1.1 anton 5202: @end example
5203:
5204: [According to @cite{Webster's New Encyclopedic Dictionary}, @dfn{then
5205: (adv.)} has the following meanings:
5206: @quotation
5207: ... 2b: following next after in order ... 3d: as a necessary consequence
5208: (if you were there, then you saw them).
5209: @end quotation
5210: Forth's @code{THEN} has the meaning 2b, whereas @code{THEN} in Pascal
5211: and many other programming languages has the meaning 3d.]
5212:
1.21 crook 5213: Gforth also provides the words @code{?DUP-IF} and @code{?DUP-0=-IF}, so
1.1 anton 5214: you can avoid using @code{?dup}. Using these alternatives is also more
1.26 crook 5215: efficient than using @code{?dup}. Definitions in ANS Forth
1.1 anton 5216: for @code{ENDIF}, @code{?DUP-IF} and @code{?DUP-0=-IF} are provided in
5217: @file{compat/control.fs}.
5218:
5219: @cindex @code{CASE} control structure
5220: @example
1.29 crook 5221: @i{n}
1.1 anton 5222: CASE
1.29 crook 5223: @i{n1} OF @i{code1} ENDOF
5224: @i{n2} OF @i{code2} ENDOF
1.1 anton 5225: @dots{}
5226: ENDCASE
5227: @end example
5228:
1.29 crook 5229: Executes the first @i{codei}, where the @i{ni} is equal to
5230: @i{n}. A default case can be added by simply writing the code after
5231: the last @code{ENDOF}. It may use @i{n}, which is on top of the stack,
1.1 anton 5232: but must not consume it.
5233:
5234: @node Simple Loops, Counted Loops, Selection, Control Structures
5235: @subsection Simple Loops
5236: @cindex simple loops
5237: @cindex loops without count
5238:
5239: @cindex @code{WHILE} loop
5240: @example
5241: BEGIN
1.29 crook 5242: @i{code1}
5243: @i{flag}
1.1 anton 5244: WHILE
1.29 crook 5245: @i{code2}
1.1 anton 5246: REPEAT
5247: @end example
5248:
1.29 crook 5249: @i{code1} is executed and @i{flag} is computed. If it is true,
5250: @i{code2} is executed and the loop is restarted; If @i{flag} is
1.1 anton 5251: false, execution continues after the @code{REPEAT}.
5252:
5253: @cindex @code{UNTIL} loop
5254: @example
5255: BEGIN
1.29 crook 5256: @i{code}
5257: @i{flag}
1.1 anton 5258: UNTIL
5259: @end example
5260:
1.29 crook 5261: @i{code} is executed. The loop is restarted if @code{flag} is false.
1.1 anton 5262:
5263: @cindex endless loop
5264: @cindex loops, endless
5265: @example
5266: BEGIN
1.29 crook 5267: @i{code}
1.1 anton 5268: AGAIN
5269: @end example
5270:
5271: This is an endless loop.
5272:
5273: @node Counted Loops, Arbitrary control structures, Simple Loops, Control Structures
5274: @subsection Counted Loops
5275: @cindex counted loops
5276: @cindex loops, counted
5277: @cindex @code{DO} loops
5278:
5279: The basic counted loop is:
5280: @example
1.29 crook 5281: @i{limit} @i{start}
1.1 anton 5282: ?DO
1.29 crook 5283: @i{body}
1.1 anton 5284: LOOP
5285: @end example
5286:
1.29 crook 5287: This performs one iteration for every integer, starting from @i{start}
5288: and up to, but excluding @i{limit}. The counter, or @i{index}, can be
1.21 crook 5289: accessed with @code{i}. For example, the loop:
1.1 anton 5290: @example
5291: 10 0 ?DO
5292: i .
5293: LOOP
5294: @end example
1.21 crook 5295: @noindent
5296: prints @code{0 1 2 3 4 5 6 7 8 9}
5297:
1.1 anton 5298: The index of the innermost loop can be accessed with @code{i}, the index
5299: of the next loop with @code{j}, and the index of the third loop with
5300: @code{k}.
5301:
1.44 crook 5302:
1.1 anton 5303: doc-i
5304: doc-j
5305: doc-k
5306:
1.44 crook 5307:
1.1 anton 5308: The loop control data are kept on the return stack, so there are some
1.21 crook 5309: restrictions on mixing return stack accesses and counted loop words. In
5310: particuler, if you put values on the return stack outside the loop, you
5311: cannot read them inside the loop@footnote{well, not in a way that is
5312: portable.}. If you put values on the return stack within a loop, you
5313: have to remove them before the end of the loop and before accessing the
5314: index of the loop.
1.1 anton 5315:
5316: There are several variations on the counted loop:
5317:
1.21 crook 5318: @itemize @bullet
5319: @item
5320: @code{LEAVE} leaves the innermost counted loop immediately; execution
5321: continues after the associated @code{LOOP} or @code{NEXT}. For example:
5322:
5323: @example
5324: 10 0 ?DO i DUP . 3 = IF LEAVE THEN LOOP
5325: @end example
5326: prints @code{0 1 2 3}
5327:
1.1 anton 5328:
1.21 crook 5329: @item
5330: @code{UNLOOP} prepares for an abnormal loop exit, e.g., via
5331: @code{EXIT}. @code{UNLOOP} removes the loop control parameters from the
5332: return stack so @code{EXIT} can get to its return address. For example:
5333:
5334: @example
5335: : demo 10 0 ?DO i DUP . 3 = IF UNLOOP EXIT THEN LOOP ." Done" ;
5336: @end example
5337: prints @code{0 1 2 3}
5338:
5339:
5340: @item
1.29 crook 5341: If @i{start} is greater than @i{limit}, a @code{?DO} loop is entered
1.1 anton 5342: (and @code{LOOP} iterates until they become equal by wrap-around
5343: arithmetic). This behaviour is usually not what you want. Therefore,
5344: Gforth offers @code{+DO} and @code{U+DO} (as replacements for
1.29 crook 5345: @code{?DO}), which do not enter the loop if @i{start} is greater than
5346: @i{limit}; @code{+DO} is for signed loop parameters, @code{U+DO} for
1.1 anton 5347: unsigned loop parameters.
5348:
1.21 crook 5349: @item
5350: @code{?DO} can be replaced by @code{DO}. @code{DO} always enters
5351: the loop, independent of the loop parameters. Do not use @code{DO}, even
5352: if you know that the loop is entered in any case. Such knowledge tends
5353: to become invalid during maintenance of a program, and then the
5354: @code{DO} will make trouble.
5355:
5356: @item
1.29 crook 5357: @code{LOOP} can be replaced with @code{@i{n} +LOOP}; this updates the
5358: index by @i{n} instead of by 1. The loop is terminated when the border
5359: between @i{limit-1} and @i{limit} is crossed. E.g.:
1.1 anton 5360:
1.21 crook 5361: @example
5362: 4 0 +DO i . 2 +LOOP
5363: @end example
5364: @noindent
5365: prints @code{0 2}
5366:
5367: @example
5368: 4 1 +DO i . 2 +LOOP
5369: @end example
5370: @noindent
5371: prints @code{1 3}
1.1 anton 5372:
5373:
5374: @cindex negative increment for counted loops
5375: @cindex counted loops with negative increment
1.29 crook 5376: The behaviour of @code{@i{n} +LOOP} is peculiar when @i{n} is negative:
1.1 anton 5377:
1.21 crook 5378: @example
5379: -1 0 ?DO i . -1 +LOOP
5380: @end example
5381: @noindent
5382: prints @code{0 -1}
1.1 anton 5383:
1.21 crook 5384: @example
5385: 0 0 ?DO i . -1 +LOOP
5386: @end example
5387: prints nothing.
1.1 anton 5388:
1.29 crook 5389: Therefore we recommend avoiding @code{@i{n} +LOOP} with negative
5390: @i{n}. One alternative is @code{@i{u} -LOOP}, which reduces the
5391: index by @i{u} each iteration. The loop is terminated when the border
5392: between @i{limit+1} and @i{limit} is crossed. Gforth also provides
1.1 anton 5393: @code{-DO} and @code{U-DO} for down-counting loops. E.g.:
5394:
1.21 crook 5395: @example
5396: -2 0 -DO i . 1 -LOOP
5397: @end example
5398: @noindent
5399: prints @code{0 -1}
1.1 anton 5400:
1.21 crook 5401: @example
5402: -1 0 -DO i . 1 -LOOP
5403: @end example
5404: @noindent
5405: prints @code{0}
5406:
5407: @example
5408: 0 0 -DO i . 1 -LOOP
5409: @end example
5410: @noindent
5411: prints nothing.
1.1 anton 5412:
1.21 crook 5413: @end itemize
1.1 anton 5414:
5415: Unfortunately, @code{+DO}, @code{U+DO}, @code{-DO}, @code{U-DO} and
1.26 crook 5416: @code{-LOOP} are not defined in ANS Forth. However, an implementation
5417: for these words that uses only standard words is provided in
5418: @file{compat/loops.fs}.
1.1 anton 5419:
5420:
5421: @cindex @code{FOR} loops
1.26 crook 5422: Another counted loop is:
1.1 anton 5423: @example
1.29 crook 5424: @i{n}
1.1 anton 5425: FOR
1.29 crook 5426: @i{body}
1.1 anton 5427: NEXT
5428: @end example
5429: This is the preferred loop of native code compiler writers who are too
1.26 crook 5430: lazy to optimize @code{?DO} loops properly. This loop structure is not
1.29 crook 5431: defined in ANS Forth. In Gforth, this loop iterates @i{n+1} times;
5432: @code{i} produces values starting with @i{n} and ending with 0. Other
1.26 crook 5433: Forth systems may behave differently, even if they support @code{FOR}
5434: loops. To avoid problems, don't use @code{FOR} loops.
1.1 anton 5435:
5436: @node Arbitrary control structures, Calls and returns, Counted Loops, Control Structures
5437: @subsection Arbitrary control structures
5438: @cindex control structures, user-defined
5439:
5440: @cindex control-flow stack
5441: ANS Forth permits and supports using control structures in a non-nested
5442: way. Information about incomplete control structures is stored on the
5443: control-flow stack. This stack may be implemented on the Forth data
5444: stack, and this is what we have done in Gforth.
5445:
5446: @cindex @code{orig}, control-flow stack item
5447: @cindex @code{dest}, control-flow stack item
5448: An @i{orig} entry represents an unresolved forward branch, a @i{dest}
5449: entry represents a backward branch target. A few words are the basis for
5450: building any control structure possible (except control structures that
5451: need storage, like calls, coroutines, and backtracking).
5452:
1.44 crook 5453:
1.1 anton 5454: doc-if
5455: doc-ahead
5456: doc-then
5457: doc-begin
5458: doc-until
5459: doc-again
5460: doc-cs-pick
5461: doc-cs-roll
5462:
1.44 crook 5463:
1.21 crook 5464: The Standard words @code{CS-PICK} and @code{CS-ROLL} allow you to
5465: manipulate the control-flow stack in a portable way. Without them, you
5466: would need to know how many stack items are occupied by a control-flow
5467: entry (many systems use one cell. In Gforth they currently take three,
5468: but this may change in the future).
5469:
1.1 anton 5470: Some standard control structure words are built from these words:
5471:
1.44 crook 5472:
1.1 anton 5473: doc-else
5474: doc-while
5475: doc-repeat
5476:
1.44 crook 5477:
5478: @noindent
1.1 anton 5479: Gforth adds some more control-structure words:
5480:
1.44 crook 5481:
1.1 anton 5482: doc-endif
5483: doc-?dup-if
5484: doc-?dup-0=-if
5485:
1.44 crook 5486:
5487: @noindent
1.1 anton 5488: Counted loop words constitute a separate group of words:
5489:
1.44 crook 5490:
1.1 anton 5491: doc-?do
5492: doc-+do
5493: doc-u+do
5494: doc--do
5495: doc-u-do
5496: doc-do
5497: doc-for
5498: doc-loop
5499: doc-+loop
5500: doc--loop
5501: doc-next
5502: doc-leave
5503: doc-?leave
5504: doc-unloop
5505: doc-done
5506:
1.44 crook 5507:
1.21 crook 5508: The standard does not allow using @code{CS-PICK} and @code{CS-ROLL} on
5509: @i{do-sys}. Gforth allows it, but it's your job to ensure that for
1.1 anton 5510: every @code{?DO} etc. there is exactly one @code{UNLOOP} on any path
5511: through the definition (@code{LOOP} etc. compile an @code{UNLOOP} on the
5512: fall-through path). Also, you have to ensure that all @code{LEAVE}s are
5513: resolved (by using one of the loop-ending words or @code{DONE}).
5514:
1.44 crook 5515: @noindent
1.26 crook 5516: Another group of control structure words are:
1.1 anton 5517:
1.44 crook 5518:
1.1 anton 5519: doc-case
5520: doc-endcase
5521: doc-of
5522: doc-endof
5523:
1.44 crook 5524:
1.21 crook 5525: @i{case-sys} and @i{of-sys} cannot be processed using @code{CS-PICK} and
5526: @code{CS-ROLL}.
1.1 anton 5527:
5528: @subsubsection Programming Style
1.47 crook 5529: @cindex control structures programming style
5530: @cindex programming style, arbitrary control structures
1.1 anton 5531:
5532: In order to ensure readability we recommend that you do not create
5533: arbitrary control structures directly, but define new control structure
5534: words for the control structure you want and use these words in your
1.26 crook 5535: program. For example, instead of writing:
1.1 anton 5536:
5537: @example
1.26 crook 5538: BEGIN
1.1 anton 5539: ...
1.26 crook 5540: IF [ 1 CS-ROLL ]
1.1 anton 5541: ...
1.26 crook 5542: AGAIN THEN
1.1 anton 5543: @end example
5544:
1.21 crook 5545: @noindent
1.1 anton 5546: we recommend defining control structure words, e.g.,
5547:
5548: @example
1.26 crook 5549: : WHILE ( DEST -- ORIG DEST )
5550: POSTPONE IF
5551: 1 CS-ROLL ; immediate
5552:
5553: : REPEAT ( orig dest -- )
5554: POSTPONE AGAIN
5555: POSTPONE THEN ; immediate
1.1 anton 5556: @end example
5557:
1.21 crook 5558: @noindent
1.1 anton 5559: and then using these to create the control structure:
5560:
5561: @example
1.26 crook 5562: BEGIN
1.1 anton 5563: ...
1.26 crook 5564: WHILE
1.1 anton 5565: ...
1.26 crook 5566: REPEAT
1.1 anton 5567: @end example
5568:
5569: That's much easier to read, isn't it? Of course, @code{REPEAT} and
5570: @code{WHILE} are predefined, so in this example it would not be
5571: necessary to define them.
5572:
5573: @node Calls and returns, Exception Handling, Arbitrary control structures, Control Structures
5574: @subsection Calls and returns
5575: @cindex calling a definition
5576: @cindex returning from a definition
5577:
1.3 anton 5578: @cindex recursive definitions
5579: A definition can be called simply be writing the name of the definition
1.26 crook 5580: to be called. Normally a definition is invisible during its own
1.3 anton 5581: definition. If you want to write a directly recursive definition, you
1.26 crook 5582: can use @code{recursive} to make the current definition visible, or
5583: @code{recurse} to call the current definition directly.
1.3 anton 5584:
1.44 crook 5585:
1.3 anton 5586: doc-recursive
5587: doc-recurse
5588:
1.44 crook 5589:
1.21 crook 5590: @comment TODO add example of the two recursion methods
1.12 anton 5591: @quotation
5592: @progstyle
5593: I prefer using @code{recursive} to @code{recurse}, because calling the
5594: definition by name is more descriptive (if the name is well-chosen) than
5595: the somewhat cryptic @code{recurse}. E.g., in a quicksort
5596: implementation, it is much better to read (and think) ``now sort the
5597: partitions'' than to read ``now do a recursive call''.
5598: @end quotation
1.3 anton 5599:
1.29 crook 5600: For mutual recursion, use @code{Defer}red words, like this:
1.3 anton 5601:
5602: @example
1.28 crook 5603: Defer foo
1.3 anton 5604:
5605: : bar ( ... -- ... )
5606: ... foo ... ;
5607:
5608: :noname ( ... -- ... )
5609: ... bar ... ;
5610: IS foo
5611: @end example
5612:
1.44 crook 5613: Deferred words are discussed in more detail in @ref{Deferred words}.
1.33 anton 5614:
1.26 crook 5615: The current definition returns control to the calling definition when
1.33 anton 5616: the end of the definition is reached or @code{EXIT} is encountered.
1.1 anton 5617:
5618: doc-exit
5619: doc-;s
5620:
1.44 crook 5621:
1.1 anton 5622: @node Exception Handling, , Calls and returns, Control Structures
5623: @subsection Exception Handling
1.26 crook 5624: @cindex exceptions
1.1 anton 5625:
1.26 crook 5626: If your program detects a fatal error condition, the simplest action
5627: that it can take is to @code{quit}. This resets the return stack and
5628: restarts the text interpreter, but does not print any error message.
1.21 crook 5629:
1.26 crook 5630: The next stage in severity is to execute @code{abort}, which has the
5631: same effect as @code{quit}, with the addition that it resets the data
5632: stack.
1.1 anton 5633:
1.26 crook 5634: A slightly more sophisticated approach is use use @code{abort"}, which
5635: compiles a string to be used as an error message and does a conditional
5636: @code{abort} at run-time. For example:
1.1 anton 5637:
1.26 crook 5638: @example
1.30 anton 5639: @kbd{: checker abort" That flag was true" ." A false flag" ;@key{RET}} ok
5640: @kbd{0 checker@key{RET}} A false flag ok
5641: @kbd{1 checker@key{RET}}
1.26 crook 5642: :1: That flag was true
5643: 1 checker
5644: ^^^^^^^
5645: $400D1648 throw
5646: $400E4660
5647: @end example
1.1 anton 5648:
1.26 crook 5649: These simple techniques allow a program to react to a fatal error
5650: condition, but they are not exactly user-friendly. The ANS Forth
5651: Exception word set provides the pair of words @code{throw} and
5652: @code{catch}, which can be used to provide sophisticated error-handling.
1.1 anton 5653:
1.26 crook 5654: @code{catch} has a similar behaviour to @code{execute}, in that it takes
1.29 crook 5655: an @i{xt} as a parameter and starts execution of the xt. However,
1.26 crook 5656: before passing control to the xt, @code{catch} pushes an
1.29 crook 5657: @dfn{exception frame} onto the @dfn{exception stack}. This exception
1.26 crook 5658: frame is used to restore the system to a known state if a detected error
5659: occurs during the execution of the xt. A typical way to use @code{catch}
5660: would be:
1.1 anton 5661:
1.26 crook 5662: @example
5663: ... ['] foo catch IF ...
5664: @end example
1.1 anton 5665:
1.33 anton 5666: @c TOS is undefined. - anton
1.44 crook 5667:
5668: @c nac-> TODO -- I need to look at this example again.
5669:
1.26 crook 5670: Whilst @code{foo} executes, it can call other words to any level of
5671: nesting, as usual. If @code{foo} (and all the words that it calls)
1.33 anton 5672: execute successfully, control will ultimately pass to the word following
5673: the @code{catch}, and there will be a 0 at TOS. However, if any word
5674: detects an error, it can terminate the execution of @code{foo} by
5675: pushing a non-zero error code onto the stack and then performing a
5676: @code{throw}. The execution of @code{throw} will pass control to the
5677: word following the @code{catch}, but this time the TOS will hold the
5678: error code. Therefore, the @code{IF} in the example can be used to
5679: determine whether @code{foo} executed successfully.
1.1 anton 5680:
1.26 crook 5681: This simple example shows how you can use @code{throw} and @code{catch}
5682: to ``take over'' exception handling from the system:
1.1 anton 5683: @example
1.26 crook 5684: : my-div ['] / catch if ." DIVIDE ERROR" else ." OK.. " . then ;
1.1 anton 5685: @end example
5686:
1.26 crook 5687: The next example is more sophisticated and shows a multi-level
5688: @code{throw} and @code{catch}. To understand this example, start at the
5689: definition of @code{top-level} and work backwards:
5690:
1.1 anton 5691: @example
1.26 crook 5692: : lowest-level ( -- c )
5693: key dup 27 = if
1.44 crook 5694: 1 throw \ ESCAPE key pressed
1.26 crook 5695: else
1.44 crook 5696: ." lowest-level successful" CR
1.26 crook 5697: then
5698: ;
5699:
5700: : lower-level ( -- c )
5701: lowest-level
5702: \ at this level consider a CTRL-U to be a fatal error
5703: dup 21 = if \ CTRL-U
1.44 crook 5704: 2 throw
1.26 crook 5705: else
1.44 crook 5706: ." lower-level successful" CR
1.26 crook 5707: then
5708: ;
5709:
5710: : low-level ( -- c )
5711: ['] lower-level catch
5712: ?dup if
1.44 crook 5713: \ error occurred - do we recognise it?
5714: dup 1 = if
5715: \ ESCAPE key pressed.. pretend it was an E
5716: [char] E
5717: else throw \ propogate the error upwards
5718: then
1.26 crook 5719: then
5720: ." low-level successfull" CR
5721: ;
5722:
5723: : top-level ( -- )
5724: CR ['] low-level catch \ CATCH is used like EXECUTE
5725: ?dup if \ error occurred..
1.44 crook 5726: ." Error " . ." occurred - contact your supplier"
1.26 crook 5727: else
1.44 crook 5728: ." The '" emit ." ' key was pressed" CR
1.26 crook 5729: then
5730: ;
1.1 anton 5731: @end example
5732:
1.26 crook 5733: The ANS Forth document assigns @code{throw} codes thus:
1.1 anton 5734:
1.26 crook 5735: @itemize @bullet
5736: @item
5737: codes in the range -1 -- -255 are reserved to be assigned by the
5738: Standard. Assignments for codes in the range -1 -- -58 are currently
5739: documented in the Standard. In particular, @code{-1 throw} is equivalent
5740: to @code{abort} and @code{-2 throw} is equivalent to @code{abort"}.
5741: @item
5742: codes in the range -256 -- -4095 are reserved to be assigned by the system.
5743: @item
5744: all other codes may be assigned by programs.
5745: @end itemize
1.1 anton 5746:
1.26 crook 5747: Gforth provides the word @code{exception} as a mechanism for assigning
5748: system throw codes to applications. This allows multiple applications to
5749: co-exist in memory without any clash of @code{throw} codes. A definition
5750: of @code{exception} in ANS Forth is provided in
5751: @file{compat/exception.fs}.
1.1 anton 5752:
1.44 crook 5753:
1.26 crook 5754: doc-quit
5755: doc-abort
5756: doc-abort"
1.1 anton 5757:
1.26 crook 5758: doc-catch
1.29 crook 5759: doc-throw
5760: doc---exception-exception
5761:
5762:
1.44 crook 5763:
1.29 crook 5764: @c -------------------------------------------------------------
1.47 crook 5765: @node Defining Words, Interpretation and Compilation Semantics, Control Structures, Words
1.29 crook 5766: @section Defining Words
5767: @cindex defining words
5768:
1.47 crook 5769: Defining words are used to extend Forth by creating new entries in the dictionary.
5770:
1.29 crook 5771: @menu
1.44 crook 5772: * CREATE::
5773: * Variables:: Variables and user variables
5774: * Constants::
5775: * Values:: Initialised variables
1.29 crook 5776: * Colon Definitions::
1.44 crook 5777: * Anonymous Definitions:: Definitions without names
1.29 crook 5778: * User-defined Defining Words::
1.44 crook 5779: * Deferred words:: Allow forward references
5780: * Aliases::
1.29 crook 5781: * Supplying names::
5782: @end menu
5783:
1.44 crook 5784: @node CREATE, Variables, Defining Words, Defining Words
5785: @subsection @code{CREATE}
1.29 crook 5786: @cindex simple defining words
5787: @cindex defining words, simple
5788:
5789: Defining words are used to create new entries in the dictionary. The
5790: simplest defining word is @code{CREATE}. @code{CREATE} is used like
5791: this:
5792:
5793: @example
5794: CREATE new-word1
5795: @end example
5796:
5797: @code{CREATE} is a parsing word that generates a dictionary entry for
5798: @code{new-word1}. When @code{new-word1} is executed, all that it does is
5799: leave an address on the stack. The address represents the value of
5800: the data space pointer (@code{HERE}) at the time that @code{new-word1}
5801: was defined. Therefore, @code{CREATE} is a way of associating a name
5802: with the address of a region of memory.
5803:
1.34 anton 5804: doc-create
5805:
1.29 crook 5806: By extending this example to reserve some memory in data space, we end
5807: up with a @i{variable}. Here are two different ways to do it:
5808:
5809: @example
5810: CREATE new-word2 1 cells allot \ reserve 1 cell - initial value undefined
5811: CREATE new-word3 4 , \ reserve 1 cell and initialise it (to 4)
5812: @end example
5813:
5814: The variable can be examined and modified using @code{@@} (``fetch'') and
5815: @code{!} (``store'') like this:
5816:
5817: @example
5818: new-word2 @@ . \ get address, fetch from it and display
5819: 1234 new-word2 ! \ new value, get address, store to it
5820: @end example
5821:
1.44 crook 5822: @cindex arrays
5823: A similar mechanism can be used to create arrays. For example, an
5824: 80-character text input buffer:
1.29 crook 5825:
5826: @example
1.44 crook 5827: CREATE text-buf 80 chars allot
5828:
5829: text-buf 0 chars c@@ \ the 1st character (offset 0)
5830: text-buf 3 chars c@@ \ the 4th character (offset 3)
5831: @end example
1.29 crook 5832:
1.44 crook 5833: You can build arbitrarily complex data structures by allocating
1.49 anton 5834: appropriate areas of memory. For further discussions of this, and to
5835: learn about some Gforth tools that make it easier, see
5836: @xref{Structures}.
1.44 crook 5837:
5838:
5839: @node Variables, Constants, CREATE, Defining Words
5840: @subsection Variables
5841: @cindex variables
5842:
5843: The previous section showed how a sequence of commands could be used to
5844: generate a variable. As a final refinement, the whole code sequence can
5845: be wrapped up in a defining word (pre-empting the subject of the next
5846: section), making it easier to create new variables:
5847:
5848: @example
5849: : myvariableX ( "name" -- a-addr ) CREATE 1 cells allot ;
5850: : myvariable0 ( "name" -- a-addr ) CREATE 0 , ;
5851:
5852: myvariableX foo \ variable foo starts off with an unknown value
5853: myvariable0 joe \ whilst joe is initialised to 0
1.29 crook 5854:
5855: 45 3 * foo ! \ set foo to 135
5856: 1234 joe ! \ set joe to 1234
5857: 3 joe +! \ increment joe by 3.. to 1237
5858: @end example
5859:
5860: Not surprisingly, there is no need to define @code{myvariable}, since
1.44 crook 5861: Forth already has a definition @code{Variable}. ANS Forth does not
5862: require a @code{Variable} to be initialised when it is created (i.e., it
5863: behaves like @code{myvariableX}). In contrast, Gforth's @code{Variable}
5864: initialises the variable to 0 (i.e., it behaves exactly like
5865: @code{myvariable0}). Forth also provides @code{2Variable} and
1.47 crook 5866: @code{fvariable} for double and floating-point variables, respectively
5867: -- both are initialised to 0 in Gforth. If you use a @code{Variable} to
5868: store a boolean, you can use @code{on} and @code{off} to toggle its
5869: state.
1.29 crook 5870:
1.34 anton 5871: doc-variable
5872: doc-2variable
5873: doc-fvariable
5874:
1.29 crook 5875: @cindex user variables
5876: @cindex user space
5877: The defining word @code{User} behaves in the same way as @code{Variable}.
5878: The difference is that it reserves space in @i{user (data) space} rather
5879: than normal data space. In a Forth system that has a multi-tasker, each
5880: task has its own set of user variables.
5881:
1.34 anton 5882: doc-user
5883:
1.29 crook 5884: @comment TODO is that stuff about user variables strictly correct? Is it
5885: @comment just terminal tasks that have user variables?
5886: @comment should document tasker.fs (with some examples) elsewhere
5887: @comment in this manual, then expand on user space and user variables.
5888:
1.44 crook 5889:
5890: @node Constants, Values, Variables, Defining Words
5891: @subsection Constants
5892: @cindex constants
5893:
5894: @code{Constant} allows you to declare a fixed value and refer to it by
5895: name. For example:
1.29 crook 5896:
5897: @example
5898: 12 Constant INCHES-PER-FOOT
5899: 3E+08 fconstant SPEED-O-LIGHT
5900: @end example
5901:
5902: A @code{Variable} can be both read and written, so its run-time
5903: behaviour is to supply an address through which its current value can be
5904: manipulated. In contrast, the value of a @code{Constant} cannot be
5905: changed once it has been declared@footnote{Well, often it can be -- but
5906: not in a Standard, portable way. It's safer to use a @code{Value} (read
5907: on).} so it's not necessary to supply the address -- it is more
5908: efficient to return the value of the constant directly. That's exactly
5909: what happens; the run-time effect of a constant is to put its value on
1.49 anton 5910: the top of the stack (You can find one
5911: way of implementing @code{Constant} in @ref{User-defined Defining Words}).
1.29 crook 5912:
5913: Gforth also provides @code{2Constant} and @code{fconstant} for defining
5914: double and floating-point constants, respectively.
5915:
1.34 anton 5916: doc-constant
5917: doc-2constant
5918: doc-fconstant
5919:
5920: @c that's too deep, and it's not necessarily true for all ANS Forths. - anton
1.44 crook 5921: @c nac-> How could that not be true in an ANS Forth? You can't define a
5922: @c constant, use it and then delete the definition of the constant..
5923: @c I agree that it's rather deep, but IMO it is an important difference
5924: @c relative to other programming languages.. often it's annoying: it
5925: @c certainly changes my programming style relative to C.
5926:
1.29 crook 5927: Constants in Forth behave differently from their equivalents in other
5928: programming languages. In other languages, a constant (such as an EQU in
5929: assembler or a #define in C) only exists at compile-time; in the
5930: executable program the constant has been translated into an absolute
5931: number and, unless you are using a symbolic debugger, it's impossible to
5932: know what abstract thing that number represents. In Forth a constant has
1.44 crook 5933: an entry in the header space and remains there after the code that uses
5934: it has been defined. In fact, it must remain in the dictionary since it
5935: has run-time duties to perform. For example:
1.29 crook 5936:
5937: @example
5938: 12 Constant INCHES-PER-FOOT
5939: : FEET-TO-INCHES ( n1 -- n2 ) INCHES-PER-FOOT * ;
5940: @end example
5941:
5942: @cindex in-lining of constants
5943: When @code{FEET-TO-INCHES} is executed, it will in turn execute the xt
5944: associated with the constant @code{INCHES-PER-FOOT}. If you use
5945: @code{see} to decompile the definition of @code{FEET-TO-INCHES}, you can
5946: see that it makes a call to @code{INCHES-PER-FOOT}. Some Forth compilers
5947: attempt to optimise constants by in-lining them where they are used. You
5948: can force Gforth to in-line a constant like this:
5949:
5950: @example
5951: : FEET-TO-INCHES ( n1 -- n2 ) [ INCHES-PER-FOOT ] LITERAL * ;
5952: @end example
5953:
5954: If you use @code{see} to decompile @i{this} version of
5955: @code{FEET-TO-INCHES}, you can see that @code{INCHES-PER-FOOT} is no
1.49 anton 5956: longer present. To understand how this works, read
5957: @ref{Interpret/Compile states}, and @ref{Literals}.
1.29 crook 5958:
5959: In-lining constants in this way might improve execution time
5960: fractionally, and can ensure that a constant is now only referenced at
5961: compile-time. However, the definition of the constant still remains in
5962: the dictionary. Some Forth compilers provide a mechanism for controlling
5963: a second dictionary for holding transient words such that this second
5964: dictionary can be deleted later in order to recover memory
5965: space. However, there is no standard way of doing this.
5966:
5967:
1.44 crook 5968: @node Values, Colon Definitions, Constants, Defining Words
5969: @subsection Values
5970: @cindex values
1.34 anton 5971:
1.44 crook 5972: A @code{Value} is like a @code{Variable} but with two important
5973: differences:
1.29 crook 5974:
5975: @itemize @bullet
5976: @item
1.44 crook 5977: A @code{Value} is initialised when it is declared; like a
5978: @code{Constant} but unlike a @code{Variable}.
1.29 crook 5979: @item
1.44 crook 5980: A @code{Value} returns its value rather than its address when it is
5981: executed; i.e., it has the same run-time behaviour as @code{Constant}.
1.29 crook 5982: @end itemize
5983:
1.44 crook 5984: A @code{Value} needs an additional word, @code{TO} to allow its value to
5985: be changed. Here are some examples:
1.29 crook 5986:
5987: @example
1.44 crook 5988: 12 Value APPLES \ Define APPLES with an initial value of 12
5989: 34 TO APPLES \ Change the value of APPLES. TO is a parsing word
5990: APPLES \ puts 34 on the top of the stack.
1.29 crook 5991: @end example
5992:
1.44 crook 5993: doc-value
5994: doc-to
1.29 crook 5995:
1.35 anton 5996:
1.44 crook 5997: @node Colon Definitions, Anonymous Definitions, Values, Defining Words
5998: @subsection Colon Definitions
5999: @cindex colon definitions
1.35 anton 6000:
6001: @example
1.44 crook 6002: : name ( ... -- ... )
6003: word1 word2 word3 ;
1.29 crook 6004: @end example
6005:
1.44 crook 6006: @noindent
6007: Creates a word called @code{name} that, upon execution, executes
6008: @code{word1 word2 word3}. @code{name} is a @dfn{(colon) definition}.
1.29 crook 6009:
1.49 anton 6010: The explanation above is somewhat superficial. For simple examples of
6011: colon definitions see @ref{Your first definition}. For an in-depth
6012: discussion of some of the issues involved, see @xref{Interpretation and
6013: Compilation Semantics}.
1.29 crook 6014:
1.44 crook 6015: doc-:
6016: doc-;
1.1 anton 6017:
1.34 anton 6018:
1.44 crook 6019: @node Anonymous Definitions, User-defined Defining Words, Colon Definitions, Defining Words
6020: @subsection Anonymous Definitions
6021: @cindex colon definitions
6022: @cindex defining words without name
1.34 anton 6023:
1.44 crook 6024: Sometimes you want to define an @dfn{anonymous word}; a word without a
6025: name. You can do this with:
1.1 anton 6026:
1.44 crook 6027: doc-:noname
1.1 anton 6028:
1.44 crook 6029: This leaves the execution token for the word on the stack after the
6030: closing @code{;}. Here's an example in which a deferred word is
6031: initialised with an @code{xt} from an anonymous colon definition:
1.1 anton 6032:
1.29 crook 6033: @example
1.44 crook 6034: Defer deferred
6035: :noname ( ... -- ... )
6036: ... ;
6037: IS deferred
1.29 crook 6038: @end example
1.26 crook 6039:
1.44 crook 6040: @noindent
6041: Gforth provides an alternative way of doing this, using two separate
6042: words:
1.27 crook 6043:
1.44 crook 6044: doc-noname
6045: @cindex execution token of last defined word
6046: doc-lastxt
1.1 anton 6047:
1.44 crook 6048: @noindent
6049: The previous example can be rewritten using @code{noname} and
6050: @code{lastxt}:
1.1 anton 6051:
1.26 crook 6052: @example
1.44 crook 6053: Defer deferred
6054: noname : ( ... -- ... )
6055: ... ;
6056: lastxt IS deferred
1.26 crook 6057: @end example
1.1 anton 6058:
1.29 crook 6059: @noindent
1.44 crook 6060: @code{noname} works with any defining word, not just @code{:}.
6061:
6062: @code{lastxt} also works when the last word was not defined as
6063: @code{noname}. It also has the useful property that is is valid as soon
6064: as the header for a definition has been built. Thus:
6065:
6066: @example
6067: lastxt . : foo [ lastxt . ] ; ' foo .
6068: @end example
1.1 anton 6069:
1.44 crook 6070: @noindent
6071: prints 3 numbers; the last two are the same.
1.26 crook 6072:
1.1 anton 6073:
1.44 crook 6074: @node User-defined Defining Words, Deferred words, Anonymous Definitions, Defining Words
1.26 crook 6075: @subsection User-defined Defining Words
6076: @cindex user-defined defining words
6077: @cindex defining words, user-defined
1.1 anton 6078:
1.29 crook 6079: You can create a new defining word by wrapping defining-time code around
6080: an existing defining word and putting the sequence in a colon
6081: definition. For example, suppose that you have a word @code{stats} that
6082: gathers statistics about colon definitions given the @i{xt} of the
6083: definition, and you want every colon definition in your application to
6084: make a call to @code{stats}. You can define and use a new version of
6085: @code{:} like this:
6086:
6087: @example
6088: : stats ( xt -- ) DUP ." (Gathering statistics for " . ." )"
6089: ... ; \ other code
6090:
6091: : my: : lastxt postpone literal ['] stats compile, ;
6092:
6093: my: foo + - ;
6094: @end example
6095:
6096: When @code{foo} is defined using @code{my:} these steps occur:
6097:
6098: @itemize @bullet
6099: @item
6100: @code{my:} is executed.
6101: @item
6102: The @code{:} within the definition (the one between @code{my:} and
6103: @code{lastxt}) is executed, and does just what it always does; it parses
6104: the input stream for a name, builds a dictionary header for the name
6105: @code{foo} and switches @code{state} from interpret to compile.
6106: @item
6107: The word @code{lastxt} is executed. It puts the @i{xt} for the word that is
6108: being defined -- @code{foo} -- onto the stack.
6109: @item
6110: The code that was produced by @code{postpone literal} is executed; this
6111: causes the value on the stack to be compiled as a literal in the code
6112: area of @code{foo}.
6113: @item
6114: The code @code{['] stats} compiles a literal into the definition of
6115: @code{my:}. When @code{compile,} is executed, that literal -- the
6116: execution token for @code{stats} -- is layed down in the code area of
6117: @code{foo} , following the literal@footnote{Strictly speaking, the
6118: mechanism that @code{compile,} uses to convert an @i{xt} into something
6119: in the code area is implementation-dependent. A threaded implementation
6120: might spit out the execution token directly whilst another
6121: implementation might spit out a native code sequence.}.
6122: @item
6123: At this point, the execution of @code{my:} is complete, and control
6124: returns to the text interpreter. The text interpreter is in compile
6125: state, so subsequent text @code{+ -} is compiled into the definition of
6126: @code{foo} and the @code{;} terminates the definition as always.
6127: @end itemize
6128:
6129: You can use @code{see} to decompile a word that was defined using
6130: @code{my:} and see how it is different from a normal @code{:}
6131: definition. For example:
6132:
6133: @example
6134: : bar + - ; \ like foo but using : rather than my:
6135: see bar
6136: : bar
6137: + - ;
6138: see foo
6139: : foo
6140: 107645672 stats + - ;
6141:
6142: \ use ' stats . to show that 107645672 is the xt for stats
6143: @end example
6144:
6145: You can use techniques like this to make new defining words in terms of
6146: @i{any} existing defining word.
1.1 anton 6147:
6148:
1.29 crook 6149: @cindex defining defining words
1.26 crook 6150: @cindex @code{CREATE} ... @code{DOES>}
6151: If you want the words defined with your defining words to behave
6152: differently from words defined with standard defining words, you can
6153: write your defining word like this:
1.1 anton 6154:
6155: @example
1.26 crook 6156: : def-word ( "name" -- )
1.29 crook 6157: CREATE @i{code1}
1.26 crook 6158: DOES> ( ... -- ... )
1.29 crook 6159: @i{code2} ;
1.26 crook 6160:
6161: def-word name
1.1 anton 6162: @end example
6163:
1.29 crook 6164: @cindex child words
6165: This fragment defines a @dfn{defining word} @code{def-word} and then
6166: executes it. When @code{def-word} executes, it @code{CREATE}s a new
6167: word, @code{name}, and executes the code @i{code1}. The code @i{code2}
6168: is not executed at this time. The word @code{name} is sometimes called a
6169: @dfn{child} of @code{def-word}.
6170:
6171: When you execute @code{name}, the address of the body of @code{name} is
6172: put on the data stack and @i{code2} is executed (the address of the body
6173: of @code{name} is the address @code{HERE} returns immediately after the
6174: @code{CREATE}).
6175:
6176: @cindex atavism in child words
1.33 anton 6177: You can use @code{def-word} to define a set of child words that behave
1.29 crook 6178: differently, though atavistically; they all have a common run-time
6179: behaviour determined by @i{code2}. Typically, the @i{code1} sequence
6180: builds a data area in the body of the child word. The structure of the
6181: data is common to all children of @code{def-word}, but the data values
6182: are specific -- and private -- to each child word. When a child word is
6183: executed, the address of its private data area is passed as a parameter
6184: on TOS to be used and manipulated@footnote{It is legitimate both to read
6185: and write to this data area.} by @i{code2}.
6186:
6187: The two fragments of code that make up the defining words act (are
6188: executed) at two completely separate times:
1.1 anton 6189:
1.29 crook 6190: @itemize @bullet
6191: @item
6192: At @i{define time}, the defining word executes @i{code1} to generate a
6193: child word
6194: @item
6195: At @i{child execution time}, when a child word is invoked, @i{code2}
6196: is executed, using parameters (data) that are private and specific to
6197: the child word.
6198: @end itemize
6199:
1.44 crook 6200: Another way of understanding the behaviour of @code{def-word} and
6201: @code{name} is to say that, if you make the following definitions:
1.33 anton 6202: @example
6203: : def-word1 ( "name" -- )
6204: CREATE @i{code1} ;
6205:
6206: : action1 ( ... -- ... )
6207: @i{code2} ;
6208:
6209: def-word1 name1
6210: @end example
6211:
1.44 crook 6212: @noindent
6213: Then using @code{name1 action1} is equivalent to using @code{name}.
1.1 anton 6214:
1.29 crook 6215: The classic example is that you can define @code{CONSTANT} in this way:
1.26 crook 6216:
1.1 anton 6217: @example
1.29 crook 6218: : CONSTANT ( w "name" -- )
6219: CREATE ,
1.26 crook 6220: DOES> ( -- w )
6221: @@ ;
1.1 anton 6222: @end example
6223:
1.29 crook 6224: @comment There is a beautiful description of how this works and what
6225: @comment it does in the Forthwrite 100th edition.. as well as an elegant
6226: @comment commentary on the Counting Fruits problem.
6227:
6228: When you create a constant with @code{5 CONSTANT five}, a set of
6229: define-time actions take place; first a new word @code{five} is created,
6230: then the value 5 is laid down in the body of @code{five} with
1.44 crook 6231: @code{,}. When @code{five} is executed, the address of the body is put on
1.29 crook 6232: the stack, and @code{@@} retrieves the value 5. The word @code{five} has
6233: no code of its own; it simply contains a data field and a pointer to the
6234: code that follows @code{DOES>} in its defining word. That makes words
6235: created in this way very compact.
6236:
6237: The final example in this section is intended to remind you that space
6238: reserved in @code{CREATE}d words is @i{data} space and therefore can be
6239: both read and written by a Standard program@footnote{Exercise: use this
6240: example as a starting point for your own implementation of @code{Value}
6241: and @code{TO} -- if you get stuck, investigate the behaviour of @code{'} and
6242: @code{[']}.}:
6243:
6244: @example
6245: : foo ( "name" -- )
6246: CREATE -1 ,
6247: DOES> ( -- )
1.33 anton 6248: @@ . ;
1.29 crook 6249:
6250: foo first-word
6251: foo second-word
6252:
6253: 123 ' first-word >BODY !
6254: @end example
6255:
6256: If @code{first-word} had been a @code{CREATE}d word, we could simply
6257: have executed it to get the address of its data field. However, since it
6258: was defined to have @code{DOES>} actions, its execution semantics are to
6259: perform those @code{DOES>} actions. To get the address of its data field
6260: it's necessary to use @code{'} to get its xt, then @code{>BODY} to
6261: translate the xt into the address of the data field. When you execute
6262: @code{first-word}, it will display @code{123}. When you execute
6263: @code{second-word} it will display @code{-1}.
1.26 crook 6264:
6265: @cindex stack effect of @code{DOES>}-parts
6266: @cindex @code{DOES>}-parts, stack effect
1.29 crook 6267: In the examples above the stack comment after the @code{DOES>} specifies
1.26 crook 6268: the stack effect of the defined words, not the stack effect of the
6269: following code (the following code expects the address of the body on
6270: the top of stack, which is not reflected in the stack comment). This is
6271: the convention that I use and recommend (it clashes a bit with using
6272: locals declarations for stack effect specification, though).
1.1 anton 6273:
1.53 anton 6274: @menu
6275: * CREATE..DOES> applications::
6276: * CREATE..DOES> details::
1.62 ! crook 6277: * Advanced does> usage::
1.53 anton 6278: @end menu
6279:
6280: @node CREATE..DOES> applications, CREATE..DOES> details, User-defined Defining Words, User-defined Defining Words
1.26 crook 6281: @subsubsection Applications of @code{CREATE..DOES>}
6282: @cindex @code{CREATE} ... @code{DOES>}, applications
1.1 anton 6283:
1.26 crook 6284: You may wonder how to use this feature. Here are some usage patterns:
1.1 anton 6285:
1.26 crook 6286: @cindex factoring similar colon definitions
6287: When you see a sequence of code occurring several times, and you can
6288: identify a meaning, you will factor it out as a colon definition. When
6289: you see similar colon definitions, you can factor them using
6290: @code{CREATE..DOES>}. E.g., an assembler usually defines several words
6291: that look very similar:
1.1 anton 6292: @example
1.26 crook 6293: : ori, ( reg-target reg-source n -- )
6294: 0 asm-reg-reg-imm ;
6295: : andi, ( reg-target reg-source n -- )
6296: 1 asm-reg-reg-imm ;
1.1 anton 6297: @end example
6298:
1.26 crook 6299: @noindent
6300: This could be factored with:
6301: @example
6302: : reg-reg-imm ( op-code -- )
6303: CREATE ,
6304: DOES> ( reg-target reg-source n -- )
6305: @@ asm-reg-reg-imm ;
6306:
6307: 0 reg-reg-imm ori,
6308: 1 reg-reg-imm andi,
6309: @end example
1.1 anton 6310:
1.26 crook 6311: @cindex currying
6312: Another view of @code{CREATE..DOES>} is to consider it as a crude way to
6313: supply a part of the parameters for a word (known as @dfn{currying} in
6314: the functional language community). E.g., @code{+} needs two
6315: parameters. Creating versions of @code{+} with one parameter fixed can
6316: be done like this:
1.1 anton 6317: @example
1.26 crook 6318: : curry+ ( n1 -- )
6319: CREATE ,
6320: DOES> ( n2 -- n1+n2 )
6321: @@ + ;
6322:
6323: 3 curry+ 3+
6324: -2 curry+ 2-
1.1 anton 6325: @end example
6326:
1.62 ! crook 6327: @node CREATE..DOES> details, Advanced does> usage, CREATE..DOES> applications, User-defined Defining Words
1.26 crook 6328: @subsubsection The gory details of @code{CREATE..DOES>}
6329: @cindex @code{CREATE} ... @code{DOES>}, details
1.1 anton 6330:
1.26 crook 6331: doc-does>
1.1 anton 6332:
1.26 crook 6333: @cindex @code{DOES>} in a separate definition
6334: This means that you need not use @code{CREATE} and @code{DOES>} in the
6335: same definition; you can put the @code{DOES>}-part in a separate
1.29 crook 6336: definition. This allows us to, e.g., select among different @code{DOES>}-parts:
1.26 crook 6337: @example
6338: : does1
6339: DOES> ( ... -- ... )
1.44 crook 6340: ... ;
6341:
6342: : does2
6343: DOES> ( ... -- ... )
6344: ... ;
6345:
6346: : def-word ( ... -- ... )
6347: create ...
6348: IF
6349: does1
6350: ELSE
6351: does2
6352: ENDIF ;
6353: @end example
6354:
6355: In this example, the selection of whether to use @code{does1} or
6356: @code{does2} is made at compile-time; at the time that the child word is
6357: @code{CREATE}d.
6358:
6359: @cindex @code{DOES>} in interpretation state
6360: In a standard program you can apply a @code{DOES>}-part only if the last
6361: word was defined with @code{CREATE}. In Gforth, the @code{DOES>}-part
6362: will override the behaviour of the last word defined in any case. In a
6363: standard program, you can use @code{DOES>} only in a colon
6364: definition. In Gforth, you can also use it in interpretation state, in a
6365: kind of one-shot mode; for example:
6366: @example
6367: CREATE name ( ... -- ... )
6368: @i{initialization}
6369: DOES>
6370: @i{code} ;
6371: @end example
6372:
6373: @noindent
6374: is equivalent to the standard:
6375: @example
6376: :noname
6377: DOES>
6378: @i{code} ;
6379: CREATE name EXECUTE ( ... -- ... )
6380: @i{initialization}
6381: @end example
6382:
1.53 anton 6383: doc->body
6384:
1.62 ! crook 6385: @node Advanced does> usage, , CREATE..DOES> details, User-defined Defining Words
! 6386: @subsubsection Advanced does> usage
1.44 crook 6387:
6388:
6389:
6390: @node Deferred words, Aliases, User-defined Defining Words, Defining Words
6391: @subsection Deferred words
6392: @cindex deferred words
6393:
6394: The defining word @code{Defer} allows you to define a word by name
6395: without defining its behaviour; the definition of its behaviour is
6396: deferred. Here are two situation where this can be useful:
6397:
6398: @itemize @bullet
6399: @item
6400: Where you want to allow the behaviour of a word to be altered later, and
6401: for all precompiled references to the word to change when its behaviour
6402: is changed.
6403: @item
6404: For mutual recursion; @xref{Calls and returns}.
6405: @end itemize
6406:
6407: In the following example, @code{foo} always invokes the version of
6408: @code{greet} that prints ``@code{Good morning}'' whilst @code{bar}
6409: always invokes the version that prints ``@code{Hello}''. There is no way
6410: of getting @code{foo} to use the later version without re-ordering the
6411: source code and recompiling it.
6412:
6413: @example
6414: : greet ." Good morning" ;
6415: : foo ... greet ... ;
6416: : greet ." Hello" ;
6417: : bar ... greet ... ;
6418: @end example
6419:
6420: This problem can be solved by defining @code{greet} as a @code{Defer}red
6421: word. The behaviour of a @code{Defer}red word can be defined and
6422: redefined at any time by using @code{IS} to associate the xt of a
6423: previously-defined word with it. The previous example becomes:
6424:
6425: @example
6426: Defer greet
6427: : foo ... greet ... ;
6428: : bar ... greet ... ;
6429: : greet1 ." Good morning" ;
6430: : greet2 ." Hello" ;
6431: ' greet2 <IS> greet \ make greet behave like greet2
6432: @end example
6433:
6434: A deferred word can be used to improve the statistics-gathering example
6435: from @ref{User-defined Defining Words}; rather than edit the
6436: application's source code to change every @code{:} to a @code{my:}, do
6437: this:
6438:
6439: @example
6440: : real: : ; \ retain access to the original
6441: defer : \ redefine as a deferred word
6442: ' my: IS : \ use special version of :
6443: \
6444: \ load application here
6445: \
6446: ' real: IS : \ go back to the original
6447: @end example
6448:
6449:
6450: One thing to note is that @code{<IS>} consumes its name when it is
6451: executed. If you want to specify the name at compile time, use
6452: @code{[IS]}:
6453:
6454: @example
6455: : set-greet ( xt -- )
6456: [IS] greet ;
6457:
6458: ' greet1 set-greet
6459: @end example
6460:
6461: A deferred word can only inherit default semantics from the xt (because
1.49 anton 6462: that is all that an xt can represent -- for more discussion of this
6463: @pxref{Tokens for Words}). However, the semantics of the deferred word
1.44 crook 6464: itself can be modified at the time that it is defined. For example:
6465:
6466: @example
6467: : bar .... ; compile-only
6468: Defer fred immediate
6469: Defer jim
6470:
6471: ' bar <IS> jim \ jim has default semantics
6472: ' bar <IS> fred \ fred is immediate
6473: @end example
6474:
6475: doc-defer
6476: doc-<is>
6477: doc-[is]
6478: doc-is
6479: @comment TODO document these: what's defers [is]
6480: doc-what's
6481: doc-defers
6482:
6483: @c Use @code{words-deferred} to see a list of deferred words.
6484:
6485: Definitions in ANS Forth for @code{defer}, @code{<is>} and @code{[is]}
6486: are provided in @file{compat/defer.fs}.
6487:
6488:
6489: @node Aliases, Supplying names, Deferred words, Defining Words
6490: @subsection Aliases
6491: @cindex aliases
1.1 anton 6492:
1.44 crook 6493: The defining word @code{Alias} allows you to define a word by name that
6494: has the same behaviour as some other word. Here are two situation where
6495: this can be useful:
1.1 anton 6496:
1.44 crook 6497: @itemize @bullet
6498: @item
6499: When you want access to a word's definition from a different word list
6500: (for an example of this, see the definition of the @code{Root} word list
6501: in the Gforth source).
6502: @item
6503: When you want to create a synonym; a definition that can be known by
6504: either of two names (for example, @code{THEN} and @code{ENDIF} are
6505: aliases).
6506: @end itemize
1.1 anton 6507:
1.44 crook 6508: The word whose behaviour the alias is to inherit is represented by an
6509: xt. Therefore, the alias only inherits default semantics from its
6510: ancestor. The semantics of the alias itself can be modified at the time
6511: that it is defined. For example:
1.1 anton 6512:
6513: @example
1.44 crook 6514: : foo ... ; immediate
6515:
6516: ' foo Alias bar \ bar is not an immediate word
6517: ' foo Alias fooby immediate \ fooby is an immediate word
1.1 anton 6518: @end example
6519:
1.44 crook 6520: Words that are aliases have the same xt, different headers in the
6521: dictionary, and consequently different name tokens (@pxref{Tokens for
6522: Words}) and possibly different immediate flags. An alias can only have
6523: default or immediate compilation semantics; you can define aliases for
6524: combined words with @code{interpret/compile:} -- see @ref{Combined words}.
1.1 anton 6525:
1.44 crook 6526: doc-alias
1.26 crook 6527:
1.1 anton 6528:
1.52 anton 6529: @node Supplying names, , Aliases, Defining Words
1.29 crook 6530: @subsection Supplying the name of a defined word
1.26 crook 6531: @cindex names for defined words
1.44 crook 6532: @cindex defining words, name given in a string
1.1 anton 6533:
1.29 crook 6534: By default, a defining word takes the name for the defined word from the
1.26 crook 6535: input stream. Sometimes you want to supply the name from a string. You
6536: can do this with:
1.1 anton 6537:
1.26 crook 6538: doc-nextname
1.1 anton 6539:
1.26 crook 6540: For example:
1.1 anton 6541:
1.26 crook 6542: @example
6543: s" foo" nextname create
6544: @end example
1.44 crook 6545:
1.26 crook 6546: @noindent
6547: is equivalent to:
1.44 crook 6548:
1.26 crook 6549: @example
6550: create foo
6551: @end example
1.1 anton 6552:
1.29 crook 6553: @noindent
1.44 crook 6554: @code{nextname} works with any defining word, not just @code{:}.
1.1 anton 6555:
6556:
1.47 crook 6557: @node Interpretation and Compilation Semantics, Tokens for Words, Defining Words, Words
6558: @section Interpretation and Compilation Semantics
1.26 crook 6559: @cindex semantics, interpretation and compilation
1.1 anton 6560:
1.26 crook 6561: @cindex interpretation semantics
6562: The @dfn{interpretation semantics} of a word are what the text
6563: interpreter does when it encounters the word in interpret state. It also
6564: appears in some other contexts, e.g., the execution token returned by
1.29 crook 6565: @code{' @i{word}} identifies the interpretation semantics of
6566: @i{word} (in other words, @code{' @i{word} execute} is equivalent to
6567: interpret-state text interpretation of @code{@i{word}}).
1.1 anton 6568:
1.26 crook 6569: @cindex compilation semantics
6570: The @dfn{compilation semantics} of a word are what the text interpreter
6571: does when it encounters the word in compile state. It also appears in
1.29 crook 6572: other contexts, e.g, @code{POSTPONE @i{word}} compiles@footnote{In
1.26 crook 6573: standard terminology, ``appends to the current definition''.} the
1.29 crook 6574: compilation semantics of @i{word}.
1.1 anton 6575:
1.26 crook 6576: @cindex execution semantics
6577: The standard also talks about @dfn{execution semantics}. They are used
6578: only for defining the interpretation and compilation semantics of many
6579: words. By default, the interpretation semantics of a word are to
6580: @code{execute} its execution semantics, and the compilation semantics of
6581: a word are to @code{compile,} its execution semantics.@footnote{In
6582: standard terminology: The default interpretation semantics are its
6583: execution semantics; the default compilation semantics are to append its
6584: execution semantics to the execution semantics of the current
6585: definition.}
6586:
6587: @comment TODO expand, make it co-operate with new sections on text interpreter.
6588:
6589: @cindex immediate words
6590: @cindex compile-only words
6591: You can change the semantics of the most-recently defined word:
6592:
1.44 crook 6593:
1.26 crook 6594: doc-immediate
6595: doc-compile-only
6596: doc-restrict
6597:
1.44 crook 6598:
1.26 crook 6599: Note that ticking (@code{'}) a compile-only word gives an error
6600: (``Interpreting a compile-only word'').
1.1 anton 6601:
1.47 crook 6602: @menu
6603: * Combined words::
6604: @end menu
1.44 crook 6605:
1.48 anton 6606: @node Combined words, , Interpretation and Compilation Semantics, Interpretation and Compilation Semantics
1.44 crook 6607: @subsection Combined Words
6608: @cindex combined words
6609:
6610: Gforth allows you to define @dfn{combined words} -- words that have an
6611: arbitrary combination of interpretation and compilation semantics.
6612:
1.1 anton 6613:
1.26 crook 6614: doc-interpret/compile:
1.1 anton 6615:
1.44 crook 6616:
1.26 crook 6617: This feature was introduced for implementing @code{TO} and @code{S"}. I
6618: recommend that you do not define such words, as cute as they may be:
6619: they make it hard to get at both parts of the word in some contexts.
6620: E.g., assume you want to get an execution token for the compilation
6621: part. Instead, define two words, one that embodies the interpretation
6622: part, and one that embodies the compilation part. Once you have done
6623: that, you can define a combined word with @code{interpret/compile:} for
6624: the convenience of your users.
1.1 anton 6625:
1.26 crook 6626: You might try to use this feature to provide an optimizing
6627: implementation of the default compilation semantics of a word. For
6628: example, by defining:
1.1 anton 6629: @example
1.26 crook 6630: :noname
6631: foo bar ;
6632: :noname
6633: POSTPONE foo POSTPONE bar ;
1.29 crook 6634: interpret/compile: opti-foobar
1.1 anton 6635: @end example
1.26 crook 6636:
1.23 crook 6637: @noindent
1.26 crook 6638: as an optimizing version of:
6639:
1.1 anton 6640: @example
1.26 crook 6641: : foobar
6642: foo bar ;
1.1 anton 6643: @end example
6644:
1.26 crook 6645: Unfortunately, this does not work correctly with @code{[compile]},
6646: because @code{[compile]} assumes that the compilation semantics of all
6647: @code{interpret/compile:} words are non-default. I.e., @code{[compile]
1.29 crook 6648: opti-foobar} would compile compilation semantics, whereas
6649: @code{[compile] foobar} would compile interpretation semantics.
1.1 anton 6650:
1.26 crook 6651: @cindex state-smart words (are a bad idea)
1.29 crook 6652: Some people try to use @dfn{state-smart} words to emulate the feature provided
1.26 crook 6653: by @code{interpret/compile:} (words are state-smart if they check
6654: @code{STATE} during execution). E.g., they would try to code
6655: @code{foobar} like this:
1.1 anton 6656:
1.26 crook 6657: @example
6658: : foobar
6659: STATE @@
6660: IF ( compilation state )
6661: POSTPONE foo POSTPONE bar
6662: ELSE
6663: foo bar
6664: ENDIF ; immediate
6665: @end example
1.1 anton 6666:
1.26 crook 6667: Although this works if @code{foobar} is only processed by the text
6668: interpreter, it does not work in other contexts (like @code{'} or
6669: @code{POSTPONE}). E.g., @code{' foobar} will produce an execution token
6670: for a state-smart word, not for the interpretation semantics of the
6671: original @code{foobar}; when you execute this execution token (directly
6672: with @code{EXECUTE} or indirectly through @code{COMPILE,}) in compile
6673: state, the result will not be what you expected (i.e., it will not
6674: perform @code{foo bar}). State-smart words are a bad idea. Simply don't
6675: write them@footnote{For a more detailed discussion of this topic, see
6676: @cite{@code{State}-smartness -- Why it is Evil and How to Exorcise it} by Anton
6677: Ertl; presented at EuroForth '98 and available from
1.47 crook 6678: @*@uref{http://www.complang.tuwien.ac.at/papers/ertl98.ps.gz}}!
1.1 anton 6679:
1.26 crook 6680: @cindex defining words with arbitrary semantics combinations
6681: It is also possible to write defining words that define words with
6682: arbitrary combinations of interpretation and compilation semantics. In
6683: general, they look like this:
1.1 anton 6684:
1.26 crook 6685: @example
6686: : def-word
6687: create-interpret/compile
1.29 crook 6688: @i{code1}
1.26 crook 6689: interpretation>
1.29 crook 6690: @i{code2}
1.26 crook 6691: <interpretation
6692: compilation>
1.29 crook 6693: @i{code3}
1.26 crook 6694: <compilation ;
6695: @end example
1.1 anton 6696:
1.29 crook 6697: For a @i{word} defined with @code{def-word}, the interpretation
6698: semantics are to push the address of the body of @i{word} and perform
6699: @i{code2}, and the compilation semantics are to push the address of
6700: the body of @i{word} and perform @i{code3}. E.g., @code{constant}
1.26 crook 6701: can also be defined like this (except that the defined constants don't
6702: behave correctly when @code{[compile]}d):
1.1 anton 6703:
1.26 crook 6704: @example
6705: : constant ( n "name" -- )
6706: create-interpret/compile
6707: ,
6708: interpretation> ( -- n )
6709: @@
6710: <interpretation
6711: compilation> ( compilation. -- ; run-time. -- n )
6712: @@ postpone literal
6713: <compilation ;
6714: @end example
1.1 anton 6715:
1.44 crook 6716:
1.26 crook 6717: doc-create-interpret/compile
6718: doc-interpretation>
6719: doc-<interpretation
6720: doc-compilation>
6721: doc-<compilation
1.1 anton 6722:
1.44 crook 6723:
1.29 crook 6724: Words defined with @code{interpret/compile:} and
1.26 crook 6725: @code{create-interpret/compile} have an extended header structure that
6726: differs from other words; however, unless you try to access them with
6727: plain address arithmetic, you should not notice this. Words for
6728: accessing the header structure usually know how to deal with this; e.g.,
1.29 crook 6729: @code{'} @i{word} @code{>body} also gives you the body of a word created
6730: with @code{create-interpret/compile}.
1.1 anton 6731:
1.44 crook 6732:
1.27 crook 6733: doc-postpone
1.44 crook 6734:
1.29 crook 6735: @comment TODO -- expand glossary text for POSTPONE
1.27 crook 6736:
1.47 crook 6737:
6738: @c -------------------------------------------------------------
6739: @node Tokens for Words, The Text Interpreter, Interpretation and Compilation Semantics, Words
6740: @section Tokens for Words
6741: @cindex tokens for words
6742:
6743: This section describes the creation and use of tokens that represent
6744: words.
6745:
6746: Named words have information stored in their header space entries to
6747: indicate any non-default semantics (@pxref{Interpretation and
6748: Compilation Semantics}). The semantics can be modified, using
6749: @code{immediate} and/or @code{compile-only}, at the time that the words
6750: are defined. Unnamed words have (by definition) no header space
6751: entry, and therefore must have default semantics.
6752:
6753: Named words have interpretation and compilation semantics. Unnamed words
6754: just have execution semantics.
6755:
6756: @cindex xt
6757: @cindex execution token
6758: The execution semantics of an unnamed word are represented by an
6759: @dfn{execution token} (@i{xt}). As explained in @ref{Supplying names},
6760: the execution token of the last word defined can be produced with
6761: @code{lastxt}.
6762:
6763: The interpretation semantics of a named word are also represented by an
6764: execution token. You can produce the execution token using @code{'} or
6765: @code{[']}. A simple example shows the difference between the two:
6766:
6767: @example
6768: : greet ( -- ) ." Hello" ;
6769: : foo ( -- xt ) ['] greet execute ; \ ['] parses greet at compile-time
6770: : bar ( -- ) ' execute ; \ ' parses at run-time
6771:
6772: \ the next four lines all do the same thing
6773: foo
6774: bar greet
6775: greet
6776: ' greet EXECUTE
6777: @end example
6778:
6779: An execution token occupies one cell.
6780: @cindex code field address
6781: @cindex CFA
6782: In Gforth, the abstract data type @i{execution token} is implemented
6783: as a code field address (CFA).
6784: @comment TODO note that the standard does not say what it represents..
6785: @comment and you cannot necessarily compile it in all Forths (eg native
6786: @comment compilers?).
6787:
6788: For literals, use @code{'} in interpreted code and @code{[']} in
6789: compiled code. Gforth's @code{'} and @code{[']} behave somewhat
6790: unusually by complaining about compile-only words. To get the execution
6791: token for a compile-only word @i{name}, use @code{COMP' @i{name} DROP}
6792: or @code{[COMP'] @i{name} DROP}.
6793:
6794: @cindex compilation token
6795: The compilation semantics of a named word are represented by a
6796: @dfn{compilation token} consisting of two cells: @i{w xt}. The top cell
6797: @i{xt} is an execution token. The compilation semantics represented by
6798: the compilation token can be performed with @code{execute}, which
6799: consumes the whole compilation token, with an additional stack effect
6800: determined by the represented compilation semantics.
6801:
6802: At present, the @i{w} part of a compilation token is an execution token,
6803: and the @i{xt} part represents either @code{execute} or
6804: @code{compile,}@footnote{Depending upon the compilation semantics of the
6805: word. If the word has default compilation semantics, the @i{xt} will
6806: represent @code{compile,}. Otherwise (e.g., for immediate words), the
6807: @i{xt} will represent @code{execute}.}. However, don't rely on that
6808: knowledge, unless necessary; future versions of Gforth may introduce
6809: unusual compilation tokens (e.g., a compilation token that represents
6810: the compilation semantics of a literal).
6811:
6812: You can compile the compilation semantics with @code{postpone,}. I.e.,
6813: @code{COMP' @i{word} postpone,} is equivalent to @code{postpone
6814: @i{word}}.
6815:
6816: @cindex name token
6817: @cindex name field address
6818: @cindex NFA
6819: Named words are also represented by the @dfn{name token}, (@i{nt}). In
6820: Gforth, the abstract data type @emph{name token} is implemented as a
6821: name field address (NFA).
6822:
6823:
6824: doc-execute
6825: doc-perform
6826: doc-compile,
6827: doc-[']
6828: doc-'
6829: doc-[comp']
6830: doc-comp'
6831: doc-postpone,
6832:
6833: doc-find-name
6834: doc-name>int
6835: doc-name?int
6836: doc-name>comp
6837: doc-name>string
6838:
6839:
1.26 crook 6840: @c ----------------------------------------------------------
1.47 crook 6841: @node The Text Interpreter, Word Lists, Tokens for Words, Words
1.26 crook 6842: @section The Text Interpreter
6843: @cindex interpreter - outer
6844: @cindex text interpreter
6845: @cindex outer interpreter
1.1 anton 6846:
1.34 anton 6847: @c Should we really describe all these ugly details? IMO the text
6848: @c interpreter should be much cleaner, but that may not be possible within
6849: @c ANS Forth. - anton
1.44 crook 6850: @c nac-> I wanted to explain how it works to show how you can exploit
6851: @c it in your own programs. When I was writing a cross-compiler, figuring out
6852: @c some of these gory details was very helpful to me. None of the textbooks
6853: @c I've seen cover it, and the most modern Forth textbook -- Forth Inc's,
6854: @c seems to positively avoid going into too much detail for some of
6855: @c the internals.
1.34 anton 6856:
1.29 crook 6857: The text interpreter@footnote{This is an expanded version of the
6858: material in @ref{Introducing the Text Interpreter}.} is an endless loop
1.34 anton 6859: that processes input from the current input device. It is also called
6860: the outer interpreter, in contrast to the inner interpreter
6861: (@pxref{Engine}) which executes the compiled Forth code on interpretive
6862: implementations.
1.27 crook 6863:
1.29 crook 6864: @cindex interpret state
6865: @cindex compile state
6866: The text interpreter operates in one of two states: @dfn{interpret
6867: state} and @dfn{compile state}. The current state is defined by the
6868: aptly-named variable, @code{state}.
6869:
6870: This section starts by describing how the text interpreter behaves when
6871: it is in interpret state, processing input from the user input device --
6872: the keyboard. This is the mode that a Forth system is in after it starts
6873: up.
6874:
6875: @cindex input buffer
6876: @cindex terminal input buffer
6877: The text interpreter works from an area of memory called the @dfn{input
6878: buffer}@footnote{When the text interpreter is processing input from the
6879: keyboard, this area of memory is called the @dfn{terminal input buffer}
6880: (TIB) and is addressed by the (obsolescent) words @code{TIB} and
6881: @code{#TIB}.}, which stores your keyboard input when you press the
1.30 anton 6882: @key{RET} key. Starting at the beginning of the input buffer, it skips
1.29 crook 6883: leading spaces (called @dfn{delimiters}) then parses a string (a
6884: sequence of non-space characters) until it reaches either a space
6885: character or the end of the buffer. Having parsed a string, it makes two
6886: attempts to process it:
1.27 crook 6887:
1.29 crook 6888: @cindex dictionary
1.27 crook 6889: @itemize @bullet
6890: @item
1.29 crook 6891: It looks for the string in a @dfn{dictionary} of definitions. If the
6892: string is found, the string names a @dfn{definition} (also known as a
6893: @dfn{word}) and the dictionary search returns information that allows
6894: the text interpreter to perform the word's @dfn{interpretation
6895: semantics}. In most cases, this simply means that the word will be
6896: executed.
1.27 crook 6897: @item
6898: If the string is not found in the dictionary, the text interpreter
1.29 crook 6899: attempts to treat it as a number, using the rules described in
6900: @ref{Number Conversion}. If the string represents a legal number in the
6901: current radix, the number is pushed onto a parameter stack (the data
6902: stack for integers, the floating-point stack for floating-point
6903: numbers).
6904: @end itemize
6905:
6906: If both attempts fail, or if the word is found in the dictionary but has
6907: no interpretation semantics@footnote{This happens if the word was
6908: defined as @code{COMPILE-ONLY}.} the text interpreter discards the
6909: remainder of the input buffer, issues an error message and waits for
6910: more input. If one of the attempts succeeds, the text interpreter
6911: repeats the parsing process until the whole of the input buffer has been
6912: processed, at which point it prints the status message ``@code{ ok}''
6913: and waits for more input.
6914:
6915: @cindex parse area
6916: The text interpreter keeps track of its position in the input buffer by
6917: updating a variable called @code{>IN} (pronounced ``to-in''). The value
6918: of @code{>IN} starts out as 0, indicating an offset of 0 from the start
6919: of the input buffer. The region from offset @code{>IN @@} to the end of
6920: the input buffer is called the @dfn{parse area}@footnote{In other words,
6921: the text interpreter processes the contents of the input buffer by
6922: parsing strings from the parse area until the parse area is empty.}.
6923: This example shows how @code{>IN} changes as the text interpreter parses
6924: the input buffer:
6925:
6926: @example
6927: : remaining >IN @@ SOURCE 2 PICK - -ROT + SWAP
6928: CR ." ->" TYPE ." <-" ; IMMEDIATE
6929:
6930: 1 2 3 remaining + remaining .
6931:
6932: : foo 1 2 3 remaining SWAP remaining ;
6933: @end example
6934:
6935: @noindent
6936: The result is:
6937:
6938: @example
6939: ->+ remaining .<-
6940: ->.<-5 ok
6941:
6942: ->SWAP remaining ;-<
6943: ->;<- ok
6944: @end example
6945:
6946: @cindex parsing words
6947: The value of @code{>IN} can also be modified by a word in the input
6948: buffer that is executed by the text interpreter. This means that a word
6949: can ``trick'' the text interpreter into either skipping a section of the
6950: input buffer@footnote{This is how parsing words work.} or into parsing a
6951: section twice. For example:
1.27 crook 6952:
1.29 crook 6953: @example
6954: : lat ." <<lat>>" ;
6955: : flat ." <<flat>>" >IN DUP @@ 3 - SWAP ! ;
6956: @end example
6957:
6958: @noindent
6959: When @code{flat} is executed, this output is produced@footnote{Exercise
6960: for the reader: what would happen if the @code{3} were replaced with
6961: @code{4}?}:
6962:
6963: @example
6964: <<flat>><<lat>>
6965: @end example
6966:
6967: @noindent
6968: Two important notes about the behaviour of the text interpreter:
1.27 crook 6969:
6970: @itemize @bullet
6971: @item
6972: It processes each input string to completion before parsing additional
1.29 crook 6973: characters from the input buffer.
6974: @item
6975: It treats the input buffer as a read-only region (and so must your code).
6976: @end itemize
6977:
6978: @noindent
6979: When the text interpreter is in compile state, its behaviour changes in
6980: these ways:
6981:
6982: @itemize @bullet
6983: @item
6984: If a parsed string is found in the dictionary, the text interpreter will
6985: perform the word's @dfn{compilation semantics}. In most cases, this
6986: simply means that the execution semantics of the word will be appended
6987: to the current definition.
1.27 crook 6988: @item
1.29 crook 6989: When a number is encountered, it is compiled into the current definition
6990: (as a literal) rather than being pushed onto a parameter stack.
6991: @item
6992: If an error occurs, @code{state} is modified to put the text interpreter
6993: back into interpret state.
6994: @item
6995: Each time a line is entered from the keyboard, Gforth prints
6996: ``@code{ compiled}'' rather than `` @code{ok}''.
6997: @end itemize
6998:
6999: @cindex text interpreter - input sources
7000: When the text interpreter is using an input device other than the
7001: keyboard, its behaviour changes in these ways:
7002:
7003: @itemize @bullet
7004: @item
7005: When the parse area is empty, the text interpreter attempts to refill
7006: the input buffer from the input source. When the input source is
7007: exhausted, the input source is set back to the user input device.
7008: @item
7009: It doesn't print out ``@code{ ok}'' or ``@code{ compiled}'' messages each
7010: time the parse area is emptied.
7011: @item
7012: If an error occurs, the input source is set back to the user input
7013: device.
1.27 crook 7014: @end itemize
1.21 crook 7015:
1.49 anton 7016: You can read about this in more detail in @ref{Input Sources}.
1.44 crook 7017:
1.26 crook 7018: doc->in
1.27 crook 7019: doc-source
7020:
1.26 crook 7021: doc-tib
7022: doc-#tib
1.1 anton 7023:
1.44 crook 7024:
1.26 crook 7025: @menu
1.29 crook 7026: * Input Sources::
1.26 crook 7027: * Number Conversion::
7028: * Interpret/Compile states::
7029: * Literals::
7030: * Interpreter Directives::
7031: @end menu
1.1 anton 7032:
1.29 crook 7033: @node Input Sources, Number Conversion, The Text Interpreter, The Text Interpreter
7034: @subsection Input Sources
7035: @cindex input sources
7036: @cindex text interpreter - input sources
7037:
1.44 crook 7038: By default, the text interpreter processes input from the user input
1.29 crook 7039: device (the keyboard) when Forth starts up. The text interpreter can
7040: process input from any of these sources:
7041:
7042: @itemize @bullet
7043: @item
7044: The user input device -- the keyboard.
7045: @item
7046: A file, using the words described in @ref{Forth source files}.
7047: @item
7048: A block, using the words described in @ref{Blocks}.
7049: @item
7050: A text string, using @code{evaluate}.
7051: @end itemize
7052:
7053: A program can identify the current input device from the values of
7054: @code{source-id} and @code{blk}.
7055:
1.44 crook 7056:
1.29 crook 7057: doc-source-id
7058: doc-blk
7059:
7060: doc-save-input
7061: doc-restore-input
7062:
7063: doc-evaluate
1.1 anton 7064:
1.29 crook 7065:
1.44 crook 7066:
1.29 crook 7067: @node Number Conversion, Interpret/Compile states, Input Sources, The Text Interpreter
1.26 crook 7068: @subsection Number Conversion
7069: @cindex number conversion
7070: @cindex double-cell numbers, input format
7071: @cindex input format for double-cell numbers
7072: @cindex single-cell numbers, input format
7073: @cindex input format for single-cell numbers
7074: @cindex floating-point numbers, input format
7075: @cindex input format for floating-point numbers
1.1 anton 7076:
1.29 crook 7077: This section describes the rules that the text interpreter uses when it
7078: tries to convert a string into a number.
1.1 anton 7079:
1.26 crook 7080: Let <digit> represent any character that is a legal digit in the current
1.29 crook 7081: number base@footnote{For example, 0-9 when the number base is decimal or
7082: 0-9, A-F when the number base is hexadecimal.}.
1.1 anton 7083:
1.26 crook 7084: Let <decimal digit> represent any character in the range 0-9.
1.1 anton 7085:
1.29 crook 7086: Let @{@i{a b}@} represent the @i{optional} presence of any of the characters
7087: in the braces (@i{a} or @i{b} or neither).
1.1 anton 7088:
1.26 crook 7089: Let * represent any number of instances of the previous character
7090: (including none).
1.1 anton 7091:
1.26 crook 7092: Let any other character represent itself.
1.1 anton 7093:
1.29 crook 7094: @noindent
1.26 crook 7095: Now, the conversion rules are:
1.21 crook 7096:
1.26 crook 7097: @itemize @bullet
7098: @item
7099: A string of the form <digit><digit>* is treated as a single-precision
1.29 crook 7100: (cell-sized) positive integer. Examples are 0 123 6784532 32343212343456 42
1.26 crook 7101: @item
7102: A string of the form -<digit><digit>* is treated as a single-precision
1.29 crook 7103: (cell-sized) negative integer, and is represented using 2's-complement
1.26 crook 7104: arithmetic. Examples are -45 -5681 -0
7105: @item
7106: A string of the form <digit><digit>*.<digit>* is treated as a double-precision
1.29 crook 7107: (double-cell-sized) positive integer. Examples are 3465. 3.465 34.65
7108: (all three of these represent the same number).
1.26 crook 7109: @item
7110: A string of the form -<digit><digit>*.<digit>* is treated as a
1.29 crook 7111: double-precision (double-cell-sized) negative integer, and is
1.26 crook 7112: represented using 2's-complement arithmetic. Examples are -3465. -3.465
1.29 crook 7113: -34.65 (all three of these represent the same number).
1.26 crook 7114: @item
1.29 crook 7115: A string of the form @{+ -@}<decimal digit>@{.@}<decimal digit>*@{e
7116: E@}@{+ -@}<decimal digit><decimal digit>* is treated as a floating-point
1.35 anton 7117: number. Examples are 1e 1e0 1.e 1.e0 +1e+0 (which all represent the same
1.29 crook 7118: number) +12.E-4
1.26 crook 7119: @end itemize
1.1 anton 7120:
1.26 crook 7121: By default, the number base used for integer number conversion is given
1.35 anton 7122: by the contents of the variable @code{base}. Note that a lot of
7123: confusion can result from unexpected values of @code{base}. If you
7124: change @code{base} anywhere, make sure to save the old value and restore
7125: it afterwards. In general I recommend keeping @code{base} decimal, and
7126: using the prefixes described below for the popular non-decimal bases.
1.1 anton 7127:
1.29 crook 7128: doc-dpl
1.26 crook 7129: doc-base
7130: doc-hex
7131: doc-decimal
1.1 anton 7132:
1.44 crook 7133:
1.26 crook 7134: @cindex '-prefix for character strings
7135: @cindex &-prefix for decimal numbers
7136: @cindex %-prefix for binary numbers
7137: @cindex $-prefix for hexadecimal numbers
1.35 anton 7138: Gforth allows you to override the value of @code{base} by using a
1.29 crook 7139: prefix@footnote{Some Forth implementations provide a similar scheme by
7140: implementing @code{$} etc. as parsing words that process the subsequent
7141: number in the input stream and push it onto the stack. For example, see
7142: @cite{Number Conversion and Literals}, by Wil Baden; Forth Dimensions
7143: 20(3) pages 26--27. In such implementations, unlike in Gforth, a space
7144: is required between the prefix and the number.} before the first digit
7145: of an (integer) number. Four prefixes are supported:
1.1 anton 7146:
1.26 crook 7147: @itemize @bullet
7148: @item
1.35 anton 7149: @code{&} -- decimal
1.26 crook 7150: @item
1.35 anton 7151: @code{%} -- binary
1.26 crook 7152: @item
1.35 anton 7153: @code{$} -- hexadecimal
1.26 crook 7154: @item
1.35 anton 7155: @code{'} -- base @code{max-char+1}
1.26 crook 7156: @end itemize
1.1 anton 7157:
1.26 crook 7158: Here are some examples, with the equivalent decimal number shown after
7159: in braces:
1.1 anton 7160:
1.26 crook 7161: -$41 (-65), %1001101 (205), %1001.0001 (145 - a double-precision number),
7162: 'AB (16706; ascii A is 65, ascii B is 66, number is 65*256 + 66),
7163: 'ab (24930; ascii a is 97, ascii B is 98, number is 97*256 + 98),
7164: &905 (905), $abc (2478), $ABC (2478).
1.1 anton 7165:
1.26 crook 7166: @cindex number conversion - traps for the unwary
1.29 crook 7167: @noindent
1.26 crook 7168: Number conversion has a number of traps for the unwary:
1.1 anton 7169:
1.26 crook 7170: @itemize @bullet
7171: @item
7172: You cannot determine the current number base using the code sequence
1.35 anton 7173: @code{base @@ .} -- the number base is always 10 in the current number
7174: base. Instead, use something like @code{base @@ dec.}
1.26 crook 7175: @item
7176: If the number base is set to a value greater than 14 (for example,
7177: hexadecimal), the number 123E4 is ambiguous; the conversion rules allow
7178: it to be intepreted as either a single-precision integer or a
7179: floating-point number (Gforth treats it as an integer). The ambiguity
7180: can be resolved by explicitly stating the sign of the mantissa and/or
7181: exponent: 123E+4 or +123E4 -- if the number base is decimal, no
7182: ambiguity arises; either representation will be treated as a
7183: floating-point number.
7184: @item
1.29 crook 7185: There is a word @code{bin} but it does @i{not} set the number base!
1.26 crook 7186: It is used to specify file types.
7187: @item
7188: ANS Forth requires the @code{.} of a double-precision number to
7189: be the final character in the string. Allowing the @code{.} to be
7190: anywhere after the first digit is a Gforth extension.
7191: @item
7192: The number conversion process does not check for overflow.
7193: @item
7194: In Gforth, number conversion to floating-point numbers always use base
1.35 anton 7195: 10, irrespective of the value of @code{base}. In ANS Forth,
1.26 crook 7196: conversion to floating-point numbers whilst the value of
1.35 anton 7197: @code{base} is not 10 is an ambiguous condition.
1.26 crook 7198: @end itemize
1.1 anton 7199:
1.49 anton 7200: You can read numbers into your programs with the words described in
7201: @ref{Input}.
1.1 anton 7202:
1.26 crook 7203: @node Interpret/Compile states, Literals, Number Conversion, The Text Interpreter
7204: @subsection Interpret/Compile states
7205: @cindex Interpret/Compile states
1.1 anton 7206:
1.29 crook 7207: A standard program is not permitted to change @code{state}
7208: explicitly. However, it can change @code{state} implicitly, using the
7209: words @code{[} and @code{]}. When @code{[} is executed it switches
7210: @code{state} to interpret state, and therefore the text interpreter
7211: starts interpreting. When @code{]} is executed it switches @code{state}
7212: to compile state and therefore the text interpreter starts
1.44 crook 7213: compiling. The most common usage for these words is for switching into
7214: interpret state and back from within a colon definition; this technique
1.49 anton 7215: can be used to compile a literal (for an example, @pxref{Literals}) or
7216: for conditional compilation (for an example, @pxref{Interpreter
7217: Directives}).
1.44 crook 7218:
1.35 anton 7219:
7220: @c This is a bad example: It's non-standard, and it's not necessary.
7221: @c However, I can't think of a good example for switching into compile
7222: @c state when there is no current word (@code{state}-smart words are not a
7223: @c good reason). So maybe we should use an example for switching into
7224: @c interpret @code{state} in a colon def. - anton
1.44 crook 7225: @c nac-> I agree. I started out by putting in the example, then realised
7226: @c that it was non-ANS, so wrote more words around it. I hope this
7227: @c re-written version is acceptable to you. I do want to keep the example
7228: @c as it is helpful for showing what is and what is not portable, particularly
7229: @c where it outlaws a style in common use.
7230:
1.35 anton 7231:
1.44 crook 7232: @code{[} and @code{]} also give you the ability to switch into compile
7233: state and back, but we cannot think of any useful Standard application
7234: for this ability. Pre-ANS Forth textbooks have examples like this:
1.29 crook 7235:
7236: @example
7237: : AA ." this is A" ;
7238: : BB ." this is B" ;
7239: : CC ." this is C" ;
7240:
1.44 crook 7241: create table ] aa bb cc [
7242:
1.29 crook 7243: : go ( n -- ) \ n is offset into table.. 0 for 1st entry
7244: cells table + @ execute ;
7245: @end example
7246:
1.44 crook 7247: This example builds a jump table; @code{0 go} will display ``@code{this
7248: is A}''. Using @code{[} and @code{]} in this example is equivalent to
7249: defining @code{table} like this:
1.29 crook 7250:
7251: @example
1.44 crook 7252: create table ' aa COMPILE, ' bb COMPILE, ' cc COMPILE,
1.29 crook 7253: @end example
7254:
1.44 crook 7255: The problem with this code is that the definition of @code{table} is not
7256: portable -- it @i{compile}s execution tokens into code space. Whilst it
7257: @i{may} work on systems where code space and data space co-incide, the
1.29 crook 7258: Standard only allows data space to be assigned for a @code{CREATE}d
7259: word. In addition, the Standard only allows @code{@@} to access data
7260: space, whilst this example is using it to access code space. The only
7261: portable, Standard way to build this table is to build it in data space,
7262: like this:
7263:
7264: @example
7265: create table ' aa , ' bb , ' cc ,
7266: @end example
7267:
1.26 crook 7268: doc-state
7269: doc-[
7270: doc-]
1.1 anton 7271:
1.44 crook 7272:
1.26 crook 7273: @node Literals, Interpreter Directives, Interpret/Compile states, The Text Interpreter
7274: @subsection Literals
7275: @cindex Literals
1.21 crook 7276:
1.29 crook 7277: Often, you want to use a number within a colon definition. When you do
7278: this, the text interpreter automatically compiles the number as a
7279: @i{literal}. A literal is a number whose run-time effect is to be pushed
7280: onto the stack. If you had to do some maths to generate the number, you
7281: might write it like this:
7282:
7283: @example
7284: : HOUR-TO-SEC ( n1 -- n2 )
7285: 60 * \ to minutes
7286: 60 * ; \ to seconds
7287: @end example
7288:
7289: It is very clear what this definition is doing, but it's inefficient
7290: since it is performing 2 multiples at run-time. An alternative would be
7291: to write:
7292:
7293: @example
7294: : HOUR-TO-SEC ( n1 -- n2 )
7295: 3600 * ; \ to seconds
7296: @end example
7297:
7298: Which does the same thing, and has the advantage of using a single
7299: multiply. Ideally, we'd like the efficiency of the second with the
7300: readability of the first.
7301:
7302: @code{Literal} allows us to achieve that. It takes a number from the
7303: stack and lays it down in the current definition just as though the
7304: number had been typed directly into the definition. Our first attempt
7305: might look like this:
7306:
7307: @example
7308: 60 \ mins per hour
7309: 60 * \ seconds per minute
7310: : HOUR-TO-SEC ( n1 -- n2 )
7311: Literal * ; \ to seconds
7312: @end example
7313:
7314: But this produces the error message @code{unstructured}. What happened?
7315: The stack notation for @code{:} is (@i{ -- colon-sys}) and the size of
7316: @i{colon-sys} is implementation-defined. In other words, once we start a
7317: colon definition we can't portably access anything that was on the stack
7318: before the definition began@footnote{@cite{Two Problems in ANS Forth},
7319: by Thomas Worthington; Forth Dimensions 20(2) pages 32--34 describes
7320: some situations where you might want to access stack items above
7321: colon-sys, and provides a solution to the problem.}. The correct way of
7322: solving this problem in this instance is to use @code{[ ]} like this:
7323:
7324: @example
7325: : HOUR-TO-SEC ( n1 -- n2 )
7326: [ 60 \ minutes per hour
7327: 60 * ] \ seconds per minute
7328: LITERAL * ; \ to seconds
7329: @end example
1.23 crook 7330:
1.44 crook 7331:
1.26 crook 7332: doc-literal
7333: doc-]L
7334: doc-2literal
7335: doc-fliteral
1.1 anton 7336:
1.44 crook 7337:
1.48 anton 7338: @node Interpreter Directives, , Literals, The Text Interpreter
1.26 crook 7339: @subsection Interpreter Directives
7340: @cindex interpreter directives
1.1 anton 7341:
1.29 crook 7342: These words are usually used in interpret state; typically to control
7343: which parts of a source file are processed by the text
1.26 crook 7344: interpreter. There are only a few ANS Forth Standard words, but Gforth
7345: supplements these with a rich set of immediate control structure words
7346: to compensate for the fact that the non-immediate versions can only be
1.29 crook 7347: used in compile state (@pxref{Control Structures}). Typical usages:
7348:
7349: @example
7350: FALSE Constant ASSEMBLER
7351: .
7352: .
7353: ASSEMBLER [IF]
7354: : ASSEMBLER-FEATURE
7355: ...
7356: ;
7357: [ENDIF]
7358: .
7359: .
7360: : SEE
7361: ... \ general-purpose SEE code
7362: [ ASSEMBLER [IF] ]
7363: ... \ assembler-specific SEE code
7364: [ [ENDIF] ]
7365: ;
7366: @end example
1.1 anton 7367:
1.44 crook 7368:
1.26 crook 7369: doc-[IF]
7370: doc-[ELSE]
7371: doc-[THEN]
7372: doc-[ENDIF]
1.1 anton 7373:
1.26 crook 7374: doc-[IFDEF]
7375: doc-[IFUNDEF]
1.1 anton 7376:
1.26 crook 7377: doc-[?DO]
7378: doc-[DO]
7379: doc-[FOR]
7380: doc-[LOOP]
7381: doc-[+LOOP]
7382: doc-[NEXT]
1.1 anton 7383:
1.26 crook 7384: doc-[BEGIN]
7385: doc-[UNTIL]
7386: doc-[AGAIN]
7387: doc-[WHILE]
7388: doc-[REPEAT]
1.1 anton 7389:
1.27 crook 7390:
1.26 crook 7391: @c -------------------------------------------------------------
1.47 crook 7392: @node Word Lists, Environmental Queries, The Text Interpreter, Words
1.26 crook 7393: @section Word Lists
7394: @cindex word lists
1.32 anton 7395: @cindex header space
1.1 anton 7396:
1.36 anton 7397: A wordlist is a list of named words; you can add new words and look up
7398: words by name (and you can remove words in a restricted way with
7399: markers). Every named (and @code{reveal}ed) word is in one wordlist.
7400:
7401: @cindex search order stack
7402: The text interpreter searches the wordlists present in the search order
7403: (a stack of wordlists), from the top to the bottom. Within each
7404: wordlist, the search starts conceptually at the newest word; i.e., if
7405: two words in a wordlist have the same name, the newer word is found.
1.1 anton 7406:
1.26 crook 7407: @cindex compilation word list
1.36 anton 7408: New words are added to the @dfn{compilation wordlist} (aka current
7409: wordlist).
1.1 anton 7410:
1.36 anton 7411: @cindex wid
7412: A word list is identified by a cell-sized word list identifier (@i{wid})
7413: in much the same way as a file is identified by a file handle. The
7414: numerical value of the wid has no (portable) meaning, and might change
7415: from session to session.
1.1 anton 7416:
1.29 crook 7417: The ANS Forth ``Search order'' word set is intended to provide a set of
7418: low-level tools that allow various different schemes to be
1.26 crook 7419: implemented. Gforth provides @code{vocabulary}, a traditional Forth
7420: word. @file{compat/vocabulary.fs} provides an implementation in ANS
1.45 crook 7421: Forth.
1.1 anton 7422:
1.27 crook 7423: @comment TODO: locals section refers to here, saying that every word list (aka
7424: @comment vocabulary) has its own methods for searching etc. Need to document that.
1.1 anton 7425:
1.45 crook 7426: @comment TODO: document markers, reveal, tables, mappedwordlist
7427:
7428: @comment the gforthman- prefix is used to pick out the true definition of a
1.27 crook 7429: @comment word from the source files, rather than some alias.
1.44 crook 7430:
1.26 crook 7431: doc-forth-wordlist
7432: doc-definitions
7433: doc-get-current
7434: doc-set-current
7435: doc-get-order
1.45 crook 7436: doc---gforthman-set-order
1.26 crook 7437: doc-wordlist
1.30 anton 7438: doc-table
1.36 anton 7439: doc-push-order
7440: doc-previous
1.26 crook 7441: doc-also
1.45 crook 7442: doc---gforthman-forth
1.26 crook 7443: doc-only
1.45 crook 7444: doc---gforthman-order
1.15 anton 7445:
1.26 crook 7446: doc-find
7447: doc-search-wordlist
1.15 anton 7448:
1.26 crook 7449: doc-words
7450: doc-vlist
1.44 crook 7451: @c doc-words-deferred
1.1 anton 7452:
1.26 crook 7453: doc-mappedwordlist
7454: doc-root
7455: doc-vocabulary
7456: doc-seal
7457: doc-vocs
7458: doc-current
7459: doc-context
1.1 anton 7460:
1.44 crook 7461:
1.26 crook 7462: @menu
7463: * Why use word lists?::
7464: * Word list examples::
7465: @end menu
7466:
7467: @node Why use word lists?, Word list examples, Word Lists, Word Lists
7468: @subsection Why use word lists?
7469: @cindex word lists - why use them?
7470:
1.29 crook 7471: Here are some reasons for using multiple word lists:
1.26 crook 7472:
7473: @itemize @bullet
7474: @item
1.32 anton 7475: To improve compilation speed by reducing the number of header space
1.26 crook 7476: entries that must be searched. This is achieved by creating a new
7477: word list that contains all of the definitions that are used in the
7478: definition of a Forth system but which would not usually be used by
7479: programs running on that system. That word list would be on the search
7480: list when the Forth system was compiled but would be removed from the
7481: search list for normal operation. This can be a useful technique for
7482: low-performance systems (for example, 8-bit processors in embedded
7483: systems) but is unlikely to be necessary in high-performance desktop
7484: systems.
7485: @item
7486: To prevent a set of words from being used outside the context in which
7487: they are valid. Two classic examples of this are an integrated editor
7488: (all of the edit commands are defined in a separate word list; the
7489: search order is set to the editor word list when the editor is invoked;
7490: the old search order is restored when the editor is terminated) and an
7491: integrated assembler (the op-codes for the machine are defined in a
7492: separate word list which is used when a @code{CODE} word is defined).
7493: @item
7494: To prevent a name-space clash between multiple definitions with the same
7495: name. For example, when building a cross-compiler you might have a word
7496: @code{IF} that generates conditional code for your target system. By
7497: placing this definition in a different word list you can control whether
7498: the host system's @code{IF} or the target system's @code{IF} get used in
7499: any particular context by controlling the order of the word lists on the
7500: search order stack.
7501: @end itemize
1.1 anton 7502:
1.48 anton 7503: @node Word list examples, , Why use word lists?, Word Lists
1.26 crook 7504: @subsection Word list examples
7505: @cindex word lists - examples
1.1 anton 7506:
1.26 crook 7507: Here is an example of creating and using a new wordlist using ANS
7508: Forth Standard words:
1.1 anton 7509:
7510: @example
1.26 crook 7511: wordlist constant my-new-words-wordlist
7512: : my-new-words get-order nip my-new-words-wordlist swap set-order ;
1.21 crook 7513:
1.26 crook 7514: \ add it to the search order
7515: also my-new-words
1.21 crook 7516:
1.26 crook 7517: \ alternatively, add it to the search order and make it
7518: \ the compilation word list
7519: also my-new-words definitions
7520: \ type "order" to see the problem
1.21 crook 7521: @end example
7522:
1.26 crook 7523: The problem with this example is that @code{order} has no way to
7524: associate the name @code{my-new-words} with the wid of the word list (in
7525: Gforth, @code{order} and @code{vocs} will display @code{???} for a wid
7526: that has no associated name). There is no Standard way of associating a
7527: name with a wid.
7528:
7529: In Gforth, this example can be re-coded using @code{vocabulary}, which
7530: associates a name with a wid:
1.21 crook 7531:
1.26 crook 7532: @example
7533: vocabulary my-new-words
1.21 crook 7534:
1.26 crook 7535: \ add it to the search order
1.45 crook 7536: also my-new-words
1.21 crook 7537:
1.26 crook 7538: \ alternatively, add it to the search order and make it
7539: \ the compilation word list
7540: my-new-words definitions
7541: \ type "order" to see that the problem is solved
7542: @end example
1.23 crook 7543:
1.26 crook 7544: @c -------------------------------------------------------------
7545: @node Environmental Queries, Files, Word Lists, Words
7546: @section Environmental Queries
7547: @cindex environmental queries
1.21 crook 7548:
1.26 crook 7549: ANS Forth introduced the idea of ``environmental queries'' as a way
7550: for a program running on a system to determine certain characteristics of the system.
7551: The Standard specifies a number of strings that might be recognised by a system.
1.21 crook 7552:
1.32 anton 7553: The Standard requires that the header space used for environmental queries
7554: be distinct from the header space used for definitions.
1.21 crook 7555:
1.26 crook 7556: Typically, environmental queries are supported by creating a set of
1.29 crook 7557: definitions in a word list that is @i{only} used during environmental
1.26 crook 7558: queries; that is what Gforth does. There is no Standard way of adding
7559: definitions to the set of recognised environmental queries, but any
7560: implementation that supports the loading of optional word sets must have
7561: some mechanism for doing this (after loading the word set, the
7562: associated environmental query string must return @code{true}). In
7563: Gforth, the word list used to honour environmental queries can be
7564: manipulated just like any other word list.
1.21 crook 7565:
1.44 crook 7566:
1.26 crook 7567: doc-environment?
7568: doc-environment-wordlist
1.21 crook 7569:
1.26 crook 7570: doc-gforth
7571: doc-os-class
1.21 crook 7572:
1.44 crook 7573:
1.26 crook 7574: Note that, whilst the documentation for (e.g.) @code{gforth} shows it
7575: returning two items on the stack, querying it using @code{environment?}
7576: will return an additional item; the @code{true} flag that shows that the
7577: string was recognised.
1.21 crook 7578:
1.26 crook 7579: @comment TODO Document the standard strings or note where they are documented herein
1.21 crook 7580:
1.26 crook 7581: Here are some examples of using environmental queries:
1.21 crook 7582:
1.26 crook 7583: @example
7584: s" address-unit-bits" environment? 0=
7585: [IF]
7586: cr .( environmental attribute address-units-bits unknown... ) cr
7587: [THEN]
1.21 crook 7588:
1.26 crook 7589: s" block" environment? [IF] DROP include block.fs [THEN]
1.21 crook 7590:
1.26 crook 7591: s" gforth" environment? [IF] 2DROP include compat/vocabulary.fs [THEN]
1.21 crook 7592:
1.26 crook 7593: s" gforth" environment? [IF] .( Gforth version ) TYPE
7594: [ELSE] .( Not Gforth..) [THEN]
7595: @end example
1.21 crook 7596:
7597:
1.26 crook 7598: Here is an example of adding a definition to the environment word list:
1.21 crook 7599:
1.26 crook 7600: @example
7601: get-current environment-wordlist set-current
7602: true constant block
7603: true constant block-ext
7604: set-current
7605: @end example
1.21 crook 7606:
1.26 crook 7607: You can see what definitions are in the environment word list like this:
1.21 crook 7608:
1.26 crook 7609: @example
7610: get-order 1+ environment-wordlist swap set-order words previous
7611: @end example
1.21 crook 7612:
7613:
1.26 crook 7614: @c -------------------------------------------------------------
7615: @node Files, Blocks, Environmental Queries, Words
7616: @section Files
1.28 crook 7617: @cindex files
7618: @cindex I/O - file-handling
1.21 crook 7619:
1.26 crook 7620: Gforth provides facilities for accessing files that are stored in the
7621: host operating system's file-system. Files that are processed by Gforth
7622: can be divided into two categories:
1.21 crook 7623:
1.23 crook 7624: @itemize @bullet
7625: @item
1.29 crook 7626: Files that are processed by the Text Interpreter (@dfn{Forth source files}).
1.23 crook 7627: @item
1.29 crook 7628: Files that are processed by some other program (@dfn{general files}).
1.26 crook 7629: @end itemize
7630:
1.45 crook 7631: doc-loadfilename
7632: doc-sourcefilename
7633: doc-sourceline#
7634:
1.26 crook 7635: @menu
1.48 anton 7636: * Forth source files::
7637: * General files::
7638: * Search Paths::
1.26 crook 7639: @end menu
7640:
1.21 crook 7641:
1.26 crook 7642: @c -------------------------------------------------------------
7643: @node Forth source files, General files, Files, Files
7644: @subsection Forth source files
7645: @cindex including files
7646: @cindex Forth source files
1.21 crook 7647:
1.26 crook 7648: The simplest way to interpret the contents of a file is to use one of
7649: these two formats:
1.21 crook 7650:
1.26 crook 7651: @example
7652: include mysource.fs
7653: s" mysource.fs" included
7654: @end example
1.21 crook 7655:
1.26 crook 7656: Sometimes you want to include a file only if it is not included already
7657: (by, say, another source file). In that case, you can use one of these
1.45 crook 7658: three formats:
1.21 crook 7659:
1.26 crook 7660: @example
7661: require mysource.fs
7662: needs mysource.fs
7663: s" mysource.fs" required
7664: @end example
1.21 crook 7665:
1.26 crook 7666: @cindex stack effect of included files
7667: @cindex including files, stack effect
1.45 crook 7668: It is good practice to write your source files such that interpreting them
7669: does not change the stack. Source files designed in this way can be used with
1.26 crook 7670: @code{required} and friends without complications. For example:
1.21 crook 7671:
1.26 crook 7672: @example
7673: 1 require foo.fs drop
7674: @end example
1.21 crook 7675:
1.44 crook 7676:
1.26 crook 7677: doc-include-file
7678: doc-included
1.28 crook 7679: doc-included?
1.26 crook 7680: doc-include
7681: doc-required
7682: doc-require
7683: doc-needs
1.28 crook 7684: doc-init-included-files
1.21 crook 7685:
1.44 crook 7686:
1.26 crook 7687: A definition in ANS Forth for @code{required} is provided in
7688: @file{compat/required.fs}.
1.21 crook 7689:
1.26 crook 7690: @c -------------------------------------------------------------
7691: @node General files, Search Paths, Forth source files, Files
7692: @subsection General files
7693: @cindex general files
7694: @cindex file-handling
1.21 crook 7695:
1.26 crook 7696: Files are opened/created by name and type. The following types are
7697: recognised:
1.1 anton 7698:
1.44 crook 7699:
1.26 crook 7700: doc-r/o
7701: doc-r/w
7702: doc-w/o
7703: doc-bin
1.1 anton 7704:
1.44 crook 7705:
1.26 crook 7706: When a file is opened/created, it returns a file identifier,
1.29 crook 7707: @i{wfileid} that is used for all other file commands. All file
7708: commands also return a status value, @i{wior}, that is 0 for a
1.26 crook 7709: successful operation and an implementation-defined non-zero value in the
7710: case of an error.
1.21 crook 7711:
1.44 crook 7712:
1.26 crook 7713: doc-open-file
7714: doc-create-file
1.21 crook 7715:
1.26 crook 7716: doc-close-file
7717: doc-delete-file
7718: doc-rename-file
7719: doc-read-file
7720: doc-read-line
7721: doc-write-file
7722: doc-write-line
7723: doc-emit-file
7724: doc-flush-file
1.21 crook 7725:
1.26 crook 7726: doc-file-status
7727: doc-file-position
7728: doc-reposition-file
7729: doc-file-size
7730: doc-resize-file
1.21 crook 7731:
1.44 crook 7732:
1.26 crook 7733: @c ---------------------------------------------------------
1.48 anton 7734: @node Search Paths, , General files, Files
1.26 crook 7735: @subsection Search Paths
7736: @cindex path for @code{included}
7737: @cindex file search path
7738: @cindex @code{include} search path
7739: @cindex search path for files
1.21 crook 7740:
1.26 crook 7741: If you specify an absolute filename (i.e., a filename starting with
7742: @file{/} or @file{~}, or with @file{:} in the second position (as in
7743: @samp{C:...})) for @code{included} and friends, that file is included
7744: just as you would expect.
1.21 crook 7745:
1.26 crook 7746: For relative filenames, Gforth uses a search path similar to Forth's
7747: search order (@pxref{Word Lists}). It tries to find the given filename
7748: in the directories present in the path, and includes the first one it
7749: finds. There are separate search paths for Forth source files and
7750: general files.
1.21 crook 7751:
1.26 crook 7752: If the search path contains the directory @file{.} (as it should), this
7753: refers to the directory that the present file was @code{included}
7754: from. This allows files to include other files relative to their own
7755: position (irrespective of the current working directory or the absolute
7756: position). This feature is essential for libraries consisting of
7757: several files, where a file may include other files from the library.
7758: It corresponds to @code{#include "..."} in C. If the current input
7759: source is not a file, @file{.} refers to the directory of the innermost
7760: file being included, or, if there is no file being included, to the
7761: current working directory.
1.21 crook 7762:
1.26 crook 7763: Use @file{~+} to refer to the current working directory (as in the
7764: @code{bash}).
1.1 anton 7765:
1.26 crook 7766: If the filename starts with @file{./}, the search path is not searched
7767: (just as with absolute filenames), and the @file{.} has the same meaning
7768: as described above.
1.1 anton 7769:
1.48 anton 7770: @menu
7771: * Forth Search Paths::
7772: * General Search Paths::
7773: @end menu
7774:
1.26 crook 7775: @c ---------------------------------------------------------
1.48 anton 7776: @node Forth Search Paths, General Search Paths, Search Paths, Search Paths
1.26 crook 7777: @subsubsection Forth Search Paths
1.28 crook 7778: @cindex search path control - Forth
1.5 anton 7779:
1.26 crook 7780: The search path is initialized when you start Gforth (@pxref{Invoking
7781: Gforth}). You can display it and change it using these words:
1.5 anton 7782:
1.44 crook 7783:
1.26 crook 7784: doc-.fpath
7785: doc-fpath+
7786: doc-fpath=
7787: doc-open-fpath-file
1.5 anton 7788:
1.44 crook 7789:
7790: @noindent
1.26 crook 7791: Here is an example of using @code{fpath} and @code{require}:
1.5 anton 7792:
1.26 crook 7793: @example
7794: fpath= /usr/lib/forth/|./
7795: require timer.fs
7796: @end example
1.5 anton 7797:
1.26 crook 7798: @c ---------------------------------------------------------
1.48 anton 7799: @node General Search Paths, , Forth Search Paths, Search Paths
1.26 crook 7800: @subsubsection General Search Paths
7801: @cindex search path control - for user applications
1.5 anton 7802:
1.26 crook 7803: Your application may need to search files in several directories, like
7804: @code{included} does. To facilitate this, Gforth allows you to define
7805: and use your own search paths, by providing generic equivalents of the
7806: Forth search path words:
1.5 anton 7807:
1.44 crook 7808:
1.26 crook 7809: doc-.path
7810: doc-path+
7811: doc-path=
7812: doc-open-path-file
1.5 anton 7813:
1.44 crook 7814:
1.26 crook 7815: Here's an example of creating a search path:
1.5 anton 7816:
1.26 crook 7817: @example
7818: \ Make a buffer for the path:
7819: create mypath 100 chars , \ maximum length (is checked)
7820: 0 , \ real len
7821: 100 chars allot \ space for path
7822: @end example
1.5 anton 7823:
1.26 crook 7824: @c -------------------------------------------------------------
7825: @node Blocks, Other I/O, Files, Words
7826: @section Blocks
1.28 crook 7827: @cindex I/O - blocks
7828: @cindex blocks
7829:
7830: When you run Gforth on a modern desk-top computer, it runs under the
7831: control of an operating system which provides certain services. One of
7832: these services is @var{file services}, which allows Forth source code
7833: and data to be stored in files and read into Gforth (@pxref{Files}).
7834:
7835: Traditionally, Forth has been an important programming language on
7836: systems where it has interfaced directly to the underlying hardware with
7837: no intervening operating system. Forth provides a mechanism, called
1.29 crook 7838: @dfn{blocks}, for accessing mass storage on such systems.
1.28 crook 7839:
7840: A block is a 1024-byte data area, which can be used to hold data or
7841: Forth source code. No structure is imposed on the contents of the
7842: block. A block is identified by its number; blocks are numbered
7843: contiguously from 1 to an implementation-defined maximum.
7844:
7845: A typical system that used blocks but no operating system might use a
7846: single floppy-disk drive for mass storage, with the disks formatted to
7847: provide 256-byte sectors. Blocks would be implemented by assigning the
7848: first four sectors of the disk to block 1, the second four sectors to
7849: block 2 and so on, up to the limit of the capacity of the disk. The disk
7850: would not contain any file system information, just the set of blocks.
7851:
1.29 crook 7852: @cindex blocks file
1.28 crook 7853: On systems that do provide file services, blocks are typically
1.29 crook 7854: implemented by storing a sequence of blocks within a single @dfn{blocks
1.28 crook 7855: file}. The size of the blocks file will be an exact multiple of 1024
7856: bytes, corresponding to the number of blocks it contains. This is the
7857: mechanism that Gforth uses.
7858:
1.29 crook 7859: @cindex @file{blocks.fb}
1.28 crook 7860: Only 1 blocks file can be open at a time. If you use block words without
7861: having specified a blocks file, Gforth defaults to the blocks file
7862: @file{blocks.fb}. Gforth uses the Forth search path when attempting to
7863: locate a blocks file (@pxref{Forth Search Paths}).
7864:
1.29 crook 7865: @cindex block buffers
1.28 crook 7866: When you read and write blocks under program control, Gforth uses a
1.29 crook 7867: number of @dfn{block buffers} as intermediate storage. These buffers are
1.28 crook 7868: not used when you use @code{load} to interpret the contents of a block.
7869:
7870: The behaviour of the block buffers is directly analagous to that of a
7871: cache. Each block buffer has three states:
7872:
7873: @itemize @bullet
7874: @item
7875: Unassigned
7876: @item
7877: Assigned-clean
7878: @item
7879: Assigned-dirty
7880: @end itemize
7881:
1.29 crook 7882: Initially, all block buffers are @i{unassigned}. In order to access a
1.28 crook 7883: block, the block (specified by its block number) must be assigned to a
7884: block buffer.
7885:
7886: The assignment of a block to a block buffer is performed by @code{block}
7887: or @code{buffer}. Use @code{block} when you wish to modify the existing
7888: contents of a block. Use @code{buffer} when you don't care about the
7889: existing contents of the block@footnote{The ANS Forth definition of
1.35 anton 7890: @code{buffer} is intended not to cause disk I/O; if the data associated
1.28 crook 7891: with the particular block is already stored in a block buffer due to an
7892: earlier @code{block} command, @code{buffer} will return that block
7893: buffer and the existing contents of the block will be
7894: available. Otherwise, @code{buffer} will simply assign a new, empty
1.29 crook 7895: block buffer for the block.}.
1.28 crook 7896:
1.47 crook 7897: Once a block has been assigned to a block buffer using @code{block} or
7898: @code{buffer}, that block buffer becomes the @i{current block buffer}
7899: and its state changes to @i{assigned-clean}. Data may only be
7900: manipulated (read or written) within the current block buffer.
7901:
7902: When the contents of the current block buffer has been modified it is
1.48 anton 7903: necessary, @emph{before calling @code{block} or @code{buffer} again}, to
7904: either abandon the changes (by doing nothing) or commit the changes,
7905: using @code{update}. Using @code{update} does not change the blocks
7906: file; it simply changes a block buffer's state to @i{assigned-dirty}.
1.28 crook 7907:
1.29 crook 7908: The word @code{flush} causes all @i{assigned-dirty} blocks to be
1.28 crook 7909: written back to the blocks file on disk. Leaving Gforth using @code{bye}
7910: also causes a @code{flush} to be performed.
7911:
1.29 crook 7912: In Gforth, @code{block} and @code{buffer} use a @i{direct-mapped}
1.28 crook 7913: algorithm to assign a block buffer to a block. That means that any
7914: particular block can only be assigned to one specific block buffer,
1.29 crook 7915: called (for the particular operation) the @i{victim buffer}. If the
1.47 crook 7916: victim buffer is @i{unassigned} or @i{assigned-clean} it is allocated to
7917: the new block immediately. If it is @i{assigned-dirty} its current
7918: contents are written back to the blocks file on disk before it is
1.28 crook 7919: allocated to the new block.
7920:
7921: Although no structure is imposed on the contents of a block, it is
7922: traditional to display the contents as 16 lines each of 64 characters. A
7923: block provides a single, continuous stream of input (for example, it
7924: acts as a single parse area) -- there are no end-of-line characters
7925: within a block, and no end-of-file character at the end of a
7926: block. There are two consequences of this:
1.26 crook 7927:
1.28 crook 7928: @itemize @bullet
7929: @item
7930: The last character of one line wraps straight into the first character
7931: of the following line
7932: @item
7933: The word @code{\} -- comment to end of line -- requires special
7934: treatment; in the context of a block it causes all characters until the
7935: end of the current 64-character ``line'' to be ignored.
7936: @end itemize
7937:
7938: In Gforth, when you use @code{block} with a non-existent block number,
1.45 crook 7939: the current blocks file will be extended to the appropriate size and the
1.28 crook 7940: block buffer will be initialised with spaces.
7941:
1.47 crook 7942: Gforth includes a simple block editor (type @code{use blocked.fb 0 list}
7943: for details) but doesn't encourage the use of blocks; the mechanism is
7944: only provided for backward compatibility -- ANS Forth requires blocks to
7945: be available when files are.
1.28 crook 7946:
7947: Common techniques that are used when working with blocks include:
7948:
7949: @itemize @bullet
7950: @item
7951: A screen editor that allows you to edit blocks without leaving the Forth
7952: environment.
7953: @item
7954: Shadow screens; where every code block has an associated block
7955: containing comments (for example: code in odd block numbers, comments in
7956: even block numbers). Typically, the block editor provides a convenient
7957: mechanism to toggle between code and comments.
7958: @item
7959: Load blocks; a single block (typically block 1) contains a number of
7960: @code{thru} commands which @code{load} the whole of the application.
7961: @end itemize
1.26 crook 7962:
1.29 crook 7963: See Frank Sergeant's Pygmy Forth to see just how well blocks can be
7964: integrated into a Forth programming environment.
1.26 crook 7965:
7966: @comment TODO what about errors on open-blocks?
1.44 crook 7967:
1.26 crook 7968: doc-open-blocks
7969: doc-use
7970: doc-get-block-fid
7971: doc-block-position
1.28 crook 7972:
7973: doc-scr
7974: doc-list
7975:
1.45 crook 7976: doc---gforthman-block
1.28 crook 7977: doc-buffer
7978:
1.26 crook 7979: doc-update
1.28 crook 7980: doc-updated?
1.26 crook 7981: doc-save-buffers
7982: doc-empty-buffers
7983: doc-empty-buffer
7984: doc-flush
1.28 crook 7985:
1.26 crook 7986: doc-load
7987: doc-thru
7988: doc-+load
7989: doc-+thru
1.45 crook 7990: doc---gforthman--->
1.26 crook 7991: doc-block-included
7992:
1.44 crook 7993:
1.26 crook 7994: @c -------------------------------------------------------------
7995: @node Other I/O, Programming Tools, Blocks, Words
7996: @section Other I/O
1.28 crook 7997: @cindex I/O - keyboard and display
1.26 crook 7998:
7999: @menu
8000: * Simple numeric output:: Predefined formats
8001: * Formatted numeric output:: Formatted (pictured) output
8002: * String Formats:: How Forth stores strings in memory
8003: * Displaying characters and strings:: Other stuff
8004: * Input:: Input
8005: @end menu
8006:
8007: @node Simple numeric output, Formatted numeric output, Other I/O, Other I/O
8008: @subsection Simple numeric output
1.28 crook 8009: @cindex numeric output - simple/free-format
1.5 anton 8010:
1.26 crook 8011: The simplest output functions are those that display numbers from the
8012: data or floating-point stacks. Floating-point output is always displayed
8013: using base 10. Numbers displayed from the data stack use the value stored
8014: in @code{base}.
1.5 anton 8015:
1.44 crook 8016:
1.26 crook 8017: doc-.
8018: doc-dec.
8019: doc-hex.
8020: doc-u.
8021: doc-.r
8022: doc-u.r
8023: doc-d.
8024: doc-ud.
8025: doc-d.r
8026: doc-ud.r
8027: doc-f.
8028: doc-fe.
8029: doc-fs.
1.5 anton 8030:
1.44 crook 8031:
1.26 crook 8032: Examples of printing the number 1234.5678E23 in the different floating-point output
8033: formats are shown below:
1.5 anton 8034:
8035: @example
1.26 crook 8036: f. 123456779999999000000000000.
8037: fe. 123.456779999999E24
8038: fs. 1.23456779999999E26
1.5 anton 8039: @end example
8040:
8041:
1.26 crook 8042: @node Formatted numeric output, String Formats, Simple numeric output, Other I/O
8043: @subsection Formatted numeric output
1.28 crook 8044: @cindex formatted numeric output
1.26 crook 8045: @cindex pictured numeric output
1.28 crook 8046: @cindex numeric output - formatted
1.26 crook 8047:
1.29 crook 8048: Forth traditionally uses a technique called @dfn{pictured numeric
1.26 crook 8049: output} for formatted printing of integers. In this technique, digits
8050: are extracted from the number (using the current output radix defined by
8051: @code{base}), converted to ASCII codes and appended to a string that is
8052: built in a scratch-pad area of memory (@pxref{core-idef,
8053: Implementation-defined options, Implementation-defined
8054: options}). Arbitrary characters can be appended to the string during the
8055: extraction process. The completed string is specified by an address
8056: and length and can be manipulated (@code{TYPE}ed, copied, modified)
8057: under program control.
1.5 anton 8058:
1.26 crook 8059: All of the words described in the previous section for simple numeric
8060: output are implemented in Gforth using pictured numeric output.
1.5 anton 8061:
1.47 crook 8062: Three important things to remember about pictured numeric output:
1.5 anton 8063:
1.26 crook 8064: @itemize @bullet
8065: @item
1.28 crook 8066: It always operates on double-precision numbers; to display a
1.49 anton 8067: single-precision number, convert it first (for ways of doing this
8068: @pxref{Double precision}).
1.26 crook 8069: @item
1.28 crook 8070: It always treats the double-precision number as though it were
8071: unsigned. The examples below show ways of printing signed numbers.
1.26 crook 8072: @item
8073: The string is built up from right to left; least significant digit first.
8074: @end itemize
1.5 anton 8075:
1.44 crook 8076:
1.26 crook 8077: doc-<#
1.47 crook 8078: doc-<<#
1.26 crook 8079: doc-#
8080: doc-#s
8081: doc-hold
8082: doc-sign
8083: doc-#>
1.47 crook 8084: doc-#>>
1.5 anton 8085:
1.26 crook 8086: doc-represent
1.5 anton 8087:
1.44 crook 8088:
8089: @noindent
1.26 crook 8090: Here are some examples of using pictured numeric output:
1.5 anton 8091:
8092: @example
1.26 crook 8093: : my-u. ( u -- )
8094: \ Simplest use of pns.. behaves like Standard u.
8095: 0 \ convert to unsigned double
8096: <# \ start conversion
8097: #s \ convert all digits
8098: #> \ complete conversion
8099: TYPE SPACE ; \ display, with trailing space
1.5 anton 8100:
1.26 crook 8101: : cents-only ( u -- )
8102: 0 \ convert to unsigned double
8103: <# \ start conversion
8104: # # \ convert two least-significant digits
8105: #> \ complete conversion, discard other digits
8106: TYPE SPACE ; \ display, with trailing space
1.5 anton 8107:
1.26 crook 8108: : dollars-and-cents ( u -- )
8109: 0 \ convert to unsigned double
8110: <# \ start conversion
8111: # # \ convert two least-significant digits
8112: [char] . hold \ insert decimal point
8113: #s \ convert remaining digits
8114: [char] $ hold \ append currency symbol
8115: #> \ complete conversion
8116: TYPE SPACE ; \ display, with trailing space
1.5 anton 8117:
1.26 crook 8118: : my-. ( n -- )
8119: \ handling negatives.. behaves like Standard .
8120: s>d \ convert to signed double
8121: swap over dabs \ leave sign byte followed by unsigned double
8122: <# \ start conversion
8123: #s \ convert all digits
8124: rot sign \ get at sign byte, append "-" if needed
8125: #> \ complete conversion
8126: TYPE SPACE ; \ display, with trailing space
1.5 anton 8127:
1.26 crook 8128: : account. ( n -- )
8129: \ accountants don't like minus signs, they use braces
8130: \ for negative numbers
8131: s>d \ convert to signed double
8132: swap over dabs \ leave sign byte followed by unsigned double
8133: <# \ start conversion
8134: 2 pick \ get copy of sign byte
8135: 0< IF [char] ) hold THEN \ right-most character of output
8136: #s \ convert all digits
8137: rot \ get at sign byte
8138: 0< IF [char] ( hold THEN
8139: #> \ complete conversion
8140: TYPE SPACE ; \ display, with trailing space
1.5 anton 8141: @end example
8142:
1.26 crook 8143: Here are some examples of using these words:
1.5 anton 8144:
8145: @example
1.26 crook 8146: 1 my-u. 1
8147: hex -1 my-u. decimal FFFFFFFF
8148: 1 cents-only 01
8149: 1234 cents-only 34
8150: 2 dollars-and-cents $0.02
8151: 1234 dollars-and-cents $12.34
8152: 123 my-. 123
8153: -123 my. -123
8154: 123 account. 123
8155: -456 account. (456)
1.5 anton 8156: @end example
8157:
8158:
1.26 crook 8159: @node String Formats, Displaying characters and strings, Formatted numeric output, Other I/O
8160: @subsection String Formats
1.27 crook 8161: @cindex strings - see character strings
8162: @cindex character strings - formats
1.28 crook 8163: @cindex I/O - see character strings
1.26 crook 8164:
1.27 crook 8165: Forth commonly uses two different methods for representing character
8166: strings:
1.26 crook 8167:
8168: @itemize @bullet
8169: @item
8170: @cindex address of counted string
1.45 crook 8171: @cindex counted string
1.29 crook 8172: As a @dfn{counted string}, represented by a @i{c-addr}. The char
8173: addressed by @i{c-addr} contains a character-count, @i{n}, of the
8174: string and the string occupies the subsequent @i{n} char addresses in
1.26 crook 8175: memory.
8176: @item
1.29 crook 8177: As cell pair on the stack; @i{c-addr u}, where @i{u} is the length
8178: of the string in characters, and @i{c-addr} is the address of the
1.26 crook 8179: first byte of the string.
8180: @end itemize
8181:
8182: ANS Forth encourages the use of the second format when representing
8183: strings on the stack, whilst conceeding that the counted string format
8184: remains useful as a way of storing strings in memory.
8185:
1.44 crook 8186:
1.26 crook 8187: doc-count
8188:
1.44 crook 8189:
1.49 anton 8190: For words that move, copy and search for strings see @ref{Memory
8191: Blocks}. For words that display characters and strings see
8192: @ref{Displaying characters and strings}.
1.26 crook 8193:
8194: @node Displaying characters and strings, Input, String Formats, Other I/O
8195: @subsection Displaying characters and strings
1.27 crook 8196: @cindex characters - compiling and displaying
8197: @cindex character strings - compiling and displaying
1.26 crook 8198:
8199: This section starts with a glossary of Forth words and ends with a set
8200: of examples.
8201:
1.44 crook 8202:
1.26 crook 8203: doc-bl
8204: doc-space
8205: doc-spaces
8206: doc-emit
8207: doc-toupper
8208: doc-."
8209: doc-.(
8210: doc-type
1.44 crook 8211: doc-typewhite
1.26 crook 8212: doc-cr
1.27 crook 8213: @cindex cursor control
1.26 crook 8214: doc-at-xy
8215: doc-page
8216: doc-s"
8217: doc-c"
8218: doc-char
8219: doc-[char]
8220: doc-sliteral
8221:
1.44 crook 8222:
8223: @noindent
1.26 crook 8224: As an example, consider the following text, stored in a file @file{test.fs}:
1.5 anton 8225:
8226: @example
1.26 crook 8227: .( text-1)
8228: : my-word
8229: ." text-2" cr
8230: .( text-3)
8231: ;
8232:
8233: ." text-4"
8234:
8235: : my-char
8236: [char] ALPHABET emit
8237: char emit
8238: ;
1.5 anton 8239: @end example
8240:
1.26 crook 8241: When you load this code into Gforth, the following output is generated:
1.5 anton 8242:
1.26 crook 8243: @example
1.30 anton 8244: @kbd{include test.fs @key{RET}} text-1text-3text-4 ok
1.26 crook 8245: @end example
1.5 anton 8246:
1.26 crook 8247: @itemize @bullet
8248: @item
8249: Messages @code{text-1} and @code{text-3} are displayed because @code{.(}
8250: is an immediate word; it behaves in the same way whether it is used inside
8251: or outside a colon definition.
8252: @item
8253: Message @code{text-4} is displayed because of Gforth's added interpretation
8254: semantics for @code{."}.
8255: @item
1.29 crook 8256: Message @code{text-2} is @i{not} displayed, because the text interpreter
1.26 crook 8257: performs the compilation semantics for @code{."} within the definition of
8258: @code{my-word}.
8259: @end itemize
1.5 anton 8260:
1.26 crook 8261: Here are some examples of executing @code{my-word} and @code{my-char}:
1.5 anton 8262:
1.26 crook 8263: @example
1.30 anton 8264: @kbd{my-word @key{RET}} text-2
1.26 crook 8265: ok
1.30 anton 8266: @kbd{my-char fred @key{RET}} Af ok
8267: @kbd{my-char jim @key{RET}} Aj ok
1.26 crook 8268: @end example
1.5 anton 8269:
8270: @itemize @bullet
8271: @item
1.26 crook 8272: Message @code{text-2} is displayed because of the run-time behaviour of
8273: @code{."}.
8274: @item
8275: @code{[char]} compiles the ``A'' from ``ALPHABET'' and puts its display code
8276: on the stack at run-time. @code{emit} always displays the character
8277: when @code{my-char} is executed.
8278: @item
8279: @code{char} parses a string at run-time and the second @code{emit} displays
8280: the first character of the string.
1.5 anton 8281: @item
1.26 crook 8282: If you type @code{see my-char} you can see that @code{[char]} discarded
8283: the text ``LPHABET'' and only compiled the display code for ``A'' into the
8284: definition of @code{my-char}.
1.5 anton 8285: @end itemize
8286:
8287:
8288:
1.48 anton 8289: @node Input, , Displaying characters and strings, Other I/O
1.26 crook 8290: @subsection Input
8291: @cindex input
1.28 crook 8292: @cindex I/O - see input
8293: @cindex parsing a string
1.5 anton 8294:
1.49 anton 8295: For ways of storing character strings in memory see @ref{String Formats}.
1.5 anton 8296:
1.27 crook 8297: @comment TODO examples for >number >float accept key key? pad parse word refill
1.29 crook 8298: @comment then index them
1.27 crook 8299:
1.44 crook 8300:
1.27 crook 8301: doc-key
8302: doc-key?
1.45 crook 8303: doc-ekey
8304: doc-ekey?
8305: doc-ekey>char
1.26 crook 8306: doc->number
8307: doc->float
8308: doc-accept
1.27 crook 8309: doc-pad
8310: doc-parse
8311: doc-word
8312: doc-sword
1.44 crook 8313: doc-(name)
1.27 crook 8314: doc-refill
8315: @comment obsolescent words..
8316: doc-convert
1.26 crook 8317: doc-query
8318: doc-expect
1.27 crook 8319: doc-span
1.5 anton 8320:
8321:
1.44 crook 8322:
1.5 anton 8323: @c -------------------------------------------------------------
1.26 crook 8324: @node Programming Tools, Assembler and Code Words, Other I/O, Words
8325: @section Programming Tools
8326: @cindex programming tools
1.12 anton 8327:
8328: @menu
1.26 crook 8329: * Debugging:: Simple and quick.
8330: * Assertions:: Making your programs self-checking.
1.46 pazsan 8331: * Singlestep Debugger:: Executing your program word by word.
1.5 anton 8332: @end menu
8333:
1.26 crook 8334: @node Debugging, Assertions, Programming Tools, Programming Tools
8335: @subsection Debugging
8336: @cindex debugging
1.5 anton 8337:
1.26 crook 8338: Languages with a slow edit/compile/link/test development loop tend to
8339: require sophisticated tracing/stepping debuggers to facilate
8340: productive debugging.
1.5 anton 8341:
1.26 crook 8342: A much better (faster) way in fast-compiling languages is to add
8343: printing code at well-selected places, let the program run, look at
8344: the output, see where things went wrong, add more printing code, etc.,
8345: until the bug is found.
1.5 anton 8346:
1.26 crook 8347: The simple debugging aids provided in @file{debugs.fs}
8348: are meant to support this style of debugging. In addition, there are
8349: words for non-destructively inspecting the stack and memory:
1.5 anton 8350:
1.44 crook 8351:
1.26 crook 8352: doc-.s
8353: doc-f.s
1.5 anton 8354:
1.44 crook 8355:
1.29 crook 8356: There is a word @code{.r} but it does @i{not} display the return
1.26 crook 8357: stack! It is used for formatted numeric output.
1.5 anton 8358:
1.44 crook 8359:
1.26 crook 8360: doc-depth
8361: doc-fdepth
8362: doc-clearstack
8363: doc-?
8364: doc-dump
1.5 anton 8365:
1.44 crook 8366:
1.26 crook 8367: The word @code{~~} prints debugging information (by default the source
8368: location and the stack contents). It is easy to insert. If you use Emacs
8369: it is also easy to remove (@kbd{C-x ~} in the Emacs Forth mode to
8370: query-replace them with nothing). The deferred words
8371: @code{printdebugdata} and @code{printdebugline} control the output of
8372: @code{~~}. The default source location output format works well with
8373: Emacs' compilation mode, so you can step through the program at the
8374: source level using @kbd{C-x `} (the advantage over a stepping debugger
8375: is that you can step in any direction and you know where the crash has
8376: happened or where the strange data has occurred).
1.5 anton 8377:
1.26 crook 8378: The default actions of @code{~~} clobber the contents of the pictured
8379: numeric output string, so you should not use @code{~~}, e.g., between
8380: @code{<#} and @code{#>}.
1.5 anton 8381:
1.44 crook 8382:
1.26 crook 8383: doc-~~
8384: doc-printdebugdata
8385: doc-printdebugline
1.5 anton 8386:
1.26 crook 8387: doc-see
8388: doc-marker
1.5 anton 8389:
1.44 crook 8390:
1.26 crook 8391: Here's an example of using @code{marker} at the start of a source file
8392: that you are debugging; it ensures that you only ever have one copy of
8393: the file's definitions compiled at any time:
1.5 anton 8394:
1.26 crook 8395: @example
8396: [IFDEF] my-code
8397: my-code
8398: [ENDIF]
1.5 anton 8399:
1.26 crook 8400: marker my-code
1.28 crook 8401: init-included-files
1.5 anton 8402:
1.26 crook 8403: \ .. definitions start here
8404: \ .
8405: \ .
8406: \ end
8407: @end example
1.5 anton 8408:
8409:
8410:
1.26 crook 8411: @node Assertions, Singlestep Debugger, Debugging, Programming Tools
8412: @subsection Assertions
8413: @cindex assertions
1.5 anton 8414:
1.26 crook 8415: It is a good idea to make your programs self-checking, especially if you
8416: make an assumption that may become invalid during maintenance (for
8417: example, that a certain field of a data structure is never zero). Gforth
1.29 crook 8418: supports @dfn{assertions} for this purpose. They are used like this:
1.23 crook 8419:
1.26 crook 8420: @example
1.29 crook 8421: assert( @i{flag} )
1.26 crook 8422: @end example
1.23 crook 8423:
1.26 crook 8424: The code between @code{assert(} and @code{)} should compute a flag, that
8425: should be true if everything is alright and false otherwise. It should
8426: not change anything else on the stack. The overall stack effect of the
8427: assertion is @code{( -- )}. E.g.
1.23 crook 8428:
1.26 crook 8429: @example
8430: assert( 1 1 + 2 = ) \ what we learn in school
8431: assert( dup 0<> ) \ assert that the top of stack is not zero
8432: assert( false ) \ this code should not be reached
8433: @end example
1.23 crook 8434:
1.26 crook 8435: The need for assertions is different at different times. During
8436: debugging, we want more checking, in production we sometimes care more
8437: for speed. Therefore, assertions can be turned off, i.e., the assertion
8438: becomes a comment. Depending on the importance of an assertion and the
8439: time it takes to check it, you may want to turn off some assertions and
8440: keep others turned on. Gforth provides several levels of assertions for
8441: this purpose:
1.23 crook 8442:
1.44 crook 8443:
1.26 crook 8444: doc-assert0(
8445: doc-assert1(
8446: doc-assert2(
8447: doc-assert3(
8448: doc-assert(
8449: doc-)
1.23 crook 8450:
1.44 crook 8451:
1.26 crook 8452: The variable @code{assert-level} specifies the highest assertions that
8453: are turned on. I.e., at the default @code{assert-level} of one,
8454: @code{assert0(} and @code{assert1(} assertions perform checking, while
8455: @code{assert2(} and @code{assert3(} assertions are treated as comments.
8456:
8457: The value of @code{assert-level} is evaluated at compile-time, not at
8458: run-time. Therefore you cannot turn assertions on or off at run-time;
8459: you have to set the @code{assert-level} appropriately before compiling a
8460: piece of code. You can compile different pieces of code at different
8461: @code{assert-level}s (e.g., a trusted library at level 1 and
8462: newly-written code at level 3).
1.23 crook 8463:
1.44 crook 8464:
1.26 crook 8465: doc-assert-level
1.23 crook 8466:
1.44 crook 8467:
1.26 crook 8468: If an assertion fails, a message compatible with Emacs' compilation mode
8469: is produced and the execution is aborted (currently with @code{ABORT"}.
8470: If there is interest, we will introduce a special throw code. But if you
8471: intend to @code{catch} a specific condition, using @code{throw} is
8472: probably more appropriate than an assertion).
1.23 crook 8473:
1.26 crook 8474: Definitions in ANS Forth for these assertion words are provided
8475: in @file{compat/assert.fs}.
1.23 crook 8476:
8477:
1.48 anton 8478: @node Singlestep Debugger, , Assertions, Programming Tools
1.26 crook 8479: @subsection Singlestep Debugger
8480: @cindex singlestep Debugger
8481: @cindex debugging Singlestep
1.23 crook 8482:
1.26 crook 8483: When you create a new word there's often the need to check whether it
8484: behaves correctly or not. You can do this by typing @code{dbg
8485: badword}. A debug session might look like this:
1.23 crook 8486:
1.26 crook 8487: @example
8488: : badword 0 DO i . LOOP ; ok
8489: 2 dbg badword
8490: : badword
8491: Scanning code...
1.23 crook 8492:
1.26 crook 8493: Nesting debugger ready!
1.23 crook 8494:
1.26 crook 8495: 400D4738 8049BC4 0 -> [ 2 ] 00002 00000
8496: 400D4740 8049F68 DO -> [ 0 ]
8497: 400D4744 804A0C8 i -> [ 1 ] 00000
8498: 400D4748 400C5E60 . -> 0 [ 0 ]
8499: 400D474C 8049D0C LOOP -> [ 0 ]
8500: 400D4744 804A0C8 i -> [ 1 ] 00001
8501: 400D4748 400C5E60 . -> 1 [ 0 ]
8502: 400D474C 8049D0C LOOP -> [ 0 ]
8503: 400D4758 804B384 ; -> ok
8504: @end example
1.23 crook 8505:
1.26 crook 8506: Each line displayed is one step. You always have to hit return to
8507: execute the next word that is displayed. If you don't want to execute
8508: the next word in a whole, you have to type @kbd{n} for @code{nest}. Here is
8509: an overview what keys are available:
1.23 crook 8510:
1.26 crook 8511: @table @i
1.23 crook 8512:
1.30 anton 8513: @item @key{RET}
1.26 crook 8514: Next; Execute the next word.
1.23 crook 8515:
1.26 crook 8516: @item n
8517: Nest; Single step through next word.
1.5 anton 8518:
1.26 crook 8519: @item u
8520: Unnest; Stop debugging and execute rest of word. If we got to this word
8521: with nest, continue debugging with the calling word.
1.5 anton 8522:
1.26 crook 8523: @item d
8524: Done; Stop debugging and execute rest.
1.5 anton 8525:
1.26 crook 8526: @item s
8527: Stop; Abort immediately.
1.5 anton 8528:
1.26 crook 8529: @end table
1.5 anton 8530:
1.26 crook 8531: Debugging large application with this mechanism is very difficult, because
8532: you have to nest very deeply into the program before the interesting part
8533: begins. This takes a lot of time.
1.5 anton 8534:
1.26 crook 8535: To do it more directly put a @code{BREAK:} command into your source code.
8536: When program execution reaches @code{BREAK:} the single step debugger is
8537: invoked and you have all the features described above.
1.23 crook 8538:
1.26 crook 8539: If you have more than one part to debug it is useful to know where the
8540: program has stopped at the moment. You can do this by the
8541: @code{BREAK" string"} command. This behaves like @code{BREAK:} except that
8542: string is typed out when the ``breakpoint'' is reached.
8543:
1.44 crook 8544:
1.26 crook 8545: doc-dbg
1.45 crook 8546: doc-break:
8547: doc-break"
1.26 crook 8548:
8549:
1.44 crook 8550:
1.26 crook 8551: @c -------------------------------------------------------------
8552: @node Assembler and Code Words, Threading Words, Programming Tools, Words
8553: @section Assembler and Code Words
8554: @cindex assembler
8555: @cindex code words
1.5 anton 8556:
1.52 anton 8557: @menu
1.53 anton 8558: * Code and ;code::
8559: * Common Assembler:: Assembler Syntax
1.52 anton 8560: * Common Disassembler::
8561: * 386 Assembler:: Deviations and special cases
8562: * Alpha Assembler:: Deviations and special cases
8563: * MIPS assembler:: Deviations and special cases
1.53 anton 8564: * Other assemblers:: How to write them
1.52 anton 8565: @end menu
8566:
1.53 anton 8567: @node Code and ;code, Common Assembler, Assembler and Code Words, Assembler and Code Words
8568: @subsection @code{Code} and @code{;code}
1.52 anton 8569:
1.26 crook 8570: Gforth provides some words for defining primitives (words written in
1.29 crook 8571: machine code), and for defining the machine-code equivalent of
1.26 crook 8572: @code{DOES>}-based defining words. However, the machine-independent
8573: nature of Gforth poses a few problems: First of all, Gforth runs on
8574: several architectures, so it can provide no standard assembler. What's
8575: worse is that the register allocation not only depends on the processor,
8576: but also on the @code{gcc} version and options used.
1.5 anton 8577:
1.29 crook 8578: The words that Gforth offers encapsulate some system dependences (e.g.,
8579: the header structure), so a system-independent assembler may be used in
1.26 crook 8580: Gforth. If you do not have an assembler, you can compile machine code
1.29 crook 8581: directly with @code{,} and @code{c,}@footnote{This isn't portable,
8582: because these words emit stuff in @i{data} space; it works because
8583: Gforth has unified code/data spaces. Assembler isn't likely to be
8584: portable anyway.}.
1.5 anton 8585:
1.44 crook 8586:
1.26 crook 8587: doc-assembler
1.45 crook 8588: doc-init-asm
1.26 crook 8589: doc-code
8590: doc-end-code
8591: doc-;code
8592: doc-flush-icache
1.5 anton 8593:
1.44 crook 8594:
1.26 crook 8595: If @code{flush-icache} does not work correctly, @code{code} words
8596: etc. will not work (reliably), either.
1.5 anton 8597:
1.29 crook 8598: The typical usage of these @code{code} words can be shown most easily by
8599: analogy to the equivalent high-level defining words:
8600:
8601: @example
1.53 anton 8602: : foo code foo
8603: <high-level Forth words> <assembler>
8604: ; end-code
8605:
8606: : bar : bar
8607: <high-level Forth words> <high-level Forth words>
8608: CREATE CREATE
8609: <high-level Forth words> <high-level Forth words>
8610: DOES> ;code
8611: <high-level Forth words> <assembler>
8612: ; end-code
1.29 crook 8613: @end example
8614:
1.26 crook 8615: @code{flush-icache} is always present. The other words are rarely used
8616: and reside in @code{code.fs}, which is usually not loaded. You can load
8617: it with @code{require code.fs}.
1.5 anton 8618:
1.26 crook 8619: @cindex registers of the inner interpreter
8620: In the assembly code you will want to refer to the inner interpreter's
8621: registers (e.g., the data stack pointer) and you may want to use other
8622: registers for temporary storage. Unfortunately, the register allocation
8623: is installation-dependent.
1.5 anton 8624:
1.26 crook 8625: The easiest solution is to use explicit register declarations
8626: (@pxref{Explicit Reg Vars, , Variables in Specified Registers, gcc.info,
8627: GNU C Manual}) for all of the inner interpreter's registers: You have to
8628: compile Gforth with @code{-DFORCE_REG} (configure option
8629: @code{--enable-force-reg}) and the appropriate declarations must be
8630: present in the @code{machine.h} file (see @code{mips.h} for an example;
8631: you can find a full list of all declarable register symbols with
8632: @code{grep register engine.c}). If you give explicit registers to all
8633: variables that are declared at the beginning of @code{engine()}, you
8634: should be able to use the other caller-saved registers for temporary
8635: storage. Alternatively, you can use the @code{gcc} option
8636: @code{-ffixed-REG} (@pxref{Code Gen Options, , Options for Code
8637: Generation Conventions, gcc.info, GNU C Manual}) to reserve a register
8638: (however, this restriction on register allocation may slow Gforth
8639: significantly).
1.5 anton 8640:
1.26 crook 8641: If this solution is not viable (e.g., because @code{gcc} does not allow
8642: you to explicitly declare all the registers you need), you have to find
8643: out by looking at the code where the inner interpreter's registers
8644: reside and which registers can be used for temporary storage. You can
8645: get an assembly listing of the engine's code with @code{make engine.s}.
1.5 anton 8646:
1.26 crook 8647: In any case, it is good practice to abstract your assembly code from the
8648: actual register allocation. E.g., if the data stack pointer resides in
8649: register @code{$17}, create an alias for this register called @code{sp},
8650: and use that in your assembly code.
1.5 anton 8651:
1.26 crook 8652: @cindex code words, portable
8653: Another option for implementing normal and defining words efficiently
8654: is to add the desired functionality to the source of Gforth. For normal
8655: words you just have to edit @file{primitives} (@pxref{Automatic
8656: Generation}). Defining words (equivalent to @code{;CODE} words, for fast
8657: defined words) may require changes in @file{engine.c}, @file{kernel.fs},
8658: @file{prims2x.fs}, and possibly @file{cross.fs}.
1.5 anton 8659:
1.53 anton 8660: @node Common Assembler, Common Disassembler, Code and ;code, Assembler and Code Words
8661: @subsection Common Assembler
8662:
8663: The assemblers in Gforth generally use a postfix syntax, i.e., the
8664: instruction name follows the operands.
8665:
8666: The operands are passed in the usual order (the same that is used in the
8667: manual of the architecture). Since they all are Forth words, they have
8668: to be separated by spaces; you can also use Forth words to compute the
8669: operands.
8670:
8671: The instruction names usually end with a @code{,}. This makes it easier
8672: to visually separate instructions if you put several of them on one
8673: line; it also avoids shadowing other Forth words (e.g., @code{and}).
8674:
1.55 anton 8675: Registers are usually specified by number; e.g., (decimal) @code{11}
8676: specifies registers R11 and F11 on the Alpha architecture (which one,
8677: depends on the instruction). The usual names are also available, e.g.,
8678: @code{s2} for R11 on Alpha.
8679:
1.53 anton 8680: Control flow is specified similar to normal Forth code (@pxref{Arbitrary
8681: control structures}), with @code{if,}, @code{ahead,}, @code{then,},
8682: @code{begin,}, @code{until,}, @code{again,}, @code{cs-roll},
8683: @code{cs-pick}, @code{else,}, @code{while,}, and @code{repeat,}. The
8684: conditions are specified in a way specific to each assembler.
8685:
1.57 anton 8686: Note that the register assignments of the Gforth engine can change
8687: between Gforth versions, or even between different compilations of the
8688: same Gforth version (e.g., if you use a different GCC version). So if
8689: you want to refer to Gforth's registers (e.g., the stack pointer or
8690: TOS), I recommend defining your own words for refering to these
8691: registers, and using them later on; then you can easily adapt to a
8692: changed register assignment. The stability of the register assignment
8693: is usually better if you build Gforth with @code{--enable-force-reg}.
8694:
8695: In particular, the resturn stack pointer and the instruction pointer are
8696: in memory in @code{gforth}, and usually in registers in
8697: @code{gforth-fast}. The most common use of these registers is to
8698: dispatch to the next word (the @code{next} routine). A portable way to
8699: do this is to jump to @code{' noop >code-address} (of course, this is
8700: less efficient than integrating the @code{next} code and scheduling it
8701: well).
8702:
1.52 anton 8703: @node Common Disassembler, 386 Assembler, Common Assembler, Assembler and Code Words
8704: @subsection Common Disassembler
8705:
8706: You can disassemble a @code{code} word with @code{see}
8707: (@pxref{Debugging}). You can disassemble a section of memory with
8708:
8709: doc-disasm
8710:
8711: The disassembler generally produces output that can be fed into the
8712: assembler (i.e., same syntax, etc.). It also includes additional
1.53 anton 8713: information in comments. In particular, the address of the instruction
8714: is given in a comment before the instruction.
8715:
8716: @code{See} may display more or less than the actual code of the word,
8717: because the recognition of the end of the code is unreliable. You can
8718: use @code{disasm} if it did not display enough. It may display more, if
8719: the code word is not immediately followed by a named word. If you have
8720: something else there, you can follow the word with @code{align last @ ,}
8721: to ensure that the end is recognized.
1.52 anton 8722:
8723: @node 386 Assembler, Alpha Assembler, Common Disassembler, Assembler and Code Words
8724: @subsection 386 Assembler
8725:
8726: The 386 assembler and disassembler included in Gforth was written by
8727: Andrew McKewan; it is in the public domain.
1.57 anton 8728:
8729: The disassembler displays code in prefix Intel syntax.
8730:
8731: The assembler uses an Intel-inspired postfix syntax with reversed
8732: parameters. As usual, a @code{,} is appended to the instruction names
8733: (including @code{rep,} etc.).
8734:
8735: The assembler is somewhat meager, missing a number of instructions
8736: (including FP) and absolute memory addressing modes.
8737:
8738: The registers have their usual names @code{eax} etc. Immediate values
8739: are indicated by postfixing them with @code{#}, e.g., @code{3 #}. Here
8740: are some examples of addressing modes:
8741:
8742: @example
8743: 3 #
8744: eax
8745: 100 [edi]
8746: 4 [ebx] [ecx]
8747: 0 [edi] [eax] *4 \ base register required!
8748: @end example
8749:
8750: Some example of instructions are:
8751:
8752: @example
8753: EAX EBX MOV, \ move ebx,eax
8754: 3 # EAX MOV, \ mov eax,3
8755: 100 [EDI] EAX MOV, \ mov eax,100[edi]
8756: 4 [EBX] [ECX] EAX MOV, \ mov eax,4[ebx][ecx]
8757: 16: EAX EBX MOV, \ mov bx,ax
8758: @end example
8759:
8760: You cannot use the prefix @code{16:} with immediate operands. The
8761: following forms are supported for binary instructions:
8762:
8763: @example
8764: <reg> <reg> <inst>
8765: <n> # <reg> <inst>
8766: <mem> <reg> <inst>
8767: <reg> <mem> <inst>
8768: @end example
8769:
8770: Immediate to memory is not supported. The shift/rotate syntax is:
8771:
8772: @example
8773: <reg/mem> shl,
8774: <reg/mem> 4 shl,
8775: <reg/mem> cl shl,
8776: @end example
8777:
8778: Precede string instructions (@code{movs,} etc.) with @code{byte} to get
8779: the byte version.
8780:
8781: The control structure words @code{if,} @code{until,} etc. must be
8782: preceded by one of these conditions: @code{0= 0< u< u> < > ov ecx0<>}.
8783: You can invert the condition with @code{not} (Note that most of these
8784: words shadow some Forth words when @code{assembler} is before
8785: @code{forth} in the search path, e.g., in code words). Currently the
8786: control structure words use one stack item, so you have to use
8787: @code{roll} instead of @code{cs-roll} to shuffle them (you can also use
8788: @code{swap} etc.).
1.52 anton 8789:
8790: @node Alpha Assembler, MIPS assembler, 386 Assembler, Assembler and Code Words
8791: @subsection Alpha Assembler
8792:
1.55 anton 8793: The Alpha assembler and disassembler were originally written by Bernd
8794: Thallner.
8795:
8796: The register names @code{a0}--@code{a5} are not available to avoid
8797: shadowing hex numbers.
8798:
8799: Immediate forms of arithmetic instructions are distinguished by a
8800: @code{#} just before the @code{,}, e.g., @code{and#,} (note: @code{lda,}
8801: does not count as arithmetic instruction).
8802:
8803: You have to specify all operands to an instruction, even those that
8804: other assemblers consider optional, e.g., the destination register for
8805: @code{br,}, or the destination register and hint for @code{jmp,}.
8806:
8807: You can specify conditions for @code{if,} by removing the first @code{b}
8808: and the trailing @code{,} from a branch with a corresponding name; e.g.,
8809:
8810: @example
8811: 11 fgt if, \ if F11>0e
8812: ...
8813: endif,
1.56 anton 8814: @end example
1.55 anton 8815:
8816: @code{fbgt,} gives @code{fgt}.
1.52 anton 8817:
1.53 anton 8818: @node MIPS assembler, Other assemblers, Alpha Assembler, Assembler and Code Words
1.52 anton 8819: @subsection MIPS assembler
8820:
8821: The MIPS assembler was originally written by Christian Pirker.
8822:
8823: Currently the assembler and disassembler only cover the MIPS-I
8824: architecture (R3000), and don't support FP instructions.
8825:
1.55 anton 8826: The register names @code{$a0}--@code{$a3} are not available to avoid
8827: shadowing hex numbers.
1.52 anton 8828:
8829: Because there is no way to distinguish registers from immediate values,
8830: you have to explicitly use the immediate forms of instructions, i.e.,
8831: @code{addiu,}, not just @code{addu,} (@command{as} does this
8832: implicitly).
8833:
8834: If the architecture manual specifies several formats for the instruction
8835: (e.g., for @code{jalr,}), you usually have to use the one with more
8836: arguments (i.e., two for @code{jalr,}). When in doubt, see
8837: @code{arch/mips/testasm.fs} for an example of correct use.
8838:
1.53 anton 8839: Branches and jumps in the MIPS architecture have a delay slot. You have
8840: to fill it yourself (the simplest way is to use @code{nop,}), the
8841: assembler does not do it for you (unlike @command{as}). Even
8842: @code{if,}, @code{ahead,}, @code{until,}, @code{again,}, @code{while,},
8843: @code{else,} and @code{repeat,} need a delay slot. Since @code{begin,}
8844: and @code{then,} just specify branch targets, they are not affected.
8845:
8846: Note that you must not put branches, jumps, or @code{li,} into the delay
8847: slot: @code{li,} may expand to several instructions, and control flow
8848: instructions may not be put into the branch delay slot in any case.
1.52 anton 8849:
8850: For branches the argument specifying the target is a relative address;
8851: You have to add the address of the delay slot to get the absolute
8852: address.
1.53 anton 8853:
8854: The MIPS architecture also has load delay slots and restrictions on
8855: using @code{mfhi,} and @code{mflo,}; you have to order the instructions
8856: yourself to satisfy these restrictions, the assembler does not do it for
8857: you.
8858:
8859: You can specify the conditions for @code{if,} etc. by taking a
8860: conditional branch and leaving away the @code{b} at the start and the
8861: @code{,} at the end. E.g.,
8862:
8863: @example
8864: 4 5 eq if,
8865: ... \ do something if $4 equals $5
8866: then,
8867: @end example
8868:
8869: @node Other assemblers, , MIPS assembler, Assembler and Code Words
8870: @subsection Other assemblers
8871:
8872: If you want to contribute another assembler/disassembler, please contact
8873: us (@email{bug-gforth@@gnu.org}) to check if we have such an assembler
8874: already. If you are writing them from scratch, please use a similar
8875: syntax style as the one we use (i.e., postfix, commas at the end of the
8876: instruction names, @pxref{Common Assembler}); make the output of the
8877: disassembler be valid input for the assembler, and keep the style
8878: similar to the style we used.
8879:
8880: Hints on implementation: The most important part is to have a good test
8881: suite that contains all instructions. Once you have that, the rest is
8882: easy. For actual coding you can take a look at
8883: @file{arch/mips/disasm.fs} to get some ideas on how to use data for both
8884: the assembler and disassembler, avoiding redundancy and some potential
1.62 ! crook 8885: bugs. You can also look at that file (and @pxref{Advanced does> usage})
! 8886: to get ideas how to factor a disassembler.
1.5 anton 8887:
1.54 anton 8888: Start with the disassembler, because it's easier to reuse data from the
8889: disassembler for the assembler than the other way round.
8890:
8891: For the assembler, take a look at @file{arch/alpha/asm.fs}, which shows
8892: how simple it can be.
8893:
1.26 crook 8894: @c -------------------------------------------------------------
8895: @node Threading Words, Locals, Assembler and Code Words, Words
8896: @section Threading Words
8897: @cindex threading words
1.5 anton 8898:
1.26 crook 8899: @cindex code address
8900: These words provide access to code addresses and other threading stuff
8901: in Gforth (and, possibly, other interpretive Forths). It more or less
8902: abstracts away the differences between direct and indirect threading
8903: (and, for direct threading, the machine dependences). However, at
8904: present this wordset is still incomplete. It is also pretty low-level;
8905: some day it will hopefully be made unnecessary by an internals wordset
8906: that abstracts implementation details away completely.
1.5 anton 8907:
1.44 crook 8908:
1.26 crook 8909: doc-threading-method
8910: doc->code-address
8911: doc->does-code
8912: doc-code-address!
8913: doc-does-code!
8914: doc-does-handler!
8915: doc-/does-handler
1.5 anton 8916:
1.44 crook 8917:
1.26 crook 8918: The code addresses produced by various defining words are produced by
8919: the following words:
1.5 anton 8920:
1.44 crook 8921:
1.26 crook 8922: doc-docol:
8923: doc-docon:
8924: doc-dovar:
8925: doc-douser:
8926: doc-dodefer:
8927: doc-dofield:
1.5 anton 8928:
1.44 crook 8929:
1.26 crook 8930: You can recognize words defined by a @code{CREATE}...@code{DOES>} word
8931: with @code{>does-code}. If the word was defined in that way, the value
8932: returned is non-zero and identifies the @code{DOES>} used by the
8933: defining word.
8934: @comment TODO should that be ``identifies the xt of the DOES> ??''
1.5 anton 8935:
1.26 crook 8936: @c -------------------------------------------------------------
8937: @node Locals, Structures, Threading Words, Words
8938: @section Locals
8939: @cindex locals
1.5 anton 8940:
1.26 crook 8941: Local variables can make Forth programming more enjoyable and Forth
8942: programs easier to read. Unfortunately, the locals of ANS Forth are
8943: laden with restrictions. Therefore, we provide not only the ANS Forth
8944: locals wordset, but also our own, more powerful locals wordset (we
8945: implemented the ANS Forth locals wordset through our locals wordset).
1.5 anton 8946:
1.26 crook 8947: The ideas in this section have also been published in the paper
8948: @cite{Automatic Scoping of Local Variables} by M. Anton Ertl, presented
8949: at EuroForth '94; it is available at
1.47 crook 8950: @*@uref{http://www.complang.tuwien.ac.at/papers/ertl94l.ps.gz}.
1.5 anton 8951:
1.26 crook 8952: @menu
8953: * Gforth locals::
8954: * ANS Forth locals::
8955: @end menu
1.5 anton 8956:
1.26 crook 8957: @node Gforth locals, ANS Forth locals, Locals, Locals
8958: @subsection Gforth locals
8959: @cindex Gforth locals
8960: @cindex locals, Gforth style
1.5 anton 8961:
1.26 crook 8962: Locals can be defined with
1.5 anton 8963:
8964: @example
1.26 crook 8965: @{ local1 local2 ... -- comment @}
8966: @end example
8967: or
8968: @example
8969: @{ local1 local2 ... @}
1.5 anton 8970: @end example
8971:
1.26 crook 8972: E.g.,
1.5 anton 8973: @example
1.26 crook 8974: : max @{ n1 n2 -- n3 @}
8975: n1 n2 > if
8976: n1
8977: else
8978: n2
8979: endif ;
1.5 anton 8980: @end example
8981:
1.26 crook 8982: The similarity of locals definitions with stack comments is intended. A
8983: locals definition often replaces the stack comment of a word. The order
8984: of the locals corresponds to the order in a stack comment and everything
8985: after the @code{--} is really a comment.
1.5 anton 8986:
1.26 crook 8987: This similarity has one disadvantage: It is too easy to confuse locals
8988: declarations with stack comments, causing bugs and making them hard to
8989: find. However, this problem can be avoided by appropriate coding
8990: conventions: Do not use both notations in the same program. If you do,
8991: they should be distinguished using additional means, e.g. by position.
8992:
8993: @cindex types of locals
8994: @cindex locals types
8995: The name of the local may be preceded by a type specifier, e.g.,
8996: @code{F:} for a floating point value:
8997:
8998: @example
8999: : CX* @{ F: Ar F: Ai F: Br F: Bi -- Cr Ci @}
9000: \ complex multiplication
9001: Ar Br f* Ai Bi f* f-
9002: Ar Bi f* Ai Br f* f+ ;
9003: @end example
9004:
9005: @cindex flavours of locals
9006: @cindex locals flavours
9007: @cindex value-flavoured locals
9008: @cindex variable-flavoured locals
9009: Gforth currently supports cells (@code{W:}, @code{W^}), doubles
9010: (@code{D:}, @code{D^}), floats (@code{F:}, @code{F^}) and characters
9011: (@code{C:}, @code{C^}) in two flavours: a value-flavoured local (defined
9012: with @code{W:}, @code{D:} etc.) produces its value and can be changed
9013: with @code{TO}. A variable-flavoured local (defined with @code{W^} etc.)
9014: produces its address (which becomes invalid when the variable's scope is
9015: left). E.g., the standard word @code{emit} can be defined in terms of
9016: @code{type} like this:
1.5 anton 9017:
9018: @example
1.26 crook 9019: : emit @{ C^ char* -- @}
9020: char* 1 type ;
1.5 anton 9021: @end example
9022:
1.26 crook 9023: @cindex default type of locals
9024: @cindex locals, default type
9025: A local without type specifier is a @code{W:} local. Both flavours of
9026: locals are initialized with values from the data or FP stack.
1.5 anton 9027:
1.26 crook 9028: Currently there is no way to define locals with user-defined data
9029: structures, but we are working on it.
1.5 anton 9030:
1.26 crook 9031: Gforth allows defining locals everywhere in a colon definition. This
9032: poses the following questions:
1.5 anton 9033:
1.26 crook 9034: @menu
9035: * Where are locals visible by name?::
9036: * How long do locals live?::
9037: * Programming Style::
9038: * Implementation::
9039: @end menu
1.5 anton 9040:
1.26 crook 9041: @node Where are locals visible by name?, How long do locals live?, Gforth locals, Gforth locals
9042: @subsubsection Where are locals visible by name?
9043: @cindex locals visibility
9044: @cindex visibility of locals
9045: @cindex scope of locals
1.5 anton 9046:
1.26 crook 9047: Basically, the answer is that locals are visible where you would expect
9048: it in block-structured languages, and sometimes a little longer. If you
9049: want to restrict the scope of a local, enclose its definition in
9050: @code{SCOPE}...@code{ENDSCOPE}.
1.5 anton 9051:
1.44 crook 9052:
1.26 crook 9053: doc-scope
9054: doc-endscope
1.5 anton 9055:
1.44 crook 9056:
1.26 crook 9057: These words behave like control structure words, so you can use them
9058: with @code{CS-PICK} and @code{CS-ROLL} to restrict the scope in
9059: arbitrary ways.
1.5 anton 9060:
1.26 crook 9061: If you want a more exact answer to the visibility question, here's the
9062: basic principle: A local is visible in all places that can only be
9063: reached through the definition of the local@footnote{In compiler
9064: construction terminology, all places dominated by the definition of the
9065: local.}. In other words, it is not visible in places that can be reached
9066: without going through the definition of the local. E.g., locals defined
9067: in @code{IF}...@code{ENDIF} are visible until the @code{ENDIF}, locals
9068: defined in @code{BEGIN}...@code{UNTIL} are visible after the
9069: @code{UNTIL} (until, e.g., a subsequent @code{ENDSCOPE}).
1.5 anton 9070:
1.26 crook 9071: The reasoning behind this solution is: We want to have the locals
9072: visible as long as it is meaningful. The user can always make the
9073: visibility shorter by using explicit scoping. In a place that can
9074: only be reached through the definition of a local, the meaning of a
9075: local name is clear. In other places it is not: How is the local
9076: initialized at the control flow path that does not contain the
9077: definition? Which local is meant, if the same name is defined twice in
9078: two independent control flow paths?
1.5 anton 9079:
1.26 crook 9080: This should be enough detail for nearly all users, so you can skip the
9081: rest of this section. If you really must know all the gory details and
9082: options, read on.
1.5 anton 9083:
1.26 crook 9084: In order to implement this rule, the compiler has to know which places
9085: are unreachable. It knows this automatically after @code{AHEAD},
9086: @code{AGAIN}, @code{EXIT} and @code{LEAVE}; in other cases (e.g., after
9087: most @code{THROW}s), you can use the word @code{UNREACHABLE} to tell the
9088: compiler that the control flow never reaches that place. If
9089: @code{UNREACHABLE} is not used where it could, the only consequence is
9090: that the visibility of some locals is more limited than the rule above
9091: says. If @code{UNREACHABLE} is used where it should not (i.e., if you
9092: lie to the compiler), buggy code will be produced.
1.5 anton 9093:
1.44 crook 9094:
1.26 crook 9095: doc-unreachable
1.5 anton 9096:
1.44 crook 9097:
1.26 crook 9098: Another problem with this rule is that at @code{BEGIN}, the compiler
9099: does not know which locals will be visible on the incoming
9100: back-edge. All problems discussed in the following are due to this
9101: ignorance of the compiler (we discuss the problems using @code{BEGIN}
9102: loops as examples; the discussion also applies to @code{?DO} and other
9103: loops). Perhaps the most insidious example is:
1.5 anton 9104: @example
1.26 crook 9105: AHEAD
9106: BEGIN
9107: x
9108: [ 1 CS-ROLL ] THEN
9109: @{ x @}
9110: ...
9111: UNTIL
9112: @end example
1.5 anton 9113:
1.26 crook 9114: This should be legal according to the visibility rule. The use of
9115: @code{x} can only be reached through the definition; but that appears
9116: textually below the use.
1.5 anton 9117:
1.26 crook 9118: From this example it is clear that the visibility rules cannot be fully
9119: implemented without major headaches. Our implementation treats common
9120: cases as advertised and the exceptions are treated in a safe way: The
9121: compiler makes a reasonable guess about the locals visible after a
9122: @code{BEGIN}; if it is too pessimistic, the
9123: user will get a spurious error about the local not being defined; if the
9124: compiler is too optimistic, it will notice this later and issue a
9125: warning. In the case above the compiler would complain about @code{x}
9126: being undefined at its use. You can see from the obscure examples in
9127: this section that it takes quite unusual control structures to get the
9128: compiler into trouble, and even then it will often do fine.
1.5 anton 9129:
1.26 crook 9130: If the @code{BEGIN} is reachable from above, the most optimistic guess
9131: is that all locals visible before the @code{BEGIN} will also be
9132: visible after the @code{BEGIN}. This guess is valid for all loops that
9133: are entered only through the @code{BEGIN}, in particular, for normal
9134: @code{BEGIN}...@code{WHILE}...@code{REPEAT} and
9135: @code{BEGIN}...@code{UNTIL} loops and it is implemented in our
9136: compiler. When the branch to the @code{BEGIN} is finally generated by
9137: @code{AGAIN} or @code{UNTIL}, the compiler checks the guess and
9138: warns the user if it was too optimistic:
9139: @example
9140: IF
9141: @{ x @}
9142: BEGIN
9143: \ x ?
9144: [ 1 cs-roll ] THEN
9145: ...
9146: UNTIL
1.5 anton 9147: @end example
9148:
1.26 crook 9149: Here, @code{x} lives only until the @code{BEGIN}, but the compiler
9150: optimistically assumes that it lives until the @code{THEN}. It notices
9151: this difference when it compiles the @code{UNTIL} and issues a
9152: warning. The user can avoid the warning, and make sure that @code{x}
9153: is not used in the wrong area by using explicit scoping:
9154: @example
9155: IF
9156: SCOPE
9157: @{ x @}
9158: ENDSCOPE
9159: BEGIN
9160: [ 1 cs-roll ] THEN
9161: ...
9162: UNTIL
9163: @end example
1.5 anton 9164:
1.26 crook 9165: Since the guess is optimistic, there will be no spurious error messages
9166: about undefined locals.
1.5 anton 9167:
1.26 crook 9168: If the @code{BEGIN} is not reachable from above (e.g., after
9169: @code{AHEAD} or @code{EXIT}), the compiler cannot even make an
9170: optimistic guess, as the locals visible after the @code{BEGIN} may be
9171: defined later. Therefore, the compiler assumes that no locals are
9172: visible after the @code{BEGIN}. However, the user can use
9173: @code{ASSUME-LIVE} to make the compiler assume that the same locals are
9174: visible at the BEGIN as at the point where the top control-flow stack
9175: item was created.
1.5 anton 9176:
1.44 crook 9177:
1.26 crook 9178: doc-assume-live
1.5 anton 9179:
1.44 crook 9180:
9181: @noindent
1.26 crook 9182: E.g.,
1.5 anton 9183: @example
1.26 crook 9184: @{ x @}
9185: AHEAD
9186: ASSUME-LIVE
9187: BEGIN
9188: x
9189: [ 1 CS-ROLL ] THEN
9190: ...
9191: UNTIL
1.5 anton 9192: @end example
9193:
1.26 crook 9194: Other cases where the locals are defined before the @code{BEGIN} can be
9195: handled by inserting an appropriate @code{CS-ROLL} before the
9196: @code{ASSUME-LIVE} (and changing the control-flow stack manipulation
9197: behind the @code{ASSUME-LIVE}).
1.5 anton 9198:
1.26 crook 9199: Cases where locals are defined after the @code{BEGIN} (but should be
9200: visible immediately after the @code{BEGIN}) can only be handled by
9201: rearranging the loop. E.g., the ``most insidious'' example above can be
9202: arranged into:
1.5 anton 9203: @example
1.26 crook 9204: BEGIN
9205: @{ x @}
9206: ... 0=
9207: WHILE
9208: x
9209: REPEAT
1.5 anton 9210: @end example
9211:
1.26 crook 9212: @node How long do locals live?, Programming Style, Where are locals visible by name?, Gforth locals
9213: @subsubsection How long do locals live?
9214: @cindex locals lifetime
9215: @cindex lifetime of locals
1.5 anton 9216:
1.26 crook 9217: The right answer for the lifetime question would be: A local lives at
9218: least as long as it can be accessed. For a value-flavoured local this
9219: means: until the end of its visibility. However, a variable-flavoured
9220: local could be accessed through its address far beyond its visibility
9221: scope. Ultimately, this would mean that such locals would have to be
9222: garbage collected. Since this entails un-Forth-like implementation
9223: complexities, I adopted the same cowardly solution as some other
9224: languages (e.g., C): The local lives only as long as it is visible;
9225: afterwards its address is invalid (and programs that access it
9226: afterwards are erroneous).
1.5 anton 9227:
1.26 crook 9228: @node Programming Style, Implementation, How long do locals live?, Gforth locals
9229: @subsubsection Programming Style
9230: @cindex locals programming style
9231: @cindex programming style, locals
1.5 anton 9232:
1.26 crook 9233: The freedom to define locals anywhere has the potential to change
9234: programming styles dramatically. In particular, the need to use the
9235: return stack for intermediate storage vanishes. Moreover, all stack
9236: manipulations (except @code{PICK}s and @code{ROLL}s with run-time
9237: determined arguments) can be eliminated: If the stack items are in the
9238: wrong order, just write a locals definition for all of them; then
9239: write the items in the order you want.
1.5 anton 9240:
1.26 crook 9241: This seems a little far-fetched and eliminating stack manipulations is
9242: unlikely to become a conscious programming objective. Still, the number
9243: of stack manipulations will be reduced dramatically if local variables
1.49 anton 9244: are used liberally (e.g., compare @code{max} (@pxref{Gforth locals}) with
1.26 crook 9245: a traditional implementation of @code{max}).
1.5 anton 9246:
1.26 crook 9247: This shows one potential benefit of locals: making Forth programs more
9248: readable. Of course, this benefit will only be realized if the
9249: programmers continue to honour the principle of factoring instead of
9250: using the added latitude to make the words longer.
1.5 anton 9251:
1.26 crook 9252: @cindex single-assignment style for locals
9253: Using @code{TO} can and should be avoided. Without @code{TO},
9254: every value-flavoured local has only a single assignment and many
9255: advantages of functional languages apply to Forth. I.e., programs are
9256: easier to analyse, to optimize and to read: It is clear from the
9257: definition what the local stands for, it does not turn into something
9258: different later.
1.5 anton 9259:
1.26 crook 9260: E.g., a definition using @code{TO} might look like this:
1.5 anton 9261: @example
1.26 crook 9262: : strcmp @{ addr1 u1 addr2 u2 -- n @}
9263: u1 u2 min 0
9264: ?do
9265: addr1 c@@ addr2 c@@ -
9266: ?dup-if
9267: unloop exit
9268: then
9269: addr1 char+ TO addr1
9270: addr2 char+ TO addr2
9271: loop
9272: u1 u2 - ;
1.5 anton 9273: @end example
1.26 crook 9274: Here, @code{TO} is used to update @code{addr1} and @code{addr2} at
9275: every loop iteration. @code{strcmp} is a typical example of the
9276: readability problems of using @code{TO}. When you start reading
9277: @code{strcmp}, you think that @code{addr1} refers to the start of the
9278: string. Only near the end of the loop you realize that it is something
9279: else.
1.5 anton 9280:
1.26 crook 9281: This can be avoided by defining two locals at the start of the loop that
9282: are initialized with the right value for the current iteration.
1.5 anton 9283: @example
1.26 crook 9284: : strcmp @{ addr1 u1 addr2 u2 -- n @}
9285: addr1 addr2
9286: u1 u2 min 0
9287: ?do @{ s1 s2 @}
9288: s1 c@@ s2 c@@ -
9289: ?dup-if
9290: unloop exit
9291: then
9292: s1 char+ s2 char+
9293: loop
9294: 2drop
9295: u1 u2 - ;
1.5 anton 9296: @end example
1.26 crook 9297: Here it is clear from the start that @code{s1} has a different value
9298: in every loop iteration.
1.5 anton 9299:
1.26 crook 9300: @node Implementation, , Programming Style, Gforth locals
9301: @subsubsection Implementation
9302: @cindex locals implementation
9303: @cindex implementation of locals
1.5 anton 9304:
1.26 crook 9305: @cindex locals stack
9306: Gforth uses an extra locals stack. The most compelling reason for
9307: this is that the return stack is not float-aligned; using an extra stack
9308: also eliminates the problems and restrictions of using the return stack
9309: as locals stack. Like the other stacks, the locals stack grows toward
9310: lower addresses. A few primitives allow an efficient implementation:
1.5 anton 9311:
1.44 crook 9312:
1.26 crook 9313: doc-@local#
9314: doc-f@local#
9315: doc-laddr#
9316: doc-lp+!#
9317: doc-lp!
9318: doc->l
9319: doc-f>l
1.5 anton 9320:
1.44 crook 9321:
1.26 crook 9322: In addition to these primitives, some specializations of these
9323: primitives for commonly occurring inline arguments are provided for
9324: efficiency reasons, e.g., @code{@@local0} as specialization of
9325: @code{@@local#} for the inline argument 0. The following compiling words
9326: compile the right specialized version, or the general version, as
9327: appropriate:
1.6 pazsan 9328:
1.44 crook 9329:
1.26 crook 9330: doc-compile-@local
9331: doc-compile-f@local
9332: doc-compile-lp+!
1.12 anton 9333:
1.44 crook 9334:
1.26 crook 9335: Combinations of conditional branches and @code{lp+!#} like
9336: @code{?branch-lp+!#} (the locals pointer is only changed if the branch
9337: is taken) are provided for efficiency and correctness in loops.
1.6 pazsan 9338:
1.26 crook 9339: A special area in the dictionary space is reserved for keeping the
9340: local variable names. @code{@{} switches the dictionary pointer to this
9341: area and @code{@}} switches it back and generates the locals
9342: initializing code. @code{W:} etc.@ are normal defining words. This
9343: special area is cleared at the start of every colon definition.
1.6 pazsan 9344:
1.26 crook 9345: @cindex word list for defining locals
9346: A special feature of Gforth's dictionary is used to implement the
9347: definition of locals without type specifiers: every word list (aka
9348: vocabulary) has its own methods for searching
9349: etc. (@pxref{Word Lists}). For the present purpose we defined a word list
9350: with a special search method: When it is searched for a word, it
9351: actually creates that word using @code{W:}. @code{@{} changes the search
9352: order to first search the word list containing @code{@}}, @code{W:} etc.,
9353: and then the word list for defining locals without type specifiers.
1.12 anton 9354:
1.26 crook 9355: The lifetime rules support a stack discipline within a colon
9356: definition: The lifetime of a local is either nested with other locals
9357: lifetimes or it does not overlap them.
1.6 pazsan 9358:
1.26 crook 9359: At @code{BEGIN}, @code{IF}, and @code{AHEAD} no code for locals stack
9360: pointer manipulation is generated. Between control structure words
9361: locals definitions can push locals onto the locals stack. @code{AGAIN}
9362: is the simplest of the other three control flow words. It has to
9363: restore the locals stack depth of the corresponding @code{BEGIN}
9364: before branching. The code looks like this:
9365: @format
9366: @code{lp+!#} current-locals-size @minus{} dest-locals-size
9367: @code{branch} <begin>
9368: @end format
1.6 pazsan 9369:
1.26 crook 9370: @code{UNTIL} is a little more complicated: If it branches back, it
9371: must adjust the stack just like @code{AGAIN}. But if it falls through,
9372: the locals stack must not be changed. The compiler generates the
9373: following code:
9374: @format
9375: @code{?branch-lp+!#} <begin> current-locals-size @minus{} dest-locals-size
9376: @end format
9377: The locals stack pointer is only adjusted if the branch is taken.
1.6 pazsan 9378:
1.26 crook 9379: @code{THEN} can produce somewhat inefficient code:
9380: @format
9381: @code{lp+!#} current-locals-size @minus{} orig-locals-size
9382: <orig target>:
9383: @code{lp+!#} orig-locals-size @minus{} new-locals-size
9384: @end format
9385: The second @code{lp+!#} adjusts the locals stack pointer from the
1.29 crook 9386: level at the @i{orig} point to the level after the @code{THEN}. The
1.26 crook 9387: first @code{lp+!#} adjusts the locals stack pointer from the current
9388: level to the level at the orig point, so the complete effect is an
9389: adjustment from the current level to the right level after the
9390: @code{THEN}.
1.6 pazsan 9391:
1.26 crook 9392: @cindex locals information on the control-flow stack
9393: @cindex control-flow stack items, locals information
9394: In a conventional Forth implementation a dest control-flow stack entry
9395: is just the target address and an orig entry is just the address to be
9396: patched. Our locals implementation adds a word list to every orig or dest
9397: item. It is the list of locals visible (or assumed visible) at the point
9398: described by the entry. Our implementation also adds a tag to identify
9399: the kind of entry, in particular to differentiate between live and dead
9400: (reachable and unreachable) orig entries.
1.6 pazsan 9401:
1.26 crook 9402: A few unusual operations have to be performed on locals word lists:
1.6 pazsan 9403:
1.44 crook 9404:
1.26 crook 9405: doc-common-list
9406: doc-sub-list?
9407: doc-list-size
1.6 pazsan 9408:
1.44 crook 9409:
1.26 crook 9410: Several features of our locals word list implementation make these
9411: operations easy to implement: The locals word lists are organised as
9412: linked lists; the tails of these lists are shared, if the lists
9413: contain some of the same locals; and the address of a name is greater
9414: than the address of the names behind it in the list.
1.6 pazsan 9415:
1.26 crook 9416: Another important implementation detail is the variable
9417: @code{dead-code}. It is used by @code{BEGIN} and @code{THEN} to
9418: determine if they can be reached directly or only through the branch
9419: that they resolve. @code{dead-code} is set by @code{UNREACHABLE},
9420: @code{AHEAD}, @code{EXIT} etc., and cleared at the start of a colon
9421: definition, by @code{BEGIN} and usually by @code{THEN}.
1.6 pazsan 9422:
1.26 crook 9423: Counted loops are similar to other loops in most respects, but
9424: @code{LEAVE} requires special attention: It performs basically the same
9425: service as @code{AHEAD}, but it does not create a control-flow stack
9426: entry. Therefore the information has to be stored elsewhere;
9427: traditionally, the information was stored in the target fields of the
9428: branches created by the @code{LEAVE}s, by organizing these fields into a
9429: linked list. Unfortunately, this clever trick does not provide enough
9430: space for storing our extended control flow information. Therefore, we
9431: introduce another stack, the leave stack. It contains the control-flow
9432: stack entries for all unresolved @code{LEAVE}s.
1.6 pazsan 9433:
1.26 crook 9434: Local names are kept until the end of the colon definition, even if
9435: they are no longer visible in any control-flow path. In a few cases
9436: this may lead to increased space needs for the locals name area, but
9437: usually less than reclaiming this space would cost in code size.
1.6 pazsan 9438:
9439:
1.26 crook 9440: @node ANS Forth locals, , Gforth locals, Locals
9441: @subsection ANS Forth locals
9442: @cindex locals, ANS Forth style
1.6 pazsan 9443:
1.26 crook 9444: The ANS Forth locals wordset does not define a syntax for locals, but
9445: words that make it possible to define various syntaxes. One of the
9446: possible syntaxes is a subset of the syntax we used in the Gforth locals
9447: wordset, i.e.:
1.6 pazsan 9448:
9449: @example
1.26 crook 9450: @{ local1 local2 ... -- comment @}
1.6 pazsan 9451: @end example
1.23 crook 9452: @noindent
1.26 crook 9453: or
1.6 pazsan 9454: @example
1.26 crook 9455: @{ local1 local2 ... @}
1.6 pazsan 9456: @end example
9457:
1.26 crook 9458: The order of the locals corresponds to the order in a stack comment. The
9459: restrictions are:
1.6 pazsan 9460:
9461: @itemize @bullet
9462: @item
1.26 crook 9463: Locals can only be cell-sized values (no type specifiers are allowed).
1.6 pazsan 9464: @item
1.26 crook 9465: Locals can be defined only outside control structures.
1.6 pazsan 9466: @item
1.26 crook 9467: Locals can interfere with explicit usage of the return stack. For the
9468: exact (and long) rules, see the standard. If you don't use return stack
9469: accessing words in a definition using locals, you will be all right. The
9470: purpose of this rule is to make locals implementation on the return
9471: stack easier.
1.6 pazsan 9472: @item
1.26 crook 9473: The whole definition must be in one line.
9474: @end itemize
1.6 pazsan 9475:
1.44 crook 9476: Locals defined in this way behave like @code{VALUE}s
1.49 anton 9477: (@pxref{Values}). I.e., they are initialized from the stack. Using their
1.26 crook 9478: name produces their value. Their value can be changed using @code{TO}.
1.6 pazsan 9479:
1.26 crook 9480: Since this syntax is supported by Gforth directly, you need not do
9481: anything to use it. If you want to port a program using this syntax to
9482: another ANS Forth system, use @file{compat/anslocal.fs} to implement the
9483: syntax on the other system.
1.6 pazsan 9484:
1.26 crook 9485: Note that a syntax shown in the standard, section A.13 looks
9486: similar, but is quite different in having the order of locals
9487: reversed. Beware!
1.6 pazsan 9488:
1.26 crook 9489: The ANS Forth locals wordset itself consists of a word:
1.6 pazsan 9490:
1.44 crook 9491:
1.26 crook 9492: doc-(local)
1.6 pazsan 9493:
1.44 crook 9494:
1.26 crook 9495: The ANS Forth locals extension wordset defines a syntax using @code{locals|}, but it is so
9496: awful that we strongly recommend not to use it. We have implemented this
9497: syntax to make porting to Gforth easy, but do not document it here. The
9498: problem with this syntax is that the locals are defined in an order
9499: reversed with respect to the standard stack comment notation, making
9500: programs harder to read, and easier to misread and miswrite. The only
9501: merit of this syntax is that it is easy to implement using the ANS Forth
9502: locals wordset.
1.7 pazsan 9503:
9504:
1.26 crook 9505: @c ----------------------------------------------------------
9506: @node Structures, Object-oriented Forth, Locals, Words
9507: @section Structures
9508: @cindex structures
9509: @cindex records
1.7 pazsan 9510:
1.26 crook 9511: This section presents the structure package that comes with Gforth. A
9512: version of the package implemented in ANS Forth is available in
9513: @file{compat/struct.fs}. This package was inspired by a posting on
9514: comp.lang.forth in 1989 (unfortunately I don't remember, by whom;
9515: possibly John Hayes). A version of this section has been published in
9516: ???. Marcel Hendrix provided helpful comments.
1.7 pazsan 9517:
1.26 crook 9518: @menu
9519: * Why explicit structure support?::
9520: * Structure Usage::
9521: * Structure Naming Convention::
9522: * Structure Implementation::
9523: * Structure Glossary::
9524: @end menu
1.7 pazsan 9525:
1.26 crook 9526: @node Why explicit structure support?, Structure Usage, Structures, Structures
9527: @subsection Why explicit structure support?
1.7 pazsan 9528:
1.26 crook 9529: @cindex address arithmetic for structures
9530: @cindex structures using address arithmetic
9531: If we want to use a structure containing several fields, we could simply
9532: reserve memory for it, and access the fields using address arithmetic
1.32 anton 9533: (@pxref{Address arithmetic}). As an example, consider a structure with
1.26 crook 9534: the following fields
1.7 pazsan 9535:
1.26 crook 9536: @table @code
9537: @item a
9538: is a float
9539: @item b
9540: is a cell
9541: @item c
9542: is a float
9543: @end table
1.7 pazsan 9544:
1.26 crook 9545: Given the (float-aligned) base address of the structure we get the
9546: address of the field
1.13 pazsan 9547:
1.26 crook 9548: @table @code
9549: @item a
9550: without doing anything further.
9551: @item b
9552: with @code{float+}
9553: @item c
9554: with @code{float+ cell+ faligned}
9555: @end table
1.13 pazsan 9556:
1.26 crook 9557: It is easy to see that this can become quite tiring.
1.13 pazsan 9558:
1.26 crook 9559: Moreover, it is not very readable, because seeing a
9560: @code{cell+} tells us neither which kind of structure is
9561: accessed nor what field is accessed; we have to somehow infer the kind
9562: of structure, and then look up in the documentation, which field of
9563: that structure corresponds to that offset.
1.13 pazsan 9564:
1.26 crook 9565: Finally, this kind of address arithmetic also causes maintenance
9566: troubles: If you add or delete a field somewhere in the middle of the
9567: structure, you have to find and change all computations for the fields
9568: afterwards.
1.13 pazsan 9569:
1.26 crook 9570: So, instead of using @code{cell+} and friends directly, how
9571: about storing the offsets in constants:
1.13 pazsan 9572:
9573: @example
1.26 crook 9574: 0 constant a-offset
9575: 0 float+ constant b-offset
9576: 0 float+ cell+ faligned c-offset
1.13 pazsan 9577: @end example
9578:
1.26 crook 9579: Now we can get the address of field @code{x} with @code{x-offset
9580: +}. This is much better in all respects. Of course, you still
9581: have to change all later offset definitions if you add a field. You can
9582: fix this by declaring the offsets in the following way:
1.13 pazsan 9583:
9584: @example
1.26 crook 9585: 0 constant a-offset
9586: a-offset float+ constant b-offset
9587: b-offset cell+ faligned constant c-offset
1.13 pazsan 9588: @end example
9589:
1.26 crook 9590: Since we always use the offsets with @code{+}, we could use a defining
9591: word @code{cfield} that includes the @code{+} in the action of the
9592: defined word:
1.8 pazsan 9593:
9594: @example
1.26 crook 9595: : cfield ( n "name" -- )
9596: create ,
9597: does> ( name execution: addr1 -- addr2 )
9598: @@ + ;
1.13 pazsan 9599:
1.26 crook 9600: 0 cfield a
9601: 0 a float+ cfield b
9602: 0 b cell+ faligned cfield c
1.13 pazsan 9603: @end example
9604:
1.26 crook 9605: Instead of @code{x-offset +}, we now simply write @code{x}.
9606:
9607: The structure field words now can be used quite nicely. However,
9608: their definition is still a bit cumbersome: We have to repeat the
9609: name, the information about size and alignment is distributed before
9610: and after the field definitions etc. The structure package presented
9611: here addresses these problems.
9612:
9613: @node Structure Usage, Structure Naming Convention, Why explicit structure support?, Structures
9614: @subsection Structure Usage
9615: @cindex structure usage
1.13 pazsan 9616:
1.26 crook 9617: @cindex @code{field} usage
9618: @cindex @code{struct} usage
9619: @cindex @code{end-struct} usage
9620: You can define a structure for a (data-less) linked list with:
1.13 pazsan 9621: @example
1.26 crook 9622: struct
9623: cell% field list-next
9624: end-struct list%
1.13 pazsan 9625: @end example
9626:
1.26 crook 9627: With the address of the list node on the stack, you can compute the
9628: address of the field that contains the address of the next node with
9629: @code{list-next}. E.g., you can determine the length of a list
9630: with:
1.13 pazsan 9631:
9632: @example
1.26 crook 9633: : list-length ( list -- n )
9634: \ "list" is a pointer to the first element of a linked list
9635: \ "n" is the length of the list
9636: 0 BEGIN ( list1 n1 )
9637: over
9638: WHILE ( list1 n1 )
9639: 1+ swap list-next @@ swap
9640: REPEAT
9641: nip ;
1.13 pazsan 9642: @end example
9643:
1.26 crook 9644: You can reserve memory for a list node in the dictionary with
9645: @code{list% %allot}, which leaves the address of the list node on the
9646: stack. For the equivalent allocation on the heap you can use @code{list%
9647: %alloc} (or, for an @code{allocate}-like stack effect (i.e., with ior),
9648: use @code{list% %allocate}). You can get the the size of a list
9649: node with @code{list% %size} and its alignment with @code{list%
9650: %alignment}.
1.13 pazsan 9651:
1.26 crook 9652: Note that in ANS Forth the body of a @code{create}d word is
9653: @code{aligned} but not necessarily @code{faligned};
9654: therefore, if you do a:
1.13 pazsan 9655: @example
1.26 crook 9656: create @emph{name} foo% %allot
1.8 pazsan 9657: @end example
9658:
1.26 crook 9659: @noindent
9660: then the memory alloted for @code{foo%} is
9661: guaranteed to start at the body of @code{@emph{name}} only if
9662: @code{foo%} contains only character, cell and double fields.
1.20 pazsan 9663:
1.45 crook 9664: @cindex structures containing structures
1.26 crook 9665: You can include a structure @code{foo%} as a field of
9666: another structure, like this:
1.20 pazsan 9667: @example
1.26 crook 9668: struct
9669: ...
9670: foo% field ...
9671: ...
9672: end-struct ...
1.20 pazsan 9673: @end example
9674:
1.26 crook 9675: @cindex structure extension
9676: @cindex extended records
9677: Instead of starting with an empty structure, you can extend an
9678: existing structure. E.g., a plain linked list without data, as defined
9679: above, is hardly useful; You can extend it to a linked list of integers,
9680: like this:@footnote{This feature is also known as @emph{extended
9681: records}. It is the main innovation in the Oberon language; in other
9682: words, adding this feature to Modula-2 led Wirth to create a new
9683: language, write a new compiler etc. Adding this feature to Forth just
9684: required a few lines of code.}
1.20 pazsan 9685:
9686: @example
1.26 crook 9687: list%
9688: cell% field intlist-int
9689: end-struct intlist%
1.20 pazsan 9690: @end example
9691:
1.26 crook 9692: @code{intlist%} is a structure with two fields:
9693: @code{list-next} and @code{intlist-int}.
1.20 pazsan 9694:
1.26 crook 9695: @cindex structures containing arrays
9696: You can specify an array type containing @emph{n} elements of
9697: type @code{foo%} like this:
1.20 pazsan 9698:
9699: @example
1.26 crook 9700: foo% @emph{n} *
1.20 pazsan 9701: @end example
9702:
1.26 crook 9703: You can use this array type in any place where you can use a normal
9704: type, e.g., when defining a @code{field}, or with
9705: @code{%allot}.
1.20 pazsan 9706:
1.26 crook 9707: @cindex first field optimization
9708: The first field is at the base address of a structure and the word
9709: for this field (e.g., @code{list-next}) actually does not change
9710: the address on the stack. You may be tempted to leave it away in the
9711: interest of run-time and space efficiency. This is not necessary,
9712: because the structure package optimizes this case and compiling such
9713: words does not generate any code. So, in the interest of readability
9714: and maintainability you should include the word for the field when
9715: accessing the field.
1.20 pazsan 9716:
1.26 crook 9717: @node Structure Naming Convention, Structure Implementation, Structure Usage, Structures
9718: @subsection Structure Naming Convention
9719: @cindex structure naming convention
1.20 pazsan 9720:
1.26 crook 9721: The field names that come to (my) mind are often quite generic, and,
9722: if used, would cause frequent name clashes. E.g., many structures
9723: probably contain a @code{counter} field. The structure names
9724: that come to (my) mind are often also the logical choice for the names
9725: of words that create such a structure.
1.20 pazsan 9726:
1.26 crook 9727: Therefore, I have adopted the following naming conventions:
1.20 pazsan 9728:
1.26 crook 9729: @itemize @bullet
9730: @cindex field naming convention
9731: @item
9732: The names of fields are of the form
9733: @code{@emph{struct}-@emph{field}}, where
9734: @code{@emph{struct}} is the basic name of the structure, and
9735: @code{@emph{field}} is the basic name of the field. You can
9736: think of field words as converting the (address of the)
9737: structure into the (address of the) field.
1.20 pazsan 9738:
1.26 crook 9739: @cindex structure naming convention
9740: @item
9741: The names of structures are of the form
9742: @code{@emph{struct}%}, where
9743: @code{@emph{struct}} is the basic name of the structure.
9744: @end itemize
1.20 pazsan 9745:
1.26 crook 9746: This naming convention does not work that well for fields of extended
9747: structures; e.g., the integer list structure has a field
9748: @code{intlist-int}, but has @code{list-next}, not
9749: @code{intlist-next}.
1.20 pazsan 9750:
1.26 crook 9751: @node Structure Implementation, Structure Glossary, Structure Naming Convention, Structures
9752: @subsection Structure Implementation
9753: @cindex structure implementation
9754: @cindex implementation of structures
1.20 pazsan 9755:
1.26 crook 9756: The central idea in the implementation is to pass the data about the
9757: structure being built on the stack, not in some global
9758: variable. Everything else falls into place naturally once this design
9759: decision is made.
1.20 pazsan 9760:
1.26 crook 9761: The type description on the stack is of the form @emph{align
9762: size}. Keeping the size on the top-of-stack makes dealing with arrays
9763: very simple.
1.20 pazsan 9764:
1.26 crook 9765: @code{field} is a defining word that uses @code{Create}
9766: and @code{DOES>}. The body of the field contains the offset
9767: of the field, and the normal @code{DOES>} action is simply:
1.20 pazsan 9768:
9769: @example
1.48 anton 9770: @@ +
1.20 pazsan 9771: @end example
9772:
1.23 crook 9773: @noindent
1.26 crook 9774: i.e., add the offset to the address, giving the stack effect
1.29 crook 9775: @i{addr1 -- addr2} for a field.
1.20 pazsan 9776:
1.26 crook 9777: @cindex first field optimization, implementation
9778: This simple structure is slightly complicated by the optimization
9779: for fields with offset 0, which requires a different
9780: @code{DOES>}-part (because we cannot rely on there being
9781: something on the stack if such a field is invoked during
9782: compilation). Therefore, we put the different @code{DOES>}-parts
9783: in separate words, and decide which one to invoke based on the
9784: offset. For a zero offset, the field is basically a noop; it is
9785: immediate, and therefore no code is generated when it is compiled.
1.20 pazsan 9786:
1.26 crook 9787: @node Structure Glossary, , Structure Implementation, Structures
9788: @subsection Structure Glossary
9789: @cindex structure glossary
1.20 pazsan 9790:
1.44 crook 9791:
1.26 crook 9792: doc-%align
9793: doc-%alignment
9794: doc-%alloc
9795: doc-%allocate
9796: doc-%allot
9797: doc-cell%
9798: doc-char%
9799: doc-dfloat%
9800: doc-double%
9801: doc-end-struct
9802: doc-field
9803: doc-float%
9804: doc-naligned
9805: doc-sfloat%
9806: doc-%size
9807: doc-struct
1.23 crook 9808:
1.44 crook 9809:
1.26 crook 9810: @c -------------------------------------------------------------
9811: @node Object-oriented Forth, Passing Commands to the OS, Structures, Words
9812: @section Object-oriented Forth
1.20 pazsan 9813:
1.26 crook 9814: Gforth comes with three packages for object-oriented programming:
9815: @file{objects.fs}, @file{oof.fs}, and @file{mini-oof.fs}; none of them
9816: is preloaded, so you have to @code{include} them before use. The most
9817: important differences between these packages (and others) are discussed
9818: in @ref{Comparison with other object models}. All packages are written
9819: in ANS Forth and can be used with any other ANS Forth.
1.20 pazsan 9820:
1.26 crook 9821: @menu
1.48 anton 9822: * Why object-oriented programming?::
9823: * Object-Oriented Terminology::
9824: * Objects::
9825: * OOF::
9826: * Mini-OOF::
1.26 crook 9827: * Comparison with other object models::
9828: @end menu
1.20 pazsan 9829:
1.48 anton 9830: @c ----------------------------------------------------------------
9831: @node Why object-oriented programming?, Object-Oriented Terminology, Object-oriented Forth, Object-oriented Forth
9832: @subsection Why object-oriented programming?
1.26 crook 9833: @cindex object-oriented programming motivation
9834: @cindex motivation for object-oriented programming
1.23 crook 9835:
1.26 crook 9836: Often we have to deal with several data structures (@emph{objects}),
9837: that have to be treated similarly in some respects, but differently in
9838: others. Graphical objects are the textbook example: circles, triangles,
9839: dinosaurs, icons, and others, and we may want to add more during program
9840: development. We want to apply some operations to any graphical object,
9841: e.g., @code{draw} for displaying it on the screen. However, @code{draw}
9842: has to do something different for every kind of object.
9843: @comment TODO add some other operations eg perimeter, area
9844: @comment and tie in to concrete examples later..
1.23 crook 9845:
1.26 crook 9846: We could implement @code{draw} as a big @code{CASE}
9847: control structure that executes the appropriate code depending on the
9848: kind of object to be drawn. This would be not be very elegant, and,
9849: moreover, we would have to change @code{draw} every time we add
9850: a new kind of graphical object (say, a spaceship).
1.23 crook 9851:
1.26 crook 9852: What we would rather do is: When defining spaceships, we would tell
9853: the system: ``Here's how you @code{draw} a spaceship; you figure
9854: out the rest''.
1.23 crook 9855:
1.26 crook 9856: This is the problem that all systems solve that (rightfully) call
9857: themselves object-oriented; the object-oriented packages presented here
9858: solve this problem (and not much else).
9859: @comment TODO ?list properties of oo systems.. oo vs o-based?
1.23 crook 9860:
1.48 anton 9861: @c ------------------------------------------------------------------------
1.26 crook 9862: @node Object-Oriented Terminology, Objects, Why object-oriented programming?, Object-oriented Forth
1.48 anton 9863: @subsection Object-Oriented Terminology
1.26 crook 9864: @cindex object-oriented terminology
9865: @cindex terminology for object-oriented programming
1.23 crook 9866:
1.26 crook 9867: This section is mainly for reference, so you don't have to understand
9868: all of it right away. The terminology is mainly Smalltalk-inspired. In
9869: short:
1.23 crook 9870:
1.26 crook 9871: @table @emph
9872: @cindex class
9873: @item class
9874: a data structure definition with some extras.
1.23 crook 9875:
1.26 crook 9876: @cindex object
9877: @item object
9878: an instance of the data structure described by the class definition.
1.23 crook 9879:
1.26 crook 9880: @cindex instance variables
9881: @item instance variables
9882: fields of the data structure.
1.23 crook 9883:
1.26 crook 9884: @cindex selector
9885: @cindex method selector
9886: @cindex virtual function
9887: @item selector
9888: (or @emph{method selector}) a word (e.g.,
9889: @code{draw}) that performs an operation on a variety of data
9890: structures (classes). A selector describes @emph{what} operation to
9891: perform. In C++ terminology: a (pure) virtual function.
1.23 crook 9892:
1.26 crook 9893: @cindex method
9894: @item method
9895: the concrete definition that performs the operation
9896: described by the selector for a specific class. A method specifies
9897: @emph{how} the operation is performed for a specific class.
1.23 crook 9898:
1.26 crook 9899: @cindex selector invocation
9900: @cindex message send
9901: @cindex invoking a selector
9902: @item selector invocation
9903: a call of a selector. One argument of the call (the TOS (top-of-stack))
9904: is used for determining which method is used. In Smalltalk terminology:
9905: a message (consisting of the selector and the other arguments) is sent
9906: to the object.
1.1 anton 9907:
1.26 crook 9908: @cindex receiving object
9909: @item receiving object
9910: the object used for determining the method executed by a selector
9911: invocation. In the @file{objects.fs} model, it is the object that is on
9912: the TOS when the selector is invoked. (@emph{Receiving} comes from
9913: the Smalltalk @emph{message} terminology.)
1.1 anton 9914:
1.26 crook 9915: @cindex child class
9916: @cindex parent class
9917: @cindex inheritance
9918: @item child class
9919: a class that has (@emph{inherits}) all properties (instance variables,
9920: selectors, methods) from a @emph{parent class}. In Smalltalk
9921: terminology: The subclass inherits from the superclass. In C++
9922: terminology: The derived class inherits from the base class.
1.1 anton 9923:
1.26 crook 9924: @end table
1.21 crook 9925:
1.26 crook 9926: @c If you wonder about the message sending terminology, it comes from
9927: @c a time when each object had it's own task and objects communicated via
9928: @c message passing; eventually the Smalltalk developers realized that
9929: @c they can do most things through simple (indirect) calls. They kept the
9930: @c terminology.
1.1 anton 9931:
1.48 anton 9932: @c --------------------------------------------------------------
1.26 crook 9933: @node Objects, OOF, Object-Oriented Terminology, Object-oriented Forth
9934: @subsection The @file{objects.fs} model
9935: @cindex objects
9936: @cindex object-oriented programming
1.1 anton 9937:
1.26 crook 9938: @cindex @file{objects.fs}
9939: @cindex @file{oof.fs}
1.1 anton 9940:
1.37 anton 9941: This section describes the @file{objects.fs} package. This material also
9942: has been published in @cite{Yet Another Forth Objects Package} by Anton
9943: Ertl and appeared in Forth Dimensions 19(2), pages 37--43
1.47 crook 9944: (@uref{http://www.complang.tuwien.ac.at/forth/objects/objects.html}).
1.26 crook 9945: @c McKewan's and Zsoter's packages
1.1 anton 9946:
1.26 crook 9947: This section assumes that you have read @ref{Structures}.
1.1 anton 9948:
1.26 crook 9949: The techniques on which this model is based have been used to implement
9950: the parser generator, Gray, and have also been used in Gforth for
9951: implementing the various flavours of word lists (hashed or not,
9952: case-sensitive or not, special-purpose word lists for locals etc.).
1.1 anton 9953:
9954:
1.26 crook 9955: @menu
9956: * Properties of the Objects model::
9957: * Basic Objects Usage::
1.37 anton 9958: * The Objects base class::
1.26 crook 9959: * Creating objects::
9960: * Object-Oriented Programming Style::
9961: * Class Binding::
9962: * Method conveniences::
9963: * Classes and Scoping::
1.37 anton 9964: * Dividing classes::
1.26 crook 9965: * Object Interfaces::
9966: * Objects Implementation::
9967: * Objects Glossary::
9968: @end menu
1.1 anton 9969:
1.26 crook 9970: Marcel Hendrix provided helpful comments on this section. Andras Zsoter
9971: and Bernd Paysan helped me with the related works section.
1.1 anton 9972:
1.26 crook 9973: @node Properties of the Objects model, Basic Objects Usage, Objects, Objects
9974: @subsubsection Properties of the @file{objects.fs} model
9975: @cindex @file{objects.fs} properties
1.1 anton 9976:
1.26 crook 9977: @itemize @bullet
9978: @item
9979: It is straightforward to pass objects on the stack. Passing
9980: selectors on the stack is a little less convenient, but possible.
1.1 anton 9981:
1.26 crook 9982: @item
9983: Objects are just data structures in memory, and are referenced by their
9984: address. You can create words for objects with normal defining words
9985: like @code{constant}. Likewise, there is no difference between instance
9986: variables that contain objects and those that contain other data.
1.1 anton 9987:
1.26 crook 9988: @item
9989: Late binding is efficient and easy to use.
1.21 crook 9990:
1.26 crook 9991: @item
9992: It avoids parsing, and thus avoids problems with state-smartness
9993: and reduced extensibility; for convenience there are a few parsing
9994: words, but they have non-parsing counterparts. There are also a few
9995: defining words that parse. This is hard to avoid, because all standard
9996: defining words parse (except @code{:noname}); however, such
9997: words are not as bad as many other parsing words, because they are not
9998: state-smart.
1.21 crook 9999:
1.26 crook 10000: @item
10001: It does not try to incorporate everything. It does a few things and does
10002: them well (IMO). In particular, this model was not designed to support
10003: information hiding (although it has features that may help); you can use
10004: a separate package for achieving this.
1.21 crook 10005:
1.26 crook 10006: @item
10007: It is layered; you don't have to learn and use all features to use this
1.49 anton 10008: model. Only a few features are necessary (@pxref{Basic Objects Usage},
10009: @pxref{The Objects base class}, @pxref{Creating objects}.), the others
1.26 crook 10010: are optional and independent of each other.
1.21 crook 10011:
1.26 crook 10012: @item
10013: An implementation in ANS Forth is available.
1.21 crook 10014:
1.26 crook 10015: @end itemize
1.21 crook 10016:
10017:
1.26 crook 10018: @node Basic Objects Usage, The Objects base class, Properties of the Objects model, Objects
10019: @subsubsection Basic @file{objects.fs} Usage
10020: @cindex basic objects usage
10021: @cindex objects, basic usage
1.21 crook 10022:
1.26 crook 10023: You can define a class for graphical objects like this:
1.21 crook 10024:
1.26 crook 10025: @cindex @code{class} usage
10026: @cindex @code{end-class} usage
10027: @cindex @code{selector} usage
10028: @example
10029: object class \ "object" is the parent class
10030: selector draw ( x y graphical -- )
10031: end-class graphical
10032: @end example
1.21 crook 10033:
1.26 crook 10034: This code defines a class @code{graphical} with an
10035: operation @code{draw}. We can perform the operation
10036: @code{draw} on any @code{graphical} object, e.g.:
1.21 crook 10037:
1.26 crook 10038: @example
10039: 100 100 t-rex draw
10040: @end example
1.21 crook 10041:
1.26 crook 10042: @noindent
10043: where @code{t-rex} is a word (say, a constant) that produces a
10044: graphical object.
1.21 crook 10045:
1.29 crook 10046: @comment TODO add a 2nd operation eg perimeter.. and use for
1.26 crook 10047: @comment a concrete example
1.21 crook 10048:
1.26 crook 10049: @cindex abstract class
10050: How do we create a graphical object? With the present definitions,
10051: we cannot create a useful graphical object. The class
10052: @code{graphical} describes graphical objects in general, but not
10053: any concrete graphical object type (C++ users would call it an
10054: @emph{abstract class}); e.g., there is no method for the selector
10055: @code{draw} in the class @code{graphical}.
1.21 crook 10056:
1.26 crook 10057: For concrete graphical objects, we define child classes of the
10058: class @code{graphical}, e.g.:
1.21 crook 10059:
1.26 crook 10060: @cindex @code{overrides} usage
10061: @cindex @code{field} usage in class definition
10062: @example
10063: graphical class \ "graphical" is the parent class
10064: cell% field circle-radius
1.21 crook 10065:
1.26 crook 10066: :noname ( x y circle -- )
10067: circle-radius @@ draw-circle ;
10068: overrides draw
1.21 crook 10069:
1.26 crook 10070: :noname ( n-radius circle -- )
10071: circle-radius ! ;
10072: overrides construct
1.21 crook 10073:
1.26 crook 10074: end-class circle
1.21 crook 10075: @end example
10076:
1.26 crook 10077: Here we define a class @code{circle} as a child of @code{graphical},
10078: with field @code{circle-radius} (which behaves just like a field
10079: (@pxref{Structures}); it defines (using @code{overrides}) new methods
10080: for the selectors @code{draw} and @code{construct} (@code{construct} is
10081: defined in @code{object}, the parent class of @code{graphical}).
1.21 crook 10082:
1.26 crook 10083: Now we can create a circle on the heap (i.e.,
10084: @code{allocate}d memory) with:
1.21 crook 10085:
1.26 crook 10086: @cindex @code{heap-new} usage
1.21 crook 10087: @example
1.26 crook 10088: 50 circle heap-new constant my-circle
10089: @end example
1.21 crook 10090:
1.26 crook 10091: @noindent
10092: @code{heap-new} invokes @code{construct}, thus
10093: initializing the field @code{circle-radius} with 50. We can draw
10094: this new circle at (100,100) with:
1.21 crook 10095:
1.26 crook 10096: @example
10097: 100 100 my-circle draw
1.21 crook 10098: @end example
10099:
1.26 crook 10100: @cindex selector invocation, restrictions
10101: @cindex class definition, restrictions
10102: Note: You can only invoke a selector if the object on the TOS
10103: (the receiving object) belongs to the class where the selector was
10104: defined or one of its descendents; e.g., you can invoke
10105: @code{draw} only for objects belonging to @code{graphical}
10106: or its descendents (e.g., @code{circle}). Immediately before
10107: @code{end-class}, the search order has to be the same as
10108: immediately after @code{class}.
1.21 crook 10109:
1.26 crook 10110: @node The Objects base class, Creating objects, Basic Objects Usage, Objects
10111: @subsubsection The @file{object.fs} base class
10112: @cindex @code{object} class
1.21 crook 10113:
1.26 crook 10114: When you define a class, you have to specify a parent class. So how do
10115: you start defining classes? There is one class available from the start:
10116: @code{object}. It is ancestor for all classes and so is the
10117: only class that has no parent. It has two selectors: @code{construct}
10118: and @code{print}.
1.21 crook 10119:
1.26 crook 10120: @node Creating objects, Object-Oriented Programming Style, The Objects base class, Objects
10121: @subsubsection Creating objects
10122: @cindex creating objects
10123: @cindex object creation
10124: @cindex object allocation options
1.21 crook 10125:
1.26 crook 10126: @cindex @code{heap-new} discussion
10127: @cindex @code{dict-new} discussion
10128: @cindex @code{construct} discussion
10129: You can create and initialize an object of a class on the heap with
10130: @code{heap-new} ( ... class -- object ) and in the dictionary
10131: (allocation with @code{allot}) with @code{dict-new} (
10132: ... class -- object ). Both words invoke @code{construct}, which
10133: consumes the stack items indicated by "..." above.
1.21 crook 10134:
1.26 crook 10135: @cindex @code{init-object} discussion
10136: @cindex @code{class-inst-size} discussion
10137: If you want to allocate memory for an object yourself, you can get its
10138: alignment and size with @code{class-inst-size 2@@} ( class --
10139: align size ). Once you have memory for an object, you can initialize
10140: it with @code{init-object} ( ... class object -- );
10141: @code{construct} does only a part of the necessary work.
1.21 crook 10142:
1.26 crook 10143: @node Object-Oriented Programming Style, Class Binding, Creating objects, Objects
10144: @subsubsection Object-Oriented Programming Style
10145: @cindex object-oriented programming style
1.47 crook 10146: @cindex programming style, object-oriented
1.21 crook 10147:
1.26 crook 10148: This section is not exhaustive.
1.1 anton 10149:
1.26 crook 10150: @cindex stack effects of selectors
10151: @cindex selectors and stack effects
10152: In general, it is a good idea to ensure that all methods for the
10153: same selector have the same stack effect: when you invoke a selector,
10154: you often have no idea which method will be invoked, so, unless all
10155: methods have the same stack effect, you will not know the stack effect
10156: of the selector invocation.
1.21 crook 10157:
1.26 crook 10158: One exception to this rule is methods for the selector
10159: @code{construct}. We know which method is invoked, because we
10160: specify the class to be constructed at the same place. Actually, I
10161: defined @code{construct} as a selector only to give the users a
10162: convenient way to specify initialization. The way it is used, a
10163: mechanism different from selector invocation would be more natural
10164: (but probably would take more code and more space to explain).
1.21 crook 10165:
1.26 crook 10166: @node Class Binding, Method conveniences, Object-Oriented Programming Style, Objects
10167: @subsubsection Class Binding
10168: @cindex class binding
10169: @cindex early binding
1.21 crook 10170:
1.26 crook 10171: @cindex late binding
10172: Normal selector invocations determine the method at run-time depending
10173: on the class of the receiving object. This run-time selection is called
1.29 crook 10174: @i{late binding}.
1.21 crook 10175:
1.26 crook 10176: Sometimes it's preferable to invoke a different method. For example,
10177: you might want to use the simple method for @code{print}ing
10178: @code{object}s instead of the possibly long-winded @code{print} method
10179: of the receiver class. You can achieve this by replacing the invocation
10180: of @code{print} with:
1.21 crook 10181:
1.26 crook 10182: @cindex @code{[bind]} usage
10183: @example
10184: [bind] object print
1.21 crook 10185: @end example
10186:
1.26 crook 10187: @noindent
10188: in compiled code or:
1.21 crook 10189:
1.26 crook 10190: @cindex @code{bind} usage
1.21 crook 10191: @example
1.26 crook 10192: bind object print
1.21 crook 10193: @end example
10194:
1.26 crook 10195: @cindex class binding, alternative to
10196: @noindent
10197: in interpreted code. Alternatively, you can define the method with a
10198: name (e.g., @code{print-object}), and then invoke it through the
10199: name. Class binding is just a (often more convenient) way to achieve
10200: the same effect; it avoids name clutter and allows you to invoke
10201: methods directly without naming them first.
10202:
10203: @cindex superclass binding
10204: @cindex parent class binding
10205: A frequent use of class binding is this: When we define a method
10206: for a selector, we often want the method to do what the selector does
10207: in the parent class, and a little more. There is a special word for
10208: this purpose: @code{[parent]}; @code{[parent]
10209: @emph{selector}} is equivalent to @code{[bind] @emph{parent
10210: selector}}, where @code{@emph{parent}} is the parent
10211: class of the current class. E.g., a method definition might look like:
1.21 crook 10212:
1.26 crook 10213: @cindex @code{[parent]} usage
1.21 crook 10214: @example
1.26 crook 10215: :noname
10216: dup [parent] foo \ do parent's foo on the receiving object
10217: ... \ do some more
10218: ; overrides foo
1.21 crook 10219: @end example
10220:
1.26 crook 10221: @cindex class binding as optimization
10222: In @cite{Object-oriented programming in ANS Forth} (Forth Dimensions,
10223: March 1997), Andrew McKewan presents class binding as an optimization
10224: technique. I recommend not using it for this purpose unless you are in
10225: an emergency. Late binding is pretty fast with this model anyway, so the
10226: benefit of using class binding is small; the cost of using class binding
10227: where it is not appropriate is reduced maintainability.
1.21 crook 10228:
1.26 crook 10229: While we are at programming style questions: You should bind
10230: selectors only to ancestor classes of the receiving object. E.g., say,
10231: you know that the receiving object is of class @code{foo} or its
10232: descendents; then you should bind only to @code{foo} and its
10233: ancestors.
1.21 crook 10234:
1.26 crook 10235: @node Method conveniences, Classes and Scoping, Class Binding, Objects
10236: @subsubsection Method conveniences
10237: @cindex method conveniences
1.1 anton 10238:
1.26 crook 10239: In a method you usually access the receiving object pretty often. If
10240: you define the method as a plain colon definition (e.g., with
10241: @code{:noname}), you may have to do a lot of stack
10242: gymnastics. To avoid this, you can define the method with @code{m:
10243: ... ;m}. E.g., you could define the method for
10244: @code{draw}ing a @code{circle} with
1.20 pazsan 10245:
1.26 crook 10246: @cindex @code{this} usage
10247: @cindex @code{m:} usage
10248: @cindex @code{;m} usage
10249: @example
10250: m: ( x y circle -- )
10251: ( x y ) this circle-radius @@ draw-circle ;m
10252: @end example
1.20 pazsan 10253:
1.26 crook 10254: @cindex @code{exit} in @code{m: ... ;m}
10255: @cindex @code{exitm} discussion
10256: @cindex @code{catch} in @code{m: ... ;m}
10257: When this method is executed, the receiver object is removed from the
10258: stack; you can access it with @code{this} (admittedly, in this
10259: example the use of @code{m: ... ;m} offers no advantage). Note
10260: that I specify the stack effect for the whole method (i.e. including
10261: the receiver object), not just for the code between @code{m:}
10262: and @code{;m}. You cannot use @code{exit} in
10263: @code{m:...;m}; instead, use
10264: @code{exitm}.@footnote{Moreover, for any word that calls
10265: @code{catch} and was defined before loading
10266: @code{objects.fs}, you have to redefine it like I redefined
10267: @code{catch}: @code{: catch this >r catch r> to-this ;}}
1.20 pazsan 10268:
1.26 crook 10269: @cindex @code{inst-var} usage
10270: You will frequently use sequences of the form @code{this
10271: @emph{field}} (in the example above: @code{this
10272: circle-radius}). If you use the field only in this way, you can
10273: define it with @code{inst-var} and eliminate the
10274: @code{this} before the field name. E.g., the @code{circle}
10275: class above could also be defined with:
1.20 pazsan 10276:
1.26 crook 10277: @example
10278: graphical class
10279: cell% inst-var radius
1.20 pazsan 10280:
1.26 crook 10281: m: ( x y circle -- )
10282: radius @@ draw-circle ;m
10283: overrides draw
1.20 pazsan 10284:
1.26 crook 10285: m: ( n-radius circle -- )
10286: radius ! ;m
10287: overrides construct
1.12 anton 10288:
1.26 crook 10289: end-class circle
10290: @end example
1.12 anton 10291:
1.26 crook 10292: @code{radius} can only be used in @code{circle} and its
10293: descendent classes and inside @code{m:...;m}.
1.12 anton 10294:
1.26 crook 10295: @cindex @code{inst-value} usage
10296: You can also define fields with @code{inst-value}, which is
10297: to @code{inst-var} what @code{value} is to
10298: @code{variable}. You can change the value of such a field with
10299: @code{[to-inst]}. E.g., we could also define the class
10300: @code{circle} like this:
1.12 anton 10301:
1.26 crook 10302: @example
10303: graphical class
10304: inst-value radius
1.12 anton 10305:
1.26 crook 10306: m: ( x y circle -- )
10307: radius draw-circle ;m
10308: overrides draw
1.12 anton 10309:
1.26 crook 10310: m: ( n-radius circle -- )
10311: [to-inst] radius ;m
10312: overrides construct
1.21 crook 10313:
1.26 crook 10314: end-class circle
1.12 anton 10315: @end example
10316:
1.38 anton 10317: Finally, you can define named methods with @code{:m}. One use of this
10318: feature is the definition of words that occur only in one class and are
10319: not intended to be overridden, but which still need method context
10320: (e.g., for accessing @code{inst-var}s). Another use is for methods that
10321: would be bound frequently, if defined anonymously.
10322:
1.12 anton 10323:
1.37 anton 10324: @node Classes and Scoping, Dividing classes, Method conveniences, Objects
1.26 crook 10325: @subsubsection Classes and Scoping
10326: @cindex classes and scoping
10327: @cindex scoping and classes
1.12 anton 10328:
1.26 crook 10329: Inheritance is frequent, unlike structure extension. This exacerbates
10330: the problem with the field name convention (@pxref{Structure Naming
10331: Convention}): One always has to remember in which class the field was
10332: originally defined; changing a part of the class structure would require
10333: changes for renaming in otherwise unaffected code.
1.12 anton 10334:
1.26 crook 10335: @cindex @code{inst-var} visibility
10336: @cindex @code{inst-value} visibility
10337: To solve this problem, I added a scoping mechanism (which was not in my
10338: original charter): A field defined with @code{inst-var} (or
10339: @code{inst-value}) is visible only in the class where it is defined and in
10340: the descendent classes of this class. Using such fields only makes
10341: sense in @code{m:}-defined methods in these classes anyway.
1.12 anton 10342:
1.26 crook 10343: This scoping mechanism allows us to use the unadorned field name,
10344: because name clashes with unrelated words become much less likely.
1.12 anton 10345:
1.26 crook 10346: @cindex @code{protected} discussion
10347: @cindex @code{private} discussion
10348: Once we have this mechanism, we can also use it for controlling the
10349: visibility of other words: All words defined after
10350: @code{protected} are visible only in the current class and its
10351: descendents. @code{public} restores the compilation
10352: (i.e. @code{current}) word list that was in effect before. If you
10353: have several @code{protected}s without an intervening
10354: @code{public} or @code{set-current}, @code{public}
10355: will restore the compilation word list in effect before the first of
10356: these @code{protected}s.
1.12 anton 10357:
1.37 anton 10358: @node Dividing classes, Object Interfaces, Classes and Scoping, Objects
10359: @subsubsection Dividing classes
10360: @cindex Dividing classes
10361: @cindex @code{methods}...@code{end-methods}
10362:
10363: You may want to do the definition of methods separate from the
10364: definition of the class, its selectors, fields, and instance variables,
10365: i.e., separate the implementation from the definition. You can do this
10366: in the following way:
10367:
10368: @example
10369: graphical class
10370: inst-value radius
10371: end-class circle
10372:
10373: ... \ do some other stuff
10374:
10375: circle methods \ now we are ready
10376:
10377: m: ( x y circle -- )
10378: radius draw-circle ;m
10379: overrides draw
10380:
10381: m: ( n-radius circle -- )
10382: [to-inst] radius ;m
10383: overrides construct
10384:
10385: end-methods
10386: @end example
10387:
10388: You can use several @code{methods}...@code{end-methods} sections. The
10389: only things you can do to the class in these sections are: defining
10390: methods, and overriding the class's selectors. You must not define new
10391: selectors or fields.
10392:
10393: Note that you often have to override a selector before using it. In
10394: particular, you usually have to override @code{construct} with a new
10395: method before you can invoke @code{heap-new} and friends. E.g., you
10396: must not create a circle before the @code{overrides construct} sequence
10397: in the example above.
10398:
10399: @node Object Interfaces, Objects Implementation, Dividing classes, Objects
1.26 crook 10400: @subsubsection Object Interfaces
10401: @cindex object interfaces
10402: @cindex interfaces for objects
1.12 anton 10403:
1.26 crook 10404: In this model you can only call selectors defined in the class of the
10405: receiving objects or in one of its ancestors. If you call a selector
10406: with a receiving object that is not in one of these classes, the
10407: result is undefined; if you are lucky, the program crashes
10408: immediately.
1.12 anton 10409:
1.26 crook 10410: @cindex selectors common to hardly-related classes
10411: Now consider the case when you want to have a selector (or several)
10412: available in two classes: You would have to add the selector to a
10413: common ancestor class, in the worst case to @code{object}. You
10414: may not want to do this, e.g., because someone else is responsible for
10415: this ancestor class.
1.12 anton 10416:
1.26 crook 10417: The solution for this problem is interfaces. An interface is a
10418: collection of selectors. If a class implements an interface, the
10419: selectors become available to the class and its descendents. A class
10420: can implement an unlimited number of interfaces. For the problem
10421: discussed above, we would define an interface for the selector(s), and
10422: both classes would implement the interface.
1.12 anton 10423:
1.26 crook 10424: As an example, consider an interface @code{storage} for
10425: writing objects to disk and getting them back, and a class
10426: @code{foo} that implements it. The code would look like this:
1.12 anton 10427:
1.26 crook 10428: @cindex @code{interface} usage
10429: @cindex @code{end-interface} usage
10430: @cindex @code{implementation} usage
10431: @example
10432: interface
10433: selector write ( file object -- )
10434: selector read1 ( file object -- )
10435: end-interface storage
1.12 anton 10436:
1.26 crook 10437: bar class
10438: storage implementation
1.12 anton 10439:
1.26 crook 10440: ... overrides write
1.37 anton 10441: ... overrides read1
1.26 crook 10442: ...
10443: end-class foo
1.12 anton 10444: @end example
10445:
1.26 crook 10446: @noindent
1.29 crook 10447: (I would add a word @code{read} @i{( file -- object )} that uses
1.26 crook 10448: @code{read1} internally, but that's beyond the point illustrated
10449: here.)
1.12 anton 10450:
1.26 crook 10451: Note that you cannot use @code{protected} in an interface; and
10452: of course you cannot define fields.
1.12 anton 10453:
1.26 crook 10454: In the Neon model, all selectors are available for all classes;
10455: therefore it does not need interfaces. The price you pay in this model
10456: is slower late binding, and therefore, added complexity to avoid late
10457: binding.
1.12 anton 10458:
1.26 crook 10459: @node Objects Implementation, Objects Glossary, Object Interfaces, Objects
10460: @subsubsection @file{objects.fs} Implementation
10461: @cindex @file{objects.fs} implementation
1.12 anton 10462:
1.26 crook 10463: @cindex @code{object-map} discussion
10464: An object is a piece of memory, like one of the data structures
10465: described with @code{struct...end-struct}. It has a field
10466: @code{object-map} that points to the method map for the object's
10467: class.
1.12 anton 10468:
1.26 crook 10469: @cindex method map
10470: @cindex virtual function table
10471: The @emph{method map}@footnote{This is Self terminology; in C++
10472: terminology: virtual function table.} is an array that contains the
1.29 crook 10473: execution tokens (@i{xt}s) of the methods for the object's class. Each
1.26 crook 10474: selector contains an offset into a method map.
1.12 anton 10475:
1.26 crook 10476: @cindex @code{selector} implementation, class
10477: @code{selector} is a defining word that uses
10478: @code{CREATE} and @code{DOES>}. The body of the
1.44 crook 10479: selector contains the offset; the @code{DOES>} action for a
1.26 crook 10480: class selector is, basically:
1.21 crook 10481:
1.26 crook 10482: @example
10483: ( object addr ) @@ over object-map @@ + @@ execute
10484: @end example
1.12 anton 10485:
1.26 crook 10486: Since @code{object-map} is the first field of the object, it
10487: does not generate any code. As you can see, calling a selector has a
10488: small, constant cost.
1.12 anton 10489:
1.26 crook 10490: @cindex @code{current-interface} discussion
10491: @cindex class implementation and representation
10492: A class is basically a @code{struct} combined with a method
10493: map. During the class definition the alignment and size of the class
10494: are passed on the stack, just as with @code{struct}s, so
10495: @code{field} can also be used for defining class
10496: fields. However, passing more items on the stack would be
10497: inconvenient, so @code{class} builds a data structure in memory,
10498: which is accessed through the variable
10499: @code{current-interface}. After its definition is complete, the
10500: class is represented on the stack by a pointer (e.g., as parameter for
10501: a child class definition).
1.1 anton 10502:
1.26 crook 10503: A new class starts off with the alignment and size of its parent,
10504: and a copy of the parent's method map. Defining new fields extends the
10505: size and alignment; likewise, defining new selectors extends the
1.29 crook 10506: method map. @code{overrides} just stores a new @i{xt} in the method
1.26 crook 10507: map at the offset given by the selector.
1.20 pazsan 10508:
1.26 crook 10509: @cindex class binding, implementation
1.29 crook 10510: Class binding just gets the @i{xt} at the offset given by the selector
1.26 crook 10511: from the class's method map and @code{compile,}s (in the case of
10512: @code{[bind]}) it.
1.21 crook 10513:
1.26 crook 10514: @cindex @code{this} implementation
10515: @cindex @code{catch} and @code{this}
10516: @cindex @code{this} and @code{catch}
10517: I implemented @code{this} as a @code{value}. At the
10518: start of an @code{m:...;m} method the old @code{this} is
10519: stored to the return stack and restored at the end; and the object on
10520: the TOS is stored @code{TO this}. This technique has one
10521: disadvantage: If the user does not leave the method via
10522: @code{;m}, but via @code{throw} or @code{exit},
10523: @code{this} is not restored (and @code{exit} may
10524: crash). To deal with the @code{throw} problem, I have redefined
10525: @code{catch} to save and restore @code{this}; the same
10526: should be done with any word that can catch an exception. As for
10527: @code{exit}, I simply forbid it (as a replacement, there is
10528: @code{exitm}).
1.21 crook 10529:
1.26 crook 10530: @cindex @code{inst-var} implementation
10531: @code{inst-var} is just the same as @code{field}, with
10532: a different @code{DOES>} action:
10533: @example
10534: @@ this +
10535: @end example
10536: Similar for @code{inst-value}.
1.21 crook 10537:
1.26 crook 10538: @cindex class scoping implementation
10539: Each class also has a word list that contains the words defined with
10540: @code{inst-var} and @code{inst-value}, and its protected
10541: words. It also has a pointer to its parent. @code{class} pushes
10542: the word lists of the class and all its ancestors onto the search order stack,
10543: and @code{end-class} drops them.
1.21 crook 10544:
1.26 crook 10545: @cindex interface implementation
10546: An interface is like a class without fields, parent and protected
10547: words; i.e., it just has a method map. If a class implements an
10548: interface, its method map contains a pointer to the method map of the
10549: interface. The positive offsets in the map are reserved for class
10550: methods, therefore interface map pointers have negative
10551: offsets. Interfaces have offsets that are unique throughout the
10552: system, unlike class selectors, whose offsets are only unique for the
10553: classes where the selector is available (invokable).
1.21 crook 10554:
1.26 crook 10555: This structure means that interface selectors have to perform one
10556: indirection more than class selectors to find their method. Their body
10557: contains the interface map pointer offset in the class method map, and
10558: the method offset in the interface method map. The
10559: @code{does>} action for an interface selector is, basically:
1.21 crook 10560:
10561: @example
1.26 crook 10562: ( object selector-body )
10563: 2dup selector-interface @@ ( object selector-body object interface-offset )
10564: swap object-map @@ + @@ ( object selector-body map )
10565: swap selector-offset @@ + @@ execute
1.21 crook 10566: @end example
10567:
1.26 crook 10568: where @code{object-map} and @code{selector-offset} are
10569: first fields and generate no code.
10570:
10571: As a concrete example, consider the following code:
1.21 crook 10572:
1.26 crook 10573: @example
10574: interface
10575: selector if1sel1
10576: selector if1sel2
10577: end-interface if1
1.21 crook 10578:
1.26 crook 10579: object class
10580: if1 implementation
10581: selector cl1sel1
10582: cell% inst-var cl1iv1
1.21 crook 10583:
1.26 crook 10584: ' m1 overrides construct
10585: ' m2 overrides if1sel1
10586: ' m3 overrides if1sel2
10587: ' m4 overrides cl1sel2
10588: end-class cl1
1.21 crook 10589:
1.26 crook 10590: create obj1 object dict-new drop
10591: create obj2 cl1 dict-new drop
10592: @end example
1.21 crook 10593:
1.26 crook 10594: The data structure created by this code (including the data structure
10595: for @code{object}) is shown in the <a
10596: href="objects-implementation.eps">figure</a>, assuming a cell size of 4.
1.29 crook 10597: @comment TODO add this diagram..
1.21 crook 10598:
1.26 crook 10599: @node Objects Glossary, , Objects Implementation, Objects
10600: @subsubsection @file{objects.fs} Glossary
10601: @cindex @file{objects.fs} Glossary
1.21 crook 10602:
1.44 crook 10603:
1.26 crook 10604: doc---objects-bind
10605: doc---objects-<bind>
10606: doc---objects-bind'
10607: doc---objects-[bind]
10608: doc---objects-class
10609: doc---objects-class->map
10610: doc---objects-class-inst-size
10611: doc---objects-class-override!
10612: doc---objects-construct
10613: doc---objects-current'
10614: doc---objects-[current]
10615: doc---objects-current-interface
10616: doc---objects-dict-new
10617: doc---objects-drop-order
10618: doc---objects-end-class
10619: doc---objects-end-class-noname
10620: doc---objects-end-interface
10621: doc---objects-end-interface-noname
1.37 anton 10622: doc---objects-end-methods
1.26 crook 10623: doc---objects-exitm
10624: doc---objects-heap-new
10625: doc---objects-implementation
10626: doc---objects-init-object
10627: doc---objects-inst-value
10628: doc---objects-inst-var
10629: doc---objects-interface
1.38 anton 10630: doc---objects-m:
10631: doc---objects-:m
1.26 crook 10632: doc---objects-;m
10633: doc---objects-method
1.37 anton 10634: doc---objects-methods
1.26 crook 10635: doc---objects-object
10636: doc---objects-overrides
10637: doc---objects-[parent]
10638: doc---objects-print
10639: doc---objects-protected
10640: doc---objects-public
10641: doc---objects-push-order
10642: doc---objects-selector
10643: doc---objects-this
10644: doc---objects-<to-inst>
10645: doc---objects-[to-inst]
10646: doc---objects-to-this
10647: doc---objects-xt-new
1.21 crook 10648:
1.44 crook 10649:
1.26 crook 10650: @c -------------------------------------------------------------
10651: @node OOF, Mini-OOF, Objects, Object-oriented Forth
10652: @subsection The @file{oof.fs} model
10653: @cindex oof
10654: @cindex object-oriented programming
1.21 crook 10655:
1.26 crook 10656: @cindex @file{objects.fs}
10657: @cindex @file{oof.fs}
1.21 crook 10658:
1.26 crook 10659: This section describes the @file{oof.fs} package.
1.21 crook 10660:
1.26 crook 10661: The package described in this section has been used in bigFORTH since 1991, and
10662: used for two large applications: a chromatographic system used to
10663: create new medicaments, and a graphic user interface library (MINOS).
1.21 crook 10664:
1.26 crook 10665: You can find a description (in German) of @file{oof.fs} in @cite{Object
10666: oriented bigFORTH} by Bernd Paysan, published in @cite{Vierte Dimension}
10667: 10(2), 1994.
1.21 crook 10668:
1.26 crook 10669: @menu
10670: * Properties of the OOF model::
10671: * Basic OOF Usage::
10672: * The OOF base class::
10673: * Class Declaration::
10674: * Class Implementation::
10675: @end menu
1.21 crook 10676:
1.26 crook 10677: @node Properties of the OOF model, Basic OOF Usage, OOF, OOF
10678: @subsubsection Properties of the @file{oof.fs} model
10679: @cindex @file{oof.fs} properties
1.21 crook 10680:
1.26 crook 10681: @itemize @bullet
10682: @item
10683: This model combines object oriented programming with information
10684: hiding. It helps you writing large application, where scoping is
10685: necessary, because it provides class-oriented scoping.
1.21 crook 10686:
1.26 crook 10687: @item
10688: Named objects, object pointers, and object arrays can be created,
10689: selector invocation uses the ``object selector'' syntax. Selector invocation
10690: to objects and/or selectors on the stack is a bit less convenient, but
10691: possible.
1.21 crook 10692:
1.26 crook 10693: @item
10694: Selector invocation and instance variable usage of the active object is
10695: straightforward, since both make use of the active object.
1.21 crook 10696:
1.26 crook 10697: @item
10698: Late binding is efficient and easy to use.
1.21 crook 10699:
1.26 crook 10700: @item
10701: State-smart objects parse selectors. However, extensibility is provided
10702: using a (parsing) selector @code{postpone} and a selector @code{'}.
1.21 crook 10703:
10704: @item
1.26 crook 10705: An implementation in ANS Forth is available.
10706:
1.21 crook 10707: @end itemize
10708:
10709:
1.26 crook 10710: @node Basic OOF Usage, The OOF base class, Properties of the OOF model, OOF
10711: @subsubsection Basic @file{oof.fs} Usage
10712: @cindex @file{oof.fs} usage
10713:
10714: This section uses the same example as for @code{objects} (@pxref{Basic Objects Usage}).
1.21 crook 10715:
1.26 crook 10716: You can define a class for graphical objects like this:
1.21 crook 10717:
1.26 crook 10718: @cindex @code{class} usage
10719: @cindex @code{class;} usage
10720: @cindex @code{method} usage
10721: @example
10722: object class graphical \ "object" is the parent class
10723: method draw ( x y graphical -- )
10724: class;
10725: @end example
1.21 crook 10726:
1.26 crook 10727: This code defines a class @code{graphical} with an
10728: operation @code{draw}. We can perform the operation
10729: @code{draw} on any @code{graphical} object, e.g.:
1.21 crook 10730:
1.26 crook 10731: @example
10732: 100 100 t-rex draw
10733: @end example
1.21 crook 10734:
1.26 crook 10735: @noindent
10736: where @code{t-rex} is an object or object pointer, created with e.g.
10737: @code{graphical : t-rex}.
1.21 crook 10738:
1.26 crook 10739: @cindex abstract class
10740: How do we create a graphical object? With the present definitions,
10741: we cannot create a useful graphical object. The class
10742: @code{graphical} describes graphical objects in general, but not
10743: any concrete graphical object type (C++ users would call it an
10744: @emph{abstract class}); e.g., there is no method for the selector
10745: @code{draw} in the class @code{graphical}.
1.21 crook 10746:
1.26 crook 10747: For concrete graphical objects, we define child classes of the
10748: class @code{graphical}, e.g.:
1.21 crook 10749:
10750: @example
1.26 crook 10751: graphical class circle \ "graphical" is the parent class
10752: cell var circle-radius
10753: how:
10754: : draw ( x y -- )
10755: circle-radius @@ draw-circle ;
10756:
10757: : init ( n-radius -- (
10758: circle-radius ! ;
10759: class;
10760: @end example
10761:
10762: Here we define a class @code{circle} as a child of @code{graphical},
10763: with a field @code{circle-radius}; it defines new methods for the
10764: selectors @code{draw} and @code{init} (@code{init} is defined in
10765: @code{object}, the parent class of @code{graphical}).
1.21 crook 10766:
1.26 crook 10767: Now we can create a circle in the dictionary with:
1.21 crook 10768:
1.26 crook 10769: @example
10770: 50 circle : my-circle
1.21 crook 10771: @end example
10772:
1.26 crook 10773: @noindent
10774: @code{:} invokes @code{init}, thus initializing the field
10775: @code{circle-radius} with 50. We can draw this new circle at (100,100)
10776: with:
1.21 crook 10777:
10778: @example
1.26 crook 10779: 100 100 my-circle draw
1.21 crook 10780: @end example
10781:
1.26 crook 10782: @cindex selector invocation, restrictions
10783: @cindex class definition, restrictions
10784: Note: You can only invoke a selector if the receiving object belongs to
10785: the class where the selector was defined or one of its descendents;
10786: e.g., you can invoke @code{draw} only for objects belonging to
10787: @code{graphical} or its descendents (e.g., @code{circle}). The scoping
10788: mechanism will check if you try to invoke a selector that is not
10789: defined in this class hierarchy, so you'll get an error at compilation
10790: time.
10791:
10792:
10793: @node The OOF base class, Class Declaration, Basic OOF Usage, OOF
10794: @subsubsection The @file{oof.fs} base class
10795: @cindex @file{oof.fs} base class
10796:
10797: When you define a class, you have to specify a parent class. So how do
10798: you start defining classes? There is one class available from the start:
10799: @code{object}. You have to use it as ancestor for all classes. It is the
10800: only class that has no parent. Classes are also objects, except that
10801: they don't have instance variables; class manipulation such as
10802: inheritance or changing definitions of a class is handled through
10803: selectors of the class @code{object}.
10804:
10805: @code{object} provides a number of selectors:
10806:
1.21 crook 10807: @itemize @bullet
10808: @item
1.26 crook 10809: @code{class} for subclassing, @code{definitions} to add definitions
10810: later on, and @code{class?} to get type informations (is the class a
10811: subclass of the class passed on the stack?).
1.44 crook 10812:
1.26 crook 10813: doc---object-class
10814: doc---object-definitions
10815: doc---object-class?
10816:
1.44 crook 10817:
1.21 crook 10818: @item
1.26 crook 10819: @code{init} and @code{dispose} as constructor and destructor of the
10820: object. @code{init} is invocated after the object's memory is allocated,
10821: while @code{dispose} also handles deallocation. Thus if you redefine
10822: @code{dispose}, you have to call the parent's dispose with @code{super
10823: dispose}, too.
1.44 crook 10824:
1.26 crook 10825: doc---object-init
10826: doc---object-dispose
10827:
1.44 crook 10828:
1.21 crook 10829: @item
1.26 crook 10830: @code{new}, @code{new[]}, @code{:}, @code{ptr}, @code{asptr}, and
10831: @code{[]} to create named and unnamed objects and object arrays or
10832: object pointers.
1.44 crook 10833:
1.26 crook 10834: doc---object-new
10835: doc---object-new[]
10836: doc---object-:
10837: doc---object-ptr
10838: doc---object-asptr
10839: doc---object-[]
1.21 crook 10840:
1.44 crook 10841:
1.26 crook 10842: @item
10843: @code{::} and @code{super} for explicit scoping. You should use explicit
10844: scoping only for super classes or classes with the same set of instance
10845: variables. Explicitly-scoped selectors use early binding.
1.44 crook 10846:
1.26 crook 10847: doc---object-::
10848: doc---object-super
1.21 crook 10849:
1.44 crook 10850:
1.26 crook 10851: @item
10852: @code{self} to get the address of the object
1.44 crook 10853:
1.26 crook 10854: doc---object-self
1.21 crook 10855:
1.44 crook 10856:
1.21 crook 10857: @item
1.26 crook 10858: @code{bind}, @code{bound}, @code{link}, and @code{is} to assign object
10859: pointers and instance defers.
1.44 crook 10860:
1.26 crook 10861: doc---object-bind
10862: doc---object-bound
10863: doc---object-link
10864: doc---object-is
10865:
1.44 crook 10866:
1.21 crook 10867: @item
1.26 crook 10868: @code{'} to obtain selector tokens, @code{send} to invocate selectors
10869: form the stack, and @code{postpone} to generate selector invocation code.
1.44 crook 10870:
1.26 crook 10871: doc---object-'
10872: doc---object-postpone
10873:
1.44 crook 10874:
1.21 crook 10875: @item
1.26 crook 10876: @code{with} and @code{endwith} to select the active object from the
10877: stack, and enable its scope. Using @code{with} and @code{endwith}
10878: also allows you to create code using selector @code{postpone} without being
10879: trapped by the state-smart objects.
1.44 crook 10880:
1.26 crook 10881: doc---object-with
10882: doc---object-endwith
10883:
1.44 crook 10884:
1.21 crook 10885: @end itemize
10886:
1.26 crook 10887: @node Class Declaration, Class Implementation, The OOF base class, OOF
10888: @subsubsection Class Declaration
10889: @cindex class declaration
10890:
10891: @itemize @bullet
10892: @item
10893: Instance variables
1.44 crook 10894:
1.26 crook 10895: doc---oof-var
1.21 crook 10896:
1.44 crook 10897:
1.26 crook 10898: @item
10899: Object pointers
1.44 crook 10900:
1.26 crook 10901: doc---oof-ptr
10902: doc---oof-asptr
1.21 crook 10903:
1.44 crook 10904:
1.26 crook 10905: @item
10906: Instance defers
1.44 crook 10907:
1.26 crook 10908: doc---oof-defer
1.21 crook 10909:
1.44 crook 10910:
1.26 crook 10911: @item
10912: Method selectors
1.44 crook 10913:
1.26 crook 10914: doc---oof-early
10915: doc---oof-method
1.21 crook 10916:
1.44 crook 10917:
1.26 crook 10918: @item
10919: Class-wide variables
1.44 crook 10920:
1.26 crook 10921: doc---oof-static
1.21 crook 10922:
1.44 crook 10923:
1.26 crook 10924: @item
10925: End declaration
1.44 crook 10926:
1.26 crook 10927: doc---oof-how:
10928: doc---oof-class;
1.21 crook 10929:
1.44 crook 10930:
1.26 crook 10931: @end itemize
1.21 crook 10932:
1.26 crook 10933: @c -------------------------------------------------------------
10934: @node Class Implementation, , Class Declaration, OOF
10935: @subsubsection Class Implementation
10936: @cindex class implementation
1.21 crook 10937:
1.26 crook 10938: @c -------------------------------------------------------------
10939: @node Mini-OOF, Comparison with other object models, OOF, Object-oriented Forth
10940: @subsection The @file{mini-oof.fs} model
10941: @cindex mini-oof
1.1 anton 10942:
1.26 crook 10943: Gforth's third object oriented Forth package is a 12-liner. It uses a
10944: mixture of the @file{object.fs} and the @file{oof.fs} syntax,
10945: and reduces to the bare minimum of features. This is based on a posting
10946: of Bernd Paysan in comp.arch.
1.1 anton 10947:
10948: @menu
1.48 anton 10949: * Basic Mini-OOF Usage::
10950: * Mini-OOF Example::
10951: * Mini-OOF Implementation::
10952: * Comparison with other object models::
1.1 anton 10953: @end menu
10954:
1.26 crook 10955: @c -------------------------------------------------------------
1.48 anton 10956: @node Basic Mini-OOF Usage, Mini-OOF Example, Mini-OOF, Mini-OOF
1.26 crook 10957: @subsubsection Basic @file{mini-oof.fs} Usage
10958: @cindex mini-oof usage
1.1 anton 10959:
1.28 crook 10960: There is a base class (@code{class}, which allocates one cell for the
10961: object pointer) plus seven other words: to define a method, a variable,
10962: a class; to end a class, to resolve binding, to allocate an object and
10963: to compile a class method.
1.26 crook 10964: @comment TODO better description of the last one
1.1 anton 10965:
1.44 crook 10966:
1.26 crook 10967: doc-object
10968: doc-method
10969: doc-var
10970: doc-class
10971: doc-end-class
10972: doc-defines
10973: doc-new
10974: doc-::
1.1 anton 10975:
1.21 crook 10976:
1.44 crook 10977:
1.26 crook 10978: @c -------------------------------------------------------------
10979: @node Mini-OOF Example, Mini-OOF Implementation, Basic Mini-OOF Usage, Mini-OOF
10980: @subsubsection Mini-OOF Example
10981: @cindex mini-oof example
1.21 crook 10982:
1.26 crook 10983: A short example shows how to use this package. This example, in slightly
10984: extended form, is supplied as @file{moof-exm.fs}
1.29 crook 10985: @comment TODO could flesh this out with some comments from the Forthwrite article
1.21 crook 10986:
1.26 crook 10987: @example
10988: object class
10989: method init
10990: method draw
10991: end-class graphical
10992: @end example
1.21 crook 10993:
1.26 crook 10994: This code defines a class @code{graphical} with an
10995: operation @code{draw}. We can perform the operation
10996: @code{draw} on any @code{graphical} object, e.g.:
1.1 anton 10997:
1.26 crook 10998: @example
10999: 100 100 t-rex draw
11000: @end example
1.1 anton 11001:
1.26 crook 11002: where @code{t-rex} is an object or object pointer, created with e.g.
11003: @code{graphical new Constant t-rex}.
1.1 anton 11004:
1.26 crook 11005: For concrete graphical objects, we define child classes of the
11006: class @code{graphical}, e.g.:
1.21 crook 11007:
11008: @example
1.26 crook 11009: graphical class
11010: cell var circle-radius
11011: end-class circle \ "graphical" is the parent class
1.21 crook 11012:
1.26 crook 11013: :noname ( x y -- )
11014: circle-radius @@ draw-circle ; circle defines draw
11015: :noname ( r -- )
11016: circle-radius ! ; circle defines init
1.21 crook 11017: @end example
11018:
1.26 crook 11019: There is no implicit init method, so we have to define one. The creation
11020: code of the object now has to call init explicitely.
1.21 crook 11021:
1.26 crook 11022: @example
11023: circle new Constant my-circle
11024: 50 my-circle init
11025: @end example
1.21 crook 11026:
1.26 crook 11027: It is also possible to add a function to create named objects with
11028: automatic call of @code{init}, given that all objects have @code{init}
11029: on the same place:
1.1 anton 11030:
11031: @example
1.26 crook 11032: : new: ( .. o "name" -- )
11033: new dup Constant init ;
11034: 80 circle new: large-circle
1.1 anton 11035: @end example
11036:
1.26 crook 11037: We can draw this new circle at (100,100) with:
1.1 anton 11038:
11039: @example
1.26 crook 11040: 100 100 my-circle draw
1.1 anton 11041: @end example
11042:
1.48 anton 11043: @node Mini-OOF Implementation, , Mini-OOF Example, Mini-OOF
1.26 crook 11044: @subsubsection @file{mini-oof.fs} Implementation
1.1 anton 11045:
1.26 crook 11046: Object-oriented systems with late binding typically use a
11047: ``vtable''-approach: the first variable in each object is a pointer to a
11048: table, which contains the methods as function pointers. The vtable
11049: may also contain other information.
1.1 anton 11050:
1.26 crook 11051: So first, let's declare methods:
1.1 anton 11052:
1.26 crook 11053: @example
11054: : method ( m v -- m' v ) Create over , swap cell+ swap
11055: DOES> ( ... o -- ... ) @ over @ + @ execute ;
11056: @end example
1.1 anton 11057:
1.26 crook 11058: During method declaration, the number of methods and instance
11059: variables is on the stack (in address units). @code{method} creates
11060: one method and increments the method number. To execute a method, it
11061: takes the object, fetches the vtable pointer, adds the offset, and
1.29 crook 11062: executes the @i{xt} stored there. Each method takes the object it is
1.26 crook 11063: invoked from as top of stack parameter. The method itself should
11064: consume that object.
1.1 anton 11065:
1.26 crook 11066: Now, we also have to declare instance variables
1.21 crook 11067:
1.26 crook 11068: @example
11069: : var ( m v size -- m v' ) Create over , +
11070: DOES> ( o -- addr ) @ + ;
11071: @end example
1.21 crook 11072:
1.26 crook 11073: As before, a word is created with the current offset. Instance
11074: variables can have different sizes (cells, floats, doubles, chars), so
11075: all we do is take the size and add it to the offset. If your machine
11076: has alignment restrictions, put the proper @code{aligned} or
11077: @code{faligned} before the variable, to adjust the variable
11078: offset. That's why it is on the top of stack.
1.2 jwilke 11079:
1.26 crook 11080: We need a starting point (the base object) and some syntactic sugar:
1.21 crook 11081:
1.26 crook 11082: @example
11083: Create object 1 cells , 2 cells ,
11084: : class ( class -- class methods vars ) dup 2@ ;
11085: @end example
1.21 crook 11086:
1.26 crook 11087: For inheritance, the vtable of the parent object has to be
11088: copied when a new, derived class is declared. This gives all the
11089: methods of the parent class, which can be overridden, though.
1.21 crook 11090:
1.2 jwilke 11091: @example
1.26 crook 11092: : end-class ( class methods vars -- )
11093: Create here >r , dup , 2 cells ?DO ['] noop , 1 cells +LOOP
11094: cell+ dup cell+ r> rot @ 2 cells /string move ;
11095: @end example
11096:
11097: The first line creates the vtable, initialized with
11098: @code{noop}s. The second line is the inheritance mechanism, it
11099: copies the xts from the parent vtable.
1.2 jwilke 11100:
1.26 crook 11101: We still have no way to define new methods, let's do that now:
1.2 jwilke 11102:
1.26 crook 11103: @example
11104: : defines ( xt class -- ) ' >body @ + ! ;
1.2 jwilke 11105: @end example
11106:
1.26 crook 11107: To allocate a new object, we need a word, too:
1.2 jwilke 11108:
1.26 crook 11109: @example
11110: : new ( class -- o ) here over @ allot swap over ! ;
11111: @end example
1.2 jwilke 11112:
1.26 crook 11113: Sometimes derived classes want to access the method of the
11114: parent object. There are two ways to achieve this with Mini-OOF:
11115: first, you could use named words, and second, you could look up the
11116: vtable of the parent object.
1.2 jwilke 11117:
1.26 crook 11118: @example
11119: : :: ( class "name" -- ) ' >body @ + @ compile, ;
11120: @end example
1.2 jwilke 11121:
11122:
1.26 crook 11123: Nothing can be more confusing than a good example, so here is
11124: one. First let's declare a text object (called
11125: @code{button}), that stores text and position:
1.2 jwilke 11126:
1.26 crook 11127: @example
11128: object class
11129: cell var text
11130: cell var len
11131: cell var x
11132: cell var y
11133: method init
11134: method draw
11135: end-class button
11136: @end example
1.2 jwilke 11137:
1.26 crook 11138: @noindent
11139: Now, implement the two methods, @code{draw} and @code{init}:
1.2 jwilke 11140:
1.26 crook 11141: @example
11142: :noname ( o -- )
11143: >r r@ x @ r@ y @ at-xy r@ text @ r> len @ type ;
11144: button defines draw
11145: :noname ( addr u o -- )
11146: >r 0 r@ x ! 0 r@ y ! r@ len ! r> text ! ;
11147: button defines init
11148: @end example
1.2 jwilke 11149:
1.26 crook 11150: @noindent
11151: To demonstrate inheritance, we define a class @code{bold-button}, with no
11152: new data and no new methods:
1.2 jwilke 11153:
1.26 crook 11154: @example
11155: button class
11156: end-class bold-button
1.1 anton 11157:
1.26 crook 11158: : bold 27 emit ." [1m" ;
11159: : normal 27 emit ." [0m" ;
11160: @end example
1.1 anton 11161:
1.26 crook 11162: @noindent
11163: The class @code{bold-button} has a different draw method to
11164: @code{button}, but the new method is defined in terms of the draw method
11165: for @code{button}:
1.1 anton 11166:
1.26 crook 11167: @example
11168: :noname bold [ button :: draw ] normal ; bold-button defines draw
11169: @end example
1.1 anton 11170:
1.26 crook 11171: @noindent
11172: Finally, create two objects and apply methods:
1.1 anton 11173:
1.26 crook 11174: @example
11175: button new Constant foo
11176: s" thin foo" foo init
11177: page
11178: foo draw
11179: bold-button new Constant bar
11180: s" fat bar" bar init
11181: 1 bar y !
11182: bar draw
11183: @end example
1.1 anton 11184:
11185:
1.48 anton 11186: @node Comparison with other object models, , Mini-OOF, Object-oriented Forth
11187: @subsection Comparison with other object models
1.26 crook 11188: @cindex comparison of object models
11189: @cindex object models, comparison
1.1 anton 11190:
1.26 crook 11191: Many object-oriented Forth extensions have been proposed (@cite{A survey
11192: of object-oriented Forths} (SIGPLAN Notices, April 1996) by Bradford
11193: J. Rodriguez and W. F. S. Poehlman lists 17). This section discusses the
11194: relation of the object models described here to two well-known and two
11195: closely-related (by the use of method maps) models.
1.1 anton 11196:
1.26 crook 11197: @cindex Neon model
11198: The most popular model currently seems to be the Neon model (see
11199: @cite{Object-oriented programming in ANS Forth} (Forth Dimensions, March
11200: 1997) by Andrew McKewan) but this model has a number of limitations
11201: @footnote{A longer version of this critique can be
11202: found in @cite{On Standardizing Object-Oriented Forth Extensions} (Forth
11203: Dimensions, May 1997) by Anton Ertl.}:
1.1 anton 11204:
1.26 crook 11205: @itemize @bullet
11206: @item
1.48 anton 11207: It uses a @code{@emph{selector object}} syntax, which makes it unnatural
11208: to pass objects on the stack.
1.1 anton 11209:
1.26 crook 11210: @item
11211: It requires that the selector parses the input stream (at
11212: compile time); this leads to reduced extensibility and to bugs that are+
11213: hard to find.
1.1 anton 11214:
1.26 crook 11215: @item
11216: It allows using every selector to every object;
11217: this eliminates the need for classes, but makes it harder to create
11218: efficient implementations.
11219: @end itemize
1.1 anton 11220:
1.26 crook 11221: @cindex Pountain's object-oriented model
11222: Another well-known publication is @cite{Object-Oriented Forth} (Academic
11223: Press, London, 1987) by Dick Pountain. However, it is not really about
11224: object-oriented programming, because it hardly deals with late
11225: binding. Instead, it focuses on features like information hiding and
11226: overloading that are characteristic of modular languages like Ada (83).
1.1 anton 11227:
1.26 crook 11228: @cindex Zsoter's object-oriented model
1.48 anton 11229: In @cite{Does late binding have to be slow?} (Forth Dimensions 18(1)
11230: 1996, pages 31-35) Andras Zsoter describes a model that makes heavy use
11231: of an active object (like @code{this} in @file{objects.fs}): The active
11232: object is not only used for accessing all fields, but also specifies the
11233: receiving object of every selector invocation; you have to change the
11234: active object explicitly with @code{@{ ... @}}, whereas in
11235: @file{objects.fs} it changes more or less implicitly at @code{m:
11236: ... ;m}. Such a change at the method entry point is unnecessary with the
11237: Zsoter's model, because the receiving object is the active object
11238: already. On the other hand, the explicit change is absolutely necessary
11239: in that model, because otherwise no one could ever change the active
11240: object. An ANS Forth implementation of this model is available at
11241: @uref{http://www.forth.org/fig/oopf.html}.
1.1 anton 11242:
1.26 crook 11243: @cindex @file{oof.fs}, differences to other models
11244: The @file{oof.fs} model combines information hiding and overloading
11245: resolution (by keeping names in various word lists) with object-oriented
11246: programming. It sets the active object implicitly on method entry, but
11247: also allows explicit changing (with @code{>o...o>} or with
11248: @code{with...endwith}). It uses parsing and state-smart objects and
11249: classes for resolving overloading and for early binding: the object or
11250: class parses the selector and determines the method from this. If the
11251: selector is not parsed by an object or class, it performs a call to the
11252: selector for the active object (late binding), like Zsoter's model.
11253: Fields are always accessed through the active object. The big
11254: disadvantage of this model is the parsing and the state-smartness, which
11255: reduces extensibility and increases the opportunities for subtle bugs;
11256: essentially, you are only safe if you never tick or @code{postpone} an
11257: object or class (Bernd disagrees, but I (Anton) am not convinced).
1.1 anton 11258:
1.26 crook 11259: @cindex @file{mini-oof.fs}, differences to other models
1.48 anton 11260: The @file{mini-oof.fs} model is quite similar to a very stripped-down
11261: version of the @file{objects.fs} model, but syntactically it is a
11262: mixture of the @file{objects.fs} and @file{oof.fs} models.
1.1 anton 11263:
1.26 crook 11264: @c -------------------------------------------------------------
1.47 crook 11265: @node Passing Commands to the OS, Keeping track of Time, Object-oriented Forth, Words
1.21 crook 11266: @section Passing Commands to the Operating System
11267: @cindex operating system - passing commands
11268: @cindex shell commands
11269:
11270: Gforth allows you to pass an arbitrary string to the host operating
11271: system shell (if such a thing exists) for execution.
11272:
1.44 crook 11273:
1.21 crook 11274: doc-sh
11275: doc-system
11276: doc-$?
1.23 crook 11277: doc-getenv
1.21 crook 11278:
1.44 crook 11279:
1.26 crook 11280: @c -------------------------------------------------------------
1.47 crook 11281: @node Keeping track of Time, Miscellaneous Words, Passing Commands to the OS, Words
11282: @section Keeping track of Time
11283: @cindex time-related words
11284:
11285: Gforth implements time-related operations by making calls to the C
11286: library function, @code{gettimeofday}.
11287:
11288: doc-ms
11289: doc-time&date
11290:
11291:
11292:
11293: @c -------------------------------------------------------------
11294: @node Miscellaneous Words, , Keeping track of Time, Words
1.21 crook 11295: @section Miscellaneous Words
11296: @cindex miscellaneous words
11297:
1.29 crook 11298: @comment TODO find homes for these
11299:
1.26 crook 11300: These section lists the ANS Forth words that are not documented
1.21 crook 11301: elsewhere in this manual. Ultimately, they all need proper homes.
11302:
11303: doc-[compile]
11304:
1.44 crook 11305:
1.26 crook 11306: The following ANS Forth words are not currently supported by Gforth
1.27 crook 11307: (@pxref{ANS conformance}):
1.21 crook 11308:
11309: @code{EDITOR}
11310: @code{EMIT?}
11311: @code{FORGET}
11312:
1.24 anton 11313: @c ******************************************************************
11314: @node Error messages, Tools, Words, Top
11315: @chapter Error messages
11316: @cindex error messages
11317: @cindex backtrace
11318:
11319: A typical Gforth error message looks like this:
11320:
11321: @example
11322: in file included from :-1
11323: in file included from ./yyy.fs:1
11324: ./xxx.fs:4: Invalid memory address
11325: bar
11326: ^^^
1.25 anton 11327: $400E664C @@
11328: $400E6664 foo
1.24 anton 11329: @end example
11330:
11331: The message identifying the error is @code{Invalid memory address}. The
11332: error happened when text-interpreting line 4 of the file
11333: @file{./xxx.fs}. This line is given (it contains @code{bar}), and the
11334: word on the line where the error happened, is pointed out (with
11335: @code{^^^}).
11336:
11337: The file containing the error was included in line 1 of @file{./yyy.fs},
11338: and @file{yyy.fs} was included from a non-file (in this case, by giving
11339: @file{yyy.fs} as command-line parameter to Gforth).
11340:
11341: At the end of the error message you find a return stack dump that can be
11342: interpreted as a backtrace (possibly empty). On top you find the top of
11343: the return stack when the @code{throw} happened, and at the bottom you
11344: find the return stack entry just above the return stack of the topmost
11345: text interpreter.
11346:
11347: To the right of most return stack entries you see a guess for the word
11348: that pushed that return stack entry as its return address. This gives a
11349: backtrace. In our case we see that @code{bar} called @code{foo}, and
11350: @code{foo} called @code{@@} (and @code{@@} had an @emph{Invalid memory
11351: address} exception).
11352:
11353: Note that the backtrace is not perfect: We don't know which return stack
11354: entries are return addresses (so we may get false positives); and in
11355: some cases (e.g., for @code{abort"}) we cannot determine from the return
11356: address the word that pushed the return address, so for some return
11357: addresses you see no names in the return stack dump.
1.25 anton 11358:
11359: @cindex @code{catch} and backtraces
11360: The return stack dump represents the return stack at the time when a
11361: specific @code{throw} was executed. In programs that make use of
11362: @code{catch}, it is not necessarily clear which @code{throw} should be
11363: used for the return stack dump (e.g., consider one @code{throw} that
11364: indicates an error, which is caught, and during recovery another error
1.42 anton 11365: happens; which @code{throw} should be used for the stack dump?). Gforth
1.25 anton 11366: presents the return stack dump for the first @code{throw} after the last
11367: executed (not returned-to) @code{catch}; this works well in the usual
11368: case.
11369:
11370: @cindex @code{gforth-fast} and backtraces
11371: @cindex @code{gforth-fast}, difference from @code{gforth}
11372: @cindex backtraces with @code{gforth-fast}
11373: @cindex return stack dump with @code{gforth-fast}
11374: @code{gforth} is able to do a return stack dump for throws generated
11375: from primitives (e.g., invalid memory address, stack empty etc.);
11376: @code{gforth-fast} is only able to do a return stack dump from a
11377: directly called @code{throw} (including @code{abort} etc.). This is the
1.30 anton 11378: only difference (apart from a speed factor of between 1.15 (K6-2) and
11379: 1.6 (21164A)) between @code{gforth} and @code{gforth-fast}. Given an
11380: exception caused by a primitive in @code{gforth-fast}, you will
11381: typically see no return stack dump at all; however, if the exception is
11382: caught by @code{catch} (e.g., for restoring some state), and then
11383: @code{throw}n again, the return stack dump will be for the first such
11384: @code{throw}.
1.2 jwilke 11385:
1.5 anton 11386: @c ******************************************************************
1.24 anton 11387: @node Tools, ANS conformance, Error messages, Top
1.1 anton 11388: @chapter Tools
11389:
11390: @menu
11391: * ANS Report:: Report the words used, sorted by wordset.
11392: @end menu
11393:
11394: See also @ref{Emacs and Gforth}.
11395:
11396: @node ANS Report, , Tools, Tools
11397: @section @file{ans-report.fs}: Report the words used, sorted by wordset
11398: @cindex @file{ans-report.fs}
11399: @cindex report the words used in your program
11400: @cindex words used in your program
11401:
11402: If you want to label a Forth program as ANS Forth Program, you must
11403: document which wordsets the program uses; for extension wordsets, it is
11404: helpful to list the words the program requires from these wordsets
11405: (because Forth systems are allowed to provide only some words of them).
11406:
11407: The @file{ans-report.fs} tool makes it easy for you to determine which
11408: words from which wordset and which non-ANS words your application
11409: uses. You simply have to include @file{ans-report.fs} before loading the
11410: program you want to check. After loading your program, you can get the
11411: report with @code{print-ans-report}. A typical use is to run this as
11412: batch job like this:
11413: @example
11414: gforth ans-report.fs myprog.fs -e "print-ans-report bye"
11415: @end example
11416:
11417: The output looks like this (for @file{compat/control.fs}):
11418: @example
11419: The program uses the following words
11420: from CORE :
11421: : POSTPONE THEN ; immediate ?dup IF 0=
11422: from BLOCK-EXT :
11423: \
11424: from FILE :
11425: (
11426: @end example
11427:
11428: @subsection Caveats
11429:
11430: Note that @file{ans-report.fs} just checks which words are used, not whether
11431: they are used in an ANS Forth conforming way!
11432:
11433: Some words are defined in several wordsets in the
11434: standard. @file{ans-report.fs} reports them for only one of the
11435: wordsets, and not necessarily the one you expect. It depends on usage
11436: which wordset is the right one to specify. E.g., if you only use the
11437: compilation semantics of @code{S"}, it is a Core word; if you also use
11438: its interpretation semantics, it is a File word.
11439:
11440: @c ******************************************************************
11441: @node ANS conformance, Model, Tools, Top
11442: @chapter ANS conformance
11443: @cindex ANS conformance of Gforth
11444:
11445: To the best of our knowledge, Gforth is an
11446:
11447: ANS Forth System
11448: @itemize @bullet
11449: @item providing the Core Extensions word set
11450: @item providing the Block word set
11451: @item providing the Block Extensions word set
11452: @item providing the Double-Number word set
11453: @item providing the Double-Number Extensions word set
11454: @item providing the Exception word set
11455: @item providing the Exception Extensions word set
11456: @item providing the Facility word set
1.40 anton 11457: @item providing @code{EKEY}, @code{EKEY>CHAR}, @code{EKEY?}, @code{MS} and @code{TIME&DATE} from the Facility Extensions word set
1.1 anton 11458: @item providing the File Access word set
11459: @item providing the File Access Extensions word set
11460: @item providing the Floating-Point word set
11461: @item providing the Floating-Point Extensions word set
11462: @item providing the Locals word set
11463: @item providing the Locals Extensions word set
11464: @item providing the Memory-Allocation word set
11465: @item providing the Memory-Allocation Extensions word set (that one's easy)
11466: @item providing the Programming-Tools word set
11467: @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
11468: @item providing the Search-Order word set
11469: @item providing the Search-Order Extensions word set
11470: @item providing the String word set
11471: @item providing the String Extensions word set (another easy one)
11472: @end itemize
11473:
11474: @cindex system documentation
11475: In addition, ANS Forth systems are required to document certain
11476: implementation choices. This chapter tries to meet these
11477: requirements. In many cases it gives a way to ask the system for the
11478: information instead of providing the information directly, in
11479: particular, if the information depends on the processor, the operating
11480: system or the installation options chosen, or if they are likely to
11481: change during the maintenance of Gforth.
11482:
11483: @comment The framework for the rest has been taken from pfe.
11484:
11485: @menu
11486: * The Core Words::
11487: * The optional Block word set::
11488: * The optional Double Number word set::
11489: * The optional Exception word set::
11490: * The optional Facility word set::
11491: * The optional File-Access word set::
11492: * The optional Floating-Point word set::
11493: * The optional Locals word set::
11494: * The optional Memory-Allocation word set::
11495: * The optional Programming-Tools word set::
11496: * The optional Search-Order word set::
11497: @end menu
11498:
11499:
11500: @c =====================================================================
11501: @node The Core Words, The optional Block word set, ANS conformance, ANS conformance
11502: @comment node-name, next, previous, up
11503: @section The Core Words
11504: @c =====================================================================
11505: @cindex core words, system documentation
11506: @cindex system documentation, core words
11507:
11508: @menu
11509: * core-idef:: Implementation Defined Options
11510: * core-ambcond:: Ambiguous Conditions
11511: * core-other:: Other System Documentation
11512: @end menu
11513:
11514: @c ---------------------------------------------------------------------
11515: @node core-idef, core-ambcond, The Core Words, The Core Words
11516: @subsection Implementation Defined Options
11517: @c ---------------------------------------------------------------------
11518: @cindex core words, implementation-defined options
11519: @cindex implementation-defined options, core words
11520:
11521:
11522: @table @i
11523: @item (Cell) aligned addresses:
11524: @cindex cell-aligned addresses
11525: @cindex aligned addresses
11526: processor-dependent. Gforth's alignment words perform natural alignment
11527: (e.g., an address aligned for a datum of size 8 is divisible by
11528: 8). Unaligned accesses usually result in a @code{-23 THROW}.
11529:
11530: @item @code{EMIT} and non-graphic characters:
11531: @cindex @code{EMIT} and non-graphic characters
11532: @cindex non-graphic characters and @code{EMIT}
11533: The character is output using the C library function (actually, macro)
11534: @code{putc}.
11535:
11536: @item character editing of @code{ACCEPT} and @code{EXPECT}:
11537: @cindex character editing of @code{ACCEPT} and @code{EXPECT}
11538: @cindex editing in @code{ACCEPT} and @code{EXPECT}
11539: @cindex @code{ACCEPT}, editing
11540: @cindex @code{EXPECT}, editing
11541: This is modeled on the GNU readline library (@pxref{Readline
11542: Interaction, , Command Line Editing, readline, The GNU Readline
11543: Library}) with Emacs-like key bindings. @kbd{Tab} deviates a little by
11544: producing a full word completion every time you type it (instead of
1.28 crook 11545: producing the common prefix of all completions). @xref{Command-line editing}.
1.1 anton 11546:
11547: @item character set:
11548: @cindex character set
11549: The character set of your computer and display device. Gforth is
11550: 8-bit-clean (but some other component in your system may make trouble).
11551:
11552: @item Character-aligned address requirements:
11553: @cindex character-aligned address requirements
11554: installation-dependent. Currently a character is represented by a C
11555: @code{unsigned char}; in the future we might switch to @code{wchar_t}
11556: (Comments on that requested).
11557:
11558: @item character-set extensions and matching of names:
11559: @cindex character-set extensions and matching of names
1.26 crook 11560: @cindex case-sensitivity for name lookup
11561: @cindex name lookup, case-sensitivity
11562: @cindex locale and case-sensitivity
1.21 crook 11563: Any character except the ASCII NUL character can be used in a
1.1 anton 11564: name. Matching is case-insensitive (except in @code{TABLE}s). The
1.47 crook 11565: matching is performed using the C library function @code{strncasecmp}, whose
1.1 anton 11566: function is probably influenced by the locale. E.g., the @code{C} locale
11567: does not know about accents and umlauts, so they are matched
11568: case-sensitively in that locale. For portability reasons it is best to
11569: write programs such that they work in the @code{C} locale. Then one can
11570: use libraries written by a Polish programmer (who might use words
11571: containing ISO Latin-2 encoded characters) and by a French programmer
11572: (ISO Latin-1) in the same program (of course, @code{WORDS} will produce
11573: funny results for some of the words (which ones, depends on the font you
11574: are using)). Also, the locale you prefer may not be available in other
11575: operating systems. Hopefully, Unicode will solve these problems one day.
11576:
11577: @item conditions under which control characters match a space delimiter:
11578: @cindex space delimiters
11579: @cindex control characters as delimiters
11580: If @code{WORD} is called with the space character as a delimiter, all
11581: white-space characters (as identified by the C macro @code{isspace()})
11582: are delimiters. @code{PARSE}, on the other hand, treats space like other
1.44 crook 11583: delimiters. @code{SWORD} treats space like @code{WORD}, but behaves
1.1 anton 11584: like @code{PARSE} otherwise. @code{(NAME)}, which is used by the outer
11585: interpreter (aka text interpreter) by default, treats all white-space
11586: characters as delimiters.
11587:
1.26 crook 11588: @item format of the control-flow stack:
11589: @cindex control-flow stack, format
11590: The data stack is used as control-flow stack. The size of a control-flow
1.1 anton 11591: stack item in cells is given by the constant @code{cs-item-size}. At the
11592: time of this writing, an item consists of a (pointer to a) locals list
11593: (third), an address in the code (second), and a tag for identifying the
11594: item (TOS). The following tags are used: @code{defstart},
11595: @code{live-orig}, @code{dead-orig}, @code{dest}, @code{do-dest},
11596: @code{scopestart}.
11597:
11598: @item conversion of digits > 35
11599: @cindex digits > 35
11600: The characters @code{[\]^_'} are the digits with the decimal value
11601: 36@minus{}41. There is no way to input many of the larger digits.
11602:
11603: @item display after input terminates in @code{ACCEPT} and @code{EXPECT}:
11604: @cindex @code{EXPECT}, display after end of input
11605: @cindex @code{ACCEPT}, display after end of input
11606: The cursor is moved to the end of the entered string. If the input is
11607: terminated using the @kbd{Return} key, a space is typed.
11608:
11609: @item exception abort sequence of @code{ABORT"}:
11610: @cindex exception abort sequence of @code{ABORT"}
11611: @cindex @code{ABORT"}, exception abort sequence
11612: The error string is stored into the variable @code{"error} and a
11613: @code{-2 throw} is performed.
11614:
11615: @item input line terminator:
11616: @cindex input line terminator
11617: @cindex line terminator on input
1.26 crook 11618: @cindex newline character on input
1.1 anton 11619: For interactive input, @kbd{C-m} (CR) and @kbd{C-j} (LF) terminate
11620: lines. One of these characters is typically produced when you type the
11621: @kbd{Enter} or @kbd{Return} key.
11622:
11623: @item maximum size of a counted string:
11624: @cindex maximum size of a counted string
11625: @cindex counted string, maximum size
11626: @code{s" /counted-string" environment? drop .}. Currently 255 characters
11627: on all ports, but this may change.
11628:
11629: @item maximum size of a parsed string:
11630: @cindex maximum size of a parsed string
11631: @cindex parsed string, maximum size
11632: Given by the constant @code{/line}. Currently 255 characters.
11633:
11634: @item maximum size of a definition name, in characters:
11635: @cindex maximum size of a definition name, in characters
11636: @cindex name, maximum length
11637: 31
11638:
11639: @item maximum string length for @code{ENVIRONMENT?}, in characters:
11640: @cindex maximum string length for @code{ENVIRONMENT?}, in characters
11641: @cindex @code{ENVIRONMENT?} string length, maximum
11642: 31
11643:
11644: @item method of selecting the user input device:
11645: @cindex user input device, method of selecting
11646: The user input device is the standard input. There is currently no way to
11647: change it from within Gforth. However, the input can typically be
11648: redirected in the command line that starts Gforth.
11649:
11650: @item method of selecting the user output device:
11651: @cindex user output device, method of selecting
11652: @code{EMIT} and @code{TYPE} output to the file-id stored in the value
1.10 anton 11653: @code{outfile-id} (@code{stdout} by default). Gforth uses unbuffered
11654: output when the user output device is a terminal, otherwise the output
11655: is buffered.
1.1 anton 11656:
11657: @item methods of dictionary compilation:
11658: What are we expected to document here?
11659:
11660: @item number of bits in one address unit:
11661: @cindex number of bits in one address unit
11662: @cindex address unit, size in bits
11663: @code{s" address-units-bits" environment? drop .}. 8 in all current
11664: ports.
11665:
11666: @item number representation and arithmetic:
11667: @cindex number representation and arithmetic
11668: Processor-dependent. Binary two's complement on all current ports.
11669:
11670: @item ranges for integer types:
11671: @cindex ranges for integer types
11672: @cindex integer types, ranges
11673: Installation-dependent. Make environmental queries for @code{MAX-N},
11674: @code{MAX-U}, @code{MAX-D} and @code{MAX-UD}. The lower bounds for
11675: unsigned (and positive) types is 0. The lower bound for signed types on
11676: two's complement and one's complement machines machines can be computed
11677: by adding 1 to the upper bound.
11678:
11679: @item read-only data space regions:
11680: @cindex read-only data space regions
11681: @cindex data-space, read-only regions
11682: The whole Forth data space is writable.
11683:
11684: @item size of buffer at @code{WORD}:
11685: @cindex size of buffer at @code{WORD}
11686: @cindex @code{WORD} buffer size
11687: @code{PAD HERE - .}. 104 characters on 32-bit machines. The buffer is
11688: shared with the pictured numeric output string. If overwriting
11689: @code{PAD} is acceptable, it is as large as the remaining dictionary
11690: space, although only as much can be sensibly used as fits in a counted
11691: string.
11692:
11693: @item size of one cell in address units:
11694: @cindex cell size
11695: @code{1 cells .}.
11696:
11697: @item size of one character in address units:
11698: @cindex char size
11699: @code{1 chars .}. 1 on all current ports.
11700:
11701: @item size of the keyboard terminal buffer:
11702: @cindex size of the keyboard terminal buffer
11703: @cindex terminal buffer, size
11704: Varies. You can determine the size at a specific time using @code{lp@@
11705: tib - .}. It is shared with the locals stack and TIBs of files that
11706: include the current file. You can change the amount of space for TIBs
11707: and locals stack at Gforth startup with the command line option
11708: @code{-l}.
11709:
11710: @item size of the pictured numeric output buffer:
11711: @cindex size of the pictured numeric output buffer
11712: @cindex pictured numeric output buffer, size
11713: @code{PAD HERE - .}. 104 characters on 32-bit machines. The buffer is
11714: shared with @code{WORD}.
11715:
11716: @item size of the scratch area returned by @code{PAD}:
11717: @cindex size of the scratch area returned by @code{PAD}
11718: @cindex @code{PAD} size
11719: The remainder of dictionary space. @code{unused pad here - - .}.
11720:
11721: @item system case-sensitivity characteristics:
11722: @cindex case-sensitivity characteristics
1.26 crook 11723: Dictionary searches are case-insensitive (except in
1.1 anton 11724: @code{TABLE}s). However, as explained above under @i{character-set
11725: extensions}, the matching for non-ASCII characters is determined by the
11726: locale you are using. In the default @code{C} locale all non-ASCII
11727: characters are matched case-sensitively.
11728:
11729: @item system prompt:
11730: @cindex system prompt
11731: @cindex prompt
11732: @code{ ok} in interpret state, @code{ compiled} in compile state.
11733:
11734: @item division rounding:
11735: @cindex division rounding
11736: installation dependent. @code{s" floored" environment? drop .}. We leave
11737: the choice to @code{gcc} (what to use for @code{/}) and to you (whether
11738: to use @code{fm/mod}, @code{sm/rem} or simply @code{/}).
11739:
11740: @item values of @code{STATE} when true:
11741: @cindex @code{STATE} values
11742: -1.
11743:
11744: @item values returned after arithmetic overflow:
11745: On two's complement machines, arithmetic is performed modulo
11746: 2**bits-per-cell for single arithmetic and 4**bits-per-cell for double
11747: arithmetic (with appropriate mapping for signed types). Division by zero
11748: typically results in a @code{-55 throw} (Floating-point unidentified
11749: fault), although a @code{-10 throw} (divide by zero) would be more
11750: appropriate.
11751:
11752: @item whether the current definition can be found after @t{DOES>}:
11753: @cindex @t{DOES>}, visibility of current definition
11754: No.
11755:
11756: @end table
11757:
11758: @c ---------------------------------------------------------------------
11759: @node core-ambcond, core-other, core-idef, The Core Words
11760: @subsection Ambiguous conditions
11761: @c ---------------------------------------------------------------------
11762: @cindex core words, ambiguous conditions
11763: @cindex ambiguous conditions, core words
11764:
11765: @table @i
11766:
11767: @item a name is neither a word nor a number:
11768: @cindex name not found
1.26 crook 11769: @cindex undefined word
1.1 anton 11770: @code{-13 throw} (Undefined word). Actually, @code{-13 bounce}, which
11771: preserves the data and FP stack, so you don't lose more work than
11772: necessary.
11773:
11774: @item a definition name exceeds the maximum length allowed:
1.26 crook 11775: @cindex word name too long
1.1 anton 11776: @code{-19 throw} (Word name too long)
11777:
11778: @item addressing a region not inside the various data spaces of the forth system:
11779: @cindex Invalid memory address
1.32 anton 11780: The stacks, code space and header space are accessible. Machine code space is
1.1 anton 11781: typically readable. Accessing other addresses gives results dependent on
11782: the operating system. On decent systems: @code{-9 throw} (Invalid memory
11783: address).
11784:
11785: @item argument type incompatible with parameter:
1.26 crook 11786: @cindex argument type mismatch
1.1 anton 11787: This is usually not caught. Some words perform checks, e.g., the control
11788: flow words, and issue a @code{ABORT"} or @code{-12 THROW} (Argument type
11789: mismatch).
11790:
11791: @item attempting to obtain the execution token of a word with undefined execution semantics:
11792: @cindex Interpreting a compile-only word, for @code{'} etc.
11793: @cindex execution token of words with undefined execution semantics
11794: @code{-14 throw} (Interpreting a compile-only word). In some cases, you
11795: get an execution token for @code{compile-only-error} (which performs a
11796: @code{-14 throw} when executed).
11797:
11798: @item dividing by zero:
11799: @cindex dividing by zero
11800: @cindex floating point unidentified fault, integer division
1.24 anton 11801: On better platforms, this produces a @code{-10 throw} (Division by
11802: zero); on other systems, this typically results in a @code{-55 throw}
11803: (Floating-point unidentified fault).
1.1 anton 11804:
11805: @item insufficient data stack or return stack space:
11806: @cindex insufficient data stack or return stack space
11807: @cindex stack overflow
1.26 crook 11808: @cindex address alignment exception, stack overflow
1.1 anton 11809: @cindex Invalid memory address, stack overflow
11810: Depending on the operating system, the installation, and the invocation
11811: of Gforth, this is either checked by the memory management hardware, or
1.24 anton 11812: it is not checked. If it is checked, you typically get a @code{-3 throw}
11813: (Stack overflow), @code{-5 throw} (Return stack overflow), or @code{-9
11814: throw} (Invalid memory address) (depending on the platform and how you
11815: achieved the overflow) as soon as the overflow happens. If it is not
11816: checked, overflows typically result in mysterious illegal memory
11817: accesses, producing @code{-9 throw} (Invalid memory address) or
11818: @code{-23 throw} (Address alignment exception); they might also destroy
11819: the internal data structure of @code{ALLOCATE} and friends, resulting in
11820: various errors in these words.
1.1 anton 11821:
11822: @item insufficient space for loop control parameters:
11823: @cindex insufficient space for loop control parameters
11824: like other return stack overflows.
11825:
11826: @item insufficient space in the dictionary:
11827: @cindex insufficient space in the dictionary
11828: @cindex dictionary overflow
1.12 anton 11829: If you try to allot (either directly with @code{allot}, or indirectly
11830: with @code{,}, @code{create} etc.) more memory than available in the
11831: dictionary, you get a @code{-8 throw} (Dictionary overflow). If you try
11832: to access memory beyond the end of the dictionary, the results are
11833: similar to stack overflows.
1.1 anton 11834:
11835: @item interpreting a word with undefined interpretation semantics:
11836: @cindex interpreting a word with undefined interpretation semantics
11837: @cindex Interpreting a compile-only word
11838: For some words, we have defined interpretation semantics. For the
11839: others: @code{-14 throw} (Interpreting a compile-only word).
11840:
11841: @item modifying the contents of the input buffer or a string literal:
11842: @cindex modifying the contents of the input buffer or a string literal
11843: These are located in writable memory and can be modified.
11844:
11845: @item overflow of the pictured numeric output string:
11846: @cindex overflow of the pictured numeric output string
11847: @cindex pictured numeric output string, overflow
1.24 anton 11848: @code{-17 throw} (Pictured numeric ouput string overflow).
1.1 anton 11849:
11850: @item parsed string overflow:
11851: @cindex parsed string overflow
11852: @code{PARSE} cannot overflow. @code{WORD} does not check for overflow.
11853:
11854: @item producing a result out of range:
11855: @cindex result out of range
11856: On two's complement machines, arithmetic is performed modulo
11857: 2**bits-per-cell for single arithmetic and 4**bits-per-cell for double
11858: arithmetic (with appropriate mapping for signed types). Division by zero
1.24 anton 11859: typically results in a @code{-10 throw} (divide by zero) or @code{-55
11860: throw} (floating point unidentified fault). @code{convert} and
11861: @code{>number} currently overflow silently.
1.1 anton 11862:
11863: @item reading from an empty data or return stack:
11864: @cindex stack empty
11865: @cindex stack underflow
1.24 anton 11866: @cindex return stack underflow
1.1 anton 11867: The data stack is checked by the outer (aka text) interpreter after
11868: every word executed. If it has underflowed, a @code{-4 throw} (Stack
11869: underflow) is performed. Apart from that, stacks may be checked or not,
1.24 anton 11870: depending on operating system, installation, and invocation. If they are
11871: caught by a check, they typically result in @code{-4 throw} (Stack
11872: underflow), @code{-6 throw} (Return stack underflow) or @code{-9 throw}
11873: (Invalid memory address), depending on the platform and which stack
11874: underflows and by how much. Note that even if the system uses checking
11875: (through the MMU), your program may have to underflow by a significant
11876: number of stack items to trigger the reaction (the reason for this is
11877: that the MMU, and therefore the checking, works with a page-size
11878: granularity). If there is no checking, the symptoms resulting from an
11879: underflow are similar to those from an overflow. Unbalanced return
11880: stack errors result in a variaty of symptoms, including @code{-9 throw}
11881: (Invalid memory address) and Illegal Instruction (typically @code{-260
11882: throw}).
1.1 anton 11883:
11884: @item unexpected end of the input buffer, resulting in an attempt to use a zero-length string as a name:
11885: @cindex unexpected end of the input buffer
11886: @cindex zero-length string as a name
11887: @cindex Attempt to use zero-length string as a name
11888: @code{Create} and its descendants perform a @code{-16 throw} (Attempt to
11889: use zero-length string as a name). Words like @code{'} probably will not
11890: find what they search. Note that it is possible to create zero-length
11891: names with @code{nextname} (should it not?).
11892:
11893: @item @code{>IN} greater than input buffer:
11894: @cindex @code{>IN} greater than input buffer
11895: The next invocation of a parsing word returns a string with length 0.
11896:
11897: @item @code{RECURSE} appears after @code{DOES>}:
11898: @cindex @code{RECURSE} appears after @code{DOES>}
11899: Compiles a recursive call to the defining word, not to the defined word.
11900:
11901: @item argument input source different than current input source for @code{RESTORE-INPUT}:
11902: @cindex argument input source different than current input source for @code{RESTORE-INPUT}
1.26 crook 11903: @cindex argument type mismatch, @code{RESTORE-INPUT}
1.1 anton 11904: @cindex @code{RESTORE-INPUT}, Argument type mismatch
11905: @code{-12 THROW}. Note that, once an input file is closed (e.g., because
11906: the end of the file was reached), its source-id may be
11907: reused. Therefore, restoring an input source specification referencing a
11908: closed file may lead to unpredictable results instead of a @code{-12
11909: THROW}.
11910:
11911: In the future, Gforth may be able to restore input source specifications
11912: from other than the current input source.
11913:
11914: @item data space containing definitions gets de-allocated:
11915: @cindex data space containing definitions gets de-allocated
11916: Deallocation with @code{allot} is not checked. This typically results in
11917: memory access faults or execution of illegal instructions.
11918:
11919: @item data space read/write with incorrect alignment:
11920: @cindex data space read/write with incorrect alignment
11921: @cindex alignment faults
1.26 crook 11922: @cindex address alignment exception
1.1 anton 11923: Processor-dependent. Typically results in a @code{-23 throw} (Address
1.12 anton 11924: alignment exception). Under Linux-Intel on a 486 or later processor with
1.1 anton 11925: alignment turned on, incorrect alignment results in a @code{-9 throw}
11926: (Invalid memory address). There are reportedly some processors with
1.12 anton 11927: alignment restrictions that do not report violations.
1.1 anton 11928:
11929: @item data space pointer not properly aligned, @code{,}, @code{C,}:
11930: @cindex data space pointer not properly aligned, @code{,}, @code{C,}
11931: Like other alignment errors.
11932:
11933: @item less than u+2 stack items (@code{PICK} and @code{ROLL}):
11934: Like other stack underflows.
11935:
11936: @item loop control parameters not available:
11937: @cindex loop control parameters not available
11938: Not checked. The counted loop words simply assume that the top of return
11939: stack items are loop control parameters and behave accordingly.
11940:
11941: @item most recent definition does not have a name (@code{IMMEDIATE}):
11942: @cindex most recent definition does not have a name (@code{IMMEDIATE})
11943: @cindex last word was headerless
11944: @code{abort" last word was headerless"}.
11945:
11946: @item name not defined by @code{VALUE} used by @code{TO}:
11947: @cindex name not defined by @code{VALUE} used by @code{TO}
11948: @cindex @code{TO} on non-@code{VALUE}s
11949: @cindex Invalid name argument, @code{TO}
11950: @code{-32 throw} (Invalid name argument) (unless name is a local or was
11951: defined by @code{CONSTANT}; in the latter case it just changes the constant).
11952:
11953: @item name not found (@code{'}, @code{POSTPONE}, @code{[']}, @code{[COMPILE]}):
11954: @cindex name not found (@code{'}, @code{POSTPONE}, @code{[']}, @code{[COMPILE]})
1.26 crook 11955: @cindex undefined word, @code{'}, @code{POSTPONE}, @code{[']}, @code{[COMPILE]}
1.1 anton 11956: @code{-13 throw} (Undefined word)
11957:
11958: @item parameters are not of the same type (@code{DO}, @code{?DO}, @code{WITHIN}):
11959: @cindex parameters are not of the same type (@code{DO}, @code{?DO}, @code{WITHIN})
11960: Gforth behaves as if they were of the same type. I.e., you can predict
11961: the behaviour by interpreting all parameters as, e.g., signed.
11962:
11963: @item @code{POSTPONE} or @code{[COMPILE]} applied to @code{TO}:
11964: @cindex @code{POSTPONE} or @code{[COMPILE]} applied to @code{TO}
11965: Assume @code{: X POSTPONE TO ; IMMEDIATE}. @code{X} performs the
11966: compilation semantics of @code{TO}.
11967:
11968: @item String longer than a counted string returned by @code{WORD}:
1.26 crook 11969: @cindex string longer than a counted string returned by @code{WORD}
1.1 anton 11970: @cindex @code{WORD}, string overflow
11971: Not checked. The string will be ok, but the count will, of course,
11972: contain only the least significant bits of the length.
11973:
11974: @item u greater than or equal to the number of bits in a cell (@code{LSHIFT}, @code{RSHIFT}):
11975: @cindex @code{LSHIFT}, large shift counts
11976: @cindex @code{RSHIFT}, large shift counts
11977: Processor-dependent. Typical behaviours are returning 0 and using only
11978: the low bits of the shift count.
11979:
11980: @item word not defined via @code{CREATE}:
11981: @cindex @code{>BODY} of non-@code{CREATE}d words
11982: @code{>BODY} produces the PFA of the word no matter how it was defined.
11983:
11984: @cindex @code{DOES>} of non-@code{CREATE}d words
11985: @code{DOES>} changes the execution semantics of the last defined word no
11986: matter how it was defined. E.g., @code{CONSTANT DOES>} is equivalent to
11987: @code{CREATE , DOES>}.
11988:
11989: @item words improperly used outside @code{<#} and @code{#>}:
11990: Not checked. As usual, you can expect memory faults.
11991:
11992: @end table
11993:
11994:
11995: @c ---------------------------------------------------------------------
11996: @node core-other, , core-ambcond, The Core Words
11997: @subsection Other system documentation
11998: @c ---------------------------------------------------------------------
11999: @cindex other system documentation, core words
12000: @cindex core words, other system documentation
12001:
12002: @table @i
12003: @item nonstandard words using @code{PAD}:
12004: @cindex @code{PAD} use by nonstandard words
12005: None.
12006:
12007: @item operator's terminal facilities available:
12008: @cindex operator's terminal facilities available
12009: After processing the command line, Gforth goes into interactive mode,
12010: and you can give commands to Gforth interactively. The actual facilities
12011: available depend on how you invoke Gforth.
12012:
12013: @item program data space available:
12014: @cindex program data space available
12015: @cindex data space available
12016: @code{UNUSED .} gives the remaining dictionary space. The total
12017: dictionary space can be specified with the @code{-m} switch
12018: (@pxref{Invoking Gforth}) when Gforth starts up.
12019:
12020: @item return stack space available:
12021: @cindex return stack space available
12022: You can compute the total return stack space in cells with
12023: @code{s" RETURN-STACK-CELLS" environment? drop .}. You can specify it at
12024: startup time with the @code{-r} switch (@pxref{Invoking Gforth}).
12025:
12026: @item stack space available:
12027: @cindex stack space available
12028: You can compute the total data stack space in cells with
12029: @code{s" STACK-CELLS" environment? drop .}. You can specify it at
12030: startup time with the @code{-d} switch (@pxref{Invoking Gforth}).
12031:
12032: @item system dictionary space required, in address units:
12033: @cindex system dictionary space required, in address units
12034: Type @code{here forthstart - .} after startup. At the time of this
12035: writing, this gives 80080 (bytes) on a 32-bit system.
12036: @end table
12037:
12038:
12039: @c =====================================================================
12040: @node The optional Block word set, The optional Double Number word set, The Core Words, ANS conformance
12041: @section The optional Block word set
12042: @c =====================================================================
12043: @cindex system documentation, block words
12044: @cindex block words, system documentation
12045:
12046: @menu
12047: * block-idef:: Implementation Defined Options
12048: * block-ambcond:: Ambiguous Conditions
12049: * block-other:: Other System Documentation
12050: @end menu
12051:
12052:
12053: @c ---------------------------------------------------------------------
12054: @node block-idef, block-ambcond, The optional Block word set, The optional Block word set
12055: @subsection Implementation Defined Options
12056: @c ---------------------------------------------------------------------
12057: @cindex implementation-defined options, block words
12058: @cindex block words, implementation-defined options
12059:
12060: @table @i
12061: @item the format for display by @code{LIST}:
12062: @cindex @code{LIST} display format
12063: First the screen number is displayed, then 16 lines of 64 characters,
12064: each line preceded by the line number.
12065:
12066: @item the length of a line affected by @code{\}:
12067: @cindex length of a line affected by @code{\}
12068: @cindex @code{\}, line length in blocks
12069: 64 characters.
12070: @end table
12071:
12072:
12073: @c ---------------------------------------------------------------------
12074: @node block-ambcond, block-other, block-idef, The optional Block word set
12075: @subsection Ambiguous conditions
12076: @c ---------------------------------------------------------------------
12077: @cindex block words, ambiguous conditions
12078: @cindex ambiguous conditions, block words
12079:
12080: @table @i
12081: @item correct block read was not possible:
12082: @cindex block read not possible
12083: Typically results in a @code{throw} of some OS-derived value (between
12084: -512 and -2048). If the blocks file was just not long enough, blanks are
12085: supplied for the missing portion.
12086:
12087: @item I/O exception in block transfer:
12088: @cindex I/O exception in block transfer
12089: @cindex block transfer, I/O exception
12090: Typically results in a @code{throw} of some OS-derived value (between
12091: -512 and -2048).
12092:
12093: @item invalid block number:
12094: @cindex invalid block number
12095: @cindex block number invalid
12096: @code{-35 throw} (Invalid block number)
12097:
12098: @item a program directly alters the contents of @code{BLK}:
12099: @cindex @code{BLK}, altering @code{BLK}
12100: The input stream is switched to that other block, at the same
12101: position. If the storing to @code{BLK} happens when interpreting
12102: non-block input, the system will get quite confused when the block ends.
12103:
12104: @item no current block buffer for @code{UPDATE}:
12105: @cindex @code{UPDATE}, no current block buffer
12106: @code{UPDATE} has no effect.
12107:
12108: @end table
12109:
12110: @c ---------------------------------------------------------------------
12111: @node block-other, , block-ambcond, The optional Block word set
12112: @subsection Other system documentation
12113: @c ---------------------------------------------------------------------
12114: @cindex other system documentation, block words
12115: @cindex block words, other system documentation
12116:
12117: @table @i
12118: @item any restrictions a multiprogramming system places on the use of buffer addresses:
12119: No restrictions (yet).
12120:
12121: @item the number of blocks available for source and data:
12122: depends on your disk space.
12123:
12124: @end table
12125:
12126:
12127: @c =====================================================================
12128: @node The optional Double Number word set, The optional Exception word set, The optional Block word set, ANS conformance
12129: @section The optional Double Number word set
12130: @c =====================================================================
12131: @cindex system documentation, double words
12132: @cindex double words, system documentation
12133:
12134: @menu
12135: * double-ambcond:: Ambiguous Conditions
12136: @end menu
12137:
12138:
12139: @c ---------------------------------------------------------------------
12140: @node double-ambcond, , The optional Double Number word set, The optional Double Number word set
12141: @subsection Ambiguous conditions
12142: @c ---------------------------------------------------------------------
12143: @cindex double words, ambiguous conditions
12144: @cindex ambiguous conditions, double words
12145:
12146: @table @i
1.29 crook 12147: @item @i{d} outside of range of @i{n} in @code{D>S}:
12148: @cindex @code{D>S}, @i{d} out of range of @i{n}
12149: The least significant cell of @i{d} is produced.
1.1 anton 12150:
12151: @end table
12152:
12153:
12154: @c =====================================================================
12155: @node The optional Exception word set, The optional Facility word set, The optional Double Number word set, ANS conformance
12156: @section The optional Exception word set
12157: @c =====================================================================
12158: @cindex system documentation, exception words
12159: @cindex exception words, system documentation
12160:
12161: @menu
12162: * exception-idef:: Implementation Defined Options
12163: @end menu
12164:
12165:
12166: @c ---------------------------------------------------------------------
12167: @node exception-idef, , The optional Exception word set, The optional Exception word set
12168: @subsection Implementation Defined Options
12169: @c ---------------------------------------------------------------------
12170: @cindex implementation-defined options, exception words
12171: @cindex exception words, implementation-defined options
12172:
12173: @table @i
12174: @item @code{THROW}-codes used in the system:
12175: @cindex @code{THROW}-codes used in the system
12176: The codes -256@minus{}-511 are used for reporting signals. The mapping
1.29 crook 12177: from OS signal numbers to throw codes is -256@minus{}@i{signal}. The
1.1 anton 12178: codes -512@minus{}-2047 are used for OS errors (for file and memory
12179: allocation operations). The mapping from OS error numbers to throw codes
12180: is -512@minus{}@code{errno}. One side effect of this mapping is that
12181: undefined OS errors produce a message with a strange number; e.g.,
12182: @code{-1000 THROW} results in @code{Unknown error 488} on my system.
12183: @end table
12184:
12185: @c =====================================================================
12186: @node The optional Facility word set, The optional File-Access word set, The optional Exception word set, ANS conformance
12187: @section The optional Facility word set
12188: @c =====================================================================
12189: @cindex system documentation, facility words
12190: @cindex facility words, system documentation
12191:
12192: @menu
12193: * facility-idef:: Implementation Defined Options
12194: * facility-ambcond:: Ambiguous Conditions
12195: @end menu
12196:
12197:
12198: @c ---------------------------------------------------------------------
12199: @node facility-idef, facility-ambcond, The optional Facility word set, The optional Facility word set
12200: @subsection Implementation Defined Options
12201: @c ---------------------------------------------------------------------
12202: @cindex implementation-defined options, facility words
12203: @cindex facility words, implementation-defined options
12204:
12205: @table @i
12206: @item encoding of keyboard events (@code{EKEY}):
12207: @cindex keyboard events, encoding in @code{EKEY}
12208: @cindex @code{EKEY}, encoding of keyboard events
1.40 anton 12209: Keys corresponding to ASCII characters are encoded as ASCII characters.
1.41 anton 12210: Other keys are encoded with the constants @code{k-left}, @code{k-right},
12211: @code{k-up}, @code{k-down}, @code{k-home}, @code{k-end}, @code{k1},
12212: @code{k2}, @code{k3}, @code{k4}, @code{k5}, @code{k6}, @code{k7},
12213: @code{k8}, @code{k9}, @code{k10}, @code{k11}, @code{k12}.
1.40 anton 12214:
1.1 anton 12215:
12216: @item duration of a system clock tick:
12217: @cindex duration of a system clock tick
12218: @cindex clock tick duration
12219: System dependent. With respect to @code{MS}, the time is specified in
12220: microseconds. How well the OS and the hardware implement this, is
12221: another question.
12222:
12223: @item repeatability to be expected from the execution of @code{MS}:
12224: @cindex repeatability to be expected from the execution of @code{MS}
12225: @cindex @code{MS}, repeatability to be expected
12226: System dependent. On Unix, a lot depends on load. If the system is
12227: lightly loaded, and the delay is short enough that Gforth does not get
12228: swapped out, the performance should be acceptable. Under MS-DOS and
12229: other single-tasking systems, it should be good.
12230:
12231: @end table
12232:
12233:
12234: @c ---------------------------------------------------------------------
12235: @node facility-ambcond, , facility-idef, The optional Facility word set
12236: @subsection Ambiguous conditions
12237: @c ---------------------------------------------------------------------
12238: @cindex facility words, ambiguous conditions
12239: @cindex ambiguous conditions, facility words
12240:
12241: @table @i
12242: @item @code{AT-XY} can't be performed on user output device:
12243: @cindex @code{AT-XY} can't be performed on user output device
12244: Largely terminal dependent. No range checks are done on the arguments.
12245: No errors are reported. You may see some garbage appearing, you may see
12246: simply nothing happen.
12247:
12248: @end table
12249:
12250:
12251: @c =====================================================================
12252: @node The optional File-Access word set, The optional Floating-Point word set, The optional Facility word set, ANS conformance
12253: @section The optional File-Access word set
12254: @c =====================================================================
12255: @cindex system documentation, file words
12256: @cindex file words, system documentation
12257:
12258: @menu
12259: * file-idef:: Implementation Defined Options
12260: * file-ambcond:: Ambiguous Conditions
12261: @end menu
12262:
12263: @c ---------------------------------------------------------------------
12264: @node file-idef, file-ambcond, The optional File-Access word set, The optional File-Access word set
12265: @subsection Implementation Defined Options
12266: @c ---------------------------------------------------------------------
12267: @cindex implementation-defined options, file words
12268: @cindex file words, implementation-defined options
12269:
12270: @table @i
12271: @item file access methods used:
12272: @cindex file access methods used
12273: @code{R/O}, @code{R/W} and @code{BIN} work as you would
12274: expect. @code{W/O} translates into the C file opening mode @code{w} (or
12275: @code{wb}): The file is cleared, if it exists, and created, if it does
12276: not (with both @code{open-file} and @code{create-file}). Under Unix
12277: @code{create-file} creates a file with 666 permissions modified by your
12278: umask.
12279:
12280: @item file exceptions:
12281: @cindex file exceptions
12282: The file words do not raise exceptions (except, perhaps, memory access
12283: faults when you pass illegal addresses or file-ids).
12284:
12285: @item file line terminator:
12286: @cindex file line terminator
12287: System-dependent. Gforth uses C's newline character as line
12288: terminator. What the actual character code(s) of this are is
12289: system-dependent.
12290:
12291: @item file name format:
12292: @cindex file name format
12293: System dependent. Gforth just uses the file name format of your OS.
12294:
12295: @item information returned by @code{FILE-STATUS}:
12296: @cindex @code{FILE-STATUS}, returned information
12297: @code{FILE-STATUS} returns the most powerful file access mode allowed
12298: for the file: Either @code{R/O}, @code{W/O} or @code{R/W}. If the file
12299: cannot be accessed, @code{R/O BIN} is returned. @code{BIN} is applicable
12300: along with the returned mode.
12301:
12302: @item input file state after an exception when including source:
12303: @cindex exception when including source
12304: All files that are left via the exception are closed.
12305:
1.29 crook 12306: @item @i{ior} values and meaning:
12307: @cindex @i{ior} values and meaning
12308: The @i{ior}s returned by the file and memory allocation words are
1.1 anton 12309: intended as throw codes. They typically are in the range
12310: -512@minus{}-2047 of OS errors. The mapping from OS error numbers to
1.29 crook 12311: @i{ior}s is -512@minus{}@i{errno}.
1.1 anton 12312:
12313: @item maximum depth of file input nesting:
12314: @cindex maximum depth of file input nesting
12315: @cindex file input nesting, maximum depth
12316: limited by the amount of return stack, locals/TIB stack, and the number
12317: of open files available. This should not give you troubles.
12318:
12319: @item maximum size of input line:
12320: @cindex maximum size of input line
12321: @cindex input line size, maximum
12322: @code{/line}. Currently 255.
12323:
12324: @item methods of mapping block ranges to files:
12325: @cindex mapping block ranges to files
12326: @cindex files containing blocks
12327: @cindex blocks in files
12328: By default, blocks are accessed in the file @file{blocks.fb} in the
12329: current working directory. The file can be switched with @code{USE}.
12330:
12331: @item number of string buffers provided by @code{S"}:
12332: @cindex @code{S"}, number of string buffers
12333: 1
12334:
12335: @item size of string buffer used by @code{S"}:
12336: @cindex @code{S"}, size of string buffer
12337: @code{/line}. currently 255.
12338:
12339: @end table
12340:
12341: @c ---------------------------------------------------------------------
12342: @node file-ambcond, , file-idef, The optional File-Access word set
12343: @subsection Ambiguous conditions
12344: @c ---------------------------------------------------------------------
12345: @cindex file words, ambiguous conditions
12346: @cindex ambiguous conditions, file words
12347:
12348: @table @i
12349: @item attempting to position a file outside its boundaries:
12350: @cindex @code{REPOSITION-FILE}, outside the file's boundaries
12351: @code{REPOSITION-FILE} is performed as usual: Afterwards,
12352: @code{FILE-POSITION} returns the value given to @code{REPOSITION-FILE}.
12353:
12354: @item attempting to read from file positions not yet written:
12355: @cindex reading from file positions not yet written
12356: End-of-file, i.e., zero characters are read and no error is reported.
12357:
1.29 crook 12358: @item @i{file-id} is invalid (@code{INCLUDE-FILE}):
12359: @cindex @code{INCLUDE-FILE}, @i{file-id} is invalid
1.1 anton 12360: An appropriate exception may be thrown, but a memory fault or other
12361: problem is more probable.
12362:
1.29 crook 12363: @item I/O exception reading or closing @i{file-id} (@code{INCLUDE-FILE}, @code{INCLUDED}):
12364: @cindex @code{INCLUDE-FILE}, I/O exception reading or closing @i{file-id}
12365: @cindex @code{INCLUDED}, I/O exception reading or closing @i{file-id}
12366: The @i{ior} produced by the operation, that discovered the problem, is
1.1 anton 12367: thrown.
12368:
12369: @item named file cannot be opened (@code{INCLUDED}):
12370: @cindex @code{INCLUDED}, named file cannot be opened
1.29 crook 12371: The @i{ior} produced by @code{open-file} is thrown.
1.1 anton 12372:
12373: @item requesting an unmapped block number:
12374: @cindex unmapped block numbers
12375: There are no unmapped legal block numbers. On some operating systems,
12376: writing a block with a large number may overflow the file system and
12377: have an error message as consequence.
12378:
12379: @item using @code{source-id} when @code{blk} is non-zero:
12380: @cindex @code{SOURCE-ID}, behaviour when @code{BLK} is non-zero
12381: @code{source-id} performs its function. Typically it will give the id of
12382: the source which loaded the block. (Better ideas?)
12383:
12384: @end table
12385:
12386:
12387: @c =====================================================================
12388: @node The optional Floating-Point word set, The optional Locals word set, The optional File-Access word set, ANS conformance
12389: @section The optional Floating-Point word set
12390: @c =====================================================================
12391: @cindex system documentation, floating-point words
12392: @cindex floating-point words, system documentation
12393:
12394: @menu
12395: * floating-idef:: Implementation Defined Options
12396: * floating-ambcond:: Ambiguous Conditions
12397: @end menu
12398:
12399:
12400: @c ---------------------------------------------------------------------
12401: @node floating-idef, floating-ambcond, The optional Floating-Point word set, The optional Floating-Point word set
12402: @subsection Implementation Defined Options
12403: @c ---------------------------------------------------------------------
12404: @cindex implementation-defined options, floating-point words
12405: @cindex floating-point words, implementation-defined options
12406:
12407: @table @i
12408: @item format and range of floating point numbers:
12409: @cindex format and range of floating point numbers
12410: @cindex floating point numbers, format and range
12411: System-dependent; the @code{double} type of C.
12412:
1.29 crook 12413: @item results of @code{REPRESENT} when @i{float} is out of range:
12414: @cindex @code{REPRESENT}, results when @i{float} is out of range
1.1 anton 12415: System dependent; @code{REPRESENT} is implemented using the C library
12416: function @code{ecvt()} and inherits its behaviour in this respect.
12417:
12418: @item rounding or truncation of floating-point numbers:
12419: @cindex rounding of floating-point numbers
12420: @cindex truncation of floating-point numbers
12421: @cindex floating-point numbers, rounding or truncation
12422: System dependent; the rounding behaviour is inherited from the hosting C
12423: compiler. IEEE-FP-based (i.e., most) systems by default round to
12424: nearest, and break ties by rounding to even (i.e., such that the last
12425: bit of the mantissa is 0).
12426:
12427: @item size of floating-point stack:
12428: @cindex floating-point stack size
12429: @code{s" FLOATING-STACK" environment? drop .} gives the total size of
12430: the floating-point stack (in floats). You can specify this on startup
12431: with the command-line option @code{-f} (@pxref{Invoking Gforth}).
12432:
12433: @item width of floating-point stack:
12434: @cindex floating-point stack width
12435: @code{1 floats}.
12436:
12437: @end table
12438:
12439:
12440: @c ---------------------------------------------------------------------
12441: @node floating-ambcond, , floating-idef, The optional Floating-Point word set
12442: @subsection Ambiguous conditions
12443: @c ---------------------------------------------------------------------
12444: @cindex floating-point words, ambiguous conditions
12445: @cindex ambiguous conditions, floating-point words
12446:
12447: @table @i
12448: @item @code{df@@} or @code{df!} used with an address that is not double-float aligned:
12449: @cindex @code{df@@} or @code{df!} used with an address that is not double-float aligned
12450: System-dependent. Typically results in a @code{-23 THROW} like other
12451: alignment violations.
12452:
12453: @item @code{f@@} or @code{f!} used with an address that is not float aligned:
12454: @cindex @code{f@@} used with an address that is not float aligned
12455: @cindex @code{f!} used with an address that is not float aligned
12456: System-dependent. Typically results in a @code{-23 THROW} like other
12457: alignment violations.
12458:
12459: @item floating-point result out of range:
12460: @cindex floating-point result out of range
12461: System-dependent. Can result in a @code{-55 THROW} (Floating-point
12462: unidentified fault), or can produce a special value representing, e.g.,
12463: Infinity.
12464:
12465: @item @code{sf@@} or @code{sf!} used with an address that is not single-float aligned:
12466: @cindex @code{sf@@} or @code{sf!} used with an address that is not single-float aligned
12467: System-dependent. Typically results in an alignment fault like other
12468: alignment violations.
12469:
1.35 anton 12470: @item @code{base} is not decimal (@code{REPRESENT}, @code{F.}, @code{FE.}, @code{FS.}):
12471: @cindex @code{base} is not decimal (@code{REPRESENT}, @code{F.}, @code{FE.}, @code{FS.})
1.1 anton 12472: The floating-point number is converted into decimal nonetheless.
12473:
12474: @item Both arguments are equal to zero (@code{FATAN2}):
12475: @cindex @code{FATAN2}, both arguments are equal to zero
12476: System-dependent. @code{FATAN2} is implemented using the C library
12477: function @code{atan2()}.
12478:
1.29 crook 12479: @item Using @code{FTAN} on an argument @i{r1} where cos(@i{r1}) is zero:
12480: @cindex @code{FTAN} on an argument @i{r1} where cos(@i{r1}) is zero
12481: System-dependent. Anyway, typically the cos of @i{r1} will not be zero
1.1 anton 12482: because of small errors and the tan will be a very large (or very small)
12483: but finite number.
12484:
1.29 crook 12485: @item @i{d} cannot be presented precisely as a float in @code{D>F}:
12486: @cindex @code{D>F}, @i{d} cannot be presented precisely as a float
1.1 anton 12487: The result is rounded to the nearest float.
12488:
12489: @item dividing by zero:
12490: @cindex dividing by zero, floating-point
12491: @cindex floating-point dividing by zero
12492: @cindex floating-point unidentified fault, FP divide-by-zero
12493: @code{-55 throw} (Floating-point unidentified fault)
12494:
12495: @item exponent too big for conversion (@code{DF!}, @code{DF@@}, @code{SF!}, @code{SF@@}):
12496: @cindex exponent too big for conversion (@code{DF!}, @code{DF@@}, @code{SF!}, @code{SF@@})
12497: System dependent. On IEEE-FP based systems the number is converted into
12498: an infinity.
12499:
1.29 crook 12500: @item @i{float}<1 (@code{FACOSH}):
12501: @cindex @code{FACOSH}, @i{float}<1
1.1 anton 12502: @cindex floating-point unidentified fault, @code{FACOSH}
12503: @code{-55 throw} (Floating-point unidentified fault)
12504:
1.29 crook 12505: @item @i{float}=<-1 (@code{FLNP1}):
12506: @cindex @code{FLNP1}, @i{float}=<-1
1.1 anton 12507: @cindex floating-point unidentified fault, @code{FLNP1}
12508: @code{-55 throw} (Floating-point unidentified fault). On IEEE-FP systems
1.29 crook 12509: negative infinity is typically produced for @i{float}=-1.
1.1 anton 12510:
1.29 crook 12511: @item @i{float}=<0 (@code{FLN}, @code{FLOG}):
12512: @cindex @code{FLN}, @i{float}=<0
12513: @cindex @code{FLOG}, @i{float}=<0
1.1 anton 12514: @cindex floating-point unidentified fault, @code{FLN} or @code{FLOG}
12515: @code{-55 throw} (Floating-point unidentified fault). On IEEE-FP systems
1.29 crook 12516: negative infinity is typically produced for @i{float}=0.
1.1 anton 12517:
1.29 crook 12518: @item @i{float}<0 (@code{FASINH}, @code{FSQRT}):
12519: @cindex @code{FASINH}, @i{float}<0
12520: @cindex @code{FSQRT}, @i{float}<0
1.1 anton 12521: @cindex floating-point unidentified fault, @code{FASINH} or @code{FSQRT}
12522: @code{-55 throw} (Floating-point unidentified fault). @code{fasinh}
12523: produces values for these inputs on my Linux box (Bug in the C library?)
12524:
1.29 crook 12525: @item |@i{float}|>1 (@code{FACOS}, @code{FASIN}, @code{FATANH}):
12526: @cindex @code{FACOS}, |@i{float}|>1
12527: @cindex @code{FASIN}, |@i{float}|>1
12528: @cindex @code{FATANH}, |@i{float}|>1
1.1 anton 12529: @cindex floating-point unidentified fault, @code{FACOS}, @code{FASIN} or @code{FATANH}
12530: @code{-55 throw} (Floating-point unidentified fault).
12531:
1.29 crook 12532: @item integer part of float cannot be represented by @i{d} in @code{F>D}:
12533: @cindex @code{F>D}, integer part of float cannot be represented by @i{d}
1.1 anton 12534: @cindex floating-point unidentified fault, @code{F>D}
12535: @code{-55 throw} (Floating-point unidentified fault).
12536:
12537: @item string larger than pictured numeric output area (@code{f.}, @code{fe.}, @code{fs.}):
12538: @cindex string larger than pictured numeric output area (@code{f.}, @code{fe.}, @code{fs.})
12539: This does not happen.
12540: @end table
12541:
12542: @c =====================================================================
12543: @node The optional Locals word set, The optional Memory-Allocation word set, The optional Floating-Point word set, ANS conformance
12544: @section The optional Locals word set
12545: @c =====================================================================
12546: @cindex system documentation, locals words
12547: @cindex locals words, system documentation
12548:
12549: @menu
12550: * locals-idef:: Implementation Defined Options
12551: * locals-ambcond:: Ambiguous Conditions
12552: @end menu
12553:
12554:
12555: @c ---------------------------------------------------------------------
12556: @node locals-idef, locals-ambcond, The optional Locals word set, The optional Locals word set
12557: @subsection Implementation Defined Options
12558: @c ---------------------------------------------------------------------
12559: @cindex implementation-defined options, locals words
12560: @cindex locals words, implementation-defined options
12561:
12562: @table @i
12563: @item maximum number of locals in a definition:
12564: @cindex maximum number of locals in a definition
12565: @cindex locals, maximum number in a definition
12566: @code{s" #locals" environment? drop .}. Currently 15. This is a lower
12567: bound, e.g., on a 32-bit machine there can be 41 locals of up to 8
12568: characters. The number of locals in a definition is bounded by the size
12569: of locals-buffer, which contains the names of the locals.
12570:
12571: @end table
12572:
12573:
12574: @c ---------------------------------------------------------------------
12575: @node locals-ambcond, , locals-idef, The optional Locals word set
12576: @subsection Ambiguous conditions
12577: @c ---------------------------------------------------------------------
12578: @cindex locals words, ambiguous conditions
12579: @cindex ambiguous conditions, locals words
12580:
12581: @table @i
12582: @item executing a named local in interpretation state:
12583: @cindex local in interpretation state
12584: @cindex Interpreting a compile-only word, for a local
12585: Locals have no interpretation semantics. If you try to perform the
12586: interpretation semantics, you will get a @code{-14 throw} somewhere
12587: (Interpreting a compile-only word). If you perform the compilation
12588: semantics, the locals access will be compiled (irrespective of state).
12589:
1.29 crook 12590: @item @i{name} not defined by @code{VALUE} or @code{(LOCAL)} (@code{TO}):
1.1 anton 12591: @cindex name not defined by @code{VALUE} or @code{(LOCAL)} used by @code{TO}
12592: @cindex @code{TO} on non-@code{VALUE}s and non-locals
12593: @cindex Invalid name argument, @code{TO}
12594: @code{-32 throw} (Invalid name argument)
12595:
12596: @end table
12597:
12598:
12599: @c =====================================================================
12600: @node The optional Memory-Allocation word set, The optional Programming-Tools word set, The optional Locals word set, ANS conformance
12601: @section The optional Memory-Allocation word set
12602: @c =====================================================================
12603: @cindex system documentation, memory-allocation words
12604: @cindex memory-allocation words, system documentation
12605:
12606: @menu
12607: * memory-idef:: Implementation Defined Options
12608: @end menu
12609:
12610:
12611: @c ---------------------------------------------------------------------
12612: @node memory-idef, , The optional Memory-Allocation word set, The optional Memory-Allocation word set
12613: @subsection Implementation Defined Options
12614: @c ---------------------------------------------------------------------
12615: @cindex implementation-defined options, memory-allocation words
12616: @cindex memory-allocation words, implementation-defined options
12617:
12618: @table @i
1.29 crook 12619: @item values and meaning of @i{ior}:
12620: @cindex @i{ior} values and meaning
12621: The @i{ior}s returned by the file and memory allocation words are
1.1 anton 12622: intended as throw codes. They typically are in the range
12623: -512@minus{}-2047 of OS errors. The mapping from OS error numbers to
1.29 crook 12624: @i{ior}s is -512@minus{}@i{errno}.
1.1 anton 12625:
12626: @end table
12627:
12628: @c =====================================================================
12629: @node The optional Programming-Tools word set, The optional Search-Order word set, The optional Memory-Allocation word set, ANS conformance
12630: @section The optional Programming-Tools word set
12631: @c =====================================================================
12632: @cindex system documentation, programming-tools words
12633: @cindex programming-tools words, system documentation
12634:
12635: @menu
12636: * programming-idef:: Implementation Defined Options
12637: * programming-ambcond:: Ambiguous Conditions
12638: @end menu
12639:
12640:
12641: @c ---------------------------------------------------------------------
12642: @node programming-idef, programming-ambcond, The optional Programming-Tools word set, The optional Programming-Tools word set
12643: @subsection Implementation Defined Options
12644: @c ---------------------------------------------------------------------
12645: @cindex implementation-defined options, programming-tools words
12646: @cindex programming-tools words, implementation-defined options
12647:
12648: @table @i
12649: @item ending sequence for input following @code{;CODE} and @code{CODE}:
12650: @cindex @code{;CODE} ending sequence
12651: @cindex @code{CODE} ending sequence
12652: @code{END-CODE}
12653:
12654: @item manner of processing input following @code{;CODE} and @code{CODE}:
12655: @cindex @code{;CODE}, processing input
12656: @cindex @code{CODE}, processing input
12657: The @code{ASSEMBLER} vocabulary is pushed on the search order stack, and
12658: the input is processed by the text interpreter, (starting) in interpret
12659: state.
12660:
12661: @item search order capability for @code{EDITOR} and @code{ASSEMBLER}:
12662: @cindex @code{ASSEMBLER}, search order capability
12663: The ANS Forth search order word set.
12664:
12665: @item source and format of display by @code{SEE}:
12666: @cindex @code{SEE}, source and format of output
12667: The source for @code{see} is the intermediate code used by the inner
12668: interpreter. The current @code{see} tries to output Forth source code
12669: as well as possible.
12670:
12671: @end table
12672:
12673: @c ---------------------------------------------------------------------
12674: @node programming-ambcond, , programming-idef, The optional Programming-Tools word set
12675: @subsection Ambiguous conditions
12676: @c ---------------------------------------------------------------------
12677: @cindex programming-tools words, ambiguous conditions
12678: @cindex ambiguous conditions, programming-tools words
12679:
12680: @table @i
12681:
1.21 crook 12682: @item deleting the compilation word list (@code{FORGET}):
12683: @cindex @code{FORGET}, deleting the compilation word list
1.1 anton 12684: Not implemented (yet).
12685:
1.29 crook 12686: @item fewer than @i{u}+1 items on the control-flow stack (@code{CS-PICK}, @code{CS-ROLL}):
12687: @cindex @code{CS-PICK}, fewer than @i{u}+1 items on the control flow-stack
12688: @cindex @code{CS-ROLL}, fewer than @i{u}+1 items on the control flow-stack
1.1 anton 12689: @cindex control-flow stack underflow
12690: This typically results in an @code{abort"} with a descriptive error
12691: message (may change into a @code{-22 throw} (Control structure mismatch)
12692: in the future). You may also get a memory access error. If you are
12693: unlucky, this ambiguous condition is not caught.
12694:
1.29 crook 12695: @item @i{name} can't be found (@code{FORGET}):
12696: @cindex @code{FORGET}, @i{name} can't be found
1.1 anton 12697: Not implemented (yet).
12698:
1.29 crook 12699: @item @i{name} not defined via @code{CREATE}:
12700: @cindex @code{;CODE}, @i{name} not defined via @code{CREATE}
1.1 anton 12701: @code{;CODE} behaves like @code{DOES>} in this respect, i.e., it changes
12702: the execution semantics of the last defined word no matter how it was
12703: defined.
12704:
12705: @item @code{POSTPONE} applied to @code{[IF]}:
12706: @cindex @code{POSTPONE} applied to @code{[IF]}
12707: @cindex @code{[IF]} and @code{POSTPONE}
12708: After defining @code{: X POSTPONE [IF] ; IMMEDIATE}. @code{X} is
12709: equivalent to @code{[IF]}.
12710:
12711: @item reaching the end of the input source before matching @code{[ELSE]} or @code{[THEN]}:
12712: @cindex @code{[IF]}, end of the input source before matching @code{[ELSE]} or @code{[THEN]}
12713: Continue in the same state of conditional compilation in the next outer
12714: input source. Currently there is no warning to the user about this.
12715:
12716: @item removing a needed definition (@code{FORGET}):
12717: @cindex @code{FORGET}, removing a needed definition
12718: Not implemented (yet).
12719:
12720: @end table
12721:
12722:
12723: @c =====================================================================
12724: @node The optional Search-Order word set, , The optional Programming-Tools word set, ANS conformance
12725: @section The optional Search-Order word set
12726: @c =====================================================================
12727: @cindex system documentation, search-order words
12728: @cindex search-order words, system documentation
12729:
12730: @menu
12731: * search-idef:: Implementation Defined Options
12732: * search-ambcond:: Ambiguous Conditions
12733: @end menu
12734:
12735:
12736: @c ---------------------------------------------------------------------
12737: @node search-idef, search-ambcond, The optional Search-Order word set, The optional Search-Order word set
12738: @subsection Implementation Defined Options
12739: @c ---------------------------------------------------------------------
12740: @cindex implementation-defined options, search-order words
12741: @cindex search-order words, implementation-defined options
12742:
12743: @table @i
12744: @item maximum number of word lists in search order:
12745: @cindex maximum number of word lists in search order
12746: @cindex search order, maximum depth
12747: @code{s" wordlists" environment? drop .}. Currently 16.
12748:
12749: @item minimum search order:
12750: @cindex minimum search order
12751: @cindex search order, minimum
12752: @code{root root}.
12753:
12754: @end table
12755:
12756: @c ---------------------------------------------------------------------
12757: @node search-ambcond, , search-idef, The optional Search-Order word set
12758: @subsection Ambiguous conditions
12759: @c ---------------------------------------------------------------------
12760: @cindex search-order words, ambiguous conditions
12761: @cindex ambiguous conditions, search-order words
12762:
12763: @table @i
1.21 crook 12764: @item changing the compilation word list (during compilation):
12765: @cindex changing the compilation word list (during compilation)
12766: @cindex compilation word list, change before definition ends
12767: The word is entered into the word list that was the compilation word list
1.1 anton 12768: at the start of the definition. Any changes to the name field (e.g.,
12769: @code{immediate}) or the code field (e.g., when executing @code{DOES>})
12770: are applied to the latest defined word (as reported by @code{last} or
1.21 crook 12771: @code{lastxt}), if possible, irrespective of the compilation word list.
1.1 anton 12772:
12773: @item search order empty (@code{previous}):
12774: @cindex @code{previous}, search order empty
1.26 crook 12775: @cindex vocstack empty, @code{previous}
1.1 anton 12776: @code{abort" Vocstack empty"}.
12777:
12778: @item too many word lists in search order (@code{also}):
12779: @cindex @code{also}, too many word lists in search order
1.26 crook 12780: @cindex vocstack full, @code{also}
1.1 anton 12781: @code{abort" Vocstack full"}.
12782:
12783: @end table
12784:
12785: @c ***************************************************************
12786: @node Model, Integrating Gforth, ANS conformance, Top
12787: @chapter Model
12788:
12789: This chapter has yet to be written. It will contain information, on
12790: which internal structures you can rely.
12791:
12792: @c ***************************************************************
12793: @node Integrating Gforth, Emacs and Gforth, Model, Top
12794: @chapter Integrating Gforth into C programs
12795:
12796: This is not yet implemented.
12797:
12798: Several people like to use Forth as scripting language for applications
12799: that are otherwise written in C, C++, or some other language.
12800:
12801: The Forth system ATLAST provides facilities for embedding it into
12802: applications; unfortunately it has several disadvantages: most
12803: importantly, it is not based on ANS Forth, and it is apparently dead
12804: (i.e., not developed further and not supported). The facilities
1.21 crook 12805: provided by Gforth in this area are inspired by ATLAST's facilities, so
1.1 anton 12806: making the switch should not be hard.
12807:
12808: We also tried to design the interface such that it can easily be
12809: implemented by other Forth systems, so that we may one day arrive at a
12810: standardized interface. Such a standard interface would allow you to
12811: replace the Forth system without having to rewrite C code.
12812:
12813: You embed the Gforth interpreter by linking with the library
12814: @code{libgforth.a} (give the compiler the option @code{-lgforth}). All
12815: global symbols in this library that belong to the interface, have the
12816: prefix @code{forth_}. (Global symbols that are used internally have the
12817: prefix @code{gforth_}).
12818:
12819: You can include the declarations of Forth types and the functions and
12820: variables of the interface with @code{#include <forth.h>}.
12821:
12822: Types.
12823:
12824: Variables.
12825:
12826: Data and FP Stack pointer. Area sizes.
12827:
12828: functions.
12829:
12830: forth_init(imagefile)
12831: forth_evaluate(string) exceptions?
12832: forth_goto(address) (or forth_execute(xt)?)
12833: forth_continue() (a corountining mechanism)
12834:
12835: Adding primitives.
12836:
12837: No checking.
12838:
12839: Signals?
12840:
12841: Accessing the Stacks
12842:
1.26 crook 12843: @c ******************************************************************
1.1 anton 12844: @node Emacs and Gforth, Image Files, Integrating Gforth, Top
12845: @chapter Emacs and Gforth
12846: @cindex Emacs and Gforth
12847:
12848: @cindex @file{gforth.el}
12849: @cindex @file{forth.el}
12850: @cindex Rydqvist, Goran
12851: @cindex comment editing commands
12852: @cindex @code{\}, editing with Emacs
12853: @cindex debug tracer editing commands
12854: @cindex @code{~~}, removal with Emacs
12855: @cindex Forth mode in Emacs
12856: Gforth comes with @file{gforth.el}, an improved version of
12857: @file{forth.el} by Goran Rydqvist (included in the TILE package). The
1.26 crook 12858: improvements are:
12859:
12860: @itemize @bullet
12861: @item
12862: A better (but still not perfect) handling of indentation.
12863: @item
12864: Comment paragraph filling (@kbd{M-q})
12865: @item
12866: Commenting (@kbd{C-x \}) and uncommenting (@kbd{C-u C-x \}) of regions
12867: @item
12868: Removal of debugging tracers (@kbd{C-x ~}, @pxref{Debugging}).
1.41 anton 12869: @item
12870: Support of the @code{info-lookup} feature for looking up the
12871: documentation of a word.
1.26 crook 12872: @end itemize
12873:
12874: I left the stuff I do not use alone, even though some of it only makes
12875: sense for TILE. To get a description of these features, enter Forth mode
12876: and type @kbd{C-h m}.
1.1 anton 12877:
12878: @cindex source location of error or debugging output in Emacs
12879: @cindex error output, finding the source location in Emacs
12880: @cindex debugging output, finding the source location in Emacs
12881: In addition, Gforth supports Emacs quite well: The source code locations
12882: given in error messages, debugging output (from @code{~~}) and failed
12883: assertion messages are in the right format for Emacs' compilation mode
12884: (@pxref{Compilation, , Running Compilations under Emacs, emacs, Emacs
12885: Manual}) so the source location corresponding to an error or other
12886: message is only a few keystrokes away (@kbd{C-x `} for the next error,
12887: @kbd{C-c C-c} for the error under the cursor).
12888:
12889: @cindex @file{TAGS} file
12890: @cindex @file{etags.fs}
12891: @cindex viewing the source of a word in Emacs
1.43 anton 12892: @cindex @code{require}, placement in files
12893: @cindex @code{include}, placement in files
12894: Also, if you @code{require} @file{etags.fs}, a new @file{TAGS} file will
1.26 crook 12895: be produced (@pxref{Tags, , Tags Tables, emacs, Emacs Manual}) that
1.1 anton 12896: contains the definitions of all words defined afterwards. You can then
12897: find the source for a word using @kbd{M-.}. Note that emacs can use
12898: several tags files at the same time (e.g., one for the Gforth sources
12899: and one for your program, @pxref{Select Tags Table,,Selecting a Tags
12900: Table,emacs, Emacs Manual}). The TAGS file for the preloaded words is
12901: @file{$(datadir)/gforth/$(VERSION)/TAGS} (e.g.,
1.43 anton 12902: @file{/usr/local/share/gforth/0.2.0/TAGS}). To get the best behaviour
12903: with @file{etags.fs}, you should avoid putting definitions both before
12904: and after @code{require} etc., otherwise you will see the same file
12905: visited several times by commands like @code{tags-search}.
1.1 anton 12906:
1.41 anton 12907: @cindex viewing the documentation of a word in Emacs
12908: @cindex context-sensitive help
12909: Moreover, for words documented in this manual, you can look up the
12910: glossary entry quickly by using @kbd{C-h TAB}
12911: (@code{info-lookup-symbol}, see @pxref{Documentation, ,Documentation
12912: Commands, emacs, Emacs Manual}). This feature requires Emacs 20.3 or
1.42 anton 12913: later and does not work for words containing @code{:}.
1.41 anton 12914:
12915:
1.1 anton 12916: @cindex @file{.emacs}
12917: To get all these benefits, add the following lines to your @file{.emacs}
12918: file:
12919:
12920: @example
12921: (autoload 'forth-mode "gforth.el")
12922: (setq auto-mode-alist (cons '("\\.fs\\'" . forth-mode) auto-mode-alist))
12923: @end example
12924:
1.26 crook 12925: @c ******************************************************************
1.1 anton 12926: @node Image Files, Engine, Emacs and Gforth, Top
12927: @chapter Image Files
1.26 crook 12928: @cindex image file
12929: @cindex @file{.fi} files
1.1 anton 12930: @cindex precompiled Forth code
12931: @cindex dictionary in persistent form
12932: @cindex persistent form of dictionary
12933:
12934: An image file is a file containing an image of the Forth dictionary,
12935: i.e., compiled Forth code and data residing in the dictionary. By
12936: convention, we use the extension @code{.fi} for image files.
12937:
12938: @menu
1.18 anton 12939: * Image Licensing Issues:: Distribution terms for images.
12940: * Image File Background:: Why have image files?
1.29 crook 12941: * Non-Relocatable Image Files:: don't always work.
1.18 anton 12942: * Data-Relocatable Image Files:: are better.
1.29 crook 12943: * Fully Relocatable Image Files:: better yet.
1.18 anton 12944: * Stack and Dictionary Sizes:: Setting the default sizes for an image.
1.29 crook 12945: * Running Image Files:: @code{gforth -i @i{file}} or @i{file}.
1.18 anton 12946: * Modifying the Startup Sequence:: and turnkey applications.
1.1 anton 12947: @end menu
12948:
1.18 anton 12949: @node Image Licensing Issues, Image File Background, Image Files, Image Files
12950: @section Image Licensing Issues
12951: @cindex license for images
12952: @cindex image license
12953:
12954: An image created with @code{gforthmi} (@pxref{gforthmi}) or
12955: @code{savesystem} (@pxref{Non-Relocatable Image Files}) includes the
12956: original image; i.e., according to copyright law it is a derived work of
12957: the original image.
12958:
12959: Since Gforth is distributed under the GNU GPL, the newly created image
12960: falls under the GNU GPL, too. In particular, this means that if you
12961: distribute the image, you have to make all of the sources for the image
12962: available, including those you wrote. For details see @ref{License, ,
12963: GNU General Public License (Section 3)}.
12964:
12965: If you create an image with @code{cross} (@pxref{cross.fs}), the image
12966: contains only code compiled from the sources you gave it; if none of
12967: these sources is under the GPL, the terms discussed above do not apply
12968: to the image. However, if your image needs an engine (a gforth binary)
12969: that is under the GPL, you should make sure that you distribute both in
12970: a way that is at most a @emph{mere aggregation}, if you don't want the
12971: terms of the GPL to apply to the image.
12972:
12973: @node Image File Background, Non-Relocatable Image Files, Image Licensing Issues, Image Files
1.1 anton 12974: @section Image File Background
12975: @cindex image file background
12976:
12977: Our Forth system consists not only of primitives, but also of
12978: definitions written in Forth. Since the Forth compiler itself belongs to
12979: those definitions, it is not possible to start the system with the
12980: primitives and the Forth source alone. Therefore we provide the Forth
1.26 crook 12981: code as an image file in nearly executable form. When Gforth starts up,
12982: a C routine loads the image file into memory, optionally relocates the
12983: addresses, then sets up the memory (stacks etc.) according to
12984: information in the image file, and (finally) starts executing Forth
12985: code.
1.1 anton 12986:
12987: The image file variants represent different compromises between the
12988: goals of making it easy to generate image files and making them
12989: portable.
12990:
12991: @cindex relocation at run-time
1.26 crook 12992: Win32Forth 3.4 and Mitch Bradley's @code{cforth} use relocation at
1.1 anton 12993: run-time. This avoids many of the complications discussed below (image
12994: files are data relocatable without further ado), but costs performance
12995: (one addition per memory access).
12996:
12997: @cindex relocation at load-time
1.26 crook 12998: By contrast, the Gforth loader performs relocation at image load time. The
12999: loader also has to replace tokens that represent primitive calls with the
1.1 anton 13000: appropriate code-field addresses (or code addresses in the case of
13001: direct threading).
13002:
13003: There are three kinds of image files, with different degrees of
13004: relocatability: non-relocatable, data-relocatable, and fully relocatable
13005: image files.
13006:
13007: @cindex image file loader
13008: @cindex relocating loader
13009: @cindex loader for image files
13010: These image file variants have several restrictions in common; they are
13011: caused by the design of the image file loader:
13012:
13013: @itemize @bullet
13014: @item
13015: There is only one segment; in particular, this means, that an image file
13016: cannot represent @code{ALLOCATE}d memory chunks (and pointers to
1.26 crook 13017: them). The contents of the stacks are not represented, either.
1.1 anton 13018:
13019: @item
13020: The only kinds of relocation supported are: adding the same offset to
13021: all cells that represent data addresses; and replacing special tokens
13022: with code addresses or with pieces of machine code.
13023:
13024: If any complex computations involving addresses are performed, the
13025: results cannot be represented in the image file. Several applications that
13026: use such computations come to mind:
13027: @itemize @minus
13028: @item
13029: Hashing addresses (or data structures which contain addresses) for table
13030: lookup. If you use Gforth's @code{table}s or @code{wordlist}s for this
13031: purpose, you will have no problem, because the hash tables are
13032: recomputed automatically when the system is started. If you use your own
13033: hash tables, you will have to do something similar.
13034:
13035: @item
13036: There's a cute implementation of doubly-linked lists that uses
13037: @code{XOR}ed addresses. You could represent such lists as singly-linked
13038: in the image file, and restore the doubly-linked representation on
13039: startup.@footnote{In my opinion, though, you should think thrice before
13040: using a doubly-linked list (whatever implementation).}
13041:
13042: @item
13043: The code addresses of run-time routines like @code{docol:} cannot be
13044: represented in the image file (because their tokens would be replaced by
13045: machine code in direct threaded implementations). As a workaround,
13046: compute these addresses at run-time with @code{>code-address} from the
13047: executions tokens of appropriate words (see the definitions of
13048: @code{docol:} and friends in @file{kernel.fs}).
13049:
13050: @item
13051: On many architectures addresses are represented in machine code in some
13052: shifted or mangled form. You cannot put @code{CODE} words that contain
13053: absolute addresses in this form in a relocatable image file. Workarounds
13054: are representing the address in some relative form (e.g., relative to
13055: the CFA, which is present in some register), or loading the address from
13056: a place where it is stored in a non-mangled form.
13057: @end itemize
13058: @end itemize
13059:
13060: @node Non-Relocatable Image Files, Data-Relocatable Image Files, Image File Background, Image Files
13061: @section Non-Relocatable Image Files
13062: @cindex non-relocatable image files
1.26 crook 13063: @cindex image file, non-relocatable
1.1 anton 13064:
13065: These files are simple memory dumps of the dictionary. They are specific
13066: to the executable (i.e., @file{gforth} file) they were created
13067: with. What's worse, they are specific to the place on which the
13068: dictionary resided when the image was created. Now, there is no
13069: guarantee that the dictionary will reside at the same place the next
13070: time you start Gforth, so there's no guarantee that a non-relocatable
13071: image will work the next time (Gforth will complain instead of crashing,
13072: though).
13073:
13074: You can create a non-relocatable image file with
13075:
1.44 crook 13076:
1.1 anton 13077: doc-savesystem
13078:
1.44 crook 13079:
1.1 anton 13080: @node Data-Relocatable Image Files, Fully Relocatable Image Files, Non-Relocatable Image Files, Image Files
13081: @section Data-Relocatable Image Files
13082: @cindex data-relocatable image files
1.26 crook 13083: @cindex image file, data-relocatable
1.1 anton 13084:
13085: These files contain relocatable data addresses, but fixed code addresses
13086: (instead of tokens). They are specific to the executable (i.e.,
13087: @file{gforth} file) they were created with. For direct threading on some
13088: architectures (e.g., the i386), data-relocatable images do not work. You
13089: get a data-relocatable image, if you use @file{gforthmi} with a
13090: Gforth binary that is not doubly indirect threaded (@pxref{Fully
13091: Relocatable Image Files}).
13092:
13093: @node Fully Relocatable Image Files, Stack and Dictionary Sizes, Data-Relocatable Image Files, Image Files
13094: @section Fully Relocatable Image Files
13095: @cindex fully relocatable image files
1.26 crook 13096: @cindex image file, fully relocatable
1.1 anton 13097:
13098: @cindex @file{kern*.fi}, relocatability
13099: @cindex @file{gforth.fi}, relocatability
13100: These image files have relocatable data addresses, and tokens for code
13101: addresses. They can be used with different binaries (e.g., with and
13102: without debugging) on the same machine, and even across machines with
13103: the same data formats (byte order, cell size, floating point
13104: format). However, they are usually specific to the version of Gforth
13105: they were created with. The files @file{gforth.fi} and @file{kernl*.fi}
13106: are fully relocatable.
13107:
13108: There are two ways to create a fully relocatable image file:
13109:
13110: @menu
1.29 crook 13111: * gforthmi:: The normal way
1.1 anton 13112: * cross.fs:: The hard way
13113: @end menu
13114:
13115: @node gforthmi, cross.fs, Fully Relocatable Image Files, Fully Relocatable Image Files
13116: @subsection @file{gforthmi}
13117: @cindex @file{comp-i.fs}
13118: @cindex @file{gforthmi}
13119:
13120: You will usually use @file{gforthmi}. If you want to create an
1.29 crook 13121: image @i{file} that contains everything you would load by invoking
13122: Gforth with @code{gforth @i{options}}, you simply say:
1.1 anton 13123: @example
1.29 crook 13124: gforthmi @i{file} @i{options}
1.1 anton 13125: @end example
13126:
13127: E.g., if you want to create an image @file{asm.fi} that has the file
13128: @file{asm.fs} loaded in addition to the usual stuff, you could do it
13129: like this:
13130:
13131: @example
13132: gforthmi asm.fi asm.fs
13133: @end example
13134:
1.27 crook 13135: @file{gforthmi} is implemented as a sh script and works like this: It
13136: produces two non-relocatable images for different addresses and then
13137: compares them. Its output reflects this: first you see the output (if
1.62 ! crook 13138: any) of the two Gforth invocations that produce the non-relocatable image
1.27 crook 13139: files, then you see the output of the comparing program: It displays the
13140: offset used for data addresses and the offset used for code addresses;
1.1 anton 13141: moreover, for each cell that cannot be represented correctly in the
1.44 crook 13142: image files, it displays a line like this:
1.1 anton 13143:
13144: @example
13145: 78DC BFFFFA50 BFFFFA40
13146: @end example
13147:
13148: This means that at offset $78dc from @code{forthstart}, one input image
13149: contains $bffffa50, and the other contains $bffffa40. Since these cells
13150: cannot be represented correctly in the output image, you should examine
13151: these places in the dictionary and verify that these cells are dead
13152: (i.e., not read before they are written).
1.39 anton 13153:
13154: @cindex --application, @code{gforthmi} option
13155: If you insert the option @code{--application} in front of the image file
13156: name, you will get an image that uses the @code{--appl-image} option
13157: instead of the @code{--image-file} option (@pxref{Invoking
13158: Gforth}). When you execute such an image on Unix (by typing the image
13159: name as command), the Gforth engine will pass all options to the image
13160: instead of trying to interpret them as engine options.
1.1 anton 13161:
1.27 crook 13162: If you type @file{gforthmi} with no arguments, it prints some usage
13163: instructions.
13164:
1.1 anton 13165: @cindex @code{savesystem} during @file{gforthmi}
13166: @cindex @code{bye} during @file{gforthmi}
13167: @cindex doubly indirect threaded code
1.44 crook 13168: @cindex environment variables
13169: @cindex @code{GFORTHD} -- environment variable
13170: @cindex @code{GFORTH} -- environment variable
1.1 anton 13171: @cindex @code{gforth-ditc}
1.29 crook 13172: There are a few wrinkles: After processing the passed @i{options}, the
1.1 anton 13173: words @code{savesystem} and @code{bye} must be visible. A special doubly
13174: indirect threaded version of the @file{gforth} executable is used for
1.62 ! crook 13175: creating the non-relocatable images; you can pass the exact filename of
1.1 anton 13176: this executable through the environment variable @code{GFORTHD}
13177: (default: @file{gforth-ditc}); if you pass a version that is not doubly
13178: indirect threaded, you will not get a fully relocatable image, but a
1.27 crook 13179: data-relocatable image (because there is no code address offset). The
13180: normal @file{gforth} executable is used for creating the relocatable
13181: image; you can pass the exact filename of this executable through the
13182: environment variable @code{GFORTH}.
1.1 anton 13183:
13184: @node cross.fs, , gforthmi, Fully Relocatable Image Files
13185: @subsection @file{cross.fs}
13186: @cindex @file{cross.fs}
13187: @cindex cross-compiler
13188: @cindex metacompiler
1.47 crook 13189: @cindex target compiler
1.1 anton 13190:
13191: You can also use @code{cross}, a batch compiler that accepts a Forth-like
1.47 crook 13192: programming language (@pxref{Cross Compiler}).
1.1 anton 13193:
1.47 crook 13194: @code{cross} allows you to create image files for machines with
1.1 anton 13195: different data sizes and data formats than the one used for generating
13196: the image file. You can also use it to create an application image that
13197: does not contain a Forth compiler. These features are bought with
13198: restrictions and inconveniences in programming. E.g., addresses have to
13199: be stored in memory with special words (@code{A!}, @code{A,}, etc.) in
13200: order to make the code relocatable.
13201:
13202:
13203: @node Stack and Dictionary Sizes, Running Image Files, Fully Relocatable Image Files, Image Files
13204: @section Stack and Dictionary Sizes
13205: @cindex image file, stack and dictionary sizes
13206: @cindex dictionary size default
13207: @cindex stack size default
13208:
13209: If you invoke Gforth with a command line flag for the size
13210: (@pxref{Invoking Gforth}), the size you specify is stored in the
13211: dictionary. If you save the dictionary with @code{savesystem} or create
13212: an image with @file{gforthmi}, this size will become the default
13213: for the resulting image file. E.g., the following will create a
1.21 crook 13214: fully relocatable version of @file{gforth.fi} with a 1MB dictionary:
1.1 anton 13215:
13216: @example
13217: gforthmi gforth.fi -m 1M
13218: @end example
13219:
13220: In other words, if you want to set the default size for the dictionary
13221: and the stacks of an image, just invoke @file{gforthmi} with the
13222: appropriate options when creating the image.
13223:
13224: @cindex stack size, cache-friendly
13225: Note: For cache-friendly behaviour (i.e., good performance), you should
13226: make the sizes of the stacks modulo, say, 2K, somewhat different. E.g.,
13227: the default stack sizes are: data: 16k (mod 2k=0); fp: 15.5k (mod
13228: 2k=1.5k); return: 15k(mod 2k=1k); locals: 14.5k (mod 2k=0.5k).
13229:
13230: @node Running Image Files, Modifying the Startup Sequence, Stack and Dictionary Sizes, Image Files
13231: @section Running Image Files
13232: @cindex running image files
13233: @cindex invoking image files
13234: @cindex image file invocation
13235:
13236: @cindex -i, invoke image file
13237: @cindex --image file, invoke image file
1.29 crook 13238: You can invoke Gforth with an image file @i{image} instead of the
1.1 anton 13239: default @file{gforth.fi} with the @code{-i} flag (@pxref{Invoking Gforth}):
13240: @example
1.29 crook 13241: gforth -i @i{image}
1.1 anton 13242: @end example
13243:
13244: @cindex executable image file
1.26 crook 13245: @cindex image file, executable
1.1 anton 13246: If your operating system supports starting scripts with a line of the
13247: form @code{#! ...}, you just have to type the image file name to start
13248: Gforth with this image file (note that the file extension @code{.fi} is
1.29 crook 13249: just a convention). I.e., to run Gforth with the image file @i{image},
13250: you can just type @i{image} instead of @code{gforth -i @i{image}}.
1.27 crook 13251: This works because every @code{.fi} file starts with a line of this
13252: format:
13253:
13254: @example
13255: #! /usr/local/bin/gforth-0.4.0 -i
13256: @end example
13257:
13258: The file and pathname for the Gforth engine specified on this line is
13259: the specific Gforth executable that it was built against; i.e. the value
13260: of the environment variable @code{GFORTH} at the time that
13261: @file{gforthmi} was executed.
1.1 anton 13262:
1.27 crook 13263: You can make use of the same shell capability to make a Forth source
13264: file into an executable. For example, if you place this text in a file:
1.26 crook 13265:
13266: @example
13267: #! /usr/local/bin/gforth
13268:
13269: ." Hello, world" CR
13270: bye
13271: @end example
13272:
13273: @noindent
1.27 crook 13274: and then make the file executable (chmod +x in Unix), you can run it
1.26 crook 13275: directly from the command line. The sequence @code{#!} is used in two
13276: ways; firstly, it is recognised as a ``magic sequence'' by the operating
1.29 crook 13277: system@footnote{The Unix kernel actually recognises two types of files:
13278: executable files and files of data, where the data is processed by an
13279: interpreter that is specified on the ``interpreter line'' -- the first
13280: line of the file, starting with the sequence #!. There may be a small
13281: limit (e.g., 32) on the number of characters that may be specified on
13282: the interpreter line.} secondly it is treated as a comment character by
13283: Gforth. Because of the second usage, a space is required between
13284: @code{#!} and the path to the executable.
1.27 crook 13285:
13286: The disadvantage of this latter technique, compared with using
13287: @file{gforthmi}, is that it is slower; the Forth source code is compiled
13288: on-the-fly, each time the program is invoked.
13289:
1.26 crook 13290:
1.1 anton 13291: doc-#!
13292:
1.44 crook 13293:
1.1 anton 13294: @node Modifying the Startup Sequence, , Running Image Files, Image Files
13295: @section Modifying the Startup Sequence
13296: @cindex startup sequence for image file
13297: @cindex image file initialization sequence
13298: @cindex initialization sequence of image file
13299:
13300: You can add your own initialization to the startup sequence through the
1.26 crook 13301: deferred word @code{'cold}. @code{'cold} is invoked just before the
13302: image-specific command line processing (by default, loading files and
13303: evaluating (@code{-e}) strings) starts.
1.1 anton 13304:
13305: A sequence for adding your initialization usually looks like this:
13306:
13307: @example
13308: :noname
13309: Defers 'cold \ do other initialization stuff (e.g., rehashing wordlists)
13310: ... \ your stuff
13311: ; IS 'cold
13312: @end example
13313:
13314: @cindex turnkey image files
1.26 crook 13315: @cindex image file, turnkey applications
1.1 anton 13316: You can make a turnkey image by letting @code{'cold} execute a word
13317: (your turnkey application) that never returns; instead, it exits Gforth
13318: via @code{bye} or @code{throw}.
13319:
13320: @cindex command-line arguments, access
13321: @cindex arguments on the command line, access
13322: You can access the (image-specific) command-line arguments through the
1.26 crook 13323: variables @code{argc} and @code{argv}. @code{arg} provides convenient
1.1 anton 13324: access to @code{argv}.
13325:
1.26 crook 13326: If @code{'cold} exits normally, Gforth processes the command-line
13327: arguments as files to be loaded and strings to be evaluated. Therefore,
13328: @code{'cold} should remove the arguments it has used in this case.
13329:
1.44 crook 13330:
13331:
1.26 crook 13332: doc-'cold
1.1 anton 13333: doc-argc
13334: doc-argv
13335: doc-arg
13336:
13337:
1.44 crook 13338:
1.1 anton 13339: @c ******************************************************************
1.13 pazsan 13340: @node Engine, Binding to System Library, Image Files, Top
1.1 anton 13341: @chapter Engine
13342: @cindex engine
13343: @cindex virtual machine
13344:
1.26 crook 13345: Reading this chapter is not necessary for programming with Gforth. It
1.1 anton 13346: may be helpful for finding your way in the Gforth sources.
13347:
13348: The ideas in this section have also been published in the papers
13349: @cite{ANS fig/GNU/??? Forth} (in German) by Bernd Paysan, presented at
13350: the Forth-Tagung '93 and @cite{A Portable Forth Engine} by M. Anton
13351: Ertl, presented at EuroForth '93; the latter is available at
1.47 crook 13352: @*@uref{http://www.complang.tuwien.ac.at/papers/ertl93.ps.Z}.
1.1 anton 13353:
13354: @menu
13355: * Portability::
13356: * Threading::
13357: * Primitives::
13358: * Performance::
13359: @end menu
13360:
13361: @node Portability, Threading, Engine, Engine
13362: @section Portability
13363: @cindex engine portability
13364:
1.26 crook 13365: An important goal of the Gforth Project is availability across a wide
13366: range of personal machines. fig-Forth, and, to a lesser extent, F83,
13367: achieved this goal by manually coding the engine in assembly language
13368: for several then-popular processors. This approach is very
13369: labor-intensive and the results are short-lived due to progress in
13370: computer architecture.
1.1 anton 13371:
13372: @cindex C, using C for the engine
13373: Others have avoided this problem by coding in C, e.g., Mitch Bradley
13374: (cforth), Mikael Patel (TILE) and Dirk Zoller (pfe). This approach is
13375: particularly popular for UNIX-based Forths due to the large variety of
13376: architectures of UNIX machines. Unfortunately an implementation in C
13377: does not mix well with the goals of efficiency and with using
13378: traditional techniques: Indirect or direct threading cannot be expressed
13379: in C, and switch threading, the fastest technique available in C, is
13380: significantly slower. Another problem with C is that it is very
13381: cumbersome to express double integer arithmetic.
13382:
13383: @cindex GNU C for the engine
13384: @cindex long long
13385: Fortunately, there is a portable language that does not have these
13386: limitations: GNU C, the version of C processed by the GNU C compiler
13387: (@pxref{C Extensions, , Extensions to the C Language Family, gcc.info,
13388: GNU C Manual}). Its labels as values feature (@pxref{Labels as Values, ,
13389: Labels as Values, gcc.info, GNU C Manual}) makes direct and indirect
13390: threading possible, its @code{long long} type (@pxref{Long Long, ,
13391: Double-Word Integers, gcc.info, GNU C Manual}) corresponds to Forth's
13392: double numbers@footnote{Unfortunately, long longs are not implemented
13393: properly on all machines (e.g., on alpha-osf1, long longs are only 64
13394: bits, the same size as longs (and pointers), but they should be twice as
1.4 anton 13395: long according to @pxref{Long Long, , Double-Word Integers, gcc.info, GNU
1.1 anton 13396: C Manual}). So, we had to implement doubles in C after all. Still, on
13397: most machines we can use long longs and achieve better performance than
13398: with the emulation package.}. GNU C is available for free on all
13399: important (and many unimportant) UNIX machines, VMS, 80386s running
13400: MS-DOS, the Amiga, and the Atari ST, so a Forth written in GNU C can run
13401: on all these machines.
13402:
13403: Writing in a portable language has the reputation of producing code that
13404: is slower than assembly. For our Forth engine we repeatedly looked at
13405: the code produced by the compiler and eliminated most compiler-induced
13406: inefficiencies by appropriate changes in the source code.
13407:
13408: @cindex explicit register declarations
13409: @cindex --enable-force-reg, configuration flag
13410: @cindex -DFORCE_REG
13411: However, register allocation cannot be portably influenced by the
13412: programmer, leading to some inefficiencies on register-starved
13413: machines. We use explicit register declarations (@pxref{Explicit Reg
13414: Vars, , Variables in Specified Registers, gcc.info, GNU C Manual}) to
13415: improve the speed on some machines. They are turned on by using the
13416: configuration flag @code{--enable-force-reg} (@code{gcc} switch
13417: @code{-DFORCE_REG}). Unfortunately, this feature not only depends on the
13418: machine, but also on the compiler version: On some machines some
13419: compiler versions produce incorrect code when certain explicit register
13420: declarations are used. So by default @code{-DFORCE_REG} is not used.
13421:
13422: @node Threading, Primitives, Portability, Engine
13423: @section Threading
13424: @cindex inner interpreter implementation
13425: @cindex threaded code implementation
13426:
13427: @cindex labels as values
13428: GNU C's labels as values extension (available since @code{gcc-2.0},
13429: @pxref{Labels as Values, , Labels as Values, gcc.info, GNU C Manual})
1.29 crook 13430: makes it possible to take the address of @i{label} by writing
13431: @code{&&@i{label}}. This address can then be used in a statement like
13432: @code{goto *@i{address}}. I.e., @code{goto *&&x} is the same as
1.1 anton 13433: @code{goto x}.
13434:
1.26 crook 13435: @cindex @code{NEXT}, indirect threaded
1.1 anton 13436: @cindex indirect threaded inner interpreter
13437: @cindex inner interpreter, indirect threaded
1.26 crook 13438: With this feature an indirect threaded @code{NEXT} looks like:
1.1 anton 13439: @example
13440: cfa = *ip++;
13441: ca = *cfa;
13442: goto *ca;
13443: @end example
13444: @cindex instruction pointer
13445: For those unfamiliar with the names: @code{ip} is the Forth instruction
13446: pointer; the @code{cfa} (code-field address) corresponds to ANS Forths
13447: execution token and points to the code field of the next word to be
13448: executed; The @code{ca} (code address) fetched from there points to some
13449: executable code, e.g., a primitive or the colon definition handler
13450: @code{docol}.
13451:
1.26 crook 13452: @cindex @code{NEXT}, direct threaded
1.1 anton 13453: @cindex direct threaded inner interpreter
13454: @cindex inner interpreter, direct threaded
13455: Direct threading is even simpler:
13456: @example
13457: ca = *ip++;
13458: goto *ca;
13459: @end example
13460:
13461: Of course we have packaged the whole thing neatly in macros called
1.26 crook 13462: @code{NEXT} and @code{NEXT1} (the part of @code{NEXT} after fetching the cfa).
1.1 anton 13463:
13464: @menu
13465: * Scheduling::
13466: * Direct or Indirect Threaded?::
13467: * DOES>::
13468: @end menu
13469:
13470: @node Scheduling, Direct or Indirect Threaded?, Threading, Threading
13471: @subsection Scheduling
13472: @cindex inner interpreter optimization
13473:
13474: There is a little complication: Pipelined and superscalar processors,
13475: i.e., RISC and some modern CISC machines can process independent
13476: instructions while waiting for the results of an instruction. The
13477: compiler usually reorders (schedules) the instructions in a way that
13478: achieves good usage of these delay slots. However, on our first tries
13479: the compiler did not do well on scheduling primitives. E.g., for
13480: @code{+} implemented as
13481: @example
13482: n=sp[0]+sp[1];
13483: sp++;
13484: sp[0]=n;
13485: NEXT;
13486: @end example
1.26 crook 13487: the @code{NEXT} comes strictly after the other code, i.e., there is nearly no
1.1 anton 13488: scheduling. After a little thought the problem becomes clear: The
1.21 crook 13489: compiler cannot know that @code{sp} and @code{ip} point to different
13490: addresses (and the version of @code{gcc} we used would not know it even
13491: if it was possible), so it could not move the load of the cfa above the
13492: store to the TOS. Indeed the pointers could be the same, if code on or
13493: very near the top of stack were executed. In the interest of speed we
13494: chose to forbid this probably unused ``feature'' and helped the compiler
1.26 crook 13495: in scheduling: @code{NEXT} is divided into the loading part (@code{NEXT_P1})
1.21 crook 13496: and the goto part (@code{NEXT_P2}). @code{+} now looks like:
1.1 anton 13497: @example
13498: n=sp[0]+sp[1];
13499: sp++;
13500: NEXT_P1;
13501: sp[0]=n;
13502: NEXT_P2;
13503: @end example
13504: This can be scheduled optimally by the compiler.
13505:
13506: This division can be turned off with the switch @code{-DCISC_NEXT}. This
13507: switch is on by default on machines that do not profit from scheduling
13508: (e.g., the 80386), in order to preserve registers.
13509:
13510: @node Direct or Indirect Threaded?, DOES>, Scheduling, Threading
13511: @subsection Direct or Indirect Threaded?
13512: @cindex threading, direct or indirect?
13513:
13514: @cindex -DDIRECT_THREADED
13515: Both! After packaging the nasty details in macro definitions we
13516: realized that we could switch between direct and indirect threading by
13517: simply setting a compilation flag (@code{-DDIRECT_THREADED}) and
13518: defining a few machine-specific macros for the direct-threading case.
13519: On the Forth level we also offer access words that hide the
13520: differences between the threading methods (@pxref{Threading Words}).
13521:
13522: Indirect threading is implemented completely machine-independently.
13523: Direct threading needs routines for creating jumps to the executable
1.21 crook 13524: code (e.g. to @code{docol} or @code{dodoes}). These routines are inherently
13525: machine-dependent, but they do not amount to many source lines. Therefore,
13526: even porting direct threading to a new machine requires little effort.
1.1 anton 13527:
13528: @cindex --enable-indirect-threaded, configuration flag
13529: @cindex --enable-direct-threaded, configuration flag
13530: The default threading method is machine-dependent. You can enforce a
13531: specific threading method when building Gforth with the configuration
13532: flag @code{--enable-direct-threaded} or
13533: @code{--enable-indirect-threaded}. Note that direct threading is not
13534: supported on all machines.
13535:
13536: @node DOES>, , Direct or Indirect Threaded?, Threading
13537: @subsection DOES>
13538: @cindex @code{DOES>} implementation
13539:
1.26 crook 13540: @cindex @code{dodoes} routine
13541: @cindex @code{DOES>}-code
1.1 anton 13542: One of the most complex parts of a Forth engine is @code{dodoes}, i.e.,
13543: the chunk of code executed by every word defined by a
13544: @code{CREATE}...@code{DOES>} pair. The main problem here is: How to find
13545: the Forth code to be executed, i.e. the code after the
1.26 crook 13546: @code{DOES>} (the @code{DOES>}-code)? There are two solutions:
1.1 anton 13547:
1.21 crook 13548: In fig-Forth the code field points directly to the @code{dodoes} and the
1.45 crook 13549: @code{DOES>}-code address is stored in the cell after the code address (i.e. at
1.29 crook 13550: @code{@i{CFA} cell+}). It may seem that this solution is illegal in
1.1 anton 13551: the Forth-79 and all later standards, because in fig-Forth this address
13552: lies in the body (which is illegal in these standards). However, by
13553: making the code field larger for all words this solution becomes legal
13554: again. We use this approach for the indirect threaded version and for
13555: direct threading on some machines. Leaving a cell unused in most words
13556: is a bit wasteful, but on the machines we are targeting this is hardly a
13557: problem. The other reason for having a code field size of two cells is
13558: to avoid having different image files for direct and indirect threaded
13559: systems (direct threaded systems require two-cell code fields on many
13560: machines).
13561:
1.26 crook 13562: @cindex @code{DOES>}-handler
1.1 anton 13563: The other approach is that the code field points or jumps to the cell
1.26 crook 13564: after @code{DOES>}. In this variant there is a jump to @code{dodoes} at
13565: this address (the @code{DOES>}-handler). @code{dodoes} can then get the
13566: @code{DOES>}-code address by computing the code address, i.e., the address of
1.45 crook 13567: the jump to @code{dodoes}, and add the length of that jump field. A variant of
1.1 anton 13568: this is to have a call to @code{dodoes} after the @code{DOES>}; then the
13569: return address (which can be found in the return register on RISCs) is
1.26 crook 13570: the @code{DOES>}-code address. Since the two cells available in the code field
1.1 anton 13571: are used up by the jump to the code address in direct threading on many
13572: architectures, we use this approach for direct threading on these
13573: architectures. We did not want to add another cell to the code field.
13574:
13575: @node Primitives, Performance, Threading, Engine
13576: @section Primitives
13577: @cindex primitives, implementation
13578: @cindex virtual machine instructions, implementation
13579:
13580: @menu
13581: * Automatic Generation::
13582: * TOS Optimization::
13583: * Produced code::
13584: @end menu
13585:
13586: @node Automatic Generation, TOS Optimization, Primitives, Primitives
13587: @subsection Automatic Generation
13588: @cindex primitives, automatic generation
13589:
13590: @cindex @file{prims2x.fs}
13591: Since the primitives are implemented in a portable language, there is no
13592: longer any need to minimize the number of primitives. On the contrary,
13593: having many primitives has an advantage: speed. In order to reduce the
13594: number of errors in primitives and to make programming them easier, we
13595: provide a tool, the primitive generator (@file{prims2x.fs}), that
13596: automatically generates most (and sometimes all) of the C code for a
13597: primitive from the stack effect notation. The source for a primitive
13598: has the following form:
13599:
13600: @cindex primitive source format
13601: @format
1.58 anton 13602: @i{Forth-name} ( @i{stack-effect} ) @i{category} [@i{pronounc.}]
1.29 crook 13603: [@code{""}@i{glossary entry}@code{""}]
13604: @i{C code}
1.1 anton 13605: [@code{:}
1.29 crook 13606: @i{Forth code}]
1.1 anton 13607: @end format
13608:
13609: The items in brackets are optional. The category and glossary fields
13610: are there for generating the documentation, the Forth code is there
13611: for manual implementations on machines without GNU C. E.g., the source
13612: for the primitive @code{+} is:
13613: @example
1.58 anton 13614: + ( n1 n2 -- n ) core plus
1.1 anton 13615: n = n1+n2;
13616: @end example
13617:
13618: This looks like a specification, but in fact @code{n = n1+n2} is C
13619: code. Our primitive generation tool extracts a lot of information from
13620: the stack effect notations@footnote{We use a one-stack notation, even
13621: though we have separate data and floating-point stacks; The separate
13622: notation can be generated easily from the unified notation.}: The number
13623: of items popped from and pushed on the stack, their type, and by what
13624: name they are referred to in the C code. It then generates a C code
13625: prelude and postlude for each primitive. The final C code for @code{+}
13626: looks like this:
13627:
13628: @example
1.46 pazsan 13629: I_plus: /* + ( n1 n2 -- n ) */ /* label, stack effect */
1.1 anton 13630: /* */ /* documentation */
13631: @{
13632: DEF_CA /* definition of variable ca (indirect threading) */
13633: Cell n1; /* definitions of variables */
13634: Cell n2;
13635: Cell n;
13636: n1 = (Cell) sp[1]; /* input */
13637: n2 = (Cell) TOS;
13638: sp += 1; /* stack adjustment */
13639: NAME("+") /* debugging output (with -DDEBUG) */
13640: @{
13641: n = n1+n2; /* C code taken from the source */
13642: @}
13643: NEXT_P1; /* NEXT part 1 */
13644: TOS = (Cell)n; /* output */
13645: NEXT_P2; /* NEXT part 2 */
13646: @}
13647: @end example
13648:
13649: This looks long and inefficient, but the GNU C compiler optimizes quite
13650: well and produces optimal code for @code{+} on, e.g., the R3000 and the
13651: HP RISC machines: Defining the @code{n}s does not produce any code, and
13652: using them as intermediate storage also adds no cost.
13653:
1.26 crook 13654: There are also other optimizations that are not illustrated by this
13655: example: assignments between simple variables are usually for free (copy
1.1 anton 13656: propagation). If one of the stack items is not used by the primitive
13657: (e.g. in @code{drop}), the compiler eliminates the load from the stack
13658: (dead code elimination). On the other hand, there are some things that
13659: the compiler does not do, therefore they are performed by
13660: @file{prims2x.fs}: The compiler does not optimize code away that stores
13661: a stack item to the place where it just came from (e.g., @code{over}).
13662:
13663: While programming a primitive is usually easy, there are a few cases
13664: where the programmer has to take the actions of the generator into
13665: account, most notably @code{?dup}, but also words that do not (always)
1.26 crook 13666: fall through to @code{NEXT}.
1.1 anton 13667:
13668: @node TOS Optimization, Produced code, Automatic Generation, Primitives
13669: @subsection TOS Optimization
13670: @cindex TOS optimization for primitives
13671: @cindex primitives, keeping the TOS in a register
13672:
13673: An important optimization for stack machine emulators, e.g., Forth
13674: engines, is keeping one or more of the top stack items in
1.29 crook 13675: registers. If a word has the stack effect @i{in1}...@i{inx} @code{--}
13676: @i{out1}...@i{outy}, keeping the top @i{n} items in registers
1.1 anton 13677: @itemize @bullet
13678: @item
1.29 crook 13679: is better than keeping @i{n-1} items, if @i{x>=n} and @i{y>=n},
1.1 anton 13680: due to fewer loads from and stores to the stack.
1.29 crook 13681: @item is slower than keeping @i{n-1} items, if @i{x<>y} and @i{x<n} and
13682: @i{y<n}, due to additional moves between registers.
1.1 anton 13683: @end itemize
13684:
13685: @cindex -DUSE_TOS
13686: @cindex -DUSE_NO_TOS
13687: In particular, keeping one item in a register is never a disadvantage,
13688: if there are enough registers. Keeping two items in registers is a
13689: disadvantage for frequent words like @code{?branch}, constants,
13690: variables, literals and @code{i}. Therefore our generator only produces
13691: code that keeps zero or one items in registers. The generated C code
13692: covers both cases; the selection between these alternatives is made at
13693: C-compile time using the switch @code{-DUSE_TOS}. @code{TOS} in the C
13694: code for @code{+} is just a simple variable name in the one-item case,
13695: otherwise it is a macro that expands into @code{sp[0]}. Note that the
13696: GNU C compiler tries to keep simple variables like @code{TOS} in
13697: registers, and it usually succeeds, if there are enough registers.
13698:
13699: @cindex -DUSE_FTOS
13700: @cindex -DUSE_NO_FTOS
13701: The primitive generator performs the TOS optimization for the
13702: floating-point stack, too (@code{-DUSE_FTOS}). For floating-point
13703: operations the benefit of this optimization is even larger:
13704: floating-point operations take quite long on most processors, but can be
13705: performed in parallel with other operations as long as their results are
13706: not used. If the FP-TOS is kept in a register, this works. If
13707: it is kept on the stack, i.e., in memory, the store into memory has to
13708: wait for the result of the floating-point operation, lengthening the
13709: execution time of the primitive considerably.
13710:
13711: The TOS optimization makes the automatic generation of primitives a
13712: bit more complicated. Just replacing all occurrences of @code{sp[0]} by
13713: @code{TOS} is not sufficient. There are some special cases to
13714: consider:
13715: @itemize @bullet
13716: @item In the case of @code{dup ( w -- w w )} the generator must not
13717: eliminate the store to the original location of the item on the stack,
13718: if the TOS optimization is turned on.
13719: @item Primitives with stack effects of the form @code{--}
1.29 crook 13720: @i{out1}...@i{outy} must store the TOS to the stack at the start.
13721: Likewise, primitives with the stack effect @i{in1}...@i{inx} @code{--}
1.1 anton 13722: must load the TOS from the stack at the end. But for the null stack
13723: effect @code{--} no stores or loads should be generated.
13724: @end itemize
13725:
13726: @node Produced code, , TOS Optimization, Primitives
13727: @subsection Produced code
13728: @cindex primitives, assembly code listing
13729:
13730: @cindex @file{engine.s}
13731: To see what assembly code is produced for the primitives on your machine
13732: with your compiler and your flag settings, type @code{make engine.s} and
13733: look at the resulting file @file{engine.s}.
13734:
13735: @node Performance, , Primitives, Engine
13736: @section Performance
13737: @cindex performance of some Forth interpreters
13738: @cindex engine performance
13739: @cindex benchmarking Forth systems
13740: @cindex Gforth performance
13741:
13742: On RISCs the Gforth engine is very close to optimal; i.e., it is usually
13743: impossible to write a significantly faster engine.
13744:
13745: On register-starved machines like the 386 architecture processors
13746: improvements are possible, because @code{gcc} does not utilize the
13747: registers as well as a human, even with explicit register declarations;
13748: e.g., Bernd Beuster wrote a Forth system fragment in assembly language
13749: and hand-tuned it for the 486; this system is 1.19 times faster on the
13750: Sieve benchmark on a 486DX2/66 than Gforth compiled with
1.40 anton 13751: @code{gcc-2.6.3} with @code{-DFORCE_REG}. The situation has improved
13752: with gcc-2.95 and gforth-0.4.9; now the most important virtual machine
13753: registers fit in real registers (and we can even afford to use the TOS
13754: optimization), resulting in a speedup of 1.14 on the sieve over the
13755: earlier results.
1.1 anton 13756:
13757: @cindex Win32Forth performance
13758: @cindex NT Forth performance
13759: @cindex eforth performance
13760: @cindex ThisForth performance
13761: @cindex PFE performance
13762: @cindex TILE performance
1.40 anton 13763: The potential advantage of assembly language implementations
1.1 anton 13764: is not necessarily realized in complete Forth systems: We compared
1.40 anton 13765: Gforth-0.4.9 (direct threaded, compiled with @code{gcc-2.95.1} and
1.1 anton 13766: @code{-DFORCE_REG}) with Win32Forth 1.2093, LMI's NT Forth (Beta, May
13767: 1994) and Eforth (with and without peephole (aka pinhole) optimization
13768: of the threaded code); all these systems were written in assembly
13769: language. We also compared Gforth with three systems written in C:
13770: PFE-0.9.14 (compiled with @code{gcc-2.6.3} with the default
13771: configuration for Linux: @code{-O2 -fomit-frame-pointer -DUSE_REGS
1.21 crook 13772: -DUNROLL_NEXT}), ThisForth Beta (compiled with @code{gcc-2.6.3 -O3
13773: -fomit-frame-pointer}; ThisForth employs peephole optimization of the
1.1 anton 13774: threaded code) and TILE (compiled with @code{make opt}). We benchmarked
13775: Gforth, PFE, ThisForth and TILE on a 486DX2/66 under Linux. Kenneth
13776: O'Heskin kindly provided the results for Win32Forth and NT Forth on a
13777: 486DX2/66 with similar memory performance under Windows NT. Marcel
13778: Hendrix ported Eforth to Linux, then extended it to run the benchmarks,
13779: added the peephole optimizer, ran the benchmarks and reported the
13780: results.
1.40 anton 13781:
1.1 anton 13782: We used four small benchmarks: the ubiquitous Sieve; bubble-sorting and
13783: matrix multiplication come from the Stanford integer benchmarks and have
13784: been translated into Forth by Martin Fraeman; we used the versions
13785: included in the TILE Forth package, but with bigger data set sizes; and
13786: a recursive Fibonacci number computation for benchmarking calling
13787: performance. The following table shows the time taken for the benchmarks
13788: scaled by the time taken by Gforth (in other words, it shows the speedup
13789: factor that Gforth achieved over the other systems).
13790:
13791: @example
1.40 anton 13792: relative Win32- NT eforth This-
1.1 anton 13793: time Gforth Forth Forth eforth +opt PFE Forth TILE
1.40 anton 13794: sieve 1.00 1.58 1.30 1.58 0.97 1.80 3.63 9.79
13795: bubble 1.00 1.55 1.67 1.75 1.04 1.78 4.59
13796: matmul 1.00 1.67 1.53 1.66 0.84 1.79 4.63
13797: fib 1.00 1.75 1.53 1.40 0.99 1.99 3.43 4.93
1.1 anton 13798: @end example
13799:
1.26 crook 13800: You may be quite surprised by the good performance of Gforth when
13801: compared with systems written in assembly language. One important reason
13802: for the disappointing performance of these other systems is probably
13803: that they are not written optimally for the 486 (e.g., they use the
13804: @code{lods} instruction). In addition, Win32Forth uses a comfortable,
13805: but costly method for relocating the Forth image: like @code{cforth}, it
13806: computes the actual addresses at run time, resulting in two address
13807: computations per @code{NEXT} (@pxref{Image File Background}).
13808:
1.40 anton 13809: Only Eforth with the peephole optimizer performs comparable to
13810: Gforth. The speedups achieved with peephole optimization of threaded
13811: code are quite remarkable. Adding a peephole optimizer to Gforth should
13812: cause similar speedups.
1.1 anton 13813:
13814: The speedup of Gforth over PFE, ThisForth and TILE can be easily
13815: explained with the self-imposed restriction of the latter systems to
13816: standard C, which makes efficient threading impossible (however, the
1.4 anton 13817: measured implementation of PFE uses a GNU C extension: @pxref{Global Reg
1.1 anton 13818: Vars, , Defining Global Register Variables, gcc.info, GNU C Manual}).
13819: Moreover, current C compilers have a hard time optimizing other aspects
13820: of the ThisForth and the TILE source.
13821:
1.26 crook 13822: The performance of Gforth on 386 architecture processors varies widely
13823: with the version of @code{gcc} used. E.g., @code{gcc-2.5.8} failed to
13824: allocate any of the virtual machine registers into real machine
13825: registers by itself and would not work correctly with explicit register
1.40 anton 13826: declarations, giving a 1.5 times slower engine (on a 486DX2/66 running
1.26 crook 13827: the Sieve) than the one measured above.
1.1 anton 13828:
1.26 crook 13829: Note that there have been several releases of Win32Forth since the
13830: release presented here, so the results presented above may have little
1.40 anton 13831: predictive value for the performance of Win32Forth today (results for
13832: the current release on an i486DX2/66 are welcome).
1.1 anton 13833:
13834: @cindex @file{Benchres}
13835: In @cite{Translating Forth to Efficient C} by M. Anton Ertl and Martin
13836: Maierhofer (presented at EuroForth '95), an indirect threaded version of
13837: Gforth is compared with Win32Forth, NT Forth, PFE, and ThisForth; that
1.40 anton 13838: version of Gforth is slower on a 486 than the direct threaded version
13839: used here. The paper available at
1.47 crook 13840: @*@uref{http://www.complang.tuwien.ac.at/papers/ertl&maierhofer95.ps.gz};
1.1 anton 13841: it also contains numbers for some native code systems. You can find a
13842: newer version of these measurements at
1.47 crook 13843: @uref{http://www.complang.tuwien.ac.at/forth/performance.html}. You can
1.1 anton 13844: find numbers for Gforth on various machines in @file{Benchres}.
13845:
1.26 crook 13846: @c ******************************************************************
1.13 pazsan 13847: @node Binding to System Library, Cross Compiler, Engine, Top
1.14 pazsan 13848: @chapter Binding to System Library
1.13 pazsan 13849:
13850: @node Cross Compiler, Bugs, Binding to System Library, Top
1.14 pazsan 13851: @chapter Cross Compiler
1.47 crook 13852: @cindex @file{cross.fs}
13853: @cindex cross-compiler
13854: @cindex metacompiler
13855: @cindex target compiler
1.13 pazsan 13856:
1.46 pazsan 13857: The cross compiler is used to bootstrap a Forth kernel. Since Gforth is
13858: mostly written in Forth, including crucial parts like the outer
13859: interpreter and compiler, it needs compiled Forth code to get
13860: started. The cross compiler allows to create new images for other
13861: architectures, even running under another Forth system.
1.13 pazsan 13862:
13863: @menu
13864: * Using the Cross Compiler::
13865: * How the Cross Compiler Works::
13866: @end menu
13867:
1.21 crook 13868: @node Using the Cross Compiler, How the Cross Compiler Works, Cross Compiler, Cross Compiler
1.14 pazsan 13869: @section Using the Cross Compiler
1.46 pazsan 13870:
13871: The cross compiler uses a language that resembles Forth, but isn't. The
13872: main difference is that you can execute Forth code after definition,
13873: while you usually can't execute the code compiled by cross, because the
13874: code you are compiling is typically for a different computer than the
13875: one you are compiling on.
13876:
13877: The Makefile is already set up to allow you to create kernels for new
13878: architectures with a simple make command. The generic kernels using the
13879: GCC compiled virtual machine are created in the normal build process
13880: with @code{make}. To create a embedded Gforth executable for e.g. the
13881: 8086 processor (running on a DOS machine), type
13882:
13883: @example
13884: make kernl-8086.fi
13885: @end example
13886:
13887: This will use the machine description from the @file{arch/8086}
13888: directory to create a new kernel. A machine file may look like that:
13889:
13890: @example
13891: \ Parameter for target systems 06oct92py
13892:
13893: 4 Constant cell \ cell size in bytes
13894: 2 Constant cell<< \ cell shift to bytes
13895: 5 Constant cell>bit \ cell shift to bits
13896: 8 Constant bits/char \ bits per character
13897: 8 Constant bits/byte \ bits per byte [default: 8]
13898: 8 Constant float \ bytes per float
13899: 8 Constant /maxalign \ maximum alignment in bytes
13900: false Constant bigendian \ byte order
13901: ( true=big, false=little )
13902:
13903: include machpc.fs \ feature list
13904: @end example
13905:
13906: This part is obligatory for the cross compiler itself, the feature list
13907: is used by the kernel to conditionally compile some features in and out,
13908: depending on whether the target supports these features.
13909:
13910: There are some optional features, if you define your own primitives,
13911: have an assembler, or need special, nonstandard preparation to make the
13912: boot process work. @code{asm-include} include an assembler,
13913: @code{prims-include} includes primitives, and @code{>boot} prepares for
13914: booting.
13915:
13916: @example
13917: : asm-include ." Include assembler" cr
13918: s" arch/8086/asm.fs" included ;
13919:
13920: : prims-include ." Include primitives" cr
13921: s" arch/8086/prim.fs" included ;
13922:
13923: : >boot ." Prepare booting" cr
13924: s" ' boot >body into-forth 1+ !" evaluate ;
13925: @end example
13926:
13927: These words are used as sort of macro during the cross compilation in
13928: the file @file{kernel/main.fs}. Instead of using this macros, it would
13929: be possible --- but more complicated --- to write a new kernel project
13930: file, too.
13931:
13932: @file{kernel/main.fs} expects the machine description file name on the
13933: stack; the cross compiler itself (@file{cross.fs}) assumes that either
13934: @code{mach-file} leaves a counted string on the stack, or
13935: @code{machine-file} leaves an address, count pair of the filename on the
13936: stack.
13937:
13938: The feature list is typically controlled using @code{SetValue}, generic
13939: files that are used by several projects can use @code{DefaultValue}
13940: instead. Both functions work like @code{Value}, when the value isn't
13941: defined, but @code{SetValue} works like @code{to} if the value is
13942: defined, and @code{DefaultValue} doesn't set anything, if the value is
13943: defined.
13944:
13945: @example
13946: \ generic mach file for pc gforth 03sep97jaw
13947:
13948: true DefaultValue NIL \ relocating
13949:
13950: >ENVIRON
13951:
13952: true DefaultValue file \ controls the presence of the
13953: \ file access wordset
13954: true DefaultValue OS \ flag to indicate a operating system
13955:
13956: true DefaultValue prims \ true: primitives are c-code
13957:
13958: true DefaultValue floating \ floating point wordset is present
13959:
13960: true DefaultValue glocals \ gforth locals are present
13961: \ will be loaded
13962: true DefaultValue dcomps \ double number comparisons
13963:
13964: true DefaultValue hash \ hashing primitives are loaded/present
13965:
13966: true DefaultValue xconds \ used together with glocals,
13967: \ special conditionals supporting gforths'
13968: \ local variables
13969: true DefaultValue header \ save a header information
13970:
13971: true DefaultValue backtrace \ enables backtrace code
13972:
13973: false DefaultValue ec
13974: false DefaultValue crlf
13975:
13976: cell 2 = [IF] &32 [ELSE] &256 [THEN] KB DefaultValue kernel-size
13977:
13978: &16 KB DefaultValue stack-size
13979: &15 KB &512 + DefaultValue fstack-size
13980: &15 KB DefaultValue rstack-size
13981: &14 KB &512 + DefaultValue lstack-size
13982: @end example
1.13 pazsan 13983:
1.48 anton 13984: @node How the Cross Compiler Works, , Using the Cross Compiler, Cross Compiler
1.14 pazsan 13985: @section How the Cross Compiler Works
1.13 pazsan 13986:
13987: @node Bugs, Origin, Cross Compiler, Top
1.21 crook 13988: @appendix Bugs
1.1 anton 13989: @cindex bug reporting
13990:
1.21 crook 13991: Known bugs are described in the file @file{BUGS} in the Gforth distribution.
1.1 anton 13992:
13993: If you find a bug, please send a bug report to
1.33 anton 13994: @email{bug-gforth@@gnu.org}. A bug report should include this
1.21 crook 13995: information:
13996:
13997: @itemize @bullet
13998: @item
13999: The Gforth version used (it is announced at the start of an
14000: interactive Gforth session).
14001: @item
14002: The machine and operating system (on Unix
14003: systems @code{uname -a} will report this information).
14004: @item
14005: The installation options (send the file @file{config.status}).
14006: @item
14007: A complete list of changes (if any) you (or your installer) have made to the
14008: Gforth sources.
14009: @item
14010: A program (or a sequence of keyboard commands) that reproduces the bug.
14011: @item
14012: A description of what you think constitutes the buggy behaviour.
14013: @end itemize
1.1 anton 14014:
14015: For a thorough guide on reporting bugs read @ref{Bug Reporting, , How
14016: to Report Bugs, gcc.info, GNU C Manual}.
14017:
14018:
1.21 crook 14019: @node Origin, Forth-related information, Bugs, Top
14020: @appendix Authors and Ancestors of Gforth
1.1 anton 14021:
14022: @section Authors and Contributors
14023: @cindex authors of Gforth
14024: @cindex contributors to Gforth
14025:
14026: The Gforth project was started in mid-1992 by Bernd Paysan and Anton
14027: Ertl. The third major author was Jens Wilke. Lennart Benschop (who was
14028: one of Gforth's first users, in mid-1993) and Stuart Ramsden inspired us
14029: with their continuous feedback. Lennart Benshop contributed
14030: @file{glosgen.fs}, while Stuart Ramsden has been working on automatic
14031: support for calling C libraries. Helpful comments also came from Paul
14032: Kleinrubatscher, Christian Pirker, Dirk Zoller, Marcel Hendrix, John
1.58 anton 14033: Wavrik, Barrie Stott, Marc de Groot, Jorge Acerada, Bruce Hoyt, and
14034: Robert Epprecht. Since the release of Gforth-0.2.1 there were also
14035: helpful comments from many others; thank you all, sorry for not listing
14036: you here (but digging through my mailbox to extract your names is on my
14037: to-do list). Since the release of Gforth-0.4.0 Neal Crook worked on the
14038: manual.
1.1 anton 14039:
14040: Gforth also owes a lot to the authors of the tools we used (GCC, CVS,
14041: and autoconf, among others), and to the creators of the Internet: Gforth
1.21 crook 14042: was developed across the Internet, and its authors did not meet
1.20 pazsan 14043: physically for the first 4 years of development.
1.1 anton 14044:
14045: @section Pedigree
1.26 crook 14046: @cindex pedigree of Gforth
1.1 anton 14047:
1.20 pazsan 14048: Gforth descends from bigFORTH (1993) and fig-Forth. Gforth and PFE (by
1.1 anton 14049: Dirk Zoller) will cross-fertilize each other. Of course, a significant
14050: part of the design of Gforth was prescribed by ANS Forth.
14051:
1.20 pazsan 14052: Bernd Paysan wrote bigFORTH, a descendent from TurboForth, an unreleased
1.1 anton 14053: 32 bit native code version of VolksForth for the Atari ST, written
14054: mostly by Dietrich Weineck.
14055:
14056: VolksForth descends from F83. It was written by Klaus Schleisiek, Bernd
14057: Pennemann, Georg Rehfeld and Dietrich Weineck for the C64 (called
14058: UltraForth there) in the mid-80s and ported to the Atari ST in 1986.
14059:
14060: Henry Laxen and Mike Perry wrote F83 as a model implementation of the
14061: Forth-83 standard. !! Pedigree? When?
14062:
14063: A team led by Bill Ragsdale implemented fig-Forth on many processors in
14064: 1979. Robert Selzer and Bill Ragsdale developed the original
14065: implementation of fig-Forth for the 6502 based on microForth.
14066:
14067: The principal architect of microForth was Dean Sanderson. microForth was
14068: FORTH, Inc.'s first off-the-shelf product. It was developed in 1976 for
14069: the 1802, and subsequently implemented on the 8080, the 6800 and the
14070: Z80.
14071:
14072: All earlier Forth systems were custom-made, usually by Charles Moore,
14073: who discovered (as he puts it) Forth during the late 60s. The first full
14074: Forth existed in 1971.
14075:
14076: A part of the information in this section comes from @cite{The Evolution
14077: of Forth} by Elizabeth D. Rather, Donald R. Colburn and Charles
14078: H. Moore, presented at the HOPL-II conference and preprinted in SIGPLAN
14079: Notices 28(3), 1993. You can find more historical and genealogical
14080: information about Forth there.
14081:
1.21 crook 14082: @node Forth-related information, Word Index, Origin, Top
14083: @appendix Other Forth-related information
14084: @cindex Forth-related information
14085:
14086: @menu
14087: * Internet resources::
14088: * Books::
14089: * The Forth Interest Group::
14090: * Conferences::
14091: @end menu
14092:
14093:
14094: @node Internet resources, Books, Forth-related information, Forth-related information
14095: @section Internet resources
1.26 crook 14096: @cindex internet resources
1.21 crook 14097:
14098: @cindex comp.lang.forth
14099: @cindex frequently asked questions
1.45 crook 14100: There is an active news group (comp.lang.forth) discussing Forth and
1.21 crook 14101: Forth-related issues. A frequently-asked-questions (FAQ) list
1.45 crook 14102: is posted to the news group regularly, and archived at these sites:
1.21 crook 14103:
14104: @itemize @bullet
14105: @item
1.47 crook 14106: @uref{ftp://rtfm.mit.edu/pub/usenet-by-group/comp.lang.forth/}
1.21 crook 14107: @item
1.47 crook 14108: @uref{ftp://ftp.forth.org/pub/Forth/FAQ/}
1.21 crook 14109: @end itemize
14110:
14111: The FAQ list should be considered mandatory reading before posting to
1.45 crook 14112: the news group.
1.21 crook 14113:
14114: Here are some other web sites holding Forth-related material:
14115:
14116: @itemize @bullet
14117: @item
1.47 crook 14118: @uref{http://www.taygeta.com/forth.html} -- Skip Carter's Forth pages.
1.21 crook 14119: @item
1.47 crook 14120: @uref{http://www.jwdt.com/~paysan/gforth.html} -- the Gforth home page.
1.21 crook 14121: @item
1.47 crook 14122: @uref{http://www.minerva.com/uathena.htm} -- home of ANS Forth Standard.
1.21 crook 14123: @item
1.47 crook 14124: @uref{http://dec.bournemouth.ac.uk/forth/index.html} -- the Forth
1.21 crook 14125: Research page, including links to the Journal of Forth Application and
14126: Research (JFAR) and a searchable Forth bibliography.
14127: @end itemize
14128:
14129:
14130: @node Books, The Forth Interest Group, Internet resources, Forth-related information
14131: @section Books
1.26 crook 14132: @cindex books on Forth
1.21 crook 14133:
14134: As the Standard is relatively new, there are not many books out yet. It
14135: is not recommended to learn Forth by using Gforth and a book that is not
14136: written for ANS Forth, as you will not know your mistakes from the
14137: deviations of the book. However, books based on the Forth-83 standard
14138: should be ok, because ANS Forth is primarily an extension of Forth-83.
1.44 crook 14139: Refer to the Forth FAQ for details of Forth-related books.
1.21 crook 14140:
14141: @cindex standard document for ANS Forth
14142: @cindex ANS Forth document
14143: The definite reference if you want to write ANS Forth programs is, of
1.26 crook 14144: course, the ANS Forth document. It is available in printed form from the
1.21 crook 14145: National Standards Institute Sales Department (Tel.: USA (212) 642-4900;
14146: Fax.: USA (212) 302-1286) as document @cite{X3.215-1994} for about
14147: $200. You can also get it from Global Engineering Documents (Tel.: USA
14148: (800) 854-7179; Fax.: (303) 843-9880) for about $300.
14149:
14150: @cite{dpANS6}, the last draft of the standard, which was then submitted
14151: to ANSI for publication is available electronically and for free in some
14152: MS Word format, and it has been converted to HTML
1.47 crook 14153: (@uref{http://www.taygeta.com/forth/dpans.html}; this HTML version also
1.44 crook 14154: includes the answers to Requests for Interpretation (RFIs). Some
14155: pointers to these versions can be found through
1.47 crook 14156: @*@uref{http://www.complang.tuwien.ac.at/projects/forth.html}.
1.44 crook 14157:
1.21 crook 14158:
14159: @node The Forth Interest Group, Conferences, Books, Forth-related information
14160: @section The Forth Interest Group
14161: @cindex Forth interest group (FIG)
14162:
14163: The Forth Interest Group (FIG) is a world-wide, non-profit,
1.26 crook 14164: member-supported organisation. It publishes a regular magazine,
14165: @var{FORTH Dimensions}, and offers other benefits of membership. You can
14166: contact the FIG through their office email address:
14167: @email{office@@forth.org} or by visiting their web site at
1.47 crook 14168: @uref{http://www.forth.org/}. This web site also includes links to FIG
1.26 crook 14169: chapters in other countries and American cities
1.47 crook 14170: (@uref{http://www.forth.org/chapters.html}).
1.21 crook 14171:
1.48 anton 14172: @node Conferences, , The Forth Interest Group, Forth-related information
1.21 crook 14173: @section Conferences
14174: @cindex Conferences
14175:
14176: There are several regular conferences related to Forth. They are all
1.26 crook 14177: well-publicised in @var{FORTH Dimensions} and on the comp.lang.forth
1.45 crook 14178: news group:
1.21 crook 14179:
14180: @itemize @bullet
14181: @item
14182: FORML -- the Forth modification laboratory convenes every year near
14183: Monterey, California.
14184: @item
14185: The Rochester Forth Conference -- an annual conference traditionally
14186: held in Rochester, New York.
14187: @item
14188: EuroForth -- this European conference takes place annually.
14189: @end itemize
14190:
14191:
1.41 anton 14192: @node Word Index, Name Index, Forth-related information, Top
1.1 anton 14193: @unnumbered Word Index
14194:
1.26 crook 14195: This index is a list of Forth words that have ``glossary'' entries
14196: within this manual. Each word is listed with its stack effect and
14197: wordset.
1.1 anton 14198:
14199: @printindex fn
14200:
1.41 anton 14201: @node Name Index, Concept Index, Word Index, Top
14202: @unnumbered Name Index
14203:
14204: This index is a list of Forth words that have ``glossary'' entries
14205: within this manual.
14206:
14207: @printindex ky
14208:
14209: @node Concept Index, , Name Index, Top
1.1 anton 14210: @unnumbered Concept and Word Index
14211:
1.26 crook 14212: Not all entries listed in this index are present verbatim in the
14213: text. This index also duplicates, in abbreviated form, all of the words
14214: listed in the Word Index (only the names are listed for the words here).
1.1 anton 14215:
14216: @printindex cp
14217:
14218: @contents
14219: @bye
14220:
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>