Mercurial > hg > xemacs-beta
comparison src/syntax.c @ 0:376386a54a3c r19-14
Import from CVS: tag r19-14
author | cvs |
---|---|
date | Mon, 13 Aug 2007 08:45:50 +0200 |
parents | |
children | ac2d302a0011 |
comparison
equal
deleted
inserted
replaced
-1:000000000000 | 0:376386a54a3c |
---|---|
1 /* XEmacs routines to deal with syntax tables; also word and list parsing. | |
2 Copyright (C) 1985-1994 Free Software Foundation, Inc. | |
3 Copyright (C) 1995 Sun Microsystems, Inc. | |
4 | |
5 This file is part of XEmacs. | |
6 | |
7 XEmacs is free software; you can redistribute it and/or modify it | |
8 under the terms of the GNU General Public License as published by the | |
9 Free Software Foundation; either version 2, or (at your option) any | |
10 later version. | |
11 | |
12 XEmacs is distributed in the hope that it will be useful, but WITHOUT | |
13 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or | |
14 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License | |
15 for more details. | |
16 | |
17 You should have received a copy of the GNU General Public License | |
18 along with XEmacs; 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 /* Synched up with: FSF 19.28. */ | |
23 | |
24 #include <config.h> | |
25 #include "lisp.h" | |
26 | |
27 #include "buffer.h" | |
28 #include "commands.h" | |
29 #include "insdel.h" | |
30 #include "syntax.h" | |
31 | |
32 Lisp_Object Qsyntax_table_p; | |
33 | |
34 int words_include_escapes; | |
35 | |
36 int parse_sexp_ignore_comments; | |
37 | |
38 /* The following two variables are provided to tell additional information | |
39 to the regex routines. We do it this way rather than change the | |
40 arguments to re_search_2() in an attempt to maintain some call | |
41 compatibility with other versions of the regex code. */ | |
42 | |
43 /* Tell the regex routines not to QUIT. Normally there is a QUIT | |
44 each iteration in re_search_2(). */ | |
45 int no_quit_in_re_search; | |
46 | |
47 /* Tell the regex routines which buffer to access for SYNTAX() lookups | |
48 and the like. */ | |
49 struct buffer *regex_emacs_buffer; | |
50 | |
51 Lisp_Object Vstandard_syntax_table; | |
52 | |
53 Lisp_Object Vsyntax_designator_chars_string; | |
54 | |
55 /* This is the internal form of the parse state used in parse-partial-sexp. */ | |
56 | |
57 struct lisp_parse_state | |
58 { | |
59 int depth; /* Depth at end of parsing */ | |
60 Emchar instring; /* -1 if not within string, else desired terminator. */ | |
61 int incomment; /* Nonzero if within a comment at end of parsing */ | |
62 int comstyle; /* comment style a=0, or b=1 */ | |
63 int quoted; /* Nonzero if just after an escape char at end of | |
64 parsing */ | |
65 Bufpos thislevelstart;/* Char number of most recent start-of-expression | |
66 at current level */ | |
67 Bufpos prevlevelstart;/* Char number of start of containing expression */ | |
68 Bufpos location; /* Char number at which parsing stopped. */ | |
69 int mindepth; /* Minimum depth seen while scanning. */ | |
70 Bufpos comstart; /* Position just after last comment starter. */ | |
71 }; | |
72 | |
73 /* These variables are a cache for finding the start of a defun. | |
74 find_start_pos is the place for which the defun start was found. | |
75 find_start_value is the defun start position found for it. | |
76 find_start_buffer is the buffer it was found in. | |
77 find_start_begv is the BEGV value when it was found. | |
78 find_start_modiff is the value of MODIFF when it was found. */ | |
79 | |
80 static Bufpos find_start_pos; | |
81 static Bufpos find_start_value; | |
82 static struct buffer *find_start_buffer; | |
83 static Bufpos find_start_begv; | |
84 static int find_start_modiff; | |
85 | |
86 /* Find a defun-start that is the last one before POS (or nearly the last). | |
87 We record what we find, so that another call in the same area | |
88 can return the same value right away. */ | |
89 | |
90 static Bufpos | |
91 find_defun_start (struct buffer *buf, Bufpos pos) | |
92 { | |
93 Bufpos tem; | |
94 Lisp_Object table = buf->syntax_table; | |
95 | |
96 /* Use previous finding, if it's valid and applies to this inquiry. */ | |
97 if (buf == find_start_buffer | |
98 /* Reuse the defun-start even if POS is a little farther on. | |
99 POS might be in the next defun, but that's ok. | |
100 Our value may not be the best possible, but will still be usable. */ | |
101 && pos <= find_start_pos + 1000 | |
102 && pos >= find_start_value | |
103 && BUF_BEGV (buf) == find_start_begv | |
104 && BUF_MODIFF (buf) == find_start_modiff) | |
105 return find_start_value; | |
106 | |
107 /* Back up to start of line. */ | |
108 tem = find_next_newline (buf, pos, -1); | |
109 | |
110 while (tem > BUF_BEGV (buf)) | |
111 { | |
112 /* Open-paren at start of line means we found our defun-start. */ | |
113 if (SYNTAX (table, BUF_FETCH_CHAR (buf, tem)) == Sopen) | |
114 break; | |
115 /* Move to beg of previous line. */ | |
116 tem = find_next_newline (buf, tem, -2); | |
117 } | |
118 | |
119 /* Record what we found, for the next try. */ | |
120 find_start_value = tem; | |
121 find_start_buffer = buf; | |
122 find_start_modiff = BUF_MODIFF (buf); | |
123 find_start_begv = BUF_BEGV (buf); | |
124 find_start_pos = pos; | |
125 | |
126 return find_start_value; | |
127 } | |
128 | |
129 DEFUN ("syntax-table-p", Fsyntax_table_p, Ssyntax_table_p, 1, 1, 0 /* | |
130 Return t if ARG is a syntax table. | |
131 Any vector of 256 elements will do. | |
132 */ ) | |
133 (obj) | |
134 Lisp_Object obj; | |
135 { | |
136 if (VECTORP (obj) && vector_length (XVECTOR (obj)) == 0400) | |
137 return Qt; | |
138 return Qnil; | |
139 } | |
140 | |
141 static Lisp_Object | |
142 check_syntax_table (Lisp_Object obj, Lisp_Object def) | |
143 { | |
144 if (NILP (obj)) | |
145 obj = def; | |
146 while (NILP (Fsyntax_table_p (obj))) | |
147 obj = wrong_type_argument (Qsyntax_table_p, obj); | |
148 return (obj); | |
149 } | |
150 | |
151 DEFUN ("syntax-table", Fsyntax_table, Ssyntax_table, 0, 1, 0 /* | |
152 Return the current syntax table. | |
153 This is the one specified by the current buffer, or by BUFFER if it | |
154 is non-nil. | |
155 */ ) | |
156 (buffer) | |
157 Lisp_Object buffer; | |
158 { | |
159 return decode_buffer (buffer, 0)->syntax_table; | |
160 } | |
161 | |
162 DEFUN ("standard-syntax-table", Fstandard_syntax_table, | |
163 Sstandard_syntax_table, 0, 0, 0 /* | |
164 Return the standard syntax table. | |
165 This is the one used for new buffers. | |
166 */ ) | |
167 () | |
168 { | |
169 return Vstandard_syntax_table; | |
170 } | |
171 | |
172 DEFUN ("copy-syntax-table", Fcopy_syntax_table, Scopy_syntax_table, 0, 1, 0 /* | |
173 Construct a new syntax table and return it. | |
174 It is a copy of the TABLE, which defaults to the standard syntax table. | |
175 */ ) | |
176 (table) | |
177 Lisp_Object table; | |
178 { | |
179 if (NILP (Vstandard_syntax_table)) | |
180 /* Can only be null during initialization */ | |
181 return make_vector (0400, Qzero); | |
182 | |
183 table = check_syntax_table (table, Vstandard_syntax_table); | |
184 return Fcopy_sequence (table); | |
185 } | |
186 | |
187 DEFUN ("set-syntax-table", Fset_syntax_table, Sset_syntax_table, 1, 2, 0 /* | |
188 Select a new syntax table for BUFFER. | |
189 One argument, a syntax table. | |
190 BUFFER defaults to the current buffer if omitted. | |
191 */ ) | |
192 (table, buffer) | |
193 Lisp_Object table, buffer; | |
194 { | |
195 struct buffer *buf = decode_buffer (buffer, 0); | |
196 table = check_syntax_table (table, Qnil); | |
197 buf->syntax_table = table; | |
198 /* Indicate that this buffer now has a specified syntax table. */ | |
199 buf->local_var_flags |= XINT (buffer_local_flags.syntax_table); | |
200 return table; | |
201 } | |
202 | |
203 /* Convert a letter which signifies a syntax code | |
204 into the code it signifies. | |
205 This is used by modify-syntax-entry, and other things. */ | |
206 | |
207 CONST unsigned char syntax_spec_code[0400] = | |
208 { 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
209 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
210 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
211 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
212 (char) Swhitespace, 0377, (char) Sstring, 0377, | |
213 (char) Smath, 0377, 0377, (char) Squote, | |
214 (char) Sopen, (char) Sclose, 0377, 0377, | |
215 0377, (char) Swhitespace, (char) Spunct, (char) Scharquote, | |
216 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
217 0377, 0377, 0377, 0377, | |
218 (char) Scomment, 0377, (char) Sendcomment, 0377, | |
219 (char) Sinherit, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* @, A ... */ | |
220 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
221 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword, | |
222 0377, 0377, 0377, 0377, (char) Sescape, 0377, 0377, (char) Ssymbol, | |
223 0377, 0377, 0377, 0377, 0377, Sextword, 0377, 0377, /* `, a, ... */ | |
224 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
225 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword, | |
226 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377 | |
227 }; | |
228 | |
229 CONST unsigned char syntax_code_spec[] = | |
230 { | |
231 ' ', '.', 'w', '_', '(', ')', '\'', '\"', '$', '\\', '/', '<', '>', '@', | |
232 'e', '\0' | |
233 }; | |
234 | |
235 DEFUN ("syntax-designator-chars", Fsyntax_designator_chars, | |
236 Ssyntax_designator_chars, 0, 0, 0 /* | |
237 Return a string of the recognized syntax designator chars. | |
238 The chars are ordered by their internal syntax codes, which are | |
239 numbered starting at 0. | |
240 */ ) | |
241 () | |
242 { | |
243 return Vsyntax_designator_chars_string; | |
244 } | |
245 | |
246 DEFUN ("char-syntax", Fchar_syntax, Schar_syntax, 1, 2, 0 /* | |
247 Return the syntax code of CHAR, described by a character. | |
248 For example, if CHAR is a word constituent, the character `?w' is returned. | |
249 The characters that correspond to various syntax codes | |
250 are listed in the documentation of `modify-syntax-entry'. | |
251 Optional second argument TABLE defaults to the current buffer's | |
252 syntax table. | |
253 */ ) | |
254 (ch, table) | |
255 Lisp_Object ch, table; | |
256 { | |
257 CHECK_CHAR_COERCE_INT (ch); | |
258 table = check_syntax_table (table, current_buffer->syntax_table); | |
259 | |
260 return make_int (syntax_code_spec[(int) SYNTAX (table, XINT (ch))]); | |
261 } | |
262 | |
263 | |
264 Lisp_Object | |
265 syntax_match (Lisp_Object table, Emchar ch) | |
266 { | |
267 unsigned char stringterm = ((SYNTAX_CODE (table, ch) >> 8) & 0377); | |
268 | |
269 if (stringterm == 0) | |
270 return Qnil; | |
271 else | |
272 return make_char (stringterm); | |
273 } | |
274 | |
275 DEFUN ("matching-paren", Fmatching_paren, Smatching_paren, 1, 2, 0 /* | |
276 Return the matching parenthesis of CHAR, or nil if none. | |
277 Optional second argument TABLE defaults to the current buffer's | |
278 syntax table. | |
279 */ ) | |
280 (ch, table) | |
281 Lisp_Object ch, table; | |
282 { | |
283 int code; | |
284 CHECK_CHAR_COERCE_INT (ch); | |
285 | |
286 table = check_syntax_table (table, current_buffer->syntax_table); | |
287 code = SYNTAX (table, XCHAR (ch)); | |
288 if (code == Sopen || code == Sclose || code == Sstring) | |
289 return syntax_match (table, XCHAR (ch)); | |
290 return Qnil; | |
291 } | |
292 | |
293 | |
294 /* Return the position across COUNT words from FROM. | |
295 If that many words cannot be found before the end of the buffer, return 0. | |
296 COUNT negative means scan backward and stop at word beginning. */ | |
297 | |
298 Bufpos | |
299 scan_words (struct buffer *buf, Bufpos from, int count) | |
300 { | |
301 Bufpos beg = BUF_BEGV (buf); | |
302 Bufpos end = BUF_ZV (buf); | |
303 enum syntaxcode code; | |
304 Lisp_Object table = buf->syntax_table; | |
305 | |
306 while (count > 0) | |
307 { | |
308 QUIT; | |
309 | |
310 while (1) | |
311 { | |
312 Emchar ch; | |
313 if (from == end) | |
314 { | |
315 return 0; | |
316 } | |
317 ch = BUF_FETCH_CHAR (buf, from); | |
318 code = SYNTAX_UNSAFE (table, ch); | |
319 if (words_include_escapes | |
320 && (code == Sescape || code == Scharquote)) | |
321 break; | |
322 if (code == Sword || code == Sextword) | |
323 break; | |
324 from++; | |
325 } | |
326 | |
327 QUIT; | |
328 | |
329 while (1) | |
330 { | |
331 Emchar ch; | |
332 if (from == end) break; | |
333 ch = BUF_FETCH_CHAR (buf, from); | |
334 code = SYNTAX_UNSAFE (table, ch); | |
335 if (!(words_include_escapes | |
336 && (code == Sescape || code == Scharquote))) | |
337 if (code != Sword && code != Sextword) | |
338 break; | |
339 from++; | |
340 } | |
341 | |
342 count--; | |
343 } | |
344 while (count < 0) | |
345 { | |
346 QUIT; | |
347 | |
348 while (1) | |
349 { | |
350 Emchar ch; | |
351 if (from == beg) | |
352 { | |
353 return 0; | |
354 } | |
355 ch = BUF_FETCH_CHAR (buf, from - 1); | |
356 code = SYNTAX_UNSAFE (table, ch); | |
357 if (words_include_escapes | |
358 && (code == Sescape || code == Scharquote)) | |
359 break; | |
360 if (code == Sword || code == Sextword) | |
361 break; | |
362 from--; | |
363 } | |
364 | |
365 QUIT; | |
366 | |
367 while (1) | |
368 { | |
369 Emchar ch; | |
370 if (from == beg) break; | |
371 ch = BUF_FETCH_CHAR (buf, from - 1); | |
372 code = SYNTAX_UNSAFE (table, ch); | |
373 if (!(words_include_escapes | |
374 && (code == Sescape || code == Scharquote))) | |
375 if (code != Sword && code != Sextword) | |
376 break; | |
377 from--; | |
378 } | |
379 count++; | |
380 } | |
381 | |
382 return from; | |
383 } | |
384 | |
385 DEFUN ("forward-word", Fforward_word, Sforward_word, 1, 2, "_p" /* | |
386 Move point forward ARG words (backward if ARG is negative). | |
387 Normally returns t. | |
388 If an edge of the buffer is reached, point is left there | |
389 and nil is returned. | |
390 */ ) | |
391 (count, buffer) | |
392 Lisp_Object count, buffer; | |
393 { | |
394 Bufpos val; | |
395 struct buffer *buf = decode_buffer (buffer, 0); | |
396 CHECK_INT (count); | |
397 | |
398 if (!(val = scan_words (buf, BUF_PT (buf), XINT (count)))) | |
399 { | |
400 BUF_SET_PT (buf, XINT (count) > 0 ? BUF_ZV (buf) : BUF_BEGV (buf)); | |
401 return Qnil; | |
402 } | |
403 BUF_SET_PT (buf, val); | |
404 return Qt; | |
405 } | |
406 | |
407 static void scan_sexps_forward (struct buffer *buf, | |
408 struct lisp_parse_state *, | |
409 Bufpos from, Bufpos end, | |
410 int targetdepth, int stopbefore, | |
411 Lisp_Object oldstate, | |
412 int commentstop); | |
413 | |
414 static int | |
415 find_start_of_comment (struct buffer *buf, Bufpos from, Bufpos stop, int mask) | |
416 { | |
417 Emchar c; | |
418 enum syntaxcode code; | |
419 Lisp_Object table = buf->syntax_table; | |
420 | |
421 /* Look back, counting the parity of string-quotes, | |
422 and recording the comment-starters seen. | |
423 When we reach a safe place, assume that's not in a string; | |
424 then step the main scan to the earliest comment-starter seen | |
425 an even number of string quotes away from the safe place. | |
426 | |
427 OFROM[I] is position of the earliest comment-starter seen | |
428 which is I+2X quotes from the comment-end. | |
429 PARITY is current parity of quotes from the comment end. */ | |
430 int parity = 0; | |
431 Emchar my_stringend = 0; | |
432 int string_lossage = 0; | |
433 Bufpos comment_end = from; | |
434 Bufpos comstart_pos = 0; | |
435 int comstart_parity = 0; | |
436 int styles_match_p = 0; | |
437 | |
438 /* At beginning of range to scan, we're outside of strings; | |
439 that determines quote parity to the comment-end. */ | |
440 while (from != stop) | |
441 { | |
442 /* Move back and examine a character. */ | |
443 from--; | |
444 | |
445 c = BUF_FETCH_CHAR (buf, from); | |
446 code = SYNTAX_UNSAFE (table, c); | |
447 | |
448 /* is this a 1-char comment end sequence? if so, try | |
449 to see if style matches previously extracted mask */ | |
450 if (code == Sendcomment) | |
451 { | |
452 styles_match_p = SYNTAX_STYLES_MATCH_1CHAR_P (table, c, mask); | |
453 } | |
454 | |
455 /* otherwise, is this a 2-char comment end sequence? */ | |
456 else if (from >= stop | |
457 && SYNTAX_END_P (table, c, BUF_FETCH_CHAR (buf, from+1))) | |
458 { | |
459 code = Sendcomment; | |
460 styles_match_p = | |
461 SYNTAX_STYLES_MATCH_END_P (table, c, | |
462 BUF_FETCH_CHAR (buf, from+1), | |
463 mask); | |
464 } | |
465 | |
466 /* or are we looking at a 1-char comment start sequence | |
467 of the style matching mask? */ | |
468 else if (code == Scomment | |
469 && SYNTAX_STYLES_MATCH_1CHAR_P (table, c, mask)) | |
470 { | |
471 styles_match_p = 1; | |
472 } | |
473 | |
474 /* or possibly, a 2-char comment start sequence */ | |
475 else if (from >= stop | |
476 && SYNTAX_STYLES_MATCH_START_P (table, c, | |
477 BUF_FETCH_CHAR (buf, from+1), | |
478 mask)) | |
479 { | |
480 code = Scomment; | |
481 styles_match_p = 1; | |
482 } | |
483 | |
484 /* Ignore escaped characters. */ | |
485 if (char_quoted (buf, from)) | |
486 continue; | |
487 | |
488 /* Track parity of quotes. */ | |
489 if (code == Sstring) | |
490 { | |
491 parity ^= 1; | |
492 if (my_stringend == 0) | |
493 my_stringend = c; | |
494 /* If we have two kinds of string delimiters. | |
495 There's no way to grok this scanning backwards. */ | |
496 else if (my_stringend != c) | |
497 string_lossage = 1; | |
498 } | |
499 | |
500 /* Record comment-starters according to that | |
501 quote-parity to the comment-end. */ | |
502 if (code == Scomment && styles_match_p) | |
503 { | |
504 comstart_parity = parity; | |
505 comstart_pos = from; | |
506 } | |
507 | |
508 /* If we find another earlier comment-ender, | |
509 any comment-starts earlier than that don't count | |
510 (because they go with the earlier comment-ender). */ | |
511 if (code == Sendcomment && styles_match_p) | |
512 break; | |
513 | |
514 /* Assume a defun-start point is outside of strings. */ | |
515 if (code == Sopen | |
516 && (from == stop || BUF_FETCH_CHAR (buf, from - 1) == '\n')) | |
517 break; | |
518 } | |
519 | |
520 if (comstart_pos == 0) | |
521 from = comment_end; | |
522 /* If the earliest comment starter | |
523 is followed by uniform paired string quotes or none, | |
524 we know it can't be inside a string | |
525 since if it were then the comment ender would be inside one. | |
526 So it does start a comment. Skip back to it. */ | |
527 else if (comstart_parity == 0 && !string_lossage) | |
528 from = comstart_pos; | |
529 else | |
530 { | |
531 /* We had two kinds of string delimiters mixed up | |
532 together. Decode this going forwards. | |
533 Scan fwd from the previous comment ender | |
534 to the one in question; this records where we | |
535 last passed a comment starter. */ | |
536 | |
537 struct lisp_parse_state state; | |
538 scan_sexps_forward (buf, &state, find_defun_start (buf, comment_end), | |
539 comment_end - 1, -10000, 0, Qnil, 0); | |
540 if (state.incomment) | |
541 from = state.comstart; | |
542 else | |
543 /* We can't grok this as a comment; scan it normally. */ | |
544 from = comment_end; | |
545 } | |
546 return from; | |
547 } | |
548 | |
549 static Bufpos | |
550 find_end_of_comment (struct buffer *buf, Bufpos from, Bufpos stop, int mask) | |
551 { | |
552 int c; | |
553 Lisp_Object table = buf->syntax_table; | |
554 | |
555 while (1) | |
556 { | |
557 if (from == stop) | |
558 { | |
559 return -1; | |
560 } | |
561 c = BUF_FETCH_CHAR (buf, from); | |
562 if (SYNTAX_UNSAFE (table, c) == Sendcomment | |
563 && SYNTAX_STYLES_MATCH_1CHAR_P (table, c, mask)) | |
564 /* we have encountered a comment end of the same style | |
565 as the comment sequence which began this comment | |
566 section */ | |
567 break; | |
568 | |
569 from++; | |
570 if (from < stop | |
571 && SYNTAX_STYLES_MATCH_END_P (table, c, | |
572 BUF_FETCH_CHAR (buf, from), mask)) | |
573 /* we have encountered a comment end of the same style | |
574 as the comment sequence which began this comment | |
575 section */ | |
576 { from++; break; } | |
577 } | |
578 return from; | |
579 } | |
580 | |
581 | |
582 /* #### between FSF 19.23 and 19.28 there are some changes to the logic | |
583 in this function (and minor changes to find_start_of_comment(), | |
584 above, which is part of Fforward_comment() in FSF). Attempts to port | |
585 that logic made this function break, so I'm leaving it out. If anyone | |
586 ever complains about this function not working properly, take a look | |
587 at those changes. --ben */ | |
588 | |
589 DEFUN ("forward-comment", Fforward_comment, Sforward_comment, 1, 2, 0 /* | |
590 Move forward across up to N comments. If N is negative, move backward. | |
591 Stop scanning if we find something other than a comment or whitespace. | |
592 Set point to where scanning stops. | |
593 If N comments are found as expected, with nothing except whitespace | |
594 between them, return t; otherwise return nil. | |
595 Point is set in either case. | |
596 Optional argument BUFFER defaults to the current buffer. | |
597 */ ) | |
598 (n, buffer) | |
599 Lisp_Object n, buffer; | |
600 { | |
601 Bufpos from; | |
602 Bufpos stop; | |
603 Emchar c; | |
604 enum syntaxcode code; | |
605 int count; | |
606 struct buffer *buf = decode_buffer (buffer, 0); | |
607 Lisp_Object table = buf->syntax_table; | |
608 | |
609 CHECK_INT (n); | |
610 count = XINT (n); | |
611 | |
612 from = BUF_PT (buf); | |
613 | |
614 while (count > 0) | |
615 { | |
616 QUIT; | |
617 | |
618 stop = BUF_ZV (buf); | |
619 while (from < stop) | |
620 { | |
621 int mask = 0; /* mask for finding matching comment style */ | |
622 | |
623 if (char_quoted (buf, from)) | |
624 { | |
625 from++; | |
626 continue; | |
627 } | |
628 | |
629 c = BUF_FETCH_CHAR (buf, from); | |
630 code = SYNTAX (table, c); | |
631 | |
632 if (code == Scomment) | |
633 { | |
634 /* we have encountered a single character comment start | |
635 sequence, and we are ignoring all text inside comments. | |
636 we must record the comment style this character begins | |
637 so that later, only a comment end of the same style actually | |
638 ends the comment section */ | |
639 mask = SYNTAX_COMMENT_1CHAR_MASK (table, c); | |
640 } | |
641 | |
642 else if (from < stop | |
643 && SYNTAX_START_P (table, c, BUF_FETCH_CHAR (buf, from+1))) | |
644 { | |
645 /* we have encountered a 2char comment start sequence and we | |
646 are ignoring all text inside comments. we must record | |
647 the comment style this sequence begins so that later, | |
648 only a comment end of the same style actually ends | |
649 the comment section */ | |
650 code = Scomment; | |
651 mask = SYNTAX_COMMENT_MASK_START (table, c, | |
652 BUF_FETCH_CHAR (buf, from+1)); | |
653 from++; | |
654 } | |
655 | |
656 if (code == Scomment) | |
657 { | |
658 Bufpos newfrom; | |
659 | |
660 newfrom = find_end_of_comment (buf, from, stop, mask); | |
661 if (newfrom < 0) | |
662 { | |
663 /* we stopped because from==stop */ | |
664 BUF_SET_PT (buf, stop); | |
665 return Qnil; | |
666 } | |
667 from = newfrom; | |
668 | |
669 /* We have skipped one comment. */ | |
670 break; | |
671 } | |
672 else if (code != Swhitespace | |
673 && code != Sendcomment | |
674 && code != Scomment ) | |
675 { | |
676 BUF_SET_PT (buf, from); | |
677 return Qnil; | |
678 } | |
679 from++; | |
680 } | |
681 | |
682 /* End of comment reached */ | |
683 count--; | |
684 } | |
685 | |
686 while (count < 0) | |
687 { | |
688 QUIT; | |
689 | |
690 stop = BUF_BEGV (buf); | |
691 while (from > stop) | |
692 { | |
693 int mask = 0; /* mask for finding matching comment style */ | |
694 | |
695 from--; | |
696 if (char_quoted (buf, from)) | |
697 { | |
698 from--; | |
699 continue; | |
700 } | |
701 | |
702 c = BUF_FETCH_CHAR (buf, from); | |
703 code = SYNTAX (table, c); | |
704 | |
705 if (code == Sendcomment) | |
706 { | |
707 /* we have found a single char end comment. we must record | |
708 the comment style encountered so that later, we can match | |
709 only the proper comment begin sequence of the same style */ | |
710 mask = SYNTAX_COMMENT_1CHAR_MASK (table, c); | |
711 } | |
712 | |
713 else if (from > stop | |
714 && SYNTAX_END_P (table, BUF_FETCH_CHAR (buf, from - 1), c) | |
715 && !char_quoted (buf, from - 1)) | |
716 { | |
717 /* We must record the comment style encountered so that | |
718 later, we can match only the proper comment begin | |
719 sequence of the same style. */ | |
720 code = Sendcomment; | |
721 mask = SYNTAX_COMMENT_MASK_END (table, | |
722 BUF_FETCH_CHAR (buf, from - 1), | |
723 c); | |
724 from--; | |
725 } | |
726 | |
727 if (code == Sendcomment) | |
728 { | |
729 from = find_start_of_comment (buf, from, stop, mask); | |
730 break; | |
731 } | |
732 | |
733 else if (code != Swhitespace | |
734 && SYNTAX (table, c) != Scomment | |
735 && SYNTAX (table, c) != Sendcomment) | |
736 { | |
737 BUF_SET_PT (buf, from + 1); | |
738 return Qnil; | |
739 } | |
740 } | |
741 | |
742 count++; | |
743 } | |
744 | |
745 BUF_SET_PT (buf, from); | |
746 return Qt; | |
747 } | |
748 | |
749 | |
750 Lisp_Object | |
751 scan_lists (struct buffer *buf, Bufpos from, int count, int depth, | |
752 int sexpflag, int no_error) | |
753 { | |
754 Bufpos stop; | |
755 Emchar c; | |
756 int quoted; | |
757 int mathexit = 0; | |
758 enum syntaxcode code; | |
759 int min_depth = depth; /* Err out if depth gets less than this. */ | |
760 Lisp_Object table = buf->syntax_table; | |
761 | |
762 if (depth > 0) min_depth = 0; | |
763 | |
764 while (count > 0) | |
765 { | |
766 QUIT; | |
767 | |
768 stop = BUF_ZV (buf); | |
769 while (from < stop) | |
770 { | |
771 int mask = 0; /* mask for finding matching comment style */ | |
772 | |
773 c = BUF_FETCH_CHAR (buf, from); | |
774 code = SYNTAX_UNSAFE (table, c); | |
775 from++; | |
776 | |
777 /* a 1-char comment start sequence */ | |
778 if (code == Scomment && parse_sexp_ignore_comments) | |
779 { | |
780 mask = SYNTAX_COMMENT_1CHAR_MASK (table, c); | |
781 } | |
782 | |
783 /* else, a 2-char comment start sequence? */ | |
784 else if (from < stop | |
785 && SYNTAX_START_P (table, c, BUF_FETCH_CHAR (buf, from)) | |
786 && parse_sexp_ignore_comments) | |
787 { | |
788 /* we have encountered a comment start sequence and we | |
789 are ignoring all text inside comments. we must record | |
790 the comment style this sequence begins so that later, | |
791 only a comment end of the same style actually ends | |
792 the comment section */ | |
793 code = Scomment; | |
794 mask = SYNTAX_COMMENT_MASK_START (table, c, | |
795 BUF_FETCH_CHAR (buf, from)); | |
796 from++; | |
797 } | |
798 | |
799 if (SYNTAX_PREFIX_UNSAFE (table, c)) | |
800 continue; | |
801 | |
802 switch (code) | |
803 { | |
804 case Sescape: | |
805 case Scharquote: | |
806 if (from == stop) goto lose; | |
807 from++; | |
808 /* treat following character as a word constituent */ | |
809 case Sword: | |
810 case Sextword: | |
811 case Ssymbol: | |
812 if (depth || !sexpflag) break; | |
813 /* This word counts as a sexp; return at end of it. */ | |
814 while (from < stop) | |
815 { | |
816 switch (SYNTAX (table, BUF_FETCH_CHAR (buf, from))) | |
817 { | |
818 case Scharquote: | |
819 case Sescape: | |
820 from++; | |
821 if (from == stop) goto lose; | |
822 break; | |
823 case Sword: | |
824 case Sextword: | |
825 case Ssymbol: | |
826 case Squote: | |
827 break; | |
828 default: | |
829 goto done; | |
830 } | |
831 from++; | |
832 } | |
833 goto done; | |
834 | |
835 case Scomment: | |
836 if (!parse_sexp_ignore_comments) | |
837 break; | |
838 { | |
839 Bufpos newfrom = find_end_of_comment (buf, from, stop, mask); | |
840 if (newfrom < 0) | |
841 { | |
842 /* we stopped because from == stop in search forward */ | |
843 from = stop; | |
844 if (depth == 0) | |
845 goto done; | |
846 goto lose; | |
847 } | |
848 from = newfrom; | |
849 } | |
850 break; | |
851 | |
852 case Smath: | |
853 if (!sexpflag) | |
854 break; | |
855 if (from != stop && c == BUF_FETCH_CHAR (buf, from)) | |
856 from++; | |
857 if (mathexit) | |
858 { | |
859 mathexit = 0; | |
860 goto close1; | |
861 } | |
862 mathexit = 1; | |
863 | |
864 case Sopen: | |
865 if (!++depth) goto done; | |
866 break; | |
867 | |
868 case Sclose: | |
869 close1: | |
870 if (!--depth) goto done; | |
871 if (depth < min_depth) | |
872 { | |
873 if (no_error) | |
874 return Qnil; | |
875 error ("Containing expression ends prematurely"); | |
876 } | |
877 break; | |
878 | |
879 case Sstring: | |
880 { | |
881 /* XEmacs change: call syntax_match on character */ | |
882 Emchar ch = BUF_FETCH_CHAR (buf, from - 1); | |
883 Lisp_Object stermobj = syntax_match (table, ch); | |
884 Emchar stringterm; | |
885 | |
886 if (CHARP (stermobj)) | |
887 stringterm = XCHAR (stermobj); | |
888 else | |
889 stringterm = ch; | |
890 | |
891 while (1) | |
892 { | |
893 if (from >= stop) | |
894 goto lose; | |
895 if (BUF_FETCH_CHAR (buf, from) == stringterm) | |
896 break; | |
897 switch (SYNTAX (table, BUF_FETCH_CHAR (buf, from))) | |
898 { | |
899 case Scharquote: | |
900 case Sescape: | |
901 from++; | |
902 break; | |
903 default: | |
904 break; | |
905 } | |
906 from++; | |
907 } | |
908 from++; | |
909 if (!depth && sexpflag) goto done; | |
910 break; | |
911 } | |
912 | |
913 default: | |
914 break; | |
915 } | |
916 } | |
917 | |
918 /* Reached end of buffer. Error if within object, | |
919 return nil if between */ | |
920 if (depth) goto lose; | |
921 | |
922 return Qnil; | |
923 | |
924 /* End of object reached */ | |
925 done: | |
926 count--; | |
927 } | |
928 | |
929 while (count < 0) | |
930 { | |
931 QUIT; | |
932 | |
933 stop = BUF_BEGV (buf); | |
934 while (from > stop) | |
935 { | |
936 int mask = 0; /* mask for finding matching comment style */ | |
937 | |
938 from--; | |
939 quoted = char_quoted (buf, from); | |
940 if (quoted) | |
941 from--; | |
942 | |
943 c = BUF_FETCH_CHAR (buf, from); | |
944 code = SYNTAX_UNSAFE (table, c); | |
945 | |
946 if (code == Sendcomment && parse_sexp_ignore_comments) | |
947 { | |
948 /* we have found a single char end comment. we must record | |
949 the comment style encountered so that later, we can match | |
950 only the proper comment begin sequence of the same style */ | |
951 mask = SYNTAX_COMMENT_1CHAR_MASK (table, c); | |
952 } | |
953 | |
954 else if (from > stop | |
955 && SYNTAX_END_P (table, BUF_FETCH_CHAR (buf, from-1), c) | |
956 && !char_quoted (buf, from - 1) | |
957 && parse_sexp_ignore_comments) | |
958 { | |
959 /* we must record the comment style encountered so that | |
960 later, we can match only the proper comment begin | |
961 sequence of the same style */ | |
962 code = Sendcomment; | |
963 mask = SYNTAX_COMMENT_MASK_END (table, | |
964 BUF_FETCH_CHAR (buf, from - 1), | |
965 c); | |
966 from--; | |
967 } | |
968 | |
969 if (SYNTAX_PREFIX_UNSAFE (table, c)) | |
970 continue; | |
971 | |
972 switch (((quoted) ? Sword : code)) | |
973 { | |
974 case Sword: | |
975 case Sextword: | |
976 case Ssymbol: | |
977 if (depth || !sexpflag) break; | |
978 /* This word counts as a sexp; count object finished after | |
979 passing it. */ | |
980 while (from > stop) | |
981 { | |
982 enum syntaxcode syncode; | |
983 quoted = char_quoted (buf, from - 1); | |
984 | |
985 if (quoted) | |
986 from--; | |
987 if (! (quoted | |
988 || (syncode = | |
989 SYNTAX (table, BUF_FETCH_CHAR (buf, from - 1))) | |
990 == Sword | |
991 || syncode == Sextword | |
992 || syncode == Ssymbol | |
993 || syncode == Squote)) | |
994 goto done2; | |
995 from--; | |
996 } | |
997 goto done2; | |
998 | |
999 case Smath: | |
1000 if (!sexpflag) | |
1001 break; | |
1002 if (from != stop && c == BUF_FETCH_CHAR (buf, from - 1)) | |
1003 from--; | |
1004 if (mathexit) | |
1005 { | |
1006 mathexit = 0; | |
1007 goto open2; | |
1008 } | |
1009 mathexit = 1; | |
1010 | |
1011 case Sclose: | |
1012 if (!++depth) goto done2; | |
1013 break; | |
1014 | |
1015 case Sopen: | |
1016 open2: | |
1017 if (!--depth) goto done2; | |
1018 if (depth < min_depth) | |
1019 { | |
1020 if (no_error) | |
1021 return Qnil; | |
1022 error ("Containing expression ends prematurely"); | |
1023 } | |
1024 break; | |
1025 | |
1026 case Sendcomment: | |
1027 if (parse_sexp_ignore_comments) | |
1028 from = find_start_of_comment (buf, from, stop, mask); | |
1029 break; | |
1030 | |
1031 case Sstring: | |
1032 { | |
1033 /* XEmacs change: call syntax_match() on character */ | |
1034 Emchar ch = BUF_FETCH_CHAR (buf, from); | |
1035 Lisp_Object stermobj = syntax_match (table, ch); | |
1036 Emchar stringterm; | |
1037 | |
1038 if (CHARP (stermobj)) | |
1039 stringterm = XCHAR (stermobj); | |
1040 else | |
1041 stringterm = ch; | |
1042 | |
1043 while (1) | |
1044 { | |
1045 if (from == stop) goto lose; | |
1046 if (!char_quoted (buf, from - 1) | |
1047 && stringterm == BUF_FETCH_CHAR (buf, from - 1)) | |
1048 break; | |
1049 from--; | |
1050 } | |
1051 from--; | |
1052 if (!depth && sexpflag) goto done2; | |
1053 break; | |
1054 } | |
1055 } | |
1056 } | |
1057 | |
1058 /* Reached start of buffer. Error if within object, | |
1059 return nil if between */ | |
1060 if (depth) goto lose; | |
1061 | |
1062 return Qnil; | |
1063 | |
1064 done2: | |
1065 count++; | |
1066 } | |
1067 | |
1068 | |
1069 return (make_int (from)); | |
1070 | |
1071 lose: | |
1072 if (!no_error) | |
1073 error ("Unbalanced parentheses"); | |
1074 return Qnil; | |
1075 } | |
1076 | |
1077 int | |
1078 char_quoted (struct buffer *buf, Bufpos pos) | |
1079 { | |
1080 enum syntaxcode code; | |
1081 Bufpos beg = BUF_BEGV (buf); | |
1082 int quoted = 0; | |
1083 Lisp_Object table = buf->syntax_table; | |
1084 | |
1085 while (pos > beg | |
1086 && ((code = SYNTAX (table, BUF_FETCH_CHAR (buf, pos - 1))) | |
1087 == Scharquote | |
1088 || code == Sescape)) | |
1089 pos--, quoted = !quoted; | |
1090 return quoted; | |
1091 } | |
1092 | |
1093 DEFUN ("scan-lists", Fscan_lists, Sscan_lists, 3, 5, 0 /* | |
1094 Scan from character number FROM by COUNT lists. | |
1095 Returns the character number of the position thus found. | |
1096 | |
1097 If DEPTH is nonzero, paren depth begins counting from that value, | |
1098 only places where the depth in parentheses becomes zero | |
1099 are candidates for stopping; COUNT such places are counted. | |
1100 Thus, a positive value for DEPTH means go out levels. | |
1101 | |
1102 Comments are ignored if `parse-sexp-ignore-comments' is non-nil. | |
1103 | |
1104 If the beginning or end of (the accessible part of) the buffer is reached | |
1105 and the depth is wrong, an error is signaled. | |
1106 If the depth is right but the count is not used up, nil is returned. | |
1107 | |
1108 If optional arg BUFFER is non-nil, scanning occurs in that buffer instead | |
1109 of in the current buffer. | |
1110 | |
1111 If optional arg NOERROR is non-nil, scan-lists will return nil instead of | |
1112 signalling an error. | |
1113 */ ) | |
1114 (from, count, depth, buffer, no_error) | |
1115 Lisp_Object from, count, depth, buffer, no_error; | |
1116 { | |
1117 struct buffer *buf; | |
1118 | |
1119 CHECK_INT (from); | |
1120 CHECK_INT (count); | |
1121 CHECK_INT (depth); | |
1122 buf = decode_buffer (buffer, 0); | |
1123 | |
1124 return scan_lists (buf, XINT (from), XINT (count), XINT (depth), 0, | |
1125 !NILP (no_error)); | |
1126 } | |
1127 | |
1128 DEFUN ("scan-sexps", Fscan_sexps, Sscan_sexps, 2, 4, 0 /* | |
1129 Scan from character number FROM by COUNT balanced expressions. | |
1130 If COUNT is negative, scan backwards. | |
1131 Returns the character number of the position thus found. | |
1132 | |
1133 Comments are ignored if `parse-sexp-ignore-comments' is non-nil. | |
1134 | |
1135 If the beginning or end of (the accessible part of) the buffer is reached | |
1136 in the middle of a parenthetical grouping, an error is signaled. | |
1137 If the beginning or end is reached between groupings | |
1138 but before count is used up, nil is returned. | |
1139 | |
1140 If optional arg BUFFER is non-nil, scanning occurs in that buffer instead | |
1141 of in the current buffer. | |
1142 | |
1143 If optional arg NOERROR is non-nil, scan-sexps will return nil instead of | |
1144 signalling an error. | |
1145 */ ) | |
1146 (from, count, buffer, no_error) | |
1147 Lisp_Object from, count, buffer, no_error; | |
1148 { | |
1149 struct buffer *buf = decode_buffer (buffer, 0); | |
1150 CHECK_INT (from); | |
1151 CHECK_INT (count); | |
1152 | |
1153 return scan_lists (buf, XINT (from), XINT (count), 0, 1, !NILP (no_error)); | |
1154 } | |
1155 | |
1156 DEFUN ("backward-prefix-chars", Fbackward_prefix_chars, Sbackward_prefix_chars, | |
1157 0, 1, 0 /* | |
1158 Move point backward over any number of chars with prefix syntax. | |
1159 This includes chars with \"quote\" or \"prefix\" syntax (' or p). | |
1160 | |
1161 Optional arg BUFFER defaults to the current buffer. | |
1162 */ ) | |
1163 (buffer) | |
1164 Lisp_Object buffer; | |
1165 { | |
1166 struct buffer *buf = decode_buffer (buffer, 0); | |
1167 Bufpos beg = BUF_BEGV (buf); | |
1168 Bufpos pos = BUF_PT (buf); | |
1169 Lisp_Object table = buf->syntax_table; | |
1170 | |
1171 while (pos > beg && !char_quoted (buf, pos - 1) | |
1172 && (SYNTAX (table, BUF_FETCH_CHAR (buf, pos - 1)) == Squote | |
1173 || SYNTAX_PREFIX (table, BUF_FETCH_CHAR (buf, pos - 1)))) | |
1174 pos--; | |
1175 | |
1176 BUF_SET_PT (buf, pos); | |
1177 | |
1178 return Qnil; | |
1179 } | |
1180 | |
1181 /* Parse forward from FROM to END, | |
1182 assuming that FROM has state OLDSTATE (nil means FROM is start of function), | |
1183 and return a description of the state of the parse at END. | |
1184 If STOPBEFORE is nonzero, stop at the start of an atom. | |
1185 If COMMENTSTOP is nonzero, stop at the start of a comment. */ | |
1186 | |
1187 static void | |
1188 scan_sexps_forward (struct buffer *buf, struct lisp_parse_state *stateptr, | |
1189 Bufpos from, Bufpos end, | |
1190 int targetdepth, int stopbefore, | |
1191 Lisp_Object oldstate, | |
1192 int commentstop) | |
1193 { | |
1194 struct lisp_parse_state state; | |
1195 | |
1196 enum syntaxcode code; | |
1197 struct level { int last, prev; }; | |
1198 struct level levelstart[100]; | |
1199 struct level *curlevel = levelstart; | |
1200 struct level *endlevel = levelstart + 100; | |
1201 int depth; /* Paren depth of current scanning location. | |
1202 level - levelstart equals this except | |
1203 when the depth becomes negative. */ | |
1204 int mindepth; /* Lowest DEPTH value seen. */ | |
1205 int start_quoted = 0; /* Nonzero means starting after a char quote */ | |
1206 Lisp_Object table = buf->syntax_table; | |
1207 Lisp_Object tem; | |
1208 int mask; /* comment mask */ | |
1209 | |
1210 if (NILP (oldstate)) | |
1211 { | |
1212 depth = 0; | |
1213 state.instring = -1; | |
1214 state.incomment = 0; | |
1215 state.comstyle = 0; /* comment style a by default */ | |
1216 mask = SYNTAX_COMMENT_STYLE_A; | |
1217 } | |
1218 else | |
1219 { | |
1220 tem = Fcar (oldstate); /* elt 0, depth */ | |
1221 if (!NILP (tem)) | |
1222 depth = XINT (tem); | |
1223 else | |
1224 depth = 0; | |
1225 | |
1226 oldstate = Fcdr (oldstate); | |
1227 oldstate = Fcdr (oldstate); | |
1228 oldstate = Fcdr (oldstate); | |
1229 tem = Fcar (oldstate); /* elt 3, instring */ | |
1230 state.instring = !NILP (tem) ? XINT (tem) : -1; | |
1231 | |
1232 oldstate = Fcdr (oldstate); /* elt 4, incomment */ | |
1233 tem = Fcar (oldstate); | |
1234 state.incomment = !NILP (tem); | |
1235 | |
1236 oldstate = Fcdr (oldstate); | |
1237 tem = Fcar (oldstate); /* elt 5, follows-quote */ | |
1238 start_quoted = !NILP (tem); | |
1239 | |
1240 /* if the eighth element of the list is nil, we are in comment style | |
1241 a. if it is non-nil, we are in comment style b */ | |
1242 oldstate = Fcdr (oldstate); | |
1243 oldstate = Fcdr (oldstate); | |
1244 oldstate = Fcdr (oldstate); | |
1245 tem = Fcar (oldstate); /* elt 8, comment style a */ | |
1246 state.comstyle = !NILP (tem); | |
1247 mask = state.comstyle ? SYNTAX_COMMENT_STYLE_B : SYNTAX_COMMENT_STYLE_A; | |
1248 } | |
1249 state.quoted = 0; | |
1250 mindepth = depth; | |
1251 | |
1252 curlevel->prev = -1; | |
1253 curlevel->last = -1; | |
1254 | |
1255 /* Enter the loop at a place appropriate for initial state. */ | |
1256 | |
1257 if (state.incomment) goto startincomment; | |
1258 if (state.instring >= 0) | |
1259 { | |
1260 if (start_quoted) goto startquotedinstring; | |
1261 goto startinstring; | |
1262 } | |
1263 if (start_quoted) goto startquoted; | |
1264 | |
1265 while (from < end) | |
1266 { | |
1267 QUIT; | |
1268 | |
1269 code = SYNTAX (table, BUF_FETCH_CHAR (buf, from)); | |
1270 from++; | |
1271 | |
1272 if (code == Scomment) | |
1273 { | |
1274 /* record the comment style we have entered so that only the | |
1275 comment-ender sequence (or single char) of the same style | |
1276 actually terminates the comment section. */ | |
1277 mask = SYNTAX_COMMENT_1CHAR_MASK (table, | |
1278 BUF_FETCH_CHAR (buf, from-1)); | |
1279 state.comstyle = (mask == SYNTAX_COMMENT_STYLE_B); | |
1280 state.comstart = from - 1; | |
1281 } | |
1282 | |
1283 else if (from < end && | |
1284 SYNTAX_START_P (table, BUF_FETCH_CHAR (buf, from-1), | |
1285 BUF_FETCH_CHAR (buf, from))) | |
1286 { | |
1287 /* Record the comment style we have entered so that only | |
1288 the comment-end sequence of the same style actually | |
1289 terminates the comment section. */ | |
1290 code = Scomment; | |
1291 mask = SYNTAX_COMMENT_MASK_START (table, | |
1292 BUF_FETCH_CHAR (buf, from-1), | |
1293 BUF_FETCH_CHAR (buf, from)); | |
1294 state.comstyle = (mask == SYNTAX_COMMENT_STYLE_B); | |
1295 state.comstart = from-1; | |
1296 from++; | |
1297 } | |
1298 | |
1299 if (SYNTAX_PREFIX (table, BUF_FETCH_CHAR (buf, from - 1))) | |
1300 continue; | |
1301 switch (code) | |
1302 { | |
1303 case Sescape: | |
1304 case Scharquote: | |
1305 if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
1306 curlevel->last = from - 1; | |
1307 startquoted: | |
1308 if (from == end) goto endquoted; | |
1309 from++; | |
1310 goto symstarted; | |
1311 /* treat following character as a word constituent */ | |
1312 case Sword: | |
1313 case Sextword: | |
1314 case Ssymbol: | |
1315 if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
1316 curlevel->last = from - 1; | |
1317 symstarted: | |
1318 while (from < end) | |
1319 { | |
1320 switch (SYNTAX (table, BUF_FETCH_CHAR (buf, from))) | |
1321 { | |
1322 case Scharquote: | |
1323 case Sescape: | |
1324 from++; | |
1325 if (from == end) goto endquoted; | |
1326 break; | |
1327 case Sword: | |
1328 case Sextword: | |
1329 case Ssymbol: | |
1330 case Squote: | |
1331 break; | |
1332 default: | |
1333 goto symdone; | |
1334 } | |
1335 from++; | |
1336 } | |
1337 symdone: | |
1338 curlevel->prev = curlevel->last; | |
1339 break; | |
1340 | |
1341 case Scomment: | |
1342 state.incomment = 1; | |
1343 startincomment: | |
1344 if (commentstop) | |
1345 goto done; | |
1346 { | |
1347 Bufpos newfrom = find_end_of_comment (buf, from, end, mask); | |
1348 if (newfrom < 0) | |
1349 { | |
1350 /* we terminated search because from == end */ | |
1351 from = end; | |
1352 goto done; | |
1353 } | |
1354 from = newfrom; | |
1355 } | |
1356 state.incomment = 0; | |
1357 state.comstyle = 0; /* reset the comment style */ | |
1358 mask = 0; | |
1359 break; | |
1360 | |
1361 case Sopen: | |
1362 if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
1363 depth++; | |
1364 /* curlevel++->last ran into compiler bug on Apollo */ | |
1365 curlevel->last = from - 1; | |
1366 if (++curlevel == endlevel) | |
1367 error ("Nesting too deep for parser"); | |
1368 curlevel->prev = -1; | |
1369 curlevel->last = -1; | |
1370 if (!--targetdepth) goto done; | |
1371 break; | |
1372 | |
1373 case Sclose: | |
1374 depth--; | |
1375 if (depth < mindepth) | |
1376 mindepth = depth; | |
1377 if (curlevel != levelstart) | |
1378 curlevel--; | |
1379 curlevel->prev = curlevel->last; | |
1380 if (!++targetdepth) goto done; | |
1381 break; | |
1382 | |
1383 case Sstring: | |
1384 { | |
1385 Emchar ch; | |
1386 if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
1387 curlevel->last = from - 1; | |
1388 /* XEmacs change: call syntax_match() on character */ | |
1389 ch = BUF_FETCH_CHAR (buf, from - 1); | |
1390 { | |
1391 Lisp_Object stermobj = syntax_match (table, ch); | |
1392 | |
1393 if (CHARP (stermobj)) | |
1394 state.instring = XCHAR (stermobj); | |
1395 else | |
1396 state.instring = ch; | |
1397 } | |
1398 } | |
1399 startinstring: | |
1400 while (1) | |
1401 { | |
1402 if (from >= end) goto done; | |
1403 if (BUF_FETCH_CHAR (buf, from) == state.instring) break; | |
1404 switch (SYNTAX (table, BUF_FETCH_CHAR (buf, from))) | |
1405 { | |
1406 case Scharquote: | |
1407 case Sescape: | |
1408 { | |
1409 from++; | |
1410 startquotedinstring: | |
1411 if (from >= end) goto endquoted; | |
1412 break; | |
1413 } | |
1414 default: | |
1415 break; | |
1416 } | |
1417 from++; | |
1418 } | |
1419 state.instring = -1; | |
1420 curlevel->prev = curlevel->last; | |
1421 from++; | |
1422 break; | |
1423 | |
1424 case Smath: | |
1425 break; | |
1426 | |
1427 case Swhitespace: | |
1428 case Spunct: | |
1429 case Squote: | |
1430 case Sendcomment: | |
1431 case Sinherit: | |
1432 case Smax: | |
1433 break; | |
1434 } | |
1435 } | |
1436 goto done; | |
1437 | |
1438 stop: /* Here if stopping before start of sexp. */ | |
1439 from--; /* We have just fetched the char that starts it; */ | |
1440 goto done; /* but return the position before it. */ | |
1441 | |
1442 endquoted: | |
1443 state.quoted = 1; | |
1444 done: | |
1445 state.depth = depth; | |
1446 state.mindepth = mindepth; | |
1447 state.thislevelstart = curlevel->prev; | |
1448 state.prevlevelstart | |
1449 = (curlevel == levelstart) ? -1 : (curlevel - 1)->last; | |
1450 state.location = from; | |
1451 | |
1452 *stateptr = state; | |
1453 } | |
1454 | |
1455 DEFUN ("parse-partial-sexp", Fparse_partial_sexp, Sparse_partial_sexp, 2, 7, 0 /* | |
1456 Parse Lisp syntax starting at FROM until TO; return status of parse at TO. | |
1457 Parsing stops at TO or when certain criteria are met; | |
1458 point is set to where parsing stops. | |
1459 If fifth arg STATE is omitted or nil, | |
1460 parsing assumes that FROM is the beginning of a function. | |
1461 Value is a list of eight elements describing final state of parsing: | |
1462 0. depth in parens. | |
1463 1. character address of start of innermost containing list; nil if none. | |
1464 2. character address of start of last complete sexp terminated. | |
1465 3. non-nil if inside a string. | |
1466 (It is the character that will terminate the string.) | |
1467 4. t if inside a comment. | |
1468 5. t if following a quote character. | |
1469 6. the minimum paren-depth encountered during this scan. | |
1470 7. nil if in comment style a, or not in a comment; t if in comment style b | |
1471 If third arg TARGETDEPTH is non-nil, parsing stops if the depth | |
1472 in parentheses becomes equal to TARGETDEPTH. | |
1473 Fourth arg STOPBEFORE non-nil means stop when come to | |
1474 any character that starts a sexp. | |
1475 Fifth arg STATE is an eight-element list like what this function returns. | |
1476 It is used to initialize the state of the parse. Its second and third | |
1477 elements are ignored. | |
1478 Sixth arg COMMENTSTOP non-nil means stop at the start of a comment. | |
1479 */ ) | |
1480 (from, to, targetdepth, stopbefore, oldstate, commentstop, buffer) | |
1481 Lisp_Object from, to, targetdepth, stopbefore, oldstate, commentstop, buffer; | |
1482 { | |
1483 struct lisp_parse_state state; | |
1484 int target; | |
1485 Bufpos start, end; | |
1486 struct buffer *buf = decode_buffer (buffer, 0); | |
1487 | |
1488 if (!NILP (targetdepth)) | |
1489 { | |
1490 CHECK_INT (targetdepth); | |
1491 target = XINT (targetdepth); | |
1492 } | |
1493 else | |
1494 target = -100000; /* We won't reach this depth */ | |
1495 | |
1496 get_buffer_range_char (buf, from, to, &start, &end, 0); | |
1497 scan_sexps_forward (buf, &state, start, end, | |
1498 target, !NILP (stopbefore), oldstate, | |
1499 !NILP (commentstop)); | |
1500 | |
1501 BUF_SET_PT (buf, state.location); | |
1502 | |
1503 { | |
1504 /* | |
1505 * This junk is necessary because of a bug in SparcWorks cc 2.0.1. It | |
1506 * doesn't handle functions as arguments to other functions very well. | |
1507 */ | |
1508 Lisp_Object retval[8]; | |
1509 | |
1510 retval[0] = make_int (state.depth); | |
1511 retval[1] = ((state.prevlevelstart < 0) ? Qnil : | |
1512 make_int (state.prevlevelstart)); | |
1513 retval[2] = ((state.thislevelstart < 0) ? Qnil : | |
1514 make_int (state.thislevelstart)); | |
1515 retval[3] = ((state.instring >= 0) ? make_int (state.instring) : Qnil); | |
1516 retval[4] = ((state.incomment) ? Qt : Qnil); | |
1517 retval[5] = ((state.quoted) ? Qt : Qnil); | |
1518 retval[6] = make_int (state.mindepth); | |
1519 retval[7] = ((state.comstyle) ? Qt : Qnil); | |
1520 | |
1521 return (Flist (8, retval)); | |
1522 } | |
1523 } | |
1524 | |
1525 | |
1526 /************************************************************************/ | |
1527 /* initialization */ | |
1528 /************************************************************************/ | |
1529 | |
1530 void | |
1531 syms_of_syntax (void) | |
1532 { | |
1533 defsymbol (&Qsyntax_table_p, "syntax-table-p"); | |
1534 | |
1535 defsubr (&Ssyntax_table_p); | |
1536 defsubr (&Ssyntax_table); | |
1537 defsubr (&Sstandard_syntax_table); | |
1538 defsubr (&Scopy_syntax_table); | |
1539 defsubr (&Sset_syntax_table); | |
1540 defsubr (&Ssyntax_designator_chars); | |
1541 defsubr (&Schar_syntax); | |
1542 defsubr (&Smatching_paren); | |
1543 /* defsubr (&Smodify_syntax_entry); now in Lisp. */ | |
1544 /* defsubr (&Sdescribe_syntax); now in Lisp. */ | |
1545 | |
1546 defsubr (&Sforward_word); | |
1547 | |
1548 defsubr (&Sforward_comment); | |
1549 defsubr (&Sscan_lists); | |
1550 defsubr (&Sscan_sexps); | |
1551 defsubr (&Sbackward_prefix_chars); | |
1552 defsubr (&Sparse_partial_sexp); | |
1553 } | |
1554 | |
1555 void | |
1556 vars_of_syntax (void) | |
1557 { | |
1558 DEFVAR_BOOL ("parse-sexp-ignore-comments", &parse_sexp_ignore_comments /* | |
1559 Non-nil means `forward-sexp', etc., should treat comments as whitespace. | |
1560 */ ); | |
1561 | |
1562 words_include_escapes = 0; | |
1563 DEFVAR_BOOL ("words-include-escapes", &words_include_escapes /* | |
1564 Non-nil means `forward-word', etc., should treat escape chars part of words. | |
1565 */ ); | |
1566 | |
1567 no_quit_in_re_search = 0; | |
1568 } | |
1569 | |
1570 void | |
1571 complex_vars_of_syntax (void) | |
1572 { | |
1573 struct Lisp_Vector *v; | |
1574 int i; | |
1575 | |
1576 /* Set this now, so first buffer creation can refer to it. */ | |
1577 /* Make it nil before calling copy-syntax-table | |
1578 so that copy-syntax-table will know not to try to copy from garbage */ | |
1579 Vstandard_syntax_table = Qnil; | |
1580 Vstandard_syntax_table = Fcopy_syntax_table (Qnil); | |
1581 staticpro (&Vstandard_syntax_table); | |
1582 | |
1583 Vsyntax_designator_chars_string = make_pure_string (syntax_code_spec, | |
1584 Smax, Qnil, 1); | |
1585 staticpro (&Vsyntax_designator_chars_string); | |
1586 | |
1587 v = XVECTOR (Vstandard_syntax_table); | |
1588 | |
1589 for (i = 'a'; i <= 'z'; i++) | |
1590 v->contents[i] = make_int ((int) Sword); | |
1591 for (i = 'A'; i <= 'Z'; i++) | |
1592 v->contents[i] = make_int ((int) Sword); | |
1593 for (i = '0'; i <= '9'; i++) | |
1594 v->contents[i] = make_int ((int) Sword); | |
1595 v->contents['$'] = make_int ((int) Sword); | |
1596 v->contents['%'] = make_int ((int) Sword); | |
1597 | |
1598 v->contents['('] = make_int ((int) Sopen + (')' << 8)); | |
1599 v->contents[')'] = make_int ((int) Sclose + ('(' << 8)); | |
1600 v->contents['['] = make_int ((int) Sopen + (']' << 8)); | |
1601 v->contents[']'] = make_int ((int) Sclose + ('[' << 8)); | |
1602 v->contents['{'] = make_int ((int) Sopen + ('}' << 8)); | |
1603 v->contents['}'] = make_int ((int) Sclose + ('{' << 8)); | |
1604 v->contents['"'] = make_int ((int) Sstring); | |
1605 v->contents['\\'] = make_int ((int) Sescape); | |
1606 | |
1607 { | |
1608 CONST char *p; | |
1609 | |
1610 for (p = "_-+*/&|<>="; *p; p++) | |
1611 v->contents[(int) *p] = make_int ((int) Ssymbol); | |
1612 | |
1613 for (p = ".,;:?!#@~^'`"; *p; p++) | |
1614 v->contents[(int) *p] = make_int ((int) Spunct); | |
1615 } | |
1616 } |