Fix -Wimplicit warnings.
[bpt/emacs.git] / src / category.c
1 /* GNU Emacs routines to deal with category tables.
2 Copyright (C) 1995, 1997 Electrotechnical Laboratory, JAPAN.
3 Licensed to the Free Software Foundation.
4
5 This file is part of GNU Emacs.
6
7 GNU Emacs is free software; you can redistribute it and/or modify
8 it under the terms of the GNU General Public License as published by
9 the Free Software Foundation; either version 2, or (at your option)
10 any later version.
11
12 GNU Emacs is distributed in the hope that it will be useful,
13 but WITHOUT ANY WARRANTY; without even the implied warranty of
14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 GNU General Public License for more details.
16
17 You should have received a copy of the GNU General Public License
18 along with GNU Emacs; see the file COPYING. If not, write to
19 the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
20 Boston, MA 02111-1307, USA. */
21
22
23 /* Here we handle three objects: category, category set, and category
24 table. Read comments in the file category.h to understand them. */
25
26 #include <config.h>
27 #include <ctype.h>
28 #include "lisp.h"
29 #include "buffer.h"
30 #include "charset.h"
31 #include "category.h"
32
33 /* The version number of the latest category table. Each category
34 table has a unique version number. It is assigned a new number
35 also when it is modified. When a regular expression is compiled
36 into the struct re_pattern_buffer, the version number of the
37 category table (of the current buffer) at that moment is also
38 embedded in the structure.
39
40 For the moment, we are not using this feature. */
41 static int category_table_version;
42
43 Lisp_Object Qcategory_table, Qcategoryp, Qcategorysetp, Qcategory_table_p;
44
45 /* Variables to determine word boundary. */
46 Lisp_Object Vword_combining_categories, Vword_separating_categories;
47
48 /* Temporary internal variable used in macro CHAR_HAS_CATEGORY. */
49 Lisp_Object _temp_category_set;
50
51 \f
52 /* Category set staff. */
53
54 DEFUN ("make-category-set", Fmake_category_set, Smake_category_set, 1, 1, 0,
55 "Return a newly created category-set which contains CATEGORIES.\n\
56 CATEGORIES is a string of category mnemonics.\n\
57 The value is a bool-vector which has t at the indices corresponding to\n\
58 those categories.")
59 (categories)
60 Lisp_Object categories;
61 {
62 Lisp_Object val;
63 int len;
64
65 CHECK_STRING (categories, 0);
66 val = MAKE_CATEGORY_SET;
67
68 if (STRING_MULTIBYTE (categories))
69 error ("Multibyte string in make-category-set");
70
71 len = XSTRING (categories)->size;
72 while (--len >= 0)
73 {
74 Lisp_Object category;
75
76 XSETFASTINT (category, XSTRING (categories)->data[len]);
77 CHECK_CATEGORY (category, 0);
78 SET_CATEGORY_SET (val, category, Qt);
79 }
80 return val;
81 }
82
83 \f
84 /* Category staff. */
85
86 Lisp_Object check_category_table ();
87
88 DEFUN ("define-category", Fdefine_category, Sdefine_category, 2, 3, 0,
89 "Define CHAR as a category which is described by DOCSTRING.\n\
90 CHAR should be an ASCII printing character in the range ` ' to `~'.\n\
91 DOCSTRING is a documentation string of the category.\n\
92 The category is defined only in category table TABLE, which defaults to\n\
93 the current buffer's category table.")
94 (category, docstring, table)
95 Lisp_Object category, docstring, table;
96 {
97 CHECK_CATEGORY (category, 0);
98 CHECK_STRING (docstring, 1);
99 table = check_category_table (table);
100
101 if (!NILP (CATEGORY_DOCSTRING (table, XFASTINT (category))))
102 error ("Category `%c' is already defined", XFASTINT (category));
103 CATEGORY_DOCSTRING (table, XFASTINT (category)) = docstring;
104
105 return Qnil;
106 }
107
108 DEFUN ("category-docstring", Fcategory_docstring, Scategory_docstring, 1, 2, 0,
109 "Return the documentation string of CATEGORY, as defined in CATEGORY-TABLE.")
110 (category, table)
111 Lisp_Object category, table;
112 {
113 Lisp_Object doc;
114
115 CHECK_CATEGORY (category, 0);
116 table = check_category_table (table);
117
118 return CATEGORY_DOCSTRING (table, XFASTINT (category));
119 }
120
121 DEFUN ("get-unused-category", Fget_unused_category, Sget_unused_category,
122 0, 1, 0,
123 "Return a category which is not yet defined in CATEGORY-TABLE.\n\
124 If no category remains available, return nil.\n\
125 The optional argument CATEGORY-TABLE specifies which category table\n\
126 to modify; it defaults to the current buffer's category table.")
127 (table)
128 Lisp_Object table;
129 {
130 int i;
131 Lisp_Object docstring_vector;
132
133 table = check_category_table (table);
134
135 for (i = ' '; i <= '~'; i++)
136 if (NILP (CATEGORY_DOCSTRING (table, i)))
137 return make_number (i);
138
139 return Qnil;
140 }
141
142 \f
143 /* Category-table staff. */
144
145 DEFUN ("category-table-p", Fcategory_table_p, Scategory_table_p, 1, 1, 0,
146 "Return t if ARG is a category table.")
147 (arg)
148 Lisp_Object arg;
149 {
150 if (CHAR_TABLE_P (arg)
151 && EQ (XCHAR_TABLE (arg)->purpose, Qcategory_table))
152 return Qt;
153 return Qnil;
154 }
155
156 /* If TABLE is nil, return the current category table. If TABLE is
157 not nil, check the validity of TABLE as a category table. If
158 valid, return TABLE itself, but if not valid, signal an error of
159 wrong-type-argument. */
160
161 Lisp_Object
162 check_category_table (table)
163 Lisp_Object table;
164 {
165 register Lisp_Object tem;
166 if (NILP (table))
167 return current_buffer->category_table;
168 while (tem = Fcategory_table_p (table), NILP (tem))
169 table = wrong_type_argument (Qcategory_table_p, table);
170 return table;
171 }
172
173 DEFUN ("category-table", Fcategory_table, Scategory_table, 0, 0, 0,
174 "Return the current category table.\n\
175 This is the one specified by the current buffer.")
176 ()
177 {
178 return current_buffer->category_table;
179 }
180
181 DEFUN ("standard-category-table", Fstandard_category_table,
182 Sstandard_category_table, 0, 0, 0,
183 "Return the standard category table.\n\
184 This is the one used for new buffers.")
185 ()
186 {
187 return Vstandard_category_table;
188 }
189
190 /* Return a copy of category table TABLE. We can't simply use the
191 function copy-sequence because no contents should be shared between
192 the original and the copy. This function is called recursively by
193 binding TABLE to a sub char table. */
194
195 Lisp_Object
196 copy_category_table (table)
197 Lisp_Object table;
198 {
199 Lisp_Object tmp;
200 int i, to;
201
202 if (!NILP (XCHAR_TABLE (table)->top))
203 {
204 /* TABLE is a top level char table.
205 At first, make a copy of tree structure of the table. */
206 table = Fcopy_sequence (table);
207
208 /* Then, copy elements for single byte characters one by one. */
209 for (i = 0; i < CHAR_TABLE_SINGLE_BYTE_SLOTS; i++)
210 if (!NILP (tmp = XCHAR_TABLE (table)->contents[i]))
211 XCHAR_TABLE (table)->contents[i] = Fcopy_sequence (tmp);
212 to = CHAR_TABLE_ORDINARY_SLOTS;
213
214 /* Also copy the first (and sole) extra slot. It is a vector
215 containing docstring of each category. */
216 Fset_char_table_extra_slot
217 (table, make_number (0),
218 Fcopy_sequence (Fchar_table_extra_slot (table, make_number (0))));
219 }
220 else
221 {
222 i = 32;
223 to = SUB_CHAR_TABLE_ORDINARY_SLOTS;
224 }
225
226 /* If the table has non-nil default value, copy it. */
227 if (!NILP (tmp = XCHAR_TABLE (table)->defalt))
228 XCHAR_TABLE (table)->defalt = Fcopy_sequence (tmp);
229
230 /* At last, copy the remaining elements while paying attention to a
231 sub char table. */
232 for (; i < to; i++)
233 if (!NILP (tmp = XCHAR_TABLE (table)->contents[i]))
234 XCHAR_TABLE (table)->contents[i]
235 = (SUB_CHAR_TABLE_P (tmp)
236 ? copy_category_table (tmp) : Fcopy_sequence (tmp));
237
238 return table;
239 }
240
241 DEFUN ("copy-category-table", Fcopy_category_table, Scopy_category_table,
242 0, 1, 0,
243 "Construct a new category table and return it.\n\
244 It is a copy of the TABLE, which defaults to the standard category table.")
245 (table)
246 Lisp_Object table;
247 {
248 if (!NILP (table))
249 check_category_table (table);
250 else
251 table = Vstandard_category_table;
252
253 return copy_category_table (table);
254 }
255
256 DEFUN ("set-category-table", Fset_category_table, Sset_category_table, 1, 1, 0,
257 "Specify TABLE as the category table for the current buffer.")
258 (table)
259 Lisp_Object table;
260 {
261 table = check_category_table (table);
262 current_buffer->category_table = table;
263 /* Indicate that this buffer now has a specified category table. */
264 current_buffer->local_var_flags
265 |= XFASTINT (buffer_local_flags.category_table);
266 return table;
267 }
268
269 \f
270 DEFUN ("char-category-set", Fchar_category_set, Schar_category_set, 1, 1, 0,
271 "Return the category set of CHAR.")
272 (ch)
273 Lisp_Object ch;
274 {
275 Lisp_Object val;
276 int charset;
277 unsigned char c1, c2;
278
279 CHECK_NUMBER (ch, 0);
280 return CATEGORY_SET (XFASTINT (ch));
281 }
282
283 DEFUN ("category-set-mnemonics", Fcategory_set_mnemonics,
284 Scategory_set_mnemonics, 1, 1, 0,
285 "Return a string containing mnemonics of the categories in CATEGORY-SET.\n\
286 CATEGORY-SET is a bool-vector, and the categories \"in\" it are those\n\
287 that are indexes where t occurs the bool-vector.\n\
288 The return value is a string containing those same categories.")
289 (category_set)
290 Lisp_Object category_set;
291 {
292 int i, j;
293 char str[96];
294
295 CHECK_CATEGORY_SET (category_set, 0);
296
297 j = 0;
298 for (i = 32; i < 127; i++)
299 if (CATEGORY_MEMBER (i, category_set))
300 str[j++] = i;
301 str[j] = '\0';
302
303 return build_string (str);
304 }
305
306 /* Modify all category sets stored under sub char-table TABLE so that
307 they contain (SET_VALUE is t) or don't contain (SET_VALUE is nil)
308 CATEGORY. */
309
310 void
311 modify_lower_category_set (table, category, set_value)
312 Lisp_Object table, category, set_value;
313 {
314 Lisp_Object val;
315 int i;
316
317 if (NILP (XCHAR_TABLE (table)->defalt))
318 {
319 val = MAKE_CATEGORY_SET;
320 SET_CATEGORY_SET (val, category, set_value);
321 XCHAR_TABLE (table)->defalt = val;
322 }
323
324 for (i = 32; i < SUB_CHAR_TABLE_ORDINARY_SLOTS; i++)
325 {
326 val = XCHAR_TABLE (table)->contents[i];
327
328 if (CATEGORY_SET_P (val))
329 SET_CATEGORY_SET (val, category, set_value);
330 else if (SUB_CHAR_TABLE_P (val))
331 modify_lower_category_set (val, category, set_value);
332 }
333 }
334
335 void
336 set_category_set (category_set, category, val)
337 Lisp_Object category_set, category, val;
338 {
339 do {
340 int idx = XINT (category) / 8;
341 unsigned char bits = 1 << (XINT (category) % 8);
342
343 if (NILP (val))
344 XCATEGORY_SET (category_set)->data[idx] &= ~bits;
345 else
346 XCATEGORY_SET (category_set)->data[idx] |= bits;
347 } while (0);
348 }
349
350 DEFUN ("modify-category-entry", Fmodify_category_entry,
351 Smodify_category_entry, 2, 4, 0,
352 "Modify the category set of CHARACTER by adding CATEGORY to it.\n\
353 The category is changed only for table TABLE, which defaults to\n\
354 the current buffer's category table.\n\
355 If optional fourth argument RESET is non-nil,\n\
356 then delete CATEGORY from the category set instead of adding it.")
357 (character, category, table, reset)
358 Lisp_Object character, category, table, reset;
359 {
360 int c, charset, c1, c2;
361 Lisp_Object set_value; /* Actual value to be set in category sets. */
362 Lisp_Object val, category_set;
363
364 CHECK_NUMBER (character, 0);
365 c = XINT (character);
366 CHECK_CATEGORY (category, 1);
367 table = check_category_table (table);
368
369 if (NILP (CATEGORY_DOCSTRING (table, XFASTINT (category))))
370 error ("Undefined category: %c", XFASTINT (category));
371
372 set_value = NILP (reset) ? Qt : Qnil;
373
374 if (c < CHAR_TABLE_SINGLE_BYTE_SLOTS)
375 {
376 val = XCHAR_TABLE (table)->contents[c];
377 if (!CATEGORY_SET_P (val))
378 XCHAR_TABLE (table)->contents[c] = (val = MAKE_CATEGORY_SET);
379 SET_CATEGORY_SET (val, category, set_value);
380 return Qnil;
381 }
382
383 SPLIT_NON_ASCII_CHAR (c, charset, c1, c2);
384
385 /* The top level table. */
386 val = XCHAR_TABLE (table)->contents[charset + 128];
387 if (CATEGORY_SET_P (val))
388 category_set = val;
389 else if (!SUB_CHAR_TABLE_P (val))
390 {
391 category_set = val = MAKE_CATEGORY_SET;
392 XCHAR_TABLE (table)->contents[charset + 128] = category_set;
393 }
394
395 if (c1 <= 0)
396 {
397 /* Only a charset is specified. */
398 if (SUB_CHAR_TABLE_P (val))
399 /* All characters in CHARSET should be the same as for having
400 CATEGORY or not. */
401 modify_lower_category_set (val, category, set_value);
402 else
403 SET_CATEGORY_SET (category_set, category, set_value);
404 return Qnil;
405 }
406
407 /* The second level table. */
408 if (!SUB_CHAR_TABLE_P (val))
409 {
410 val = make_sub_char_table (Qnil);
411 XCHAR_TABLE (table)->contents[charset + 128] = val;
412 /* We must set default category set of CHARSET in `defalt' slot. */
413 XCHAR_TABLE (val)->defalt = category_set;
414 }
415 table = val;
416
417 val = XCHAR_TABLE (table)->contents[c1];
418 if (CATEGORY_SET_P (val))
419 category_set = val;
420 else if (!SUB_CHAR_TABLE_P (val))
421 {
422 category_set = val = Fcopy_sequence (XCHAR_TABLE (table)->defalt);
423 XCHAR_TABLE (table)->contents[c1] = category_set;
424 }
425
426 if (c2 <= 0)
427 {
428 if (SUB_CHAR_TABLE_P (val))
429 /* All characters in C1 group of CHARSET should be the same as
430 for CATEGORY. */
431 modify_lower_category_set (val, category, set_value);
432 else
433 SET_CATEGORY_SET (category_set, category, set_value);
434 return Qnil;
435 }
436
437 /* The third (bottom) level table. */
438 if (!SUB_CHAR_TABLE_P (val))
439 {
440 val = make_sub_char_table (Qnil);
441 XCHAR_TABLE (table)->contents[c1] = val;
442 /* We must set default category set of CHARSET and C1 in
443 `defalt' slot. */
444 XCHAR_TABLE (val)->defalt = category_set;
445 }
446 table = val;
447
448 val = XCHAR_TABLE (table)->contents[c2];
449 if (CATEGORY_SET_P (val))
450 category_set = val;
451 else if (!SUB_CHAR_TABLE_P (val))
452 {
453 category_set = Fcopy_sequence (XCHAR_TABLE (table)->defalt);
454 XCHAR_TABLE (table)->contents[c2] = category_set;
455 }
456 else
457 /* This should never happen. */
458 error ("Invalid category table");
459
460 SET_CATEGORY_SET (category_set, category, set_value);
461
462 return Qnil;
463 }
464 \f
465 /* Dump category table to buffer in human-readable format */
466
467 static void
468 describe_category (value)
469 Lisp_Object value;
470 {
471 Lisp_Object mnemonics;
472
473 Findent_to (make_number (16), make_number (1));
474
475 if (NILP (value))
476 {
477 insert_string ("default\n");
478 return;
479 }
480
481 if (CHAR_TABLE_P (value))
482 {
483 insert_string ("deeper char-table ...\n");
484 return;
485 }
486
487 if (!CATEGORY_SET_P (value))
488 {
489 insert_string ("invalid\n");
490 return;
491 }
492
493 mnemonics = Fcategory_set_mnemonics (value);
494 insert_from_string (mnemonics, 0, 0, XSTRING (mnemonics)->size,
495 STRING_BYTES (XSTRING (mnemonics)), 0);
496 insert_string ("\n");
497 return;
498 }
499
500 static Lisp_Object
501 describe_category_1 (vector)
502 Lisp_Object vector;
503 {
504 struct buffer *old = current_buffer;
505 set_buffer_internal (XBUFFER (Vstandard_output));
506 describe_vector (vector, Qnil, describe_category, 0, Qnil, Qnil,
507 (int *)0, 0);
508 {
509 int i;
510 Lisp_Object docs = XCHAR_TABLE (vector)->extras[0];
511 Lisp_Object elt;
512
513 if (!VECTORP (docs) || XVECTOR (docs)->size != 95)
514 {
515 insert_string ("Invalid first extra slot in this char table\n");
516 return Qnil;
517 }
518
519 insert_string ("Meanings of mnemonice characters are:\n");
520 for (i = 0; i < 95; i++)
521 {
522 elt = XVECTOR (docs)->contents[i];
523 if (NILP (elt))
524 continue;
525
526 insert_char (i + 32);
527 insert (": ", 2);
528 insert_from_string (elt, 0, 0, XSTRING (elt)->size,
529 STRING_BYTES (XSTRING (elt)), 0);
530 insert ("\n", 1);
531 }
532 }
533
534 while (! NILP (XCHAR_TABLE (vector)->parent))
535 {
536 vector = XCHAR_TABLE (vector)->parent;
537 insert_string ("\nThe parent category table is:");
538 describe_vector (vector, Qnil, describe_category, 0, Qnil, Qnil,
539 (int *) 0, 0);
540 }
541
542 call0 (intern ("help-mode"));
543 set_buffer_internal (old);
544 return Qnil;
545 }
546
547 DEFUN ("describe-categories", Fdescribe_categories, Sdescribe_categories, 0, 0, "",
548 "Describe the category specifications in the current category table.\n\
549 The descriptions are inserted in a buffer, which is then displayed.")
550 ()
551 {
552 internal_with_output_to_temp_buffer
553 ("*Help*", describe_category_1, current_buffer->category_table);
554
555 return Qnil;
556 }
557 \f
558 /* Return 1 if there is a word boundary between two word-constituent
559 characters C1 and C2 if they appear in this order, else return 0.
560 Use the macro WORD_BOUNDARY_P instead of calling this function
561 directly. */
562
563 int
564 word_boundary_p (c1, c2)
565 int c1, c2;
566 {
567 Lisp_Object category_set1, category_set2;
568 Lisp_Object tail;
569 int default_result;
570
571 if (CHAR_CHARSET (c1) == CHAR_CHARSET (c2))
572 {
573 tail = Vword_separating_categories;
574 default_result = 0;
575 }
576 else
577 {
578 tail = Vword_combining_categories;
579 default_result = 1;
580 }
581
582 category_set1 = CATEGORY_SET (c1);
583 if (NILP (category_set1))
584 return default_result;
585 category_set2 = CATEGORY_SET (c2);
586 if (NILP (category_set2))
587 return default_result;
588
589 for (; CONSP (tail); tail = XCONS (tail)->cdr)
590 {
591 Lisp_Object elt = XCONS(tail)->car;
592
593 if (CONSP (elt)
594 && CATEGORYP (XCONS (elt)->car)
595 && CATEGORYP (XCONS (elt)->cdr)
596 && CATEGORY_MEMBER (XFASTINT (XCONS (elt)->car), category_set1)
597 && CATEGORY_MEMBER (XFASTINT (XCONS (elt)->cdr), category_set2))
598 return !default_result;
599 }
600 return default_result;
601 }
602
603 \f
604 void
605 init_category_once ()
606 {
607 /* This has to be done here, before we call Fmake_char_table. */
608 Qcategory_table = intern ("category-table");
609 staticpro (&Qcategory_table);
610
611 /* Intern this now in case it isn't already done.
612 Setting this variable twice is harmless.
613 But don't staticpro it here--that is done in alloc.c. */
614 Qchar_table_extra_slots = intern ("char-table-extra-slots");
615
616 /* Now we are ready to set up this property, so we can
617 create category tables. */
618 Fput (Qcategory_table, Qchar_table_extra_slots, make_number (2));
619
620 Vstandard_category_table = Fmake_char_table (Qcategory_table, Qnil);
621 /* Set a category set which contains nothing to the default. */
622 XCHAR_TABLE (Vstandard_category_table)->defalt = MAKE_CATEGORY_SET;
623 Fset_char_table_extra_slot (Vstandard_category_table, make_number (0),
624 Fmake_vector (make_number (95), Qnil));
625 }
626
627 void
628 syms_of_category ()
629 {
630 Qcategoryp = intern ("categoryp");
631 staticpro (&Qcategoryp);
632 Qcategorysetp = intern ("categorysetp");
633 staticpro (&Qcategorysetp);
634 Qcategory_table_p = intern ("category-table-p");
635 staticpro (&Qcategory_table_p);
636
637 DEFVAR_LISP ("word-combining-categories", &Vword_combining_categories,
638 "List of pair (cons) of categories to determine word boundary.\n\
639 \n\
640 Emacs treats a sequence of word constituent characters as a single\n\
641 word (i.e. finds no word boundary between them) iff they belongs to\n\
642 the same charset. But, exceptions are allowed in the following cases.\n\
643 \n\
644 (1) The case that characters are in different charsets is controlled\n\
645 by the variable `word-combining-categories'.\n\
646 \n\
647 Emacs finds no word boundary between characters of different charsets\n\
648 if they have categories matching some element of this list.\n\
649 \n\
650 More precisely, if an element of this list is a cons of category CAT1\n\
651 and CAT2, and a multibyte character C1 which has CAT1 is followed by\n\
652 C2 which has CAT2, there's no word boundary between C1 and C2.\n\
653 \n\
654 For instance, to tell that ASCII characters and Latin-1 characters can\n\
655 form a single word, the element `(?l . ?l)' should be in this list\n\
656 because both characters have the category `l' (Latin characters).\n\
657 \n\
658 (2) The case that character are in the same charset is controlled by\n\
659 the variable `word-separating-categories'.\n\
660 \n\
661 Emacs find a word boundary between characters of the same charset\n\
662 if they have categories matching some element of this list.\n\
663 \n\
664 More precisely, if an element of this list is a cons of category CAT1\n\
665 and CAT2, and a multibyte character C1 which has CAT1 is followed by\n\
666 C2 which has CAT2, there's a word boundary between C1 and C2.\n\
667 \n\
668 For instance, to tell that there's a word boundary between Japanese\n\
669 Hiragana and Japanese Kanji (both are in the same charset), the\n\
670 element `(?H . ?C) should be in this list.");
671
672 Vword_combining_categories = Qnil;
673
674 DEFVAR_LISP ("word-separating-categories", &Vword_separating_categories,
675 "List of pair (cons) of categories to determine word boundary.\n\
676 See the documentation of the variable `word-combining-categories'.");
677
678 Vword_separating_categories = Qnil;
679
680 defsubr (&Smake_category_set);
681 defsubr (&Sdefine_category);
682 defsubr (&Scategory_docstring);
683 defsubr (&Sget_unused_category);
684 defsubr (&Scategory_table_p);
685 defsubr (&Scategory_table);
686 defsubr (&Sstandard_category_table);
687 defsubr (&Scopy_category_table);
688 defsubr (&Sset_category_table);
689 defsubr (&Schar_category_set);
690 defsubr (&Scategory_set_mnemonics);
691 defsubr (&Smodify_category_entry);
692 defsubr (&Sdescribe_categories);
693
694 category_table_version = 0;
695 }