-
Notifications
You must be signed in to change notification settings - Fork 1
/
runner-forth.dtx
5601 lines (5601 loc) · 185 KB
/
runner-forth.dtx
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
% \iffalse
%% File: runner-forth.dtx Copyright (C) 2013 Bruno Le Floch
%%
%% It may be distributed and/or modified under the conditions of the
%% LaTeX Project Public License (LPPL), either version 1.3c of this
%% license or (at your option) any later version. The latest version
%% of this license is in the file
%%
%% http://www.latex-project.org/lppl.txt
%%
%% -----------------------------------------------------------------------
% \fi
%
% \begin{runner-documentation}
%
% \newcommand{\kw}[1]{\texttt{#1}}
%
% \section{The \plang{forth} language}
%
% ^^A todo: make ACCEPT unavailable in non-interactive mode.
%
% In interactive mode, invoked as |\runner{forth}[interactive]''''|,
% this \plang{forth} interpreter is a Standard System Subset, as defined
% by the ANS Forth 1994 specification, providing the majority of the
% Core and Core Extensions word sets, as well as the Double-Number and
% Double-Number Extensions word sets, with the following exceptions.
% \begin{itemize}
% \item The core word |KEY| is not implemented.
% \item The core extension words |EXPECT|, |QUERY|, |RESTORE-INPUT|,
% |SAVE-INPUT|, |SPAN| are not implemented.\footnote{Some of those
% words, and others are marked by the ANS 1994 standard as
% obsolescent, namely \texttt{\#TIB}, \texttt{TIB}, \texttt{EXPECT},
% \texttt{QUERY}, \texttt{SPAN}, \texttt{CONVERT}, \texttt{FORGET}.}
% \item The environment variable |/COUNTED-STRING| is not recognized.
% \end{itemize}
% Additionally, the interpreter provides the Programming-Tools words
% |.S|, |?|, |DUMP|, |[IF]|, |[THEN]|, |[ELSE]|, |BYE|.
%
% ^^A The words |SEE|, |WORDS|, |;CODE|, |AHEAD|, |ASSEMBLER|, |CODE|,
% ^^A |CS-PICK|, |CS-ROLL|, |EDITOR|, |FORGET|, are not implemented.
%
% In non-interactive mode (common case within a \TeX{} document), the
% system reads \plang{forth} code and input from files, or obtains them
% from \TeX{}, hence it has no prompt. This mode is otherwise
% identical to interactive mode.
%
% Single-cell integers are \(24\) bits wide, as are characters.
% Double-cell integers are \(48\) bits wide. More details on the
% implementation-defined behaviours and features can be found in
% section~\ref{seq:forth-details}.
%
% \subsection{Introduction to \plang{forth}}
%
% This is not the place for an introduction to Forth, many can be found
% online. ^^A todo: add a bunch of refs
%
% For example, one can define the factorial function as follows.
% \begin{verbatim}
% \runner{forth}[output = \result]|
% : FACTORIAL ( +n1 -- +n2)
% DUP 2 < IF DROP 1 EXIT THEN ( 0! = 1! = 1)
% DUP 1- RECURSE * ; ( n1! = n1 * [n1 - 1]!)
% ||
% CR 0 FACTORIAL . ( prints 1)
% CR 3 FACTORIAL . ( prints 6)
% CR 24 FACTORIAL . ( prints -4194304)|
% \typeout{\result}
% \end{verbatim}
% This types three lines to the terminal, containing \(1\), \(6\) and
% \(-4194304\). The last number is incorrect, because the result
% overflows the size of a single-cell unsigned integer. We can easily
% check that it is the correct result modulo \(2^{24}\).
%
% Most Forth words receive arguments from the data stack, and place
% their results back onto this stack. The stack contents takes the form
% of \(24\)-bit cells, which can represent one of the following
% data-types, each of which is represented in the documentation below by
% a short-hand (we follow the notations of the ANS 1994 standard)
% \begin{itemize}
% \item[flag] a flag (true: \(-1\), false: \(0\)),
% \item[char] a character (in \([0,1114111]\), the range of Unicode code
% points),
% \item[n] a signed number (in \([-2^{23},2^{23}-1]\)),
% \item[+n] a non-negative number (in \([0,2^{23}-1]\)),
% \item[u] an unsigned number (in \([0,2^{24}-1]\)),
% \item[xt] an execution token (which points to some \TeX{} code
% implementing a Forth word),
% \item[addr] an address in memory (typically a small non-negative integer)
% \item[x] arbitrary cell data.
% \end{itemize}
% Pairs of cells are used by some words to manipulate integers larger
% than \(2^{24}\):
% \begin{itemize}
% \item[d] a double-cell signed number (in \([-2^{47},2^{47}-1]\)),
% \item[+d] a double-cell non-negative number (in \([0,2^{47}-1]\)),
% \item[ud] a double-cell unsigned number (in \([0,2^{48}-1]\)),
% \item[xd] arbitrary pair of cells
% \end{itemize}
% Additionally, some internal data-types are used for the implementation
% of control-flow words, namely colon-sys, do-sys, case-sys, of-sys,
% orig, dest, loop-sys, nest-sys. Additionally, i*x is used to
% represent an arbitrary number of cells.
%
% \subsection{Words}
%
% ^^A todo: check which arguments are signed/unsigned.
% ^^A todo: document all words.
%
% Arithmetic operations take arguments from the stack and push their
% result back onto the stack.
% \begin{itemize}
% \item |ABS|, |NEGATE|, |1+| and |1-| are unary: |(n -- n')|.
% \item |MAX|, |MIN|, |-|, |+|, |*|, |/|, and |MOD| are binary: |(n1 n2 -- n)|
% \item |/MOD| has stack effect |(n1 n2 -- n3 n4)| where |n4| is the
% quotient and |n3| the remainder.
% \item |*/| is ternary: |(n1 n2 n3 -- n)| where |n| is obtained from
% |n1 * n2 / n3|.
% \item |M*|, |UM*| have stack effect |(n1 n2 -- d)| where the
% double-cell integer |d| is |n1 * n2|.
% \item |FM/MOD|, |SM/REM|, |UM/MOD| do |(d n1 -- n2 n3)|, where |n3| is
% the quotient, and |n2| the remainder, of |d/n1|.
% \item |*/MOD|
% \end{itemize}
%
% Bit operations take arguments from the stack and push their result
% back onto the stack.
% \begin{itemize}
% \item |2*| and |2/| shift the top of stack by one bit to the left/right.
% \item |D2/| and |D2*| shift the topmost double-cell by one bit.
% \item |INVERT| changes all bits of the top of stack.
% \item |AND|, |OR| and |XOR| are binary: |(n1 n2 -- n3)|.
% \item |LSHIFT| and |RSHIFT| have stack effect |(x1 u -- x2)|.
% \end{itemize}
%
% Case words: |CASE|, |OF|, |ENDOF|, |ENDCASE|.
%
% Comparisons take arguments from the stack and push flags (false:
% \(0\), true: \(-1\)) onto the stack.
% \begin{itemize}
% \item |FALSE|, |TRUE| always push \(0\), \(-1\).
% \item |0=| and |0<>| have stack effect |(x -- flag)|.
% \item |0<| and |0>| have stack effect |(n -- flag)|.
% \item |=| and |<>| have stack effect |(x1 x2 -- flag)|.
% \item |U<| and |U>| have stack effect |(u1 u2 -- flag)|.
% \item |<| and |>| have stack effect |(n1 n2 -- flag)|.
% \item |WITHIN| has stack effect |(n1 n2 n3 -- flag)|.
% \item |?DUP| duplicates the top-most cell if it is not zero.
% \item |D0=| and |D0<| have stack effect |(d -- flag)|
% \item |D=|, |D<|, |DMAX|, and |DMIN| have stack effect |(d1 d2 -- flag)|.
% \item |DU<|
% \end{itemize}
%
% Control-flow words: |IF|, |ELSE|, |THEN|, |BEGIN|, |AGAIN|, |UNTIL|,
% |WHILE|, |REPEAT|.
%
% Defining words: |CONSTANT|, |2CONSTANT|, |CREATE|, |VARIABLE|,
% |2VARIABLE|, |>BODY|, |VALUE|, |TO|, |:NONAME|, |:|, |DOES>|, |;|,
% |IMMEDIATE|, |LITERAL|, |2LITERAL|, |MARKER|, |POSTPONE|, |[COMPILE]|,
% |RECURSE|, |COMPILE,|, |STATE|, |[|, |]|.
%
% Do loops: |DO|, |?DO|, |LOOP|, |+LOOP|, |I|, |J|, |LEAVE|, |UNLOOP|.
%
% Environmental queries: |PAD|, |ENVIRONMENT?|.
%
% Execution tokens: |'|, |[']|, |EXECUTE|, |FIND|.
%
% Immediate conditionals: |[IF]|, |[ELSE]|, |[THEN]|.
%
% Input: |>IN|, |SOURCE|, |TIB|, |#TIB|, |CHAR|, |[CHAR]|, |(|, |\|,
% |PARSE|, |WORD|, |COUNT|, |S"|, |EVALUATE|, |ACCEPT|, |SOURCE-ID|,
% |REFILL|. ^^A)
%
% Memory cells: |ALIGN|, |ALIGNED|, |CELL+|, |CHAR+|, |CELLS|, |CHARS|,
% |HERE|, |UNUSED|, |,|, |C,|, |@|, |C@|, |2@|, |!|, |C!|, |2!|, |+!|,
% |ALLOT|, |ERASE|, |FILL|, |MOVE|.
%
% Number conversion: |BASE|, |DECIMAL|, |HEX|, |<#|, |SIGN|, |HOLD|, |#|,
% |#S|, |#>|, |>NUMBER|.
%
% Output: |.|, |U.|, |.R|, |U.R|, |D.|, |D.R|, |.S|, |."|, |.(|, |BL|,
% |CR|, |EMIT|, |SPACE|, |SPACES|, |TYPE|, |?|, |DUMP|. ^^A)
%
% Quitting: |EXIT|, |QUIT|, |ABORT|, |ABORT"|, |BYE|.
%
% Stacks: |DEPTH|, |DROP|, |2DROP|, |NIP|, |SWAP|, |2SWAP|, |ROT|,
% |DUP|, |2DUP|, |OVER|, |2OVER|, |TUCK|, |PICK|, |ROLL|, |R>|, |2R>|,
% |R@|, |2R@|, |>R|, |2>R|, |S>D|, |D>S|, |2ROT|.
%
% \subsection{Details of the interpreter}
% \label{seq:forth-details}
%
% ^^A todo: clarify what to do with character code points >= 256.
%
% Implementation specificities are as follows.
% \begin{itemize}
% \item All addresses are aligned and are character-aligned.
% \item Behaviour of |EMIT| for non-graphic character: ?
% \item Character editing of |ACCEPT|: ?
% \item Character set: ?
% \item Character-set-extensions matching characteristics: ?
% \item Conditions under which control characters match a space
% delimiter: whenever a space delimiter (character code \(32\)) is
% called for, all characters in \([0,32]\) will match. Even in the
% case of |WORD|.
% \item Format of the control-flow stack: integers, indicating what was
% the last control-flow construct used, \(0\) for |:| or |:NONAME|,
% \(1\) for |DOES>|, \(2\) for |IF|, \(3\) for |ELSE|, \(6\) for |DO|,
% \(7\) for |?DO|, \(8\) for |BEGIN|, \(9\) for |WHILE|.
% \item Conversion of digits larger than thirty-five: this never arises,
% as the base used in conversion is automatically truncated to the
% interval \([2, 36]\).
% \item Display after input terminates in |ACCEPT|: ?
% \item Exception abort sequence (as in |ABORT"|): message to the \TeX{}
% terminal, with the non-zero value that was on the stack, and the
% compile-time string.
% \item Input line terminator: ? User input device: ?
% \item Maximum size of a counted string, in chars: ?
% \item Maximum size of a parsed string: ?
% \item Maximum size of a definition name, in chars: ?
% \item Maximum string length for |ENVIRONMENT?|, in chars: ?
% \item Method of selecting a user input device: ?
% \item Method of selecting a user output device: ?
% \item Methods of dictionary compilation: ?
% \item Number of bits in one address unit: \(24\).
% \item Number representation and arithmetic: internally represented as
% \(24\)-bit unsigned integers (well, further down as \TeX{}
% dimensions in a \tn{fontdimen} array).
% \item Integers \([-2^{23}, 2^{23} - 1]\), non-negative integers \([0,
% 2^{23} - 1]\), unsigned integers \([0, 2^{24} - 1]\), double
% integers \([-2^{47}, 2^{47} - 1]\), non-negative double integers
% \([0, 2^{47} - 1]\), unsigned double integers \([0, 2^{48} - 1]\).
% \item Read-only data-space regions: ?
% \item Size of buffer at |WORD|: ?
% \item One cell is one address unit.
% \item One character is one address unit.
% \item Size of the keyboard terminal input buffer: fixed upon starting
% the Forth interpreter, at least \(128\).
% \item Size of the pictured numeric output string buffer: fixed upon
% starting the Forth interpreter, at least \(127\).
% \item Size of the scratch area whose address is returned by |PAD|:
% fixed upon starting the Forth interpreter, at least \(128\).
% \item Case-sensitivity: yes. (?)
% \item Prompt (see |QUIT|): ?
% \item Division rounding: symmetric (the quotient is rounded towards
% zero, the remainder has the sign of the numerator).
% \item Value of |STATE| when true: all bits set.
% \item Values returned after arithmetic overflow: correct result,
% modulo \(2^{24}\).
% \item In a |: X ... DOES> ... ;| construction, the definition of |X|
% can be found after the |DOES>|.
% \end{itemize}
%
% Behaviour upon encountering ambiguous conditions.
% \begin{itemize}
% \item A name is neither a valid definition name nor a valid number
% during text interpretation triggers a \TeX{} error.
% \item There is no maximum on the length of a definition name.
% \item Addressing a region outside data space can yield arbitrary data.
% \item No type checking is performed.
% \item The execution token obtained when applying |'| or |FIND| to a
% word with undefined interpretation semantics is a valid token, whose
% code triggers a \TeX{} error.
% \item Dividing by zero in |*/|, |*/MOD|, |/|, |/MOD|, |FM/MOD|, |MOD|,
% |SM/REM|, |UM/MOD|, |M*/| triggers a \TeX{} error, and the
% corresponding division yields \(0\) or the largest allowed integer,
% depending on the operation.
% \item Stack overflow or underflow triggers a \TeX{} error, and the
% stack pointer is not updated. This includes the case of
% insufficient data-stack space or return-stack space, insufficient
% space for loop-control parameters, and overflow of the pictured
% numeric output. This also includes reading from an empty data stack
% or return stack.
% \item Insufficient space for new words crashes \TeX{}.
% \item Interpreting a word with undefined interpretation semantics
% triggers a \TeX{} error, and the word is ignored.
% \item Modifying the contents of the input buffer or a string literal
% is harmless.
% \item Parsed string overflow: ?
% \item Arithmetic that produces out of range results yields the correct
% result modulo \(2^{24}\) (size of the cell).
% \item Unexpected end of input buffer, resulting in an attempt to use a
% zero-length string as a name: ?
% \item |>IN| greater than size of input buffer: ?
% \item If |RECURSE| appears after |DOES>|, it inserts the execution
% token performing the code which follows |DOES>|. The address of the
% data field of the |CREATE|d word that is affected by |DOES>| is only
% pushed once, and not at each recursive call.
% \item Argument input source different than current input source for
% |RESTORE-INPUT|: ?
% \item Data space containing definitions can never be de-allocated, since
% it is inaccessible.
% \item Data space read/write with incorrect alignment cannot happen,
% since all addresses are aligned.
% \item Less than \(u+2\) stack items for |PICK| or |ROLL|.
% \item Loop-control parameters are not available for |+LOOP|, |I|, |J|,
% |LEAVE|, |LOOP|, |UNLOOP|: ?
% \item Most recent definition does not have a name for |IMMEDIATE|: ?
% \item Name not defined by |VALUE| used by |TO|: ?
% \item Name not found |'|, |POSTPONE|, |[']|, |[COMPILE]|: ?
% \item |POSTPONE| or |[COMPILE]| applied to |TO|: ?
% \item String longer than a counted string returned by |WORD|: ?
% \item |LSHIFT| or |RSHIFT| with \(u \geq 24\): ?
% \item Using |>BODY| or |DOES>| for a word not defined via |CREATE|: ?
% \item Using |#|, |#S|, |HOLD| or |SIGN| improperly outside |<#| and
% |#>| correctly places data into the pictured numeric buffer, but
% this is later ignored.
% \end{itemize}
%
% \end{runner-documentation}
%
% \begin{runner-implementation}
%
% \section{\plang{forth} implementation}
%
% ^^A todo: normalize newlines?
% ^^A todo: document that |EVALUATE| can lead to |^^J| in the input buffer.
%
% \subsection{Generic helpers}
%
% Useful variants.
% \begin{macrocode}
\cs_generate_variant:Nn \prop_gput:Nnn { Nx }
\cs_generate_variant:Nn \prop_get:NnNTF { Nx }
\cs_generate_variant:Nn \prop_get:NnNF { No }
\cs_generate_variant:Nn \prop_put:Nnn { Nxx }
\cs_generate_variant:Nn \tl_item:nn { o }
\cs_generate_variant:Nn \tl_if_head_eq_meaning:nNF { V }
\cs_generate_variant:Nn \tl_if_head_eq_meaning:nNTF { V }
\cs_generate_variant:Nn \str_set_convert:Nnnn { Nxnn }
% \end{macrocode}
%
% \subsection{Data space and data stack}
%
% ^^A todo: convert the control-flow stack to be a simple tl.
% ^^A todo: what is the return stack used for.
%
% The data necessary to run Forth consists of
% \begin{itemize}
% \item the data stack, manipulated by the program, contains cells,
% \item the control-flow stack, which contains ``-sys'' types, and can
% be implemented using the data stack,
% \item the return stack, which can be used by the program, but is
% mostly used by the system, for definitions, do-loops, nesting info,
% \end{itemize}
% as well as
% \begin{itemize}
% \item the name space (mapping words to execution tokens),
% \item the code space (code for the execution tokens),
% \item the data space accessible to programs, which consists in
% \begin{itemize}
% \item contiguous regions,
% \item variables,
% \item text-literal regions,
% \item input buffers,
% \item and other transient regions.
% \end{itemize}
% \end{itemize}
%
% The name space is implemented as a property list, mapping words to a
% flag in \(\{1,2,3\}\) and an execution token (integer in
% \([0,65535]\)) associated to this word.
%
% The code space is implemented using \tn{toks} registers: execution
% tokens are register numbers.
%
% Everything else is stored as one global array, \cs{g_@@_forth_array},
% of \(24\)-bit unsigned integers. See \texttt{runner.dtx} for the
% array implementation.
%
% \subsubsection{The array}
%
% The array \cs{g_@@_forth_array} is split into several regions, in the
% following order (from low-numbered to high-numbered items). For
% conciseness we refer in this list to integer variables by shorthands;
% \texttt{min} and \texttt{max} variables are global and can only change
% between Forth runs; other integers are local and change during runs.
% \begin{itemize}
% \item The pictured numeric output is from \texttt{picnum_min}
% (included) to \texttt{picnum_max} (excluded): this contains material
% from \texttt{picnum_here} (included) to \texttt{picnum_max}
% (excluded).
% \item The terminal input buffer contains text from \texttt{term_min}
% (included) to \texttt{term_end} (excluded), which can go up to
% \texttt{term_max}. The actual input source lies from
% \texttt{source_begin} (included) to \texttt{source_end} (excluded),
% which are normally equal to \texttt{term_min} and \texttt{term_end},
% unless modified by |EVALUATE|. The integer \texttt{source_parsed}
% is used temporarily to store the sum of \texttt{source_end} and of
% the value at~|>IN|.
% \item The pad is from \texttt{pad_min} (included) to \texttt{pad_max}
% (excluded).
% \item The data space lies from \texttt{data_min} (included) to
% \texttt{data_max} (excluded), and is split further.
% \begin{itemize}
% \item Contiguous regions of the data space, intermixed with
% variables, lie between \texttt{data_min} (included) and
% \texttt{data_here} (excluded).
% \item The transient regions provided by |WORD| or |#>| start at
% \texttt{data_here} (included) and end at \texttt{data_word}
% (excluded).
% \item Text literals are allocated at the end of the data space, and
% lie between \texttt{data_text} (included) and \texttt{data_max}
% (excluded).
% \end{itemize}
% \item The data stack lies from \texttt{stack_min} to
% \texttt{stack_max}, and the position where a new value would be
% pushed is at \texttt{stack_here}.
% \item The control-flow stack lies from \texttt{control_min} to
% \texttt{control_max}, and the position where a new value would be
% pushed is \texttt{control_here}.
% \item The return stack lies from \texttt{return_min} to
% \texttt{return_max}, and the position where a new value would be
% pushed is at \texttt{return_here}.
% \end{itemize}
%
% In each slot of the array, we store a \(24\) bit unsigned integer,
% \ie, a number in the range \([0, 2^{24}-1]\) (note that we could go up
% to \(26\) bits with no adverse effect). Address units are \(24\) bits
% wide. Characters are one address unit wide (note that all Unicode
% code points are less than \(2^{23}\)). Cells are one character wide.
% We do not use \(8\) bit characters (and addresses) for four reasons:
% only \(256\) cells could be addressed; counted strings would be
% bounded to \(255\) characters; extracting an \(8\)-bit part from a
% \(24\) bit value in \TeX{} is not fast; and we would need to chose an
% encoding for Unicode (this is still needed in \pdfTeX{}, but only at
% the very last step of output).
%
% \begin{variable}
% {
% \g_@@_forth_picnum_min_int,
% \l_@@_forth_picnum_here_int,
% \g_@@_forth_picnum_max_int,
% }
% The pictured numeric output, ranging from
% \cs{l_@@_forth_picnum_here_int} to \cs{g_@@_forth_picnum_max_int}.
% \begin{macrocode}
\int_new:N \g_@@_forth_picnum_min_int
\int_new:N \l_@@_forth_picnum_here_int
\int_new:N \g_@@_forth_picnum_max_int
% \end{macrocode}
% \end{variable}
%
% \begin{variable}
% {
% \g_@@_forth_term_min_int,
% \l_@@_forth_term_end_int,
% \g_@@_forth_term_max_int,
% \l_@@_forth_source_begin_int,
% \l_@@_forth_source_parsed_int,
% \l_@@_forth_source_end_int,
% \l_@@_forth_in_address_int,
% \l_@@_forth_sources_seq
% }
% The terminal input buffer lies from \cs{g_@@_forth_term_min_int}
% (included) to \cs{g_@@_forth_term_max_int} (excluded), and contains
% text until \cs{l_@@_forth_term_end_int} (excluded). The source
% specification can be changed by |EVALUATE|: parsing words read the
% buffer from \cs{l_@@_forth_source_begin_int} to
% \cs{l_@@_forth_source_end_int}. The integer
% \cs{l_@@_forth_source_parsed_int} is set to the sum of
% \cs{l_@@_forth_source_begin_int} and the value at~|>IN| (whose address
% is \cs{l_@@_forth_in_address_int} before each parsing operation.
% The stack \cs{l_@@_forth_sources_seq} stores the previous source
% specification for nested |EVALUATE| calls.
% \begin{macrocode}
\int_new:N \g_@@_forth_term_min_int
\int_new:N \l_@@_forth_term_end_int
\int_new:N \g_@@_forth_term_max_int
\int_new:N \l_@@_forth_source_begin_int
\int_new:N \l_@@_forth_source_parsed_int
\int_new:N \l_@@_forth_source_end_int
\int_new:N \l_@@_forth_in_address_int
\seq_new:N \l_@@_forth_sources_seq
% \end{macrocode}
% \end{variable}
%
% \begin{variable}{\g_@@_forth_pad_min_int, \g_@@_forth_pad_max_int}
% The pad returned by |PAD| is fixed and ranges from
% \cs{g_@@_forth_pad_min_int} (included) to
% \cs{g_@@_forth_pad_max_int} (excluded).
% \begin{macrocode}
\int_new:N \g_@@_forth_pad_min_int
\int_new:N \g_@@_forth_pad_max_int
% \end{macrocode}
% \end{variable}
%
% \begin{variable}
% {
% \g_@@_forth_data_min_int,
% \l_@@_forth_data_here_int,
% \l_@@_forth_data_word_int,
% \l_@@_forth_data_text_int,
% \g_@@_forth_data_max_int,
% }
% Allotted data, and variables, range from
% \cs{g_@@_forth_data_min_int} (included) to
% \cs{l_@@_forth_data_here_int} (excluded). The latter integer is
% what |HERE| returns. Transient regions created by |WORD| or |#>|
% lie between this address (included) and
% \cs{l_@@_forth_data_word_int} (excluded). Finally, text strings are
% stored from \cs{l_@@_forth_data_text_int} (included) to
% \cs{g_@@_forth_data_max_int} (excluded).
% \begin{macrocode}
\int_new:N \g_@@_forth_data_min_int
\int_new:N \l_@@_forth_data_here_int
\int_new:N \l_@@_forth_data_word_int
\int_new:N \l_@@_forth_data_text_int
\int_new:N \g_@@_forth_data_max_int
% \end{macrocode}
% \end{variable}
%
% \begin{variable}
% {
% \g_@@_forth_stack_min_int,
% \l_@@_forth_stack_here_int,
% \g_@@_forth_stack_max_int
% }
% The data stack lies in this range, with the bottom item at
% \cs{g_@@_forth_stack_min_int} and the top item just below
% \cs{l_@@_forth_stack_here_int}.
% \begin{macrocode}
\int_new:N \g_@@_forth_stack_min_int
\int_new:N \l_@@_forth_stack_here_int
\int_new:N \g_@@_forth_stack_max_int
% \end{macrocode}
% \end{variable}
%
% \begin{variable}
% {
% \g_@@_forth_control_min_int,
% \l_@@_forth_control_here_int,
% \g_@@_forth_control_max_int
% }
% The control flow stack lies in this range, with the bottom item at
% \cs{g_@@_forth_control_min_int} and the top item just below
% \cs{l_@@_forth_control_here_int}.
% \begin{macrocode}
\int_new:N \g_@@_forth_control_min_int
\int_new:N \l_@@_forth_control_here_int
\int_new:N \g_@@_forth_control_max_int
% \end{macrocode}
% \end{variable}
%
% \begin{variable}
% {
% \g_@@_forth_return_min_int,
% \l_@@_forth_return_here_int,
% \g_@@_forth_return_max_int
% }
% The return stack lies there, with the bottom item at
% \cs{g_@@_forth_return_min_int} and the top of the stack just below
% \cs{l_@@_forth_return_here_int}.
% \begin{macrocode}
\int_new:N \g_@@_forth_return_min_int
\int_new:N \l_@@_forth_return_here_int
\int_new:N \g_@@_forth_return_max_int
% \end{macrocode}
% \end{variable}
%
% \begin{macro}[int]{\@@_forth_init_data:}
% When starting a Forth run, one must reset the pointers that vary, to
% the appropriate boundary of their region.
% \begin{macrocode}
\cs_new_protected_nopar:Npn \@@_forth_init_data:
{
\int_set_eq:NN \l_@@_forth_picnum_here_int
\g_@@_forth_picnum_max_int
\int_set_eq:NN \l_@@_forth_term_end_int
\g_@@_forth_term_min_int
\int_set_eq:NN \l_@@_forth_source_begin_int
\g_@@_forth_term_min_int
\int_set_eq:NN \l_@@_forth_source_parsed_int
\g_@@_forth_term_min_int
\int_set_eq:NN \l_@@_forth_source_end_int
\g_@@_forth_term_min_int
\seq_clear:N \l_@@_forth_sources_seq
\int_set_eq:NN \l_@@_forth_data_here_int
\g_@@_forth_data_min_int
\int_set_eq:NN \l_@@_forth_data_word_int
\g_@@_forth_data_min_int
\int_set_eq:NN \l_@@_forth_data_text_int
\g_@@_forth_data_max_int
\int_set_eq:NN \l_@@_forth_stack_here_int
\g_@@_forth_stack_min_int
\int_set_eq:NN \l_@@_forth_control_here_int
\g_@@_forth_control_min_int
\int_set_eq:NN \l_@@_forth_return_here_int
\g_@@_forth_return_min_int
}
% \end{macrocode}
% \end{macro}
%
% ^^A todo: make the array size dynamical
% ^^A todo: ensure that picnum buffer is >= 127 cells wide.
% ^^A todo: ensure that input buffer and pad are >= 128 cells wide.
%
% \begin{variable}{\g_@@_forth_array}
% This array contains most of the data needed by Forth. We also give
% here initial values for the array parameters, all at once to avoid
% mistakes.
% \begin{macrocode}
\int_gset:Nn \g_@@_forth_picnum_min_int { 1 }
\int_gset:Nn \g_@@_forth_picnum_max_int { 128 }
\int_gset:Nn \g_@@_forth_term_min_int { 128 }
\int_gset:Nn \g_@@_forth_term_max_int { 256 }
\int_gset:Nn \g_@@_forth_pad_min_int { 256 }
\int_gset:Nn \g_@@_forth_pad_max_int { 384 }
\int_gset:Nn \g_@@_forth_data_min_int { 384 }
\int_gset:Nn \g_@@_forth_data_max_int { 65152 }
\int_gset:Nn \g_@@_forth_stack_min_int { 65152 }
\int_gset:Nn \g_@@_forth_stack_max_int { 65280 }
\int_gset:Nn \g_@@_forth_control_min_int { 65280 }
\int_gset:Nn \g_@@_forth_control_max_int { 65408 }
\int_gset:Nn \g_@@_forth_return_min_int { 65408 }
\int_gset:Nn \g_@@_forth_return_max_int { 65536 }
\@@_array_new:Nn \g_@@_forth_array { 65536 }
% \end{macrocode}
% \end{variable}
%
% \begin{variable}{\l_@@_forth_fp_seq}
% The floating point stack stores floating point numbers as token
% lists (through \cs{fp_to_tl:n}).
% \begin{macrocode}
\seq_new:N \l_@@_forth_fp_seq
% \end{macrocode}
% \end{variable}
%
% \subsubsection{Integers modulo \(2^{24}\)}
%
% \begin{variable}{\c_@@_forth_mod_int}
% Cells can take up to \(2^{24}\) values. Arithmetic is often
% performed modulo that number.
% \begin{macrocode}
\int_const:Nn \c_@@_forth_mod_int { 16777216 }
% \end{macrocode}
% ^^A todo: introduce \c_@@_forth_half_int ?
% \end{variable}
%
% \begin{macro}[int, EXP]{\@@_forth_sign_mask:N}
% Yields \(0\) when |#1| is in \([0,2^{23}-1]\) and \(2^{23}\) when
% |#1| is in \([2^{23}, 2^{24}-1]\). We make use of \eTeX{}'s
% rounding behaviour: dividing |#1| by \(2^{24}\) yields \(0\) for
% numbers less than \(2^{23}\) and \(1\) for numbers in \([2^{23},
% 2^{24}-1]\).
% \begin{macrocode}
\cs_new:Npn \@@_forth_sign_mask:N #1
{
\int_eval:n
{ #1 / \c_@@_forth_mod_int * \c_@@_forth_mod_int / \c_two }
}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}[int, EXP]{\@@_forth_signed:N}
% \begin{macro}[int]{\@@_forth_set_signed:N}
% Turn the unsigned |#1| into a signed number, by subtracting
% \(2^{24}\) if it is not less than \(2^{23}\). See
% \cs{@@_forth_sign_mask:N} for details.
% \begin{macrocode}
\cs_new:Npn \@@_forth_signed:N #1
{
\int_eval:n
{ #1 - #1 / \c_@@_forth_mod_int * \c_@@_forth_mod_int }
}
\cs_new_protected:Npn \@@_forth_set_signed:N #1
{
\int_set:Nn #1
{ #1 - #1 / \c_@@_forth_mod_int * \c_@@_forth_mod_int }
}
% \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}[aux]{\@@_forth_double_negate:NN}
% Change a double-cell integer \((a,b)\) to its opposite. The
% integers |#1| and |#2| are modified in place. In most cases, the
% result is \((2^{24}-a, 2^{24}-b-1)\) (we work modulo \(2^{48}\)),
% but this fails for \(a=0\), where the correct answer is most often
% \((0, 2^{24}-b)\), with the exception of \(b=0\), where it is simply
% \((0,0)\).
% \begin{macrocode}
\cs_new_protected:Npn \@@_forth_double_negate:NN #1#2
{
\int_compare:nNnTF #1 = \c_zero
{
\int_compare:nNnF #2 = \c_zero
{ \int_set:Nn #2 { \c_@@_forth_mod_int - #2 } }
}
{
\int_set:Nn #1 { \c_@@_forth_mod_int - #1 }
\int_set:Nn #2 { \c_@@_forth_mod_int - #2 - \c_one }
}
}
% \end{macrocode}
% \end{macro}
%
% \subsubsection{Bound checking}
%
% \begin{macro}[int]
% {\@@_forth_check_data_bottom:, \@@_forth_check_data_top:}
% After the |HERE| pointer is updated, it may lie outside the allowed
% range from \texttt{data_min} to \texttt{data_text}.
% \begin{macrocode}
\cs_new_protected_nopar:Npn \@@_forth_check_data_bottom:
{
\int_compare:nNnT
\l_@@_forth_data_here_int < \g_@@_forth_data_min_int
{ \@@_forth_abort:nx { out-of-bounds } { } }
}
\cs_new_protected_nopar:Npn \@@_forth_check_data_top:
{
\int_compare:nNnT
\l_@@_forth_data_here_int > \l_@@_forth_data_text_int
{ \@@_forth_abort:nx { out-of-memory } { } }
}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}[int]
% {\@@_forth_check_stack_bottom:n, \@@_forth_check_stack_top:n}
% Check that we can pop/push |#1| items from/onto the stack.
% \begin{macrocode}
\cs_new_protected:Npn \@@_forth_check_stack_bottom:n #1
{
\int_compare:nNnT
\l_@@_forth_stack_here_int
< { \g_@@_forth_stack_min_int + #1 }
{ \@@_forth_abort:nx { empty-stack } { } }
}
\cs_new_protected:Npn \@@_forth_check_stack_top:n #1
{
\int_compare:nNnT
{ \l_@@_forth_stack_here_int + #1 }
> \g_@@_forth_stack_max_int
{ \@@_forth_abort:nx { out-of-memory } { } }
}
% \end{macrocode}
% \end{macro}
%
% \subsubsection{Data space}
%
% \begin{macro}[int, EXP]{\@@_forth_data_get:N, \@@_forth_data_get:n}
% Retrieving unsigned integers from data space.
% \begin{macrocode}
\cs_new_nopar:Npn \@@_forth_data_get:N
{ \@@_array_item:NN \g_@@_forth_array }
\cs_new_nopar:Npn \@@_forth_data_get:n
{ \@@_array_item:Nn \g_@@_forth_array }
% \end{macrocode}
% \end{macro}
%
% \begin{macro}[int]{\@@_forth_data_gset:nn}
% Storing unsigned integers into data space.
% \begin{macrocode}
\cs_new_protected_nopar:Npn \@@_forth_data_gset:nn
{ \@@_array_gset:Nnn \g_@@_forth_array }
% \end{macrocode}
% \end{macro}
%
% \begin{macro}[int]{\@@_forth_put_here:n}
% Put a value at the position given by the \texttt{data_here} integer,
% then increment that integer. Check afterwards that we do not reach
% the region occupied by text literals.
% \begin{macrocode}
\cs_new_protected:Npn \@@_forth_put_here:n #1
{
\@@_forth_data_gset:nn { \l_@@_forth_data_here_int } {#1}
\int_incr:N \l_@@_forth_data_here_int
\@@_forth_check_data_top:
}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}[int, rEXP]{\@@_forth_interval_to_clist:nn}
% \begin{macro}[aux, rEXP]{\@@_forth_interval_to_clist_aux:n}
% Build a comma-list from a region of memory.
% \begin{macrocode}
\cs_new:Npn \@@_forth_interval_to_clist:nn #1#2
{
\int_step_function:nnnN {#1} { 1 } { #2 - 1 }
\@@_forth_interval_to_clist_aux:n
}
\cs_new:Npn \@@_forth_interval_to_clist_aux:n #1
{ \@@_array_item:Nn \g_@@_forth_array {#1} , }
% \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsubsection{Data stack}
%
% \begin{macro}[int]{\@@_forth_pop_int_unsafe:N}
% Pop an unsigned integer from the stack, with no bound-checking.
% \begin{macrocode}
\cs_new_protected:Npn \@@_forth_pop_int_unsafe:N #1
{
\int_decr:N \l_@@_forth_stack_here_int
\int_set:Nn #1
{
\@@_array_item:NN \g_@@_forth_array
\l_@@_forth_stack_here_int
}
}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}[int]{\@@_forth_push_unsafe:n}
% Push an unsigned integer onto the stack, with no bound-checking.
% \begin{macrocode}
\cs_new_protected:Npn \@@_forth_push_unsafe:n #1
{
\@@_array_gset:Nnn \g_@@_forth_array
\l_@@_forth_stack_here_int {#1}
\int_incr:N \l_@@_forth_stack_here_int
}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}[int]
% {
% \@@_forth_pop_int:N,
% \@@_forth_pop_int:NN,
% \@@_forth_pop_int:NNN,
% \@@_forth_pop_int:NNNN
% }
% Get multiple integers from stack, after checking that the stack has
% enough items. Note that things are popped backwards, so that the
% last argument gets the top of the stack.
% \begin{macrocode}
\cs_new_protected:Npn \@@_forth_pop_int:N #1
{
\@@_forth_check_stack_bottom:n \c_one
\@@_forth_pop_int_unsafe:N #1
}
\cs_new_protected:Npn \@@_forth_pop_int:NN #1#2
{
\@@_forth_check_stack_bottom:n \c_two
\@@_forth_pop_int_unsafe:N #2
\@@_forth_pop_int_unsafe:N #1
}
\cs_new_protected:Npn \@@_forth_pop_int:NNN #1#2#3
{
\@@_forth_check_stack_bottom:n \c_three
\@@_forth_pop_int_unsafe:N #3
\@@_forth_pop_int_unsafe:N #2
\@@_forth_pop_int_unsafe:N #1
}
\cs_new_protected:Npn \@@_forth_pop_int:NNNN #1#2#3#4
{
\@@_forth_check_stack_bottom:n \c_four
\@@_forth_pop_int_unsafe:N #4
\@@_forth_pop_int_unsafe:N #3
\@@_forth_pop_int_unsafe:N #2
\@@_forth_pop_int_unsafe:N #1
}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}[int]
% {
% \@@_forth_push:n,
% \@@_forth_push:nn,
% \@@_forth_push:nnn,
% \@@_forth_push:nnnn,
% \@@_forth_push:nnnnn,
% \@@_forth_push:nnnnnn,
% }
% Check that there is enough space, then push one or more unsigned
% integers onto the stack.
% \begin{macrocode}
\cs_new_protected:Npn \@@_forth_push:n #1
{
\@@_forth_check_stack_top:n \c_one
\@@_forth_push_unsafe:n {#1}
}
\cs_new_protected:Npn \@@_forth_push:nn #1#2
{
\@@_forth_check_stack_top:n \c_two
\@@_forth_push_unsafe:n {#1}
\@@_forth_push_unsafe:n {#2}
}
\cs_new_protected:Npn \@@_forth_push:nnn #1#2#3
{
\@@_forth_check_stack_top:n \c_three
\@@_forth_push_unsafe:n {#1}
\@@_forth_push_unsafe:n {#2}
\@@_forth_push_unsafe:n {#3}
}
\cs_new_protected:Npn \@@_forth_push:nnnn #1#2#3#4
{
\@@_forth_check_stack_top:n \c_four
\@@_forth_push_unsafe:n {#1}
\@@_forth_push_unsafe:n {#2}
\@@_forth_push_unsafe:n {#3}
\@@_forth_push_unsafe:n {#4}
}
\cs_new_protected:Npn \@@_forth_push:nnnnn #1#2#3#4#5
{
\@@_forth_check_stack_top:n \c_five
\@@_forth_push_unsafe:n {#1}
\@@_forth_push_unsafe:n {#2}
\@@_forth_push_unsafe:n {#3}
\@@_forth_push_unsafe:n {#4}
\@@_forth_push_unsafe:n {#5}
}
\cs_new_protected:Npn \@@_forth_push:nnnnnn #1#2#3#4#5#6
{
\@@_forth_check_stack_top:n \c_six
\@@_forth_push_unsafe:n {#1}
\@@_forth_push_unsafe:n {#2}
\@@_forth_push_unsafe:n {#3}
\@@_forth_push_unsafe:n {#4}
\@@_forth_push_unsafe:n {#5}
\@@_forth_push_unsafe:n {#6}
}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}[int]{\@@_forth_push_value:n}
% Push the value at a given address onto the stack.
% \begin{macrocode}
\cs_new_protected:Npn \@@_forth_push_value:n #1
{ \@@_forth_push:n { \@@_forth_data_get:n {#1} } }
% \end{macrocode}
% \end{macro}
%
% \begin{macro}[int]{\@@_forth_push_false:, \@@_forth_push_true:}
% The \texttt{false} value is \(0\), the true value is \(-1\), which
% we store as a positive integer by shifting by
% \cs{c_@@_forth_mod_int}.
% \begin{macrocode}
\cs_new_protected_nopar:Npn \@@_forth_push_false:
{ \@@_forth_push:n { 0 } }
\cs_new_protected_nopar:Npx \@@_forth_push_true:
{
\@@_forth_push:n
{ \int_eval:n { \c_@@_forth_mod_int - 1 } }
}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}[int]{\@@_forth_push_signed:n}
% \begin{macro}[aux]{\@@_forth_push_signed_aux:n}
% Given a signed integer in the range \([-2^{24}, 2^{24}-1]\) (note
% the extra large range), push the corresponding unsigned
% representation onto the stack. This simply requires adding
% \(2^{24}\) to negative numbers.
% \begin{macrocode}
\cs_new_protected:Npn \@@_forth_push_signed:n #1
{
\exp_args:Nf \@@_forth_push_signed_aux:n
{ \int_eval:n {#1} }
}
\cs_new_protected:Npn \@@_forth_push_signed_aux:n #1
{
\@@_forth_push:n
{
\int_compare:nNnTF {#1} < \c_zero
{ #1 + \c_@@_forth_mod_int } {#1}
}
}
% \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}[int]{\@@_forth_push_mod:n}
% Given a non-negative integer, push onto the stack its residue modulo
% \(2^{24}\).
% \begin{macrocode}
\cs_new_protected:Npn \@@_forth_push_mod:n #1
{
\@@_forth_push:n
{ \int_mod:nn {#1} { \c_@@_forth_mod_int } }
}
% \end{macrocode}
% \end{macro}
%
% \begin{variable}{\l_@@_forth_push_fp}
% This floating point variable is used when pushing a floating point
% expression converted to an integer onto the stack.
% \begin{macrocode}
\fp_new:N \l_@@_forth_push_fp
% \end{macrocode}
% \end{variable}
%
% \begin{macro}[int]{\@@_forth_push_fp:n, \@@_forth_push_fp_mod:n}
% Convert a floating point expression to an integer, or an integer
% after reducing it modulo \(2^{24}\), for use in some computations
% involving intermediate double-cell results.
% \begin{macrocode}
\cs_new_protected:Npn \@@_forth_push_fp:n #1
{ \@@_forth_push:n { \fp_to_int:n {#1} } }
\cs_new_protected:Npn \@@_forth_push_fp_mod:n #1
{
\fp_set:Nn \l_@@_forth_push_fp { round0(#1) }
\@@_forth_push_fp:n
{
\l_@@_forth_push_fp - \c_@@_forth_mod_int
* round- ( \l_@@_forth_push_fp / \c_@@_forth_mod_int )
}
}