Commit | Line | Data |
---|---|---|
8489eb67 | 1 | /* GNU Emacs routines to deal with syntax tables; also word and list parsing. |
429ab54e | 2 | Copyright (C) 1985, 1987, 1993, 1994, 1995, 1997, 1998, 1999, 2001, |
8cabe764 | 3 | 2002, 2003, 2004, 2005, 2006, 2007, 2008 |
429ab54e | 4 | Free Software Foundation, Inc. |
8489eb67 RS |
5 | |
6 | This file is part of GNU Emacs. | |
7 | ||
9ec0b715 | 8 | GNU Emacs is free software: you can redistribute it and/or modify |
8489eb67 | 9 | it under the terms of the GNU General Public License as published by |
9ec0b715 GM |
10 | the Free Software Foundation, either version 3 of the License, or |
11 | (at your option) any later version. | |
8489eb67 RS |
12 | |
13 | GNU Emacs is distributed in the hope that it will be useful, | |
14 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
15 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
16 | GNU General Public License for more details. | |
17 | ||
18 | You should have received a copy of the GNU General Public License | |
9ec0b715 | 19 | along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>. */ |
8489eb67 RS |
20 | |
21 | ||
18160b98 | 22 | #include <config.h> |
8489eb67 RS |
23 | #include <ctype.h> |
24 | #include "lisp.h" | |
25 | #include "commands.h" | |
26 | #include "buffer.h" | |
5c7b02ab | 27 | #include "character.h" |
e35f6ff7 | 28 | #include "keymap.h" |
a1bc88d4 | 29 | #include "regex.h" |
195d1361 RS |
30 | |
31 | /* Make syntax table lookup grant data in gl_state. */ | |
32 | #define SYNTAX_ENTRY_VIA_PROPERTY | |
33 | ||
8489eb67 | 34 | #include "syntax.h" |
195d1361 RS |
35 | #include "intervals.h" |
36 | ||
37 | /* We use these constants in place for comment-style and | |
38 | string-ender-char to distinguish comments/strings started by | |
39 | comment_fence and string_fence codes. */ | |
40 | ||
41 | #define ST_COMMENT_STYLE (256 + 1) | |
42 | #define ST_STRING_STYLE (256 + 2) | |
93da5fff | 43 | #include "category.h" |
8489eb67 | 44 | |
7bf5e9e4 | 45 | Lisp_Object Qsyntax_table_p, Qsyntax_table, Qscan_error; |
8489eb67 RS |
46 | |
47 | int words_include_escapes; | |
195d1361 | 48 | int parse_sexp_lookup_properties; |
8489eb67 | 49 | |
bd25db08 KH |
50 | /* Nonzero means `scan-sexps' treat all multibyte characters as symbol. */ |
51 | int multibyte_syntax_as_symbol; | |
52 | ||
8ea151b2 RS |
53 | /* Used as a temporary in SYNTAX_ENTRY and other macros in syntax.h, |
54 | if not compiled with GCC. No need to mark it, since it is used | |
55 | only very temporarily. */ | |
56 | Lisp_Object syntax_temp; | |
57 | ||
f4ed767f GM |
58 | /* Non-zero means an open parenthesis in column 0 is always considered |
59 | to be the start of a defun. Zero means an open parenthesis in | |
60 | column 0 has no special meaning. */ | |
61 | ||
62 | int open_paren_in_column_0_is_defun_start; | |
63 | ||
e5d4f4dc RS |
64 | /* This is the internal form of the parse state used in parse-partial-sexp. */ |
65 | ||
66 | struct lisp_parse_state | |
67 | { | |
195d1361 RS |
68 | int depth; /* Depth at end of parsing. */ |
69 | int instring; /* -1 if not within string, else desired terminator. */ | |
95ff8dfc | 70 | int incomment; /* -1 if in unnestable comment else comment nesting */ |
195d1361 | 71 | int comstyle; /* comment style a=0, or b=1, or ST_COMMENT_STYLE. */ |
e5d4f4dc RS |
72 | int quoted; /* Nonzero if just after an escape char at end of parsing */ |
73 | int thislevelstart; /* Char number of most recent start-of-expression at current level */ | |
74 | int prevlevelstart; /* Char number of start of containing expression */ | |
195d1361 | 75 | int location; /* Char number at which parsing stopped. */ |
e5d4f4dc | 76 | int mindepth; /* Minimum depth seen while scanning. */ |
195d1361 | 77 | int comstr_start; /* Position just after last comment/string starter. */ |
1a102a58 RS |
78 | Lisp_Object levelstarts; /* Char numbers of starts-of-expression |
79 | of levels (starting from outermost). */ | |
e5d4f4dc RS |
80 | }; |
81 | \f | |
37bef230 RS |
82 | /* These variables are a cache for finding the start of a defun. |
83 | find_start_pos is the place for which the defun start was found. | |
84 | find_start_value is the defun start position found for it. | |
6a140a74 | 85 | find_start_value_byte is the corresponding byte position. |
37bef230 RS |
86 | find_start_buffer is the buffer it was found in. |
87 | find_start_begv is the BEGV value when it was found. | |
88 | find_start_modiff is the value of MODIFF when it was found. */ | |
89 | ||
90 | static int find_start_pos; | |
91 | static int find_start_value; | |
6a140a74 | 92 | static int find_start_value_byte; |
37bef230 RS |
93 | static struct buffer *find_start_buffer; |
94 | static int find_start_begv; | |
95 | static int find_start_modiff; | |
6a140a74 RS |
96 | |
97 | ||
98 | static int find_defun_start P_ ((int, int)); | |
932e6895 SM |
99 | static int back_comment P_ ((EMACS_INT, EMACS_INT, EMACS_INT, int, int, |
100 | EMACS_INT *, EMACS_INT *)); | |
6a140a74 | 101 | static int char_quoted P_ ((int, int)); |
327719ee | 102 | static Lisp_Object skip_chars P_ ((int, Lisp_Object, Lisp_Object, int)); |
b7dbcc19 | 103 | static Lisp_Object skip_syntaxes P_ ((int, Lisp_Object, Lisp_Object)); |
932e6895 | 104 | static Lisp_Object scan_lists P_ ((EMACS_INT, EMACS_INT, EMACS_INT, int)); |
6a140a74 RS |
105 | static void scan_sexps_forward P_ ((struct lisp_parse_state *, |
106 | int, int, int, int, | |
107 | int, Lisp_Object, int)); | |
a1bc88d4 | 108 | static int in_classes P_ ((int, Lisp_Object)); |
195d1361 RS |
109 | \f |
110 | ||
111 | struct gl_state_s gl_state; /* Global state of syntax parser. */ | |
112 | ||
113 | INTERVAL interval_of (); | |
114 | #define INTERVALS_AT_ONCE 10 /* 1 + max-number of intervals | |
115 | to scan to property-change. */ | |
116 | ||
6a140a74 RS |
117 | /* Update gl_state to an appropriate interval which contains CHARPOS. The |
118 | sign of COUNT give the relative position of CHARPOS wrt the previously | |
195d1361 | 119 | valid interval. If INIT, only [be]_property fields of gl_state are |
6a140a74 | 120 | valid at start, the rest is filled basing on OBJECT. |
195d1361 | 121 | |
6a140a74 | 122 | `gl_state.*_i' are the intervals, and CHARPOS is further in the search |
195d1361 RS |
123 | direction than the intervals - or in an interval. We update the |
124 | current syntax-table basing on the property of this interval, and | |
6a140a74 | 125 | update the interval to start further than CHARPOS - or be |
195d1361 | 126 | NULL_INTERVAL. We also update lim_property to be the next value of |
6a140a74 | 127 | charpos to call this subroutine again - or be before/after the |
195d1361 RS |
128 | start/end of OBJECT. */ |
129 | ||
130 | void | |
6a140a74 RS |
131 | update_syntax_table (charpos, count, init, object) |
132 | int charpos, count, init; | |
195d1361 RS |
133 | Lisp_Object object; |
134 | { | |
135 | Lisp_Object tmp_table; | |
4ffe723b | 136 | int cnt = 0, invalidate = 1; |
5a774522 | 137 | INTERVAL i; |
195d1361 RS |
138 | |
139 | if (init) | |
140 | { | |
bb0de084 | 141 | gl_state.old_prop = Qnil; |
195d1361 RS |
142 | gl_state.start = gl_state.b_property; |
143 | gl_state.stop = gl_state.e_property; | |
bb0de084 SM |
144 | i = interval_of (charpos, object); |
145 | gl_state.backward_i = gl_state.forward_i = i; | |
195d1361 RS |
146 | invalidate = 0; |
147 | if (NULL_INTERVAL_P (i)) | |
148 | return; | |
f902a008 | 149 | /* interval_of updates only ->position of the return value, so |
d80f4cc7 | 150 | update the parents manually to speed up update_interval. */ |
bb0de084 | 151 | while (!NULL_PARENT (i)) |
d80f4cc7 RS |
152 | { |
153 | if (AM_RIGHT_CHILD (i)) | |
439d5cb4 | 154 | INTERVAL_PARENT (i)->position = i->position |
d80f4cc7 | 155 | - LEFT_TOTAL_LENGTH (i) + TOTAL_LENGTH (i) /* right end */ |
439d5cb4 KR |
156 | - TOTAL_LENGTH (INTERVAL_PARENT (i)) |
157 | + LEFT_TOTAL_LENGTH (INTERVAL_PARENT (i)); | |
d80f4cc7 | 158 | else |
439d5cb4 | 159 | INTERVAL_PARENT (i)->position = i->position - LEFT_TOTAL_LENGTH (i) |
d80f4cc7 | 160 | + TOTAL_LENGTH (i); |
439d5cb4 | 161 | i = INTERVAL_PARENT (i); |
d80f4cc7 RS |
162 | } |
163 | i = gl_state.forward_i; | |
bb0de084 | 164 | gl_state.b_property = i->position - gl_state.offset; |
ee0cdb48 | 165 | gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset; |
195d1361 RS |
166 | goto update; |
167 | } | |
5a774522 | 168 | i = count > 0 ? gl_state.forward_i : gl_state.backward_i; |
195d1361 | 169 | |
423e705d | 170 | /* We are guaranteed to be called with CHARPOS either in i, |
6a140a74 | 171 | or further off. */ |
195d1361 RS |
172 | if (NULL_INTERVAL_P (i)) |
173 | error ("Error in syntax_table logic for to-the-end intervals"); | |
6a140a74 | 174 | else if (charpos < i->position) /* Move left. */ |
195d1361 RS |
175 | { |
176 | if (count > 0) | |
f902a008 | 177 | error ("Error in syntax_table logic for intervals <-"); |
195d1361 | 178 | /* Update the interval. */ |
6a140a74 | 179 | i = update_interval (i, charpos); |
bb0de084 | 180 | if (INTERVAL_LAST_POS (i) != gl_state.b_property) |
195d1361 RS |
181 | { |
182 | invalidate = 0; | |
195d1361 | 183 | gl_state.forward_i = i; |
ee0cdb48 | 184 | gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset; |
195d1361 | 185 | } |
423e705d | 186 | } |
6a140a74 | 187 | else if (charpos >= INTERVAL_LAST_POS (i)) /* Move right. */ |
195d1361 RS |
188 | { |
189 | if (count < 0) | |
f902a008 | 190 | error ("Error in syntax_table logic for intervals ->"); |
195d1361 | 191 | /* Update the interval. */ |
6a140a74 | 192 | i = update_interval (i, charpos); |
bb0de084 | 193 | if (i->position != gl_state.e_property) |
195d1361 RS |
194 | { |
195 | invalidate = 0; | |
195d1361 | 196 | gl_state.backward_i = i; |
bb0de084 | 197 | gl_state.b_property = i->position - gl_state.offset; |
195d1361 RS |
198 | } |
199 | } | |
195d1361 RS |
200 | |
201 | update: | |
202 | tmp_table = textget (i->plist, Qsyntax_table); | |
203 | ||
204 | if (invalidate) | |
205 | invalidate = !EQ (tmp_table, gl_state.old_prop); /* Need to invalidate? */ | |
7d0393cf | 206 | |
423e705d SM |
207 | if (invalidate) /* Did not get to adjacent interval. */ |
208 | { /* with the same table => */ | |
209 | /* invalidate the old range. */ | |
195d1361 RS |
210 | if (count > 0) |
211 | { | |
212 | gl_state.backward_i = i; | |
bb0de084 SM |
213 | gl_state.b_property = i->position - gl_state.offset; |
214 | } | |
215 | else | |
195d1361 | 216 | { |
bb0de084 | 217 | gl_state.forward_i = i; |
ee0cdb48 | 218 | gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset; |
195d1361 RS |
219 | } |
220 | } | |
37bef230 | 221 | |
bb0de084 | 222 | if (!EQ (tmp_table, gl_state.old_prop)) |
195d1361 | 223 | { |
bb0de084 SM |
224 | gl_state.current_syntax_table = tmp_table; |
225 | gl_state.old_prop = tmp_table; | |
226 | if (EQ (Fsyntax_table_p (tmp_table), Qt)) | |
227 | { | |
228 | gl_state.use_global = 0; | |
7d0393cf | 229 | } |
bb0de084 SM |
230 | else if (CONSP (tmp_table)) |
231 | { | |
232 | gl_state.use_global = 1; | |
233 | gl_state.global_code = tmp_table; | |
234 | } | |
7d0393cf | 235 | else |
bb0de084 SM |
236 | { |
237 | gl_state.use_global = 0; | |
238 | gl_state.current_syntax_table = current_buffer->syntax_table; | |
239 | } | |
195d1361 RS |
240 | } |
241 | ||
242 | while (!NULL_INTERVAL_P (i)) | |
243 | { | |
244 | if (cnt && !EQ (tmp_table, textget (i->plist, Qsyntax_table))) | |
245 | { | |
246 | if (count > 0) | |
bb0de084 SM |
247 | { |
248 | gl_state.e_property = i->position - gl_state.offset; | |
249 | gl_state.forward_i = i; | |
250 | } | |
251 | else | |
252 | { | |
5a774522 SM |
253 | gl_state.b_property |
254 | = i->position + LENGTH (i) - gl_state.offset; | |
bb0de084 SM |
255 | gl_state.backward_i = i; |
256 | } | |
257 | return; | |
195d1361 | 258 | } |
7d0393cf | 259 | else if (cnt == INTERVALS_AT_ONCE) |
195d1361 RS |
260 | { |
261 | if (count > 0) | |
bb0de084 | 262 | { |
5a774522 SM |
263 | gl_state.e_property |
264 | = i->position + LENGTH (i) - gl_state.offset | |
265 | /* e_property at EOB is not set to ZV but to ZV+1, so that | |
266 | we can do INC(from);UPDATE_SYNTAX_TABLE_FORWARD without | |
267 | having to check eob between the two. */ | |
268 | + (NULL_INTERVAL_P (next_interval (i)) ? 1 : 0); | |
bb0de084 SM |
269 | gl_state.forward_i = i; |
270 | } | |
271 | else | |
272 | { | |
273 | gl_state.b_property = i->position - gl_state.offset; | |
274 | gl_state.backward_i = i; | |
275 | } | |
276 | return; | |
195d1361 RS |
277 | } |
278 | cnt++; | |
279 | i = count > 0 ? next_interval (i) : previous_interval (i); | |
280 | } | |
bb0de084 SM |
281 | eassert (NULL_INTERVAL_P (i)); /* This property goes to the end. */ |
282 | if (count > 0) | |
283 | gl_state.e_property = gl_state.stop; | |
284 | else | |
285 | gl_state.b_property = gl_state.start; | |
195d1361 RS |
286 | } |
287 | \f | |
6a140a74 RS |
288 | /* Returns TRUE if char at CHARPOS is quoted. |
289 | Global syntax-table data should be set up already to be good at CHARPOS | |
290 | or after. On return global syntax data is good for lookup at CHARPOS. */ | |
195d1361 RS |
291 | |
292 | static int | |
6a140a74 RS |
293 | char_quoted (charpos, bytepos) |
294 | register int charpos, bytepos; | |
195d1361 RS |
295 | { |
296 | register enum syntaxcode code; | |
297 | register int beg = BEGV; | |
298 | register int quoted = 0; | |
6a140a74 RS |
299 | int orig = charpos; |
300 | ||
edfda783 AR |
301 | #ifdef HAVE_NS |
302 | /* For some reason keeps getting called w/both 1, then segfaulting | |
303 | due to the definitions of DEC_BOTH and DEC_POS in character.h, | |
304 | which lead to decrementing below initial address and then examining | |
305 | character there. Need to investigate further.. */ | |
306 | if (charpos < 2 || bytepos < 2) | |
307 | { | |
308 | //fprintf(stderr,"Returning because charpos = %d, bytepos = %d\n",charpos, bytepos); | |
309 | return 0; | |
310 | } | |
311 | #endif | |
312 | ||
6a140a74 | 313 | DEC_BOTH (charpos, bytepos); |
195d1361 | 314 | |
709a47ce | 315 | while (charpos >= beg) |
195d1361 | 316 | { |
ab229fdd AS |
317 | int c; |
318 | ||
6a140a74 | 319 | UPDATE_SYNTAX_TABLE_BACKWARD (charpos); |
327719ee | 320 | c = FETCH_CHAR_AS_MULTIBYTE (bytepos); |
ab229fdd | 321 | code = SYNTAX (c); |
6a140a74 RS |
322 | if (! (code == Scharquote || code == Sescape)) |
323 | break; | |
324 | ||
325 | DEC_BOTH (charpos, bytepos); | |
326 | quoted = !quoted; | |
195d1361 | 327 | } |
6a140a74 RS |
328 | |
329 | UPDATE_SYNTAX_TABLE (orig); | |
195d1361 RS |
330 | return quoted; |
331 | } | |
6a140a74 RS |
332 | |
333 | /* Return the bytepos one character after BYTEPOS. | |
334 | We assume that BYTEPOS is not at the end of the buffer. */ | |
335 | ||
336 | INLINE int | |
337 | inc_bytepos (bytepos) | |
338 | int bytepos; | |
339 | { | |
ef316cf0 RS |
340 | if (NILP (current_buffer->enable_multibyte_characters)) |
341 | return bytepos + 1; | |
342 | ||
6a140a74 RS |
343 | INC_POS (bytepos); |
344 | return bytepos; | |
345 | } | |
346 | ||
347 | /* Return the bytepos one character before BYTEPOS. | |
348 | We assume that BYTEPOS is not at the start of the buffer. */ | |
349 | ||
350 | INLINE int | |
351 | dec_bytepos (bytepos) | |
352 | int bytepos; | |
353 | { | |
ef316cf0 RS |
354 | if (NILP (current_buffer->enable_multibyte_characters)) |
355 | return bytepos - 1; | |
356 | ||
6a140a74 RS |
357 | DEC_POS (bytepos); |
358 | return bytepos; | |
359 | } | |
195d1361 | 360 | \f |
7cc80f0a RS |
361 | /* Return a defun-start position before before POS and not too far before. |
362 | It should be the last one before POS, or nearly the last. | |
363 | ||
364 | When open_paren_in_column_0_is_defun_start is nonzero, | |
1fd1cc2f | 365 | only the beginning of the buffer is treated as a defun-start. |
7cc80f0a RS |
366 | |
367 | We record the information about where the scan started | |
368 | and what its result was, so that another call in the same area | |
369 | can return the same value very quickly. | |
195d1361 RS |
370 | |
371 | There is no promise at which position the global syntax data is | |
372 | valid on return from the subroutine, so the caller should explicitly | |
373 | update the global data. */ | |
37bef230 RS |
374 | |
375 | static int | |
6a140a74 RS |
376 | find_defun_start (pos, pos_byte) |
377 | int pos, pos_byte; | |
37bef230 | 378 | { |
6a140a74 | 379 | int opoint = PT, opoint_byte = PT_BYTE; |
37bef230 | 380 | |
1fd1cc2f RS |
381 | if (!open_paren_in_column_0_is_defun_start) |
382 | { | |
383 | find_start_value_byte = BEGV_BYTE; | |
384 | return BEGV; | |
385 | } | |
386 | ||
37bef230 RS |
387 | /* Use previous finding, if it's valid and applies to this inquiry. */ |
388 | if (current_buffer == find_start_buffer | |
389 | /* Reuse the defun-start even if POS is a little farther on. | |
390 | POS might be in the next defun, but that's ok. | |
391 | Our value may not be the best possible, but will still be usable. */ | |
392 | && pos <= find_start_pos + 1000 | |
393 | && pos >= find_start_value | |
394 | && BEGV == find_start_begv | |
395 | && MODIFF == find_start_modiff) | |
396 | return find_start_value; | |
397 | ||
398 | /* Back up to start of line. */ | |
6a140a74 | 399 | scan_newline (pos, pos_byte, BEGV, BEGV_BYTE, -1, 1); |
37bef230 | 400 | |
195d1361 RS |
401 | /* We optimize syntax-table lookup for rare updates. Thus we accept |
402 | only those `^\s(' which are good in global _and_ text-property | |
403 | syntax-tables. */ | |
404 | gl_state.current_syntax_table = current_buffer->syntax_table; | |
405 | gl_state.use_global = 0; | |
1fd1cc2f | 406 | while (PT > BEGV) |
37bef230 | 407 | { |
ab229fdd AS |
408 | int c; |
409 | ||
1fd1cc2f RS |
410 | /* Open-paren at start of line means we may have found our |
411 | defun-start. */ | |
327719ee | 412 | c = FETCH_CHAR_AS_MULTIBYTE (PT_BYTE); |
ab229fdd | 413 | if (SYNTAX (c) == Sopen) |
195d1361 | 414 | { |
1fd1cc2f | 415 | SETUP_SYNTAX_TABLE (PT + 1, -1); /* Try again... */ |
327719ee | 416 | c = FETCH_CHAR_AS_MULTIBYTE (PT_BYTE); |
ab229fdd | 417 | if (SYNTAX (c) == Sopen) |
1fd1cc2f RS |
418 | break; |
419 | /* Now fallback to the default value. */ | |
420 | gl_state.current_syntax_table = current_buffer->syntax_table; | |
421 | gl_state.use_global = 0; | |
195d1361 | 422 | } |
1fd1cc2f RS |
423 | /* Move to beg of previous line. */ |
424 | scan_newline (PT, PT_BYTE, BEGV, BEGV_BYTE, -2, 1); | |
37bef230 RS |
425 | } |
426 | ||
427 | /* Record what we found, for the next try. */ | |
6a140a74 RS |
428 | find_start_value = PT; |
429 | find_start_value_byte = PT_BYTE; | |
37bef230 RS |
430 | find_start_buffer = current_buffer; |
431 | find_start_modiff = MODIFF; | |
432 | find_start_begv = BEGV; | |
433 | find_start_pos = pos; | |
434 | ||
6a140a74 RS |
435 | TEMP_SET_PT_BOTH (opoint, opoint_byte); |
436 | ||
37bef230 RS |
437 | return find_start_value; |
438 | } | |
439 | \f | |
f902a008 RS |
440 | /* Return the SYNTAX_COMEND_FIRST of the character before POS, POS_BYTE. */ |
441 | ||
442 | static int | |
443 | prev_char_comend_first (pos, pos_byte) | |
444 | int pos, pos_byte; | |
445 | { | |
446 | int c, val; | |
447 | ||
448 | DEC_BOTH (pos, pos_byte); | |
449 | UPDATE_SYNTAX_TABLE_BACKWARD (pos); | |
450 | c = FETCH_CHAR (pos_byte); | |
451 | val = SYNTAX_COMEND_FIRST (c); | |
452 | UPDATE_SYNTAX_TABLE_FORWARD (pos + 1); | |
453 | return val; | |
454 | } | |
455 | ||
456 | /* Return the SYNTAX_COMSTART_FIRST of the character before POS, POS_BYTE. */ | |
457 | ||
3f679f55 SM |
458 | /* static int |
459 | * prev_char_comstart_first (pos, pos_byte) | |
460 | * int pos, pos_byte; | |
461 | * { | |
462 | * int c, val; | |
7d0393cf | 463 | * |
3f679f55 SM |
464 | * DEC_BOTH (pos, pos_byte); |
465 | * UPDATE_SYNTAX_TABLE_BACKWARD (pos); | |
466 | * c = FETCH_CHAR (pos_byte); | |
467 | * val = SYNTAX_COMSTART_FIRST (c); | |
468 | * UPDATE_SYNTAX_TABLE_FORWARD (pos + 1); | |
469 | * return val; | |
470 | * } */ | |
f902a008 | 471 | |
6a140a74 RS |
472 | /* Checks whether charpos FROM is at the end of a comment. |
473 | FROM_BYTE is the bytepos corresponding to FROM. | |
474 | Do not move back before STOP. | |
475 | ||
476 | Return a positive value if we find a comment ending at FROM/FROM_BYTE; | |
477 | return -1 otherwise. | |
478 | ||
479 | If successful, store the charpos of the comment's beginning | |
480 | into *CHARPOS_PTR, and the bytepos into *BYTEPOS_PTR. | |
527a32d9 KH |
481 | |
482 | Global syntax data remains valid for backward search starting at | |
483 | the returned value (or at FROM, if the search was not successful). */ | |
195d1361 RS |
484 | |
485 | static int | |
95ff8dfc | 486 | back_comment (from, from_byte, stop, comnested, comstyle, charpos_ptr, bytepos_ptr) |
932e6895 | 487 | EMACS_INT from, from_byte, stop; |
95ff8dfc | 488 | int comnested, comstyle; |
932e6895 | 489 | EMACS_INT *charpos_ptr, *bytepos_ptr; |
195d1361 RS |
490 | { |
491 | /* Look back, counting the parity of string-quotes, | |
492 | and recording the comment-starters seen. | |
493 | When we reach a safe place, assume that's not in a string; | |
494 | then step the main scan to the earliest comment-starter seen | |
495 | an even number of string quotes away from the safe place. | |
496 | ||
497 | OFROM[I] is position of the earliest comment-starter seen | |
498 | which is I+2X quotes from the comment-end. | |
499 | PARITY is current parity of quotes from the comment end. */ | |
3ee5041c | 500 | int string_style = -1; /* Presumed outside of any string. */ |
195d1361 | 501 | int string_lossage = 0; |
f7c436c1 | 502 | /* Not a real lossage: indicates that we have passed a matching comment |
7d0393cf | 503 | starter plus a non-matching comment-ender, meaning that any matching |
f7c436c1 SM |
504 | comment-starter we might see later could be a false positive (hidden |
505 | inside another comment). | |
506 | Test case: { a (* b } c (* d *) */ | |
507 | int comment_lossage = 0; | |
195d1361 | 508 | int comment_end = from; |
6a140a74 | 509 | int comment_end_byte = from_byte; |
195d1361 | 510 | int comstart_pos = 0; |
6a140a74 | 511 | int comstart_byte; |
eb35b628 RS |
512 | /* Place where the containing defun starts, |
513 | or 0 if we didn't come across it yet. */ | |
514 | int defun_start = 0; | |
515 | int defun_start_byte = 0; | |
195d1361 | 516 | register enum syntaxcode code; |
95ff8dfc | 517 | int nesting = 1; /* current comment nesting */ |
ea315ed6 | 518 | int c; |
3f679f55 SM |
519 | int syntax = 0; |
520 | ||
521 | /* FIXME: A }} comment-ender style leads to incorrect behavior | |
522 | in the case of {{ c }}} because we ignore the last two chars which are | |
523 | assumed to be comment-enders although they aren't. */ | |
195d1361 RS |
524 | |
525 | /* At beginning of range to scan, we're outside of strings; | |
526 | that determines quote parity to the comment-end. */ | |
527 | while (from != stop) | |
528 | { | |
3f679f55 SM |
529 | int temp_byte, prev_syntax; |
530 | int com2start, com2end; | |
6a140a74 | 531 | |
195d1361 | 532 | /* Move back and examine a character. */ |
6a140a74 | 533 | DEC_BOTH (from, from_byte); |
195d1361 RS |
534 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
535 | ||
3f679f55 | 536 | prev_syntax = syntax; |
b7dbcc19 | 537 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
3f679f55 | 538 | syntax = SYNTAX_WITH_FLAGS (c); |
195d1361 RS |
539 | code = SYNTAX (c); |
540 | ||
3f679f55 SM |
541 | /* Check for 2-char comment markers. */ |
542 | com2start = (SYNTAX_FLAGS_COMSTART_FIRST (syntax) | |
543 | && SYNTAX_FLAGS_COMSTART_SECOND (prev_syntax) | |
544 | && comstyle == SYNTAX_FLAGS_COMMENT_STYLE (prev_syntax) | |
545 | && (SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax) | |
546 | || SYNTAX_FLAGS_COMMENT_NESTED (syntax)) == comnested); | |
547 | com2end = (SYNTAX_FLAGS_COMEND_FIRST (syntax) | |
548 | && SYNTAX_FLAGS_COMEND_SECOND (prev_syntax)); | |
549 | ||
550 | /* Nasty cases with overlapping 2-char comment markers: | |
551 | - snmp-mode: -- c -- foo -- c -- | |
552 | --- c -- | |
553 | ------ c -- | |
554 | - c-mode: *||* | |
555 | |* *|* *| | |
556 | |*| |* |*| | |
557 | /// */ | |
558 | ||
559 | /* If a 2-char comment sequence partly overlaps with another, | |
560 | we don't try to be clever. */ | |
561 | if (from > stop && (com2end || com2start)) | |
195d1361 | 562 | { |
3f679f55 SM |
563 | int next = from, next_byte = from_byte, next_c, next_syntax; |
564 | DEC_BOTH (next, next_byte); | |
565 | UPDATE_SYNTAX_TABLE_BACKWARD (next); | |
b7dbcc19 | 566 | next_c = FETCH_CHAR_AS_MULTIBYTE (next_byte); |
3f679f55 SM |
567 | next_syntax = SYNTAX_WITH_FLAGS (next_c); |
568 | if (((com2start || comnested) | |
569 | && SYNTAX_FLAGS_COMEND_SECOND (syntax) | |
570 | && SYNTAX_FLAGS_COMEND_FIRST (next_syntax)) | |
571 | || ((com2end || comnested) | |
572 | && SYNTAX_FLAGS_COMSTART_SECOND (syntax) | |
573 | && comstyle == SYNTAX_FLAGS_COMMENT_STYLE (syntax) | |
574 | && SYNTAX_FLAGS_COMSTART_FIRST (next_syntax))) | |
575 | goto lossage; | |
576 | /* UPDATE_SYNTAX_TABLE_FORWARD (next + 1); */ | |
f902a008 | 577 | } |
3f679f55 SM |
578 | |
579 | if (com2start && comstart_pos == 0) | |
580 | /* We're looking at a comment starter. But it might be a comment | |
581 | ender as well (see snmp-mode). The first time we see one, we | |
582 | need to consider it as a comment starter, | |
583 | and the subsequent times as a comment ender. */ | |
584 | com2end = 0; | |
585 | ||
586 | /* Turn a 2-char comment sequences into the appropriate syntax. */ | |
587 | if (com2end) | |
588 | code = Sendcomment; | |
589 | else if (com2start) | |
590 | code = Scomment; | |
591 | /* Ignore comment starters of a different style. */ | |
592 | else if (code == Scomment | |
593 | && (comstyle != SYNTAX_FLAGS_COMMENT_STYLE (syntax) | |
594 | || SYNTAX_FLAGS_COMMENT_NESTED (syntax) != comnested)) | |
1aa963c8 | 595 | continue; |
195d1361 | 596 | |
9828a477 KH |
597 | /* Ignore escaped characters, except comment-enders. */ |
598 | if (code != Sendcomment && char_quoted (from, from_byte)) | |
195d1361 RS |
599 | continue; |
600 | ||
3ee5041c | 601 | switch (code) |
195d1361 | 602 | { |
3ee5041c SM |
603 | case Sstring_fence: |
604 | case Scomment_fence: | |
605 | c = (code == Sstring_fence ? ST_STRING_STYLE : ST_COMMENT_STYLE); | |
606 | case Sstring: | |
607 | /* Track parity of quotes. */ | |
608 | if (string_style == -1) | |
609 | /* Entering a string. */ | |
610 | string_style = c; | |
611 | else if (string_style == c) | |
612 | /* Leaving the string. */ | |
613 | string_style = -1; | |
614 | else | |
615 | /* If we have two kinds of string delimiters. | |
616 | There's no way to grok this scanning backwards. */ | |
195d1361 | 617 | string_lossage = 1; |
3ee5041c | 618 | break; |
7d0393cf | 619 | |
3ee5041c SM |
620 | case Scomment: |
621 | /* We've already checked that it is the relevant comstyle. */ | |
f7c436c1 | 622 | if (string_style != -1 || comment_lossage || string_lossage) |
3ee5041c SM |
623 | /* There are odd string quotes involved, so let's be careful. |
624 | Test case in Pascal: " { " a { " } */ | |
625 | goto lossage; | |
626 | ||
f7c436c1 SM |
627 | if (!comnested) |
628 | { | |
629 | /* Record best comment-starter so far. */ | |
630 | comstart_pos = from; | |
631 | comstart_byte = from_byte; | |
632 | } | |
633 | else if (--nesting <= 0) | |
95ff8dfc RS |
634 | /* nested comments have to be balanced, so we don't need to |
635 | keep looking for earlier ones. We use here the same (slightly | |
636 | incorrect) reasoning as below: since it is followed by uniform | |
637 | paired string quotes, this comment-start has to be outside of | |
638 | strings, else the comment-end itself would be inside a string. */ | |
639 | goto done; | |
3ee5041c SM |
640 | break; |
641 | ||
02010917 | 642 | case Sendcomment: |
3f679f55 | 643 | if (SYNTAX_FLAGS_COMMENT_STYLE (syntax) == comstyle |
d070eb22 | 644 | && ((com2end && SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax)) |
3f679f55 | 645 | || SYNTAX_FLAGS_COMMENT_NESTED (syntax)) == comnested) |
02010917 SM |
646 | /* This is the same style of comment ender as ours. */ |
647 | { | |
648 | if (comnested) | |
649 | nesting++; | |
650 | else | |
651 | /* Anything before that can't count because it would match | |
652 | this comment-ender rather than ours. */ | |
653 | from = stop; /* Break out of the loop. */ | |
654 | } | |
f7c436c1 SM |
655 | else if (comstart_pos != 0 || c != '\n') |
656 | /* We're mixing comment styles here, so we'd better be careful. | |
657 | The (comstart_pos != 0 || c != '\n') check is not quite correct | |
658 | (we should just always set comment_lossage), but removing it | |
659 | would imply that any multiline comment in C would go through | |
660 | lossage, which seems overkill. | |
661 | The failure should only happen in the rare cases such as | |
662 | { (* } *) */ | |
663 | comment_lossage = 1; | |
02010917 | 664 | break; |
195d1361 | 665 | |
02010917 SM |
666 | case Sopen: |
667 | /* Assume a defun-start point is outside of strings. */ | |
668 | if (open_paren_in_column_0_is_defun_start | |
669 | && (from == stop | |
670 | || (temp_byte = dec_bytepos (from_byte), | |
671 | FETCH_CHAR (temp_byte) == '\n'))) | |
672 | { | |
673 | defun_start = from; | |
674 | defun_start_byte = from_byte; | |
675 | from = stop; /* Break out of the loop. */ | |
676 | } | |
eb35b628 | 677 | break; |
02010917 SM |
678 | |
679 | default: | |
f7c436c1 | 680 | break; |
eb35b628 | 681 | } |
195d1361 RS |
682 | } |
683 | ||
684 | if (comstart_pos == 0) | |
685 | { | |
686 | from = comment_end; | |
6a140a74 | 687 | from_byte = comment_end_byte; |
195d1361 RS |
688 | UPDATE_SYNTAX_TABLE_FORWARD (comment_end - 1); |
689 | } | |
f7c436c1 SM |
690 | /* If comstart_pos is set and we get here (ie. didn't jump to `lossage' |
691 | or `done'), then we've found the beginning of the non-nested comment. */ | |
692 | else if (1) /* !comnested */ | |
195d1361 RS |
693 | { |
694 | from = comstart_pos; | |
6a140a74 | 695 | from_byte = comstart_byte; |
789f3320 | 696 | UPDATE_SYNTAX_TABLE_FORWARD (from - 1); |
195d1361 RS |
697 | } |
698 | else | |
699 | { | |
3ee5041c SM |
700 | struct lisp_parse_state state; |
701 | lossage: | |
195d1361 RS |
702 | /* We had two kinds of string delimiters mixed up |
703 | together. Decode this going forwards. | |
02010917 | 704 | Scan fwd from a known safe place (beginning-of-defun) |
195d1361 RS |
705 | to the one in question; this records where we |
706 | last passed a comment starter. */ | |
eb35b628 RS |
707 | /* If we did not already find the defun start, find it now. */ |
708 | if (defun_start == 0) | |
709 | { | |
710 | defun_start = find_defun_start (comment_end, comment_end_byte); | |
711 | defun_start_byte = find_start_value_byte; | |
712 | } | |
02010917 | 713 | do |
195d1361 | 714 | { |
02010917 SM |
715 | scan_sexps_forward (&state, |
716 | defun_start, defun_start_byte, | |
717 | comment_end, -10000, 0, Qnil, 0); | |
718 | defun_start = comment_end; | |
719 | if (state.incomment == (comnested ? 1 : -1) | |
720 | && state.comstyle == comstyle) | |
721 | from = state.comstr_start; | |
722 | else | |
723 | { | |
724 | from = comment_end; | |
725 | if (state.incomment) | |
726 | /* If comment_end is inside some other comment, maybe ours | |
727 | is nested, so we need to try again from within the | |
728 | surrounding comment. Example: { a (* " *) */ | |
729 | { | |
730 | /* FIXME: We should advance by one or two chars. */ | |
731 | defun_start = state.comstr_start + 2; | |
732 | defun_start_byte = CHAR_TO_BYTE (defun_start); | |
733 | } | |
734 | } | |
735 | } while (defun_start < comment_end); | |
736 | ||
6a140a74 | 737 | from_byte = CHAR_TO_BYTE (from); |
195d1361 RS |
738 | UPDATE_SYNTAX_TABLE_FORWARD (from - 1); |
739 | } | |
7d0393cf | 740 | |
95ff8dfc | 741 | done: |
6a140a74 RS |
742 | *charpos_ptr = from; |
743 | *bytepos_ptr = from_byte; | |
744 | ||
1aa963c8 | 745 | return (from == comment_end) ? -1 : from; |
195d1361 RS |
746 | } |
747 | \f | |
8489eb67 | 748 | DEFUN ("syntax-table-p", Fsyntax_table_p, Ssyntax_table_p, 1, 1, 0, |
fdb82f93 PJ |
749 | doc: /* Return t if OBJECT is a syntax table. |
750 | Currently, any char-table counts as a syntax table. */) | |
751 | (object) | |
2203e1e8 | 752 | Lisp_Object object; |
8489eb67 | 753 | { |
2203e1e8 | 754 | if (CHAR_TABLE_P (object) |
e704cb4b | 755 | && EQ (XCHAR_TABLE (object)->purpose, Qsyntax_table)) |
8489eb67 RS |
756 | return Qt; |
757 | return Qnil; | |
758 | } | |
759 | ||
8ea151b2 | 760 | static void |
8489eb67 RS |
761 | check_syntax_table (obj) |
762 | Lisp_Object obj; | |
763 | { | |
47f5f6ae KS |
764 | CHECK_TYPE (CHAR_TABLE_P (obj) && EQ (XCHAR_TABLE (obj)->purpose, Qsyntax_table), |
765 | Qsyntax_table_p, obj); | |
7d0393cf | 766 | } |
8489eb67 | 767 | |
8489eb67 | 768 | DEFUN ("syntax-table", Fsyntax_table, Ssyntax_table, 0, 0, 0, |
fdb82f93 PJ |
769 | doc: /* Return the current syntax table. |
770 | This is the one specified by the current buffer. */) | |
771 | () | |
8489eb67 RS |
772 | { |
773 | return current_buffer->syntax_table; | |
774 | } | |
775 | ||
776 | DEFUN ("standard-syntax-table", Fstandard_syntax_table, | |
777 | Sstandard_syntax_table, 0, 0, 0, | |
fdb82f93 PJ |
778 | doc: /* Return the standard syntax table. |
779 | This is the one used for new buffers. */) | |
780 | () | |
8489eb67 RS |
781 | { |
782 | return Vstandard_syntax_table; | |
783 | } | |
784 | ||
785 | DEFUN ("copy-syntax-table", Fcopy_syntax_table, Scopy_syntax_table, 0, 1, 0, | |
fdb82f93 PJ |
786 | doc: /* Construct a new syntax table and return it. |
787 | It is a copy of the TABLE, which defaults to the standard syntax table. */) | |
788 | (table) | |
8489eb67 RS |
789 | Lisp_Object table; |
790 | { | |
8ea151b2 RS |
791 | Lisp_Object copy; |
792 | ||
265a9e55 | 793 | if (!NILP (table)) |
8ea151b2 RS |
794 | check_syntax_table (table); |
795 | else | |
796 | table = Vstandard_syntax_table; | |
797 | ||
798 | copy = Fcopy_sequence (table); | |
0f867324 RS |
799 | |
800 | /* Only the standard syntax table should have a default element. | |
801 | Other syntax tables should inherit from parents instead. */ | |
802 | XCHAR_TABLE (copy)->defalt = Qnil; | |
803 | ||
804 | /* Copied syntax tables should all have parents. | |
805 | If we copied one with no parent, such as the standard syntax table, | |
806 | use the standard syntax table as the copy's parent. */ | |
807 | if (NILP (XCHAR_TABLE (copy)->parent)) | |
808 | Fset_char_table_parent (copy, Vstandard_syntax_table); | |
8ea151b2 | 809 | return copy; |
8489eb67 RS |
810 | } |
811 | ||
812 | DEFUN ("set-syntax-table", Fset_syntax_table, Sset_syntax_table, 1, 1, 0, | |
fdb82f93 PJ |
813 | doc: /* Select a new syntax table for the current buffer. |
814 | One argument, a syntax table. */) | |
815 | (table) | |
8489eb67 RS |
816 | Lisp_Object table; |
817 | { | |
4e4dfa78 | 818 | int idx; |
8ea151b2 | 819 | check_syntax_table (table); |
8489eb67 RS |
820 | current_buffer->syntax_table = table; |
821 | /* Indicate that this buffer now has a specified syntax table. */ | |
f6cd0527 GM |
822 | idx = PER_BUFFER_VAR_IDX (syntax_table); |
823 | SET_PER_BUFFER_VALUE_P (current_buffer, idx, 1); | |
8489eb67 RS |
824 | return table; |
825 | } | |
826 | \f | |
827 | /* Convert a letter which signifies a syntax code | |
828 | into the code it signifies. | |
195d1361 | 829 | This is used by modify-syntax-entry, and other things. */ |
8489eb67 RS |
830 | |
831 | unsigned char syntax_spec_code[0400] = | |
832 | { 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
833 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
834 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
835 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
195d1361 | 836 | (char) Swhitespace, (char) Scomment_fence, (char) Sstring, 0377, |
8489eb67 RS |
837 | (char) Smath, 0377, 0377, (char) Squote, |
838 | (char) Sopen, (char) Sclose, 0377, 0377, | |
839 | 0377, (char) Swhitespace, (char) Spunct, (char) Scharquote, | |
840 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
841 | 0377, 0377, 0377, 0377, | |
842 | (char) Scomment, 0377, (char) Sendcomment, 0377, | |
6cb71bf6 | 843 | (char) Sinherit, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* @, A ... */ |
8489eb67 RS |
844 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, |
845 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword, | |
846 | 0377, 0377, 0377, 0377, (char) Sescape, 0377, 0377, (char) Ssymbol, | |
847 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* `, a, ... */ | |
848 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
849 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword, | |
195d1361 | 850 | 0377, 0377, 0377, 0377, (char) Sstring_fence, 0377, 0377, 0377 |
8489eb67 RS |
851 | }; |
852 | ||
195d1361 | 853 | /* Indexed by syntax code, give the letter that describes it. */ |
8489eb67 | 854 | |
195d1361 | 855 | char syntax_code_spec[16] = |
8489eb67 | 856 | { |
195d1361 RS |
857 | ' ', '.', 'w', '_', '(', ')', '\'', '\"', '$', '\\', '/', '<', '>', '@', |
858 | '!', '|' | |
8489eb67 | 859 | }; |
93da5fff KH |
860 | |
861 | /* Indexed by syntax code, give the object (cons of syntax code and | |
862 | nil) to be stored in syntax table. Since these objects can be | |
863 | shared among syntax tables, we generate them in advance. By | |
864 | sharing objects, the function `describe-syntax' can give a more | |
865 | compact listing. */ | |
866 | static Lisp_Object Vsyntax_code_object; | |
867 | ||
8489eb67 RS |
868 | \f |
869 | DEFUN ("char-syntax", Fchar_syntax, Schar_syntax, 1, 1, 0, | |
fdb82f93 PJ |
870 | doc: /* Return the syntax code of CHARACTER, described by a character. |
871 | For example, if CHARACTER is a word constituent, | |
872 | the character `w' is returned. | |
873 | The characters that correspond to various syntax codes | |
874 | are listed in the documentation of `modify-syntax-entry'. */) | |
875 | (character) | |
2203e1e8 | 876 | Lisp_Object character; |
8489eb67 | 877 | { |
8ea151b2 | 878 | int char_int; |
195d1361 RS |
879 | gl_state.current_syntax_table = current_buffer->syntax_table; |
880 | ||
881 | gl_state.use_global = 0; | |
b7826503 | 882 | CHECK_NUMBER (character); |
2203e1e8 | 883 | char_int = XINT (character); |
8ea151b2 | 884 | return make_number (syntax_code_spec[(int) SYNTAX (char_int)]); |
beefa22e RS |
885 | } |
886 | ||
887 | DEFUN ("matching-paren", Fmatching_paren, Smatching_paren, 1, 1, 0, | |
fdb82f93 PJ |
888 | doc: /* Return the matching parenthesis of CHARACTER, or nil if none. */) |
889 | (character) | |
2203e1e8 | 890 | Lisp_Object character; |
beefa22e | 891 | { |
8ea151b2 | 892 | int char_int, code; |
195d1361 RS |
893 | gl_state.current_syntax_table = current_buffer->syntax_table; |
894 | gl_state.use_global = 0; | |
b7826503 | 895 | CHECK_NUMBER (character); |
2203e1e8 | 896 | char_int = XINT (character); |
8ea151b2 | 897 | code = SYNTAX (char_int); |
a8bd7cd8 | 898 | if (code == Sopen || code == Sclose) |
2e34157c | 899 | return SYNTAX_MATCH (char_int); |
beefa22e | 900 | return Qnil; |
8489eb67 RS |
901 | } |
902 | ||
c65adb44 | 903 | DEFUN ("string-to-syntax", Fstring_to_syntax, Sstring_to_syntax, 1, 1, 0, |
fdb82f93 PJ |
904 | doc: /* Convert a syntax specification STRING into syntax cell form. |
905 | STRING should be a string as it is allowed as argument of | |
906 | `modify-syntax-entry'. Value is the equivalent cons cell | |
bce3dc36 | 907 | \(CODE . MATCHING-CHAR) that can be used as value of a `syntax-table' |
fdb82f93 PJ |
908 | text property. */) |
909 | (string) | |
c65adb44 | 910 | Lisp_Object string; |
8489eb67 | 911 | { |
2e567bd3 | 912 | register const unsigned char *p; |
8489eb67 | 913 | register enum syntaxcode code; |
247e20a8 | 914 | int val; |
8ea151b2 | 915 | Lisp_Object match; |
8489eb67 | 916 | |
b7826503 | 917 | CHECK_STRING (string); |
8489eb67 | 918 | |
d5db4077 | 919 | p = SDATA (string); |
8489eb67 RS |
920 | code = (enum syntaxcode) syntax_spec_code[*p++]; |
921 | if (((int) code & 0377) == 0377) | |
b764a653 | 922 | error ("Invalid syntax description letter: %c", p[-1]); |
8489eb67 | 923 | |
8ea151b2 | 924 | if (code == Sinherit) |
c65adb44 | 925 | return Qnil; |
8ea151b2 RS |
926 | |
927 | if (*p) | |
d1be9f0f | 928 | { |
93da5fff | 929 | int len; |
7c19e8e1 | 930 | int character = (STRING_CHAR_AND_LENGTH |
d5db4077 | 931 | (p, SBYTES (string) - 1, len)); |
93da5fff | 932 | XSETINT (match, character); |
d1be9f0f RS |
933 | if (XFASTINT (match) == ' ') |
934 | match = Qnil; | |
93da5fff | 935 | p += len; |
d1be9f0f RS |
936 | } |
937 | else | |
8ea151b2 | 938 | match = Qnil; |
8489eb67 | 939 | |
8ea151b2 | 940 | val = (int) code; |
8489eb67 RS |
941 | while (*p) |
942 | switch (*p++) | |
943 | { | |
944 | case '1': | |
247e20a8 | 945 | val |= 1 << 16; |
8489eb67 RS |
946 | break; |
947 | ||
948 | case '2': | |
247e20a8 | 949 | val |= 1 << 17; |
8489eb67 RS |
950 | break; |
951 | ||
952 | case '3': | |
247e20a8 | 953 | val |= 1 << 18; |
8489eb67 RS |
954 | break; |
955 | ||
956 | case '4': | |
247e20a8 | 957 | val |= 1 << 19; |
8489eb67 RS |
958 | break; |
959 | ||
960 | case 'p': | |
247e20a8 | 961 | val |= 1 << 20; |
8489eb67 | 962 | break; |
e5d4f4dc RS |
963 | |
964 | case 'b': | |
247e20a8 | 965 | val |= 1 << 21; |
e5d4f4dc | 966 | break; |
95ff8dfc RS |
967 | |
968 | case 'n': | |
969 | val |= 1 << 22; | |
970 | break; | |
8489eb67 | 971 | } |
7d0393cf | 972 | |
93da5fff | 973 | if (val < XVECTOR (Vsyntax_code_object)->size && NILP (match)) |
c65adb44 | 974 | return XVECTOR (Vsyntax_code_object)->contents[val]; |
93da5fff KH |
975 | else |
976 | /* Since we can't use a shared object, let's make a new one. */ | |
c65adb44 SM |
977 | return Fcons (make_number (val), match); |
978 | } | |
979 | ||
fdb82f93 | 980 | /* I really don't know why this is interactive |
2c9e1900 | 981 | help-form should at least be made useful whilst reading the second arg. */ |
7d0393cf | 982 | DEFUN ("modify-syntax-entry", Fmodify_syntax_entry, Smodify_syntax_entry, 2, 3, |
c65adb44 | 983 | "cSet syntax for character: \nsSet syntax for %s to: ", |
a50a10a0 | 984 | doc: /* Set syntax for character CHAR according to string NEWENTRY. |
d7ee9fab | 985 | The syntax is changed only for table SYNTAX-TABLE, which defaults to |
fdb82f93 | 986 | the current buffer's syntax table. |
5c7b02ab KH |
987 | CHAR may be a cons (MIN . MAX), in which case, syntaxes of all characters |
988 | in the range MIN and MAX are changed. | |
fdb82f93 PJ |
989 | The first character of NEWENTRY should be one of the following: |
990 | Space or - whitespace syntax. w word constituent. | |
991 | _ symbol constituent. . punctuation. | |
992 | ( open-parenthesis. ) close-parenthesis. | |
993 | " string quote. \\ escape. | |
994 | $ paired delimiter. ' expression quote or prefix operator. | |
995 | < comment starter. > comment ender. | |
996 | / character-quote. @ inherit from `standard-syntax-table'. | |
997 | | generic string fence. ! generic comment fence. | |
998 | ||
999 | Only single-character comment start and end sequences are represented thus. | |
1000 | Two-character sequences are represented as described below. | |
1001 | The second character of NEWENTRY is the matching parenthesis, | |
1002 | used only if the first character is `(' or `)'. | |
1003 | Any additional characters are flags. | |
1004 | Defined flags are the characters 1, 2, 3, 4, b, p, and n. | |
a50a10a0 PJ |
1005 | 1 means CHAR is the start of a two-char comment start sequence. |
1006 | 2 means CHAR is the second character of such a sequence. | |
1007 | 3 means CHAR is the start of a two-char comment end sequence. | |
1008 | 4 means CHAR is the second character of such a sequence. | |
fdb82f93 PJ |
1009 | |
1010 | There can be up to two orthogonal comment sequences. This is to support | |
1011 | language modes such as C++. By default, all comment sequences are of style | |
1012 | a, but you can set the comment sequence style to b (on the second character | |
1013 | of a comment-start, or the first character of a comment-end sequence) using | |
1014 | this flag: | |
a50a10a0 PJ |
1015 | b means CHAR is part of comment sequence b. |
1016 | n means CHAR is part of a nestable comment sequence. | |
fdb82f93 | 1017 | |
a50a10a0 | 1018 | p means CHAR is a prefix character for `backward-prefix-chars'; |
fdb82f93 | 1019 | such characters are treated as whitespace when they occur |
a50a10a0 PJ |
1020 | between expressions. |
1021 | usage: (modify-syntax-entry CHAR NEWENTRY &optional SYNTAX-TABLE) */) | |
fdb82f93 | 1022 | (c, newentry, syntax_table) |
c65adb44 SM |
1023 | Lisp_Object c, newentry, syntax_table; |
1024 | { | |
5c7b02ab KH |
1025 | if (CONSP (c)) |
1026 | { | |
8f924df7 KH |
1027 | CHECK_CHARACTER_CAR (c); |
1028 | CHECK_CHARACTER_CDR (c); | |
5c7b02ab KH |
1029 | } |
1030 | else | |
1031 | CHECK_CHARACTER (c); | |
c65adb44 SM |
1032 | |
1033 | if (NILP (syntax_table)) | |
1034 | syntax_table = current_buffer->syntax_table; | |
1035 | else | |
1036 | check_syntax_table (syntax_table); | |
8489eb67 | 1037 | |
5c7b02ab KH |
1038 | newentry = Fstring_to_syntax (newentry); |
1039 | if (CONSP (c)) | |
1040 | SET_RAW_SYNTAX_ENTRY_RANGE (syntax_table, c, newentry); | |
1041 | else | |
1042 | SET_RAW_SYNTAX_ENTRY (syntax_table, XINT (c), newentry); | |
e5b94d44 CY |
1043 | |
1044 | /* We clear the regexp cache, since character classes can now have | |
1045 | different values from those in the compiled regexps.*/ | |
1046 | clear_regexp_cache (); | |
1047 | ||
8489eb67 RS |
1048 | return Qnil; |
1049 | } | |
1050 | \f | |
1051 | /* Dump syntax table to buffer in human-readable format */ | |
1052 | ||
62abe9cb SM |
1053 | DEFUN ("internal-describe-syntax-value", Finternal_describe_syntax_value, |
1054 | Sinternal_describe_syntax_value, 1, 1, 0, | |
1055 | doc: /* Insert a description of the internal syntax description SYNTAX at point. */) | |
1056 | (syntax) | |
1057 | Lisp_Object syntax; | |
8489eb67 RS |
1058 | { |
1059 | register enum syntaxcode code; | |
e6365855 | 1060 | char desc, start1, start2, end1, end2, prefix, comstyle, comnested; |
8489eb67 | 1061 | char str[2]; |
62abe9cb | 1062 | Lisp_Object first, match_lisp, value = syntax; |
8489eb67 | 1063 | |
8ea151b2 RS |
1064 | if (NILP (value)) |
1065 | { | |
62abe9cb SM |
1066 | insert_string ("default"); |
1067 | return syntax; | |
8ea151b2 RS |
1068 | } |
1069 | ||
908b7fea KH |
1070 | if (CHAR_TABLE_P (value)) |
1071 | { | |
62abe9cb SM |
1072 | insert_string ("deeper char-table ..."); |
1073 | return syntax; | |
908b7fea KH |
1074 | } |
1075 | ||
8ea151b2 RS |
1076 | if (!CONSP (value)) |
1077 | { | |
62abe9cb SM |
1078 | insert_string ("invalid"); |
1079 | return syntax; | |
8ea151b2 RS |
1080 | } |
1081 | ||
c1d497be KR |
1082 | first = XCAR (value); |
1083 | match_lisp = XCDR (value); | |
8ea151b2 RS |
1084 | |
1085 | if (!INTEGERP (first) || !(NILP (match_lisp) || INTEGERP (match_lisp))) | |
8489eb67 | 1086 | { |
62abe9cb SM |
1087 | insert_string ("invalid"); |
1088 | return syntax; | |
8489eb67 RS |
1089 | } |
1090 | ||
e704cb4b | 1091 | code = (enum syntaxcode) (XINT (first) & 0377); |
8ea151b2 RS |
1092 | start1 = (XINT (first) >> 16) & 1; |
1093 | start2 = (XINT (first) >> 17) & 1; | |
1094 | end1 = (XINT (first) >> 18) & 1; | |
1095 | end2 = (XINT (first) >> 19) & 1; | |
1096 | prefix = (XINT (first) >> 20) & 1; | |
1097 | comstyle = (XINT (first) >> 21) & 1; | |
e6365855 | 1098 | comnested = (XINT (first) >> 22) & 1; |
8489eb67 RS |
1099 | |
1100 | if ((int) code < 0 || (int) code >= (int) Smax) | |
1101 | { | |
1102 | insert_string ("invalid"); | |
62abe9cb | 1103 | return syntax; |
8489eb67 RS |
1104 | } |
1105 | desc = syntax_code_spec[(int) code]; | |
1106 | ||
1107 | str[0] = desc, str[1] = 0; | |
1108 | insert (str, 1); | |
1109 | ||
93da5fff KH |
1110 | if (NILP (match_lisp)) |
1111 | insert (" ", 1); | |
1112 | else | |
1113 | insert_char (XINT (match_lisp)); | |
8489eb67 | 1114 | |
8489eb67 RS |
1115 | if (start1) |
1116 | insert ("1", 1); | |
1117 | if (start2) | |
1118 | insert ("2", 1); | |
1119 | ||
1120 | if (end1) | |
1121 | insert ("3", 1); | |
1122 | if (end2) | |
1123 | insert ("4", 1); | |
1124 | ||
1125 | if (prefix) | |
1126 | insert ("p", 1); | |
e5d4f4dc RS |
1127 | if (comstyle) |
1128 | insert ("b", 1); | |
e6365855 SM |
1129 | if (comnested) |
1130 | insert ("n", 1); | |
8489eb67 RS |
1131 | |
1132 | insert_string ("\twhich means: "); | |
1133 | ||
0220c518 | 1134 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
1135 | { |
1136 | case Swhitespace: | |
1137 | insert_string ("whitespace"); break; | |
1138 | case Spunct: | |
1139 | insert_string ("punctuation"); break; | |
1140 | case Sword: | |
1141 | insert_string ("word"); break; | |
1142 | case Ssymbol: | |
1143 | insert_string ("symbol"); break; | |
1144 | case Sopen: | |
1145 | insert_string ("open"); break; | |
1146 | case Sclose: | |
1147 | insert_string ("close"); break; | |
1148 | case Squote: | |
d671382e | 1149 | insert_string ("prefix"); break; |
8489eb67 RS |
1150 | case Sstring: |
1151 | insert_string ("string"); break; | |
1152 | case Smath: | |
1153 | insert_string ("math"); break; | |
1154 | case Sescape: | |
1155 | insert_string ("escape"); break; | |
1156 | case Scharquote: | |
1157 | insert_string ("charquote"); break; | |
1158 | case Scomment: | |
1159 | insert_string ("comment"); break; | |
1160 | case Sendcomment: | |
1161 | insert_string ("endcomment"); break; | |
d671382e SM |
1162 | case Sinherit: |
1163 | insert_string ("inherit"); break; | |
1164 | case Scomment_fence: | |
1165 | insert_string ("comment fence"); break; | |
1166 | case Sstring_fence: | |
1167 | insert_string ("string fence"); break; | |
8489eb67 RS |
1168 | default: |
1169 | insert_string ("invalid"); | |
62abe9cb | 1170 | return syntax; |
8489eb67 RS |
1171 | } |
1172 | ||
8ea151b2 | 1173 | if (!NILP (match_lisp)) |
8489eb67 RS |
1174 | { |
1175 | insert_string (", matches "); | |
8ea151b2 | 1176 | insert_char (XINT (match_lisp)); |
8489eb67 RS |
1177 | } |
1178 | ||
1179 | if (start1) | |
1180 | insert_string (",\n\t is the first character of a comment-start sequence"); | |
1181 | if (start2) | |
1182 | insert_string (",\n\t is the second character of a comment-start sequence"); | |
1183 | ||
1184 | if (end1) | |
1185 | insert_string (",\n\t is the first character of a comment-end sequence"); | |
1186 | if (end2) | |
1187 | insert_string (",\n\t is the second character of a comment-end sequence"); | |
e5d4f4dc RS |
1188 | if (comstyle) |
1189 | insert_string (" (comment style b)"); | |
e6365855 SM |
1190 | if (comnested) |
1191 | insert_string (" (nestable)"); | |
e5d4f4dc | 1192 | |
8489eb67 RS |
1193 | if (prefix) |
1194 | insert_string (",\n\t is a prefix character for `backward-prefix-chars'"); | |
1195 | ||
62abe9cb SM |
1196 | return syntax; |
1197 | } | |
8489eb67 | 1198 | \f |
195d1361 RS |
1199 | int parse_sexp_ignore_comments; |
1200 | ||
8f924df7 KH |
1201 | /* Char-table of functions that find the next or previous word |
1202 | boundary. */ | |
1203 | Lisp_Object Vfind_word_boundary_function_table; | |
869bb237 | 1204 | |
8489eb67 RS |
1205 | /* Return the position across COUNT words from FROM. |
1206 | If that many words cannot be found before the end of the buffer, return 0. | |
1207 | COUNT negative means scan backward and stop at word beginning. */ | |
1208 | ||
dfcf069d | 1209 | int |
8489eb67 RS |
1210 | scan_words (from, count) |
1211 | register int from, count; | |
1212 | { | |
1213 | register int beg = BEGV; | |
1214 | register int end = ZV; | |
6a140a74 | 1215 | register int from_byte = CHAR_TO_BYTE (from); |
93da5fff KH |
1216 | register enum syntaxcode code; |
1217 | int ch0, ch1; | |
8f924df7 | 1218 | Lisp_Object func, script, pos; |
8489eb67 RS |
1219 | |
1220 | immediate_quit = 1; | |
1221 | QUIT; | |
1222 | ||
195d1361 RS |
1223 | SETUP_SYNTAX_TABLE (from, count); |
1224 | ||
8489eb67 RS |
1225 | while (count > 0) |
1226 | { | |
1227 | while (1) | |
1228 | { | |
1229 | if (from == end) | |
1230 | { | |
1231 | immediate_quit = 0; | |
1232 | return 0; | |
1233 | } | |
195d1361 | 1234 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 1235 | ch0 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
93da5fff | 1236 | code = SYNTAX (ch0); |
6a140a74 | 1237 | INC_BOTH (from, from_byte); |
8489eb67 RS |
1238 | if (words_include_escapes |
1239 | && (code == Sescape || code == Scharquote)) | |
1240 | break; | |
1241 | if (code == Sword) | |
1242 | break; | |
8489eb67 | 1243 | } |
93da5fff KH |
1244 | /* Now CH0 is a character which begins a word and FROM is the |
1245 | position of the next character. */ | |
8f924df7 | 1246 | func = CHAR_TABLE_REF (Vfind_word_boundary_function_table, ch0); |
869bb237 | 1247 | if (! NILP (Ffboundp (func))) |
8489eb67 | 1248 | { |
869bb237 | 1249 | pos = call2 (func, make_number (from - 1), make_number (end)); |
8f924df7 KH |
1250 | if (INTEGERP (pos) && XINT (pos) > from) |
1251 | { | |
1252 | from = XINT (pos); | |
1253 | from_byte = CHAR_TO_BYTE (from); | |
1254 | } | |
8489eb67 | 1255 | } |
869bb237 | 1256 | else |
8cb8232a | 1257 | { |
8cb8232a KH |
1258 | script = CHAR_TABLE_REF (Vchar_script_table, ch0); |
1259 | while (1) | |
1260 | { | |
1261 | if (from == end) break; | |
1262 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
b7dbcc19 | 1263 | ch1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8cb8232a KH |
1264 | code = SYNTAX (ch1); |
1265 | if ((code != Sword | |
1266 | && (! words_include_escapes | |
1267 | || (code != Sescape && code != Scharquote))) | |
1268 | || ! EQ (CHAR_TABLE_REF (Vchar_script_table, ch1), script)) | |
869bb237 | 1269 | break; |
8cb8232a KH |
1270 | INC_BOTH (from, from_byte); |
1271 | ch0 = ch1; | |
1272 | } | |
8489eb67 RS |
1273 | } |
1274 | count--; | |
1275 | } | |
1276 | while (count < 0) | |
1277 | { | |
1278 | while (1) | |
1279 | { | |
1280 | if (from == beg) | |
1281 | { | |
1282 | immediate_quit = 0; | |
1283 | return 0; | |
1284 | } | |
6a140a74 | 1285 | DEC_BOTH (from, from_byte); |
195d1361 | 1286 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 1287 | ch1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
93da5fff | 1288 | code = SYNTAX (ch1); |
8489eb67 RS |
1289 | if (words_include_escapes |
1290 | && (code == Sescape || code == Scharquote)) | |
1291 | break; | |
1292 | if (code == Sword) | |
1293 | break; | |
8489eb67 | 1294 | } |
93da5fff KH |
1295 | /* Now CH1 is a character which ends a word and FROM is the |
1296 | position of it. */ | |
8f924df7 | 1297 | func = CHAR_TABLE_REF (Vfind_word_boundary_function_table, ch1); |
869bb237 | 1298 | if (! NILP (Ffboundp (func))) |
8f924df7 | 1299 | { |
869bb237 | 1300 | pos = call2 (func, make_number (from), make_number (beg)); |
8f924df7 KH |
1301 | if (INTEGERP (pos) && XINT (pos) < from) |
1302 | { | |
1303 | from = XINT (pos); | |
1304 | from_byte = CHAR_TO_BYTE (from); | |
1305 | } | |
8489eb67 | 1306 | } |
869bb237 | 1307 | else |
8489eb67 | 1308 | { |
8cb8232a KH |
1309 | script = CHAR_TABLE_REF (Vchar_script_table, ch1); |
1310 | while (1) | |
1311 | { | |
8cb8232a KH |
1312 | if (from == beg) |
1313 | break; | |
262be72a | 1314 | DEC_BOTH (from, from_byte); |
8cb8232a | 1315 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
262be72a | 1316 | ch0 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8cb8232a KH |
1317 | code = SYNTAX (ch0); |
1318 | if ((code != Sword | |
1319 | && (! words_include_escapes | |
1320 | || (code != Sescape && code != Scharquote))) | |
1321 | || ! EQ (CHAR_TABLE_REF (Vchar_script_table, ch0), script)) | |
262be72a MB |
1322 | { |
1323 | INC_BOTH (from, from_byte); | |
1324 | break; | |
1325 | } | |
8cb8232a KH |
1326 | ch1 = ch0; |
1327 | } | |
8489eb67 RS |
1328 | } |
1329 | count++; | |
1330 | } | |
1331 | ||
1332 | immediate_quit = 0; | |
1333 | ||
1334 | return from; | |
1335 | } | |
1336 | ||
1e9dbb5f | 1337 | DEFUN ("forward-word", Fforward_word, Sforward_word, 0, 1, "^p", |
fdb82f93 PJ |
1338 | doc: /* Move point forward ARG words (backward if ARG is negative). |
1339 | Normally returns t. | |
1340 | If an edge of the buffer or a field boundary is reached, point is left there | |
1341 | and the function returns nil. Field boundaries are not noticed if | |
1342 | `inhibit-field-text-motion' is non-nil. */) | |
839966f3 KH |
1343 | (arg) |
1344 | Lisp_Object arg; | |
8489eb67 | 1345 | { |
c096ae4d | 1346 | Lisp_Object tmp; |
2c6ea900 | 1347 | int orig_val, val; |
8489eb67 | 1348 | |
839966f3 KH |
1349 | if (NILP (arg)) |
1350 | XSETFASTINT (arg, 1); | |
1351 | else | |
1352 | CHECK_NUMBER (arg); | |
1353 | ||
1354 | val = orig_val = scan_words (PT, XINT (arg)); | |
2c6ea900 | 1355 | if (! orig_val) |
839966f3 | 1356 | val = XINT (arg) > 0 ? ZV : BEGV; |
5878ee6f | 1357 | |
b8855607 | 1358 | /* Avoid jumping out of an input field. */ |
c096ae4d SM |
1359 | tmp = Fconstrain_to_field (make_number (val), make_number (PT), |
1360 | Qt, Qnil, Qnil); | |
1361 | val = XFASTINT (tmp); | |
7d0393cf | 1362 | |
8489eb67 | 1363 | SET_PT (val); |
e6d8341f | 1364 | return val == orig_val ? Qt : Qnil; |
8489eb67 RS |
1365 | } |
1366 | \f | |
195d1361 RS |
1367 | Lisp_Object skip_chars (); |
1368 | ||
1369 | DEFUN ("skip-chars-forward", Fskip_chars_forward, Sskip_chars_forward, 1, 2, 0, | |
fdb82f93 PJ |
1370 | doc: /* Move point forward, stopping before a char not in STRING, or at pos LIM. |
1371 | STRING is like the inside of a `[...]' in a regular expression | |
1372 | except that `]' is never special and `\\' quotes `^', `-' or `\\' | |
1373 | (but not as the end of a range; quoting is never needed there). | |
1374 | Thus, with arg "a-zA-Z", this skips letters stopping before first nonletter. | |
1375 | With arg "^a-zA-Z", skips nonletters stopping before first letter. | |
a1bc88d4 RS |
1376 | Char classes, e.g. `[:alpha:]', are supported. |
1377 | ||
1378 | Returns the distance traveled, either zero or positive. */) | |
fdb82f93 | 1379 | (string, lim) |
195d1361 RS |
1380 | Lisp_Object string, lim; |
1381 | { | |
327719ee | 1382 | return skip_chars (1, string, lim, 1); |
195d1361 RS |
1383 | } |
1384 | ||
1385 | DEFUN ("skip-chars-backward", Fskip_chars_backward, Sskip_chars_backward, 1, 2, 0, | |
fdb82f93 PJ |
1386 | doc: /* Move point backward, stopping after a char not in STRING, or at pos LIM. |
1387 | See `skip-chars-forward' for details. | |
1388 | Returns the distance traveled, either zero or negative. */) | |
1389 | (string, lim) | |
195d1361 RS |
1390 | Lisp_Object string, lim; |
1391 | { | |
327719ee | 1392 | return skip_chars (0, string, lim, 1); |
195d1361 RS |
1393 | } |
1394 | ||
1395 | DEFUN ("skip-syntax-forward", Fskip_syntax_forward, Sskip_syntax_forward, 1, 2, 0, | |
fdb82f93 PJ |
1396 | doc: /* Move point forward across chars in specified syntax classes. |
1397 | SYNTAX is a string of syntax code characters. | |
1398 | Stop before a char whose syntax is not in SYNTAX, or at position LIM. | |
1399 | If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX. | |
1400 | This function returns the distance traveled, either zero or positive. */) | |
1401 | (syntax, lim) | |
195d1361 RS |
1402 | Lisp_Object syntax, lim; |
1403 | { | |
b7dbcc19 | 1404 | return skip_syntaxes (1, syntax, lim); |
195d1361 RS |
1405 | } |
1406 | ||
1407 | DEFUN ("skip-syntax-backward", Fskip_syntax_backward, Sskip_syntax_backward, 1, 2, 0, | |
fdb82f93 PJ |
1408 | doc: /* Move point backward across chars in specified syntax classes. |
1409 | SYNTAX is a string of syntax code characters. | |
1410 | Stop on reaching a char whose syntax is not in SYNTAX, or at position LIM. | |
1411 | If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX. | |
1412 | This function returns the distance traveled, either zero or negative. */) | |
1413 | (syntax, lim) | |
195d1361 RS |
1414 | Lisp_Object syntax, lim; |
1415 | { | |
b7dbcc19 | 1416 | return skip_syntaxes (0, syntax, lim); |
195d1361 RS |
1417 | } |
1418 | ||
6a140a74 | 1419 | static Lisp_Object |
327719ee | 1420 | skip_chars (forwardp, string, lim, handle_iso_classes) |
b7dbcc19 | 1421 | int forwardp; |
195d1361 | 1422 | Lisp_Object string, lim; |
a1bc88d4 | 1423 | int handle_iso_classes; |
195d1361 | 1424 | { |
195d1361 RS |
1425 | register unsigned int c; |
1426 | unsigned char fastmap[0400]; | |
b7dbcc19 | 1427 | /* Store the ranges of non-ASCII characters. */ |
5df4982e | 1428 | int *char_ranges; |
7e68b0ea | 1429 | int n_char_ranges = 0; |
195d1361 | 1430 | int negate = 0; |
4101e6fe | 1431 | register int i, i_byte; |
b7dbcc19 KH |
1432 | /* Set to 1 if the current buffer is multibyte and the region |
1433 | contains non-ASCII chars. */ | |
1434 | int multibyte; | |
1435 | /* Set to 1 if STRING is multibyte and it contains non-ASCII | |
1436 | chars. */ | |
1674d9a2 RS |
1437 | int string_multibyte; |
1438 | int size_byte; | |
2e567bd3 | 1439 | const unsigned char *str; |
82d497fc | 1440 | int len; |
a1bc88d4 | 1441 | Lisp_Object iso_classes; |
195d1361 | 1442 | |
b7826503 | 1443 | CHECK_STRING (string); |
a1bc88d4 | 1444 | iso_classes = Qnil; |
82d497fc | 1445 | |
195d1361 RS |
1446 | if (NILP (lim)) |
1447 | XSETINT (lim, forwardp ? ZV : BEGV); | |
1448 | else | |
b7826503 | 1449 | CHECK_NUMBER_COERCE_MARKER (lim); |
195d1361 RS |
1450 | |
1451 | /* In any case, don't allow scan outside bounds of buffer. */ | |
195d1361 RS |
1452 | if (XINT (lim) > ZV) |
1453 | XSETFASTINT (lim, ZV); | |
1454 | if (XINT (lim) < BEGV) | |
1455 | XSETFASTINT (lim, BEGV); | |
1456 | ||
b7dbcc19 | 1457 | multibyte = (!NILP (current_buffer->enable_multibyte_characters) |
92eaa22e | 1458 | && (XINT (lim) - PT != CHAR_TO_BYTE (XINT (lim)) - PT_BYTE)); |
8f924df7 | 1459 | string_multibyte = SBYTES (string) > SCHARS (string); |
b7dbcc19 | 1460 | |
195d1361 RS |
1461 | bzero (fastmap, sizeof fastmap); |
1462 | ||
8f924df7 KH |
1463 | str = SDATA (string); |
1464 | size_byte = SBYTES (string); | |
4101e6fe | 1465 | |
82d497fc | 1466 | i_byte = 0; |
13090112 | 1467 | if (i_byte < size_byte |
d5db4077 | 1468 | && SREF (string, 0) == '^') |
195d1361 | 1469 | { |
82d497fc | 1470 | negate = 1; i_byte++; |
195d1361 RS |
1471 | } |
1472 | ||
1473 | /* Find the characters specified and set their elements of fastmap. | |
b7dbcc19 | 1474 | Handle backslashes and ranges specially. |
195d1361 | 1475 | |
b7dbcc19 KH |
1476 | If STRING contains non-ASCII characters, setup char_ranges for |
1477 | them and use fastmap only for their leading codes. */ | |
4101e6fe | 1478 | |
b7dbcc19 | 1479 | if (! string_multibyte) |
195d1361 | 1480 | { |
b7dbcc19 | 1481 | int string_has_eight_bit = 0; |
4101e6fe | 1482 | |
b7dbcc19 KH |
1483 | /* At first setup fastmap. */ |
1484 | while (i_byte < size_byte) | |
1485 | { | |
1486 | c = str[i_byte++]; | |
1487 | ||
a1bc88d4 RS |
1488 | if (handle_iso_classes && c == '[' |
1489 | && i_byte < size_byte | |
327719ee | 1490 | && str[i_byte] == ':') |
a1bc88d4 RS |
1491 | { |
1492 | const unsigned char *class_beg = str + i_byte + 1; | |
1493 | const unsigned char *class_end = class_beg; | |
b3bda4fd | 1494 | const unsigned char *class_limit = str + size_byte - 2; |
db9cd97a | 1495 | /* Leave room for the null. */ |
a1bc88d4 RS |
1496 | unsigned char class_name[CHAR_CLASS_MAX_LENGTH + 1]; |
1497 | re_wctype_t cc; | |
1498 | ||
1499 | if (class_limit - class_beg > CHAR_CLASS_MAX_LENGTH) | |
1500 | class_limit = class_beg + CHAR_CLASS_MAX_LENGTH; | |
1501 | ||
b3bda4fd KS |
1502 | while (class_end < class_limit |
1503 | && *class_end >= 'a' && *class_end <= 'z') | |
a1bc88d4 RS |
1504 | class_end++; |
1505 | ||
b3bda4fd KS |
1506 | if (class_end == class_beg |
1507 | || *class_end != ':' || class_end[1] != ']') | |
1508 | goto not_a_class_name; | |
a1bc88d4 RS |
1509 | |
1510 | bcopy (class_beg, class_name, class_end - class_beg); | |
1511 | class_name[class_end - class_beg] = 0; | |
1512 | ||
1513 | cc = re_wctype (class_name); | |
1514 | if (cc == 0) | |
1515 | error ("Invalid ISO C character class"); | |
1516 | ||
1517 | iso_classes = Fcons (make_number (cc), iso_classes); | |
1518 | ||
1519 | i_byte = class_end + 2 - str; | |
1520 | continue; | |
1521 | } | |
1522 | ||
b3bda4fd | 1523 | not_a_class_name: |
b7dbcc19 KH |
1524 | if (c == '\\') |
1525 | { | |
13090112 | 1526 | if (i_byte == size_byte) |
4101e6fe RS |
1527 | break; |
1528 | ||
5c7b02ab | 1529 | c = str[i_byte++]; |
195d1361 | 1530 | } |
839966f3 KH |
1531 | /* Treat `-' as range character only if another character |
1532 | follows. */ | |
1533 | if (i_byte + 1 < size_byte | |
82d497fc | 1534 | && str[i_byte] == '-') |
195d1361 | 1535 | { |
9690d026 | 1536 | unsigned int c2; |
4101e6fe RS |
1537 | |
1538 | /* Skip over the dash. */ | |
82d497fc | 1539 | i_byte++; |
4101e6fe | 1540 | |
4101e6fe | 1541 | /* Get the end of the range. */ |
5c7b02ab | 1542 | c2 = str[i_byte++]; |
b7dbcc19 KH |
1543 | if (c2 == '\\' |
1544 | && i_byte < size_byte) | |
1545 | c2 = str[i_byte++]; | |
1546 | ||
8f924df7 KH |
1547 | if (c <= c2) |
1548 | { | |
1549 | while (c <= c2) | |
1550 | fastmap[c++] = 1; | |
1551 | if (! ASCII_CHAR_P (c2)) | |
1552 | string_has_eight_bit = 1; | |
1553 | } | |
195d1361 RS |
1554 | } |
1555 | else | |
b7dbcc19 KH |
1556 | { |
1557 | fastmap[c] = 1; | |
1558 | if (! ASCII_CHAR_P (c)) | |
1559 | string_has_eight_bit = 1; | |
1560 | } | |
1561 | } | |
1562 | ||
1563 | /* If the current range is multibyte and STRING contains | |
1564 | eight-bit chars, arrange fastmap and setup char_ranges for | |
1565 | the corresponding multibyte chars. */ | |
1566 | if (multibyte && string_has_eight_bit) | |
1567 | { | |
1568 | unsigned char fastmap2[0400]; | |
1569 | int range_start_byte, range_start_char; | |
1570 | ||
1571 | bcopy (fastmap2 + 0200, fastmap + 0200, 0200); | |
1572 | bzero (fastmap + 0200, 0200); | |
1573 | /* We are sure that this loop stops. */ | |
1574 | for (i = 0200; ! fastmap2[i]; i++); | |
1575 | c = unibyte_char_to_multibyte (i); | |
1576 | fastmap[CHAR_LEADING_CODE (c)] = 1; | |
1577 | range_start_byte = i; | |
1578 | range_start_char = c; | |
f003ca54 | 1579 | char_ranges = (int *) alloca (sizeof (int) * 128 * 2); |
b7dbcc19 KH |
1580 | for (i = 129; i < 0400; i++) |
1581 | { | |
1582 | c = unibyte_char_to_multibyte (i); | |
1583 | fastmap[CHAR_LEADING_CODE (c)] = 1; | |
1584 | if (i - range_start_byte != c - range_start_char) | |
1585 | { | |
1586 | char_ranges[n_char_ranges++] = range_start_char; | |
1587 | char_ranges[n_char_ranges++] = ((i - 1 - range_start_byte) | |
1588 | + range_start_char); | |
1589 | range_start_byte = i; | |
1590 | range_start_char = c; | |
8f924df7 | 1591 | } |
b7dbcc19 KH |
1592 | } |
1593 | char_ranges[n_char_ranges++] = range_start_char; | |
1594 | char_ranges[n_char_ranges++] = ((i - 1 - range_start_byte) | |
1595 | + range_start_char); | |
195d1361 RS |
1596 | } |
1597 | } | |
f003ca54 | 1598 | else /* STRING is multibyte */ |
b7dbcc19 | 1599 | { |
f003ca54 KH |
1600 | char_ranges = (int *) alloca (sizeof (int) * SCHARS (string) * 2); |
1601 | ||
b7dbcc19 | 1602 | while (i_byte < size_byte) |
195d1361 | 1603 | { |
b7dbcc19 | 1604 | unsigned char leading_code; |
5c7b02ab | 1605 | |
b7dbcc19 KH |
1606 | leading_code = str[i_byte]; |
1607 | c = STRING_CHAR_AND_LENGTH (str + i_byte, size_byte-i_byte, len); | |
1608 | i_byte += len; | |
5c7b02ab | 1609 | |
327719ee MB |
1610 | if (handle_iso_classes && c == '[' |
1611 | && i_byte < size_byte | |
1612 | && STRING_CHAR (str + i_byte, size_byte - i_byte) == ':') | |
1613 | { | |
1614 | const unsigned char *class_beg = str + i_byte + 1; | |
1615 | const unsigned char *class_end = class_beg; | |
1616 | const unsigned char *class_limit = str + size_byte - 2; | |
1617 | /* Leave room for the null. */ | |
1618 | unsigned char class_name[CHAR_CLASS_MAX_LENGTH + 1]; | |
1619 | re_wctype_t cc; | |
1620 | ||
1621 | if (class_limit - class_beg > CHAR_CLASS_MAX_LENGTH) | |
1622 | class_limit = class_beg + CHAR_CLASS_MAX_LENGTH; | |
1623 | ||
1624 | while (class_end < class_limit | |
1625 | && *class_end >= 'a' && *class_end <= 'z') | |
1626 | class_end++; | |
1627 | ||
1628 | if (class_end == class_beg | |
1629 | || *class_end != ':' || class_end[1] != ']') | |
1630 | goto not_a_class_name_multibyte; | |
1631 | ||
1632 | bcopy (class_beg, class_name, class_end - class_beg); | |
1633 | class_name[class_end - class_beg] = 0; | |
1634 | ||
1635 | cc = re_wctype (class_name); | |
1636 | if (cc == 0) | |
1637 | error ("Invalid ISO C character class"); | |
1638 | ||
1639 | iso_classes = Fcons (make_number (cc), iso_classes); | |
1640 | ||
1641 | i_byte = class_end + 2 - str; | |
1642 | continue; | |
1643 | } | |
1644 | ||
1645 | not_a_class_name_multibyte: | |
195d1361 RS |
1646 | if (c == '\\') |
1647 | { | |
13090112 | 1648 | if (i_byte == size_byte) |
4101e6fe RS |
1649 | break; |
1650 | ||
b7dbcc19 | 1651 | leading_code = str[i_byte]; |
839966f3 KH |
1652 | c = STRING_CHAR_AND_LENGTH (str + i_byte, |
1653 | size_byte - i_byte, len); | |
82d497fc | 1654 | i_byte += len; |
195d1361 | 1655 | } |
839966f3 KH |
1656 | /* Treat `-' as range character only if another character |
1657 | follows. */ | |
1658 | if (i_byte + 1 < size_byte | |
82d497fc | 1659 | && str[i_byte] == '-') |
195d1361 | 1660 | { |
9690d026 | 1661 | unsigned int c2; |
b7dbcc19 | 1662 | unsigned char leading_code2; |
4101e6fe RS |
1663 | |
1664 | /* Skip over the dash. */ | |
82d497fc | 1665 | i_byte++; |
4101e6fe | 1666 | |
4101e6fe | 1667 | /* Get the end of the range. */ |
b7dbcc19 | 1668 | leading_code2 = str[i_byte]; |
839966f3 KH |
1669 | c2 = STRING_CHAR_AND_LENGTH (str + i_byte, |
1670 | size_byte - i_byte, len); | |
82d497fc | 1671 | i_byte += len; |
7e68b0ea | 1672 | |
b7dbcc19 KH |
1673 | if (c2 == '\\' |
1674 | && i_byte < size_byte) | |
1675 | { | |
1676 | leading_code2 = str[i_byte]; | |
1677 | c2 =STRING_CHAR_AND_LENGTH (str + i_byte, size_byte-i_byte, len); | |
1678 | i_byte += len; | |
1679 | } | |
1680 | ||
f003ca54 KH |
1681 | if (c > c2) |
1682 | continue; | |
b7dbcc19 | 1683 | if (ASCII_CHAR_P (c)) |
06274af5 | 1684 | { |
5c7b02ab KH |
1685 | while (c <= c2 && c < 0x80) |
1686 | fastmap[c++] = 1; | |
b7dbcc19 KH |
1687 | leading_code = CHAR_LEADING_CODE (c); |
1688 | } | |
1689 | if (! ASCII_CHAR_P (c)) | |
1690 | { | |
1691 | while (leading_code <= leading_code2) | |
1692 | fastmap[leading_code++] = 1; | |
1693 | if (c <= c2) | |
e39091b6 | 1694 | { |
b7dbcc19 | 1695 | char_ranges[n_char_ranges++] = c; |
e39091b6 | 1696 | char_ranges[n_char_ranges++] = c2; |
06274af5 KH |
1697 | } |
1698 | } | |
195d1361 RS |
1699 | } |
1700 | else | |
7e68b0ea | 1701 | { |
b7dbcc19 | 1702 | if (ASCII_CHAR_P (c)) |
e39091b6 KH |
1703 | fastmap[c] = 1; |
1704 | else | |
7e68b0ea | 1705 | { |
b7dbcc19 | 1706 | fastmap[leading_code] = 1; |
4101e6fe RS |
1707 | char_ranges[n_char_ranges++] = c; |
1708 | char_ranges[n_char_ranges++] = c; | |
7e68b0ea RS |
1709 | } |
1710 | } | |
195d1361 | 1711 | } |
b7dbcc19 KH |
1712 | |
1713 | /* If the current range is unibyte and STRING contains non-ASCII | |
1714 | chars, arrange fastmap for the corresponding unibyte | |
1715 | chars. */ | |
1716 | ||
1717 | if (! multibyte && n_char_ranges > 0) | |
1718 | { | |
1719 | bzero (fastmap + 0200, 0200); | |
1720 | for (i = 0; i < n_char_ranges; i += 2) | |
1721 | { | |
1722 | int c1 = char_ranges[i]; | |
1723 | int c2 = char_ranges[i + 1]; | |
1724 | ||
1725 | for (; c1 <= c2; c1++) | |
2afc21f5 SM |
1726 | { |
1727 | int b = CHAR_TO_BYTE_SAFE (c1); | |
1728 | if (b >= 0) | |
1729 | fastmap[b] = 1; | |
1730 | } | |
b7dbcc19 KH |
1731 | } |
1732 | } | |
195d1361 RS |
1733 | } |
1734 | ||
9690d026 | 1735 | /* If ^ was the first character, complement the fastmap. */ |
195d1361 | 1736 | if (negate) |
b7dbcc19 KH |
1737 | { |
1738 | if (! multibyte) | |
1739 | for (i = 0; i < sizeof fastmap; i++) | |
1740 | fastmap[i] ^= 1; | |
1741 | else | |
1742 | { | |
1743 | for (i = 0; i < 0200; i++) | |
1744 | fastmap[i] ^= 1; | |
1745 | /* All non-ASCII chars possibly match. */ | |
1746 | for (; i < sizeof fastmap; i++) | |
1747 | fastmap[i] = 1; | |
1748 | } | |
1749 | } | |
195d1361 RS |
1750 | |
1751 | { | |
1752 | int start_point = PT; | |
1753 | int pos = PT; | |
6a140a74 | 1754 | int pos_byte = PT_BYTE; |
9af7511a KH |
1755 | unsigned char *p = PT_ADDR, *endp, *stop; |
1756 | ||
1757 | if (forwardp) | |
1758 | { | |
1759 | endp = (XINT (lim) == GPT) ? GPT_ADDR : CHAR_POS_ADDR (XINT (lim)); | |
d7ee9fab | 1760 | stop = (pos < GPT && GPT < XINT (lim)) ? GPT_ADDR : endp; |
9af7511a KH |
1761 | } |
1762 | else | |
1763 | { | |
1764 | endp = CHAR_POS_ADDR (XINT (lim)); | |
d7ee9fab | 1765 | stop = (pos >= GPT && GPT > XINT (lim)) ? GAP_END_ADDR : endp; |
9af7511a | 1766 | } |
195d1361 RS |
1767 | |
1768 | immediate_quit = 1; | |
b7dbcc19 | 1769 | if (forwardp) |
195d1361 | 1770 | { |
b7dbcc19 | 1771 | if (multibyte) |
8f924df7 | 1772 | while (1) |
b7dbcc19 | 1773 | { |
8f924df7 | 1774 | int nbytes; |
9af7511a | 1775 | |
8f924df7 | 1776 | if (p >= stop) |
9af7511a | 1777 | { |
8f924df7 | 1778 | if (p >= endp) |
9af7511a | 1779 | break; |
8f924df7 KH |
1780 | p = GAP_END_ADDR; |
1781 | stop = endp; | |
9af7511a | 1782 | } |
327719ee MB |
1783 | c = STRING_CHAR_AND_LENGTH (p, MAX_MULTIBYTE_LENGTH, nbytes); |
1784 | if (! NILP (iso_classes) && in_classes (c, iso_classes)) | |
9af7511a | 1785 | { |
327719ee | 1786 | if (negate) |
9af7511a | 1787 | break; |
327719ee MB |
1788 | else |
1789 | goto fwd_ok; | |
9af7511a | 1790 | } |
9af7511a | 1791 | |
8f924df7 | 1792 | if (! fastmap[*p]) |
b7dbcc19 KH |
1793 | break; |
1794 | if (! ASCII_CHAR_P (c)) | |
9af7511a | 1795 | { |
b7dbcc19 KH |
1796 | /* As we are looking at a multibyte character, we |
1797 | must look up the character in the table | |
1798 | CHAR_RANGES. If there's no data in the table, | |
1799 | that character is not what we want to skip. */ | |
1800 | ||
1801 | /* The following code do the right thing even if | |
1802 | n_char_ranges is zero (i.e. no data in | |
1803 | CHAR_RANGES). */ | |
1804 | for (i = 0; i < n_char_ranges; i += 2) | |
1805 | if (c >= char_ranges[i] && c <= char_ranges[i + 1]) | |
9af7511a | 1806 | break; |
b7dbcc19 KH |
1807 | if (!(negate ^ (i < n_char_ranges))) |
1808 | break; | |
9af7511a | 1809 | } |
327719ee | 1810 | fwd_ok: |
8f924df7 | 1811 | p += nbytes, pos++, pos_byte += nbytes; |
b7dbcc19 KH |
1812 | } |
1813 | else | |
8f924df7 KH |
1814 | while (1) |
1815 | { | |
1816 | if (p >= stop) | |
9af7511a | 1817 | { |
8f924df7 | 1818 | if (p >= endp) |
9af7511a | 1819 | break; |
8f924df7 KH |
1820 | p = GAP_END_ADDR; |
1821 | stop = endp; | |
9af7511a | 1822 | } |
327719ee MB |
1823 | |
1824 | if (!NILP (iso_classes) && in_classes (*p, iso_classes)) | |
1825 | { | |
1826 | if (negate) | |
1827 | break; | |
1828 | else | |
1829 | goto fwd_unibyte_ok; | |
1830 | } | |
1831 | ||
8f924df7 KH |
1832 | if (!fastmap[*p]) |
1833 | break; | |
327719ee | 1834 | fwd_unibyte_ok: |
8f924df7 KH |
1835 | p++, pos++, pos_byte++; |
1836 | } | |
195d1361 RS |
1837 | } |
1838 | else | |
1839 | { | |
b7dbcc19 | 1840 | if (multibyte) |
8f924df7 | 1841 | while (1) |
b7dbcc19 | 1842 | { |
8f924df7 | 1843 | unsigned char *prev_p; |
9af7511a | 1844 | |
8f924df7 | 1845 | if (p <= stop) |
9af7511a | 1846 | { |
8f924df7 | 1847 | if (p <= endp) |
9af7511a | 1848 | break; |
8f924df7 KH |
1849 | p = GPT_ADDR; |
1850 | stop = endp; | |
9af7511a | 1851 | } |
8f924df7 KH |
1852 | prev_p = p; |
1853 | while (--p >= stop && ! CHAR_HEAD_P (*p)); | |
327719ee | 1854 | c = STRING_CHAR (p, MAX_MULTIBYTE_LENGTH); |
a1bc88d4 | 1855 | |
327719ee MB |
1856 | if (! NILP (iso_classes) && in_classes (c, iso_classes)) |
1857 | { | |
1858 | if (negate) | |
1859 | break; | |
9690d026 | 1860 | else |
327719ee | 1861 | goto back_ok; |
7e68b0ea | 1862 | } |
a1bc88d4 | 1863 | |
8f924df7 | 1864 | if (! fastmap[*p]) |
b7dbcc19 KH |
1865 | break; |
1866 | if (! ASCII_CHAR_P (c)) | |
7e68b0ea | 1867 | { |
b7dbcc19 KH |
1868 | /* See the comment in the previous similar code. */ |
1869 | for (i = 0; i < n_char_ranges; i += 2) | |
1870 | if (c >= char_ranges[i] && c <= char_ranges[i + 1]) | |
1871 | break; | |
1872 | if (!(negate ^ (i < n_char_ranges))) | |
1873 | break; | |
7e68b0ea | 1874 | } |
327719ee | 1875 | back_ok: |
8f924df7 | 1876 | pos--, pos_byte -= prev_p - p; |
b7dbcc19 KH |
1877 | } |
1878 | else | |
8f924df7 KH |
1879 | while (1) |
1880 | { | |
1881 | if (p <= stop) | |
9af7511a | 1882 | { |
8f924df7 | 1883 | if (p <= endp) |
9af7511a | 1884 | break; |
8f924df7 KH |
1885 | p = GPT_ADDR; |
1886 | stop = endp; | |
9af7511a | 1887 | } |
e39091b6 | 1888 | |
327719ee MB |
1889 | if (! NILP (iso_classes) && in_classes (p[-1], iso_classes)) |
1890 | { | |
1891 | if (negate) | |
1892 | break; | |
9af7511a | 1893 | else |
327719ee MB |
1894 | goto back_unibyte_ok; |
1895 | } | |
a1bc88d4 | 1896 | |
8f924df7 KH |
1897 | if (!fastmap[p[-1]]) |
1898 | break; | |
327719ee | 1899 | back_unibyte_ok: |
8f924df7 KH |
1900 | p--, pos--, pos_byte--; |
1901 | } | |
195d1361 | 1902 | } |
7e68b0ea | 1903 | |
b7dbcc19 KH |
1904 | SET_PT_BOTH (pos, pos_byte); |
1905 | immediate_quit = 0; | |
1906 | ||
1907 | return make_number (PT - start_point); | |
1908 | } | |
1909 | } | |
7e68b0ea | 1910 | |
b7dbcc19 KH |
1911 | |
1912 | static Lisp_Object | |
1913 | skip_syntaxes (forwardp, string, lim) | |
1914 | int forwardp; | |
1915 | Lisp_Object string, lim; | |
1916 | { | |
1917 | register unsigned int c; | |
1918 | unsigned char fastmap[0400]; | |
1919 | int negate = 0; | |
1920 | register int i, i_byte; | |
1921 | int multibyte; | |
1922 | int size_byte; | |
1923 | unsigned char *str; | |
1924 | ||
1925 | CHECK_STRING (string); | |
1926 | ||
1927 | if (NILP (lim)) | |
1928 | XSETINT (lim, forwardp ? ZV : BEGV); | |
1929 | else | |
1930 | CHECK_NUMBER_COERCE_MARKER (lim); | |
1931 | ||
1932 | /* In any case, don't allow scan outside bounds of buffer. */ | |
1933 | if (XINT (lim) > ZV) | |
1934 | XSETFASTINT (lim, ZV); | |
1935 | if (XINT (lim) < BEGV) | |
1936 | XSETFASTINT (lim, BEGV); | |
1937 | ||
ade8ee9e | 1938 | if (forwardp ? (PT >= XFASTINT (lim)) : (PT <= XFASTINT (lim))) |
49feb1cd | 1939 | return make_number (0); |
8c6e735b | 1940 | |
b7dbcc19 | 1941 | multibyte = (!NILP (current_buffer->enable_multibyte_characters) |
92eaa22e | 1942 | && (XINT (lim) - PT != CHAR_TO_BYTE (XINT (lim)) - PT_BYTE)); |
b7dbcc19 KH |
1943 | |
1944 | bzero (fastmap, sizeof fastmap); | |
1945 | ||
8f924df7 KH |
1946 | if (SBYTES (string) > SCHARS (string)) |
1947 | /* As this is very rare case (syntax spec is ASCII only), don't | |
1948 | consider efficiency. */ | |
bc796a59 KH |
1949 | string = string_make_unibyte (string); |
1950 | ||
8f924df7 KH |
1951 | str = SDATA (string); |
1952 | size_byte = SBYTES (string); | |
bc796a59 | 1953 | |
b7dbcc19 KH |
1954 | i_byte = 0; |
1955 | if (i_byte < size_byte | |
8f924df7 | 1956 | && SREF (string, 0) == '^') |
b7dbcc19 KH |
1957 | { |
1958 | negate = 1; i_byte++; | |
1959 | } | |
1960 | ||
b7dbcc19 KH |
1961 | /* Find the syntaxes specified and set their elements of fastmap. */ |
1962 | ||
1963 | while (i_byte < size_byte) | |
1964 | { | |
1965 | c = str[i_byte++]; | |
8f924df7 | 1966 | fastmap[syntax_spec_code[c]] = 1; |
b7dbcc19 | 1967 | } |
195d1361 | 1968 | |
9690d026 | 1969 | /* If ^ was the first character, complement the fastmap. */ |
195d1361 RS |
1970 | if (negate) |
1971 | for (i = 0; i < sizeof fastmap; i++) | |
9690d026 | 1972 | fastmap[i] ^= 1; |
195d1361 RS |
1973 | |
1974 | { | |
1975 | int start_point = PT; | |
1976 | int pos = PT; | |
6a140a74 | 1977 | int pos_byte = PT_BYTE; |
8f924df7 KH |
1978 | unsigned char *p = PT_ADDR, *endp, *stop; |
1979 | ||
1980 | if (forwardp) | |
1981 | { | |
1982 | endp = (XINT (lim) == GPT) ? GPT_ADDR : CHAR_POS_ADDR (XINT (lim)); | |
1983 | stop = (pos < GPT && GPT < XINT (lim)) ? GPT_ADDR : endp; | |
1984 | } | |
1985 | else | |
1986 | { | |
1987 | endp = CHAR_POS_ADDR (XINT (lim)); | |
1988 | stop = (pos >= GPT && GPT > XINT (lim)) ? GAP_END_ADDR : endp; | |
1989 | } | |
195d1361 RS |
1990 | |
1991 | immediate_quit = 1; | |
b7dbcc19 KH |
1992 | SETUP_SYNTAX_TABLE (pos, forwardp ? 1 : -1); |
1993 | if (forwardp) | |
195d1361 | 1994 | { |
b7dbcc19 | 1995 | if (multibyte) |
195d1361 | 1996 | { |
8f924df7 | 1997 | while (1) |
8c6e735b | 1998 | { |
8f924df7 KH |
1999 | int nbytes; |
2000 | ||
2001 | if (p >= stop) | |
2002 | { | |
2003 | if (p >= endp) | |
2004 | break; | |
2005 | p = GAP_END_ADDR; | |
2006 | stop = endp; | |
2007 | } | |
2008 | c = STRING_CHAR_AND_LENGTH (p, MAX_MULTIBYTE_LENGTH, nbytes); | |
2009 | if (! fastmap[(int) SYNTAX (c)]) | |
8c6e735b | 2010 | break; |
8f924df7 | 2011 | p += nbytes, pos++, pos_byte += nbytes; |
8c6e735b KH |
2012 | UPDATE_SYNTAX_TABLE_FORWARD (pos); |
2013 | } | |
195d1361 RS |
2014 | } |
2015 | else | |
2016 | { | |
8c6e735b | 2017 | while (1) |
7e68b0ea | 2018 | { |
8f924df7 KH |
2019 | if (p >= stop) |
2020 | { | |
2021 | if (p >= endp) | |
2022 | break; | |
2023 | p = GAP_END_ADDR; | |
2024 | stop = endp; | |
2025 | } | |
2026 | if (! fastmap[(int) SYNTAX (*p)]) | |
8c6e735b | 2027 | break; |
8f924df7 | 2028 | p++, pos++, pos_byte++; |
b7dbcc19 | 2029 | UPDATE_SYNTAX_TABLE_FORWARD (pos); |
195d1361 RS |
2030 | } |
2031 | } | |
2032 | } | |
2033 | else | |
2034 | { | |
b7dbcc19 | 2035 | if (multibyte) |
195d1361 | 2036 | { |
8c6e735b | 2037 | while (1) |
b7dbcc19 | 2038 | { |
8f924df7 KH |
2039 | unsigned char *prev_p; |
2040 | ||
2041 | if (p <= stop) | |
b7dbcc19 | 2042 | { |
8f924df7 KH |
2043 | if (p <= endp) |
2044 | break; | |
2045 | p = GPT_ADDR; | |
2046 | stop = endp; | |
b7dbcc19 | 2047 | } |
262be72a | 2048 | UPDATE_SYNTAX_TABLE_BACKWARD (pos - 1); |
8f924df7 KH |
2049 | prev_p = p; |
2050 | while (--p >= stop && ! CHAR_HEAD_P (*p)); | |
2051 | c = STRING_CHAR (p, MAX_MULTIBYTE_LENGTH); | |
2052 | if (! fastmap[(int) SYNTAX (c)]) | |
2053 | break; | |
2054 | pos--, pos_byte -= prev_p - p; | |
b7dbcc19 | 2055 | } |
195d1361 RS |
2056 | } |
2057 | else | |
2058 | { | |
8c6e735b KH |
2059 | while (1) |
2060 | { | |
8f924df7 KH |
2061 | if (p <= stop) |
2062 | { | |
2063 | if (p <= endp) | |
2064 | break; | |
2065 | p = GPT_ADDR; | |
2066 | stop = endp; | |
2067 | } | |
102f6132 | 2068 | UPDATE_SYNTAX_TABLE_BACKWARD (pos - 1); |
8f924df7 | 2069 | if (! fastmap[(int) SYNTAX (p[-1])]) |
8c6e735b | 2070 | break; |
8f924df7 | 2071 | p--, pos--, pos_byte--; |
8c6e735b | 2072 | } |
195d1361 RS |
2073 | } |
2074 | } | |
6a140a74 RS |
2075 | |
2076 | SET_PT_BOTH (pos, pos_byte); | |
195d1361 RS |
2077 | immediate_quit = 0; |
2078 | ||
2079 | return make_number (PT - start_point); | |
2080 | } | |
2081 | } | |
a1bc88d4 RS |
2082 | |
2083 | /* Return 1 if character C belongs to one of the ISO classes | |
2084 | in the list ISO_CLASSES. Each class is represented by an | |
2085 | integer which is its type according to re_wctype. */ | |
2086 | ||
2087 | static int | |
2088 | in_classes (c, iso_classes) | |
2089 | int c; | |
2090 | Lisp_Object iso_classes; | |
2091 | { | |
2092 | int fits_class = 0; | |
2093 | ||
2094 | while (! NILP (iso_classes)) | |
2095 | { | |
2096 | Lisp_Object elt; | |
2097 | elt = XCAR (iso_classes); | |
2098 | iso_classes = XCDR (iso_classes); | |
2099 | ||
2100 | if (re_iswctype (c, XFASTINT (elt))) | |
2101 | fits_class = 1; | |
2102 | } | |
2103 | ||
2104 | return fits_class; | |
2105 | } | |
195d1361 | 2106 | \f |
95ff8dfc RS |
2107 | /* Jump over a comment, assuming we are at the beginning of one. |
2108 | FROM is the current position. | |
2109 | FROM_BYTE is the bytepos corresponding to FROM. | |
2110 | Do not move past STOP (a charpos). | |
2111 | The comment over which we have to jump is of style STYLE | |
2112 | (either SYNTAX_COMMENT_STYLE(foo) or ST_COMMENT_STYLE). | |
2113 | NESTING should be positive to indicate the nesting at the beginning | |
2114 | for nested comments and should be zero or negative else. | |
2115 | ST_COMMENT_STYLE cannot be nested. | |
2116 | PREV_SYNTAX is the SYNTAX_WITH_FLAGS of the previous character | |
2117 | (or 0 If the search cannot start in the middle of a two-character). | |
2118 | ||
2119 | If successful, return 1 and store the charpos of the comment's end | |
2120 | into *CHARPOS_PTR and the corresponding bytepos into *BYTEPOS_PTR. | |
2121 | Else, return 0 and store the charpos STOP into *CHARPOS_PTR, the | |
2122 | corresponding bytepos into *BYTEPOS_PTR and the current nesting | |
2123 | (as defined for state.incomment) in *INCOMMENT_PTR. | |
2124 | ||
2125 | The comment end is the last character of the comment rather than the | |
2126 | character just after the comment. | |
2127 | ||
2128 | Global syntax data is assumed to initially be valid for FROM and | |
2129 | remains valid for forward search starting at the returned position. */ | |
2130 | ||
2131 | static int | |
2132 | forw_comment (from, from_byte, stop, nesting, style, prev_syntax, | |
2133 | charpos_ptr, bytepos_ptr, incomment_ptr) | |
a0aa7fcf | 2134 | EMACS_INT from, from_byte, stop; |
95ff8dfc | 2135 | int nesting, style, prev_syntax; |
a0aa7fcf SM |
2136 | EMACS_INT *charpos_ptr, *bytepos_ptr; |
2137 | int *incomment_ptr; | |
95ff8dfc RS |
2138 | { |
2139 | register int c, c1; | |
2140 | register enum syntaxcode code; | |
2141 | register int syntax; | |
2142 | ||
2143 | if (nesting <= 0) nesting = -1; | |
2144 | ||
2145 | /* Enter the loop in the middle so that we find | |
2146 | a 2-char comment ender if we start in the middle of it. */ | |
2147 | syntax = prev_syntax; | |
2148 | if (syntax != 0) goto forw_incomment; | |
2149 | ||
2150 | while (1) | |
2151 | { | |
2152 | if (from == stop) | |
2153 | { | |
2154 | *incomment_ptr = nesting; | |
2155 | *charpos_ptr = from; | |
2156 | *bytepos_ptr = from_byte; | |
2157 | return 0; | |
2158 | } | |
8f924df7 | 2159 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2160 | syntax = SYNTAX_WITH_FLAGS (c); |
2161 | code = syntax & 0xff; | |
2162 | if (code == Sendcomment | |
2163 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style | |
abf8a9ff SM |
2164 | && (SYNTAX_FLAGS_COMMENT_NESTED (syntax) ? |
2165 | (nesting > 0 && --nesting == 0) : nesting < 0)) | |
95ff8dfc RS |
2166 | /* we have encountered a comment end of the same style |
2167 | as the comment sequence which began this comment | |
2168 | section */ | |
2169 | break; | |
2170 | if (code == Scomment_fence | |
2171 | && style == ST_COMMENT_STYLE) | |
2172 | /* we have encountered a comment end of the same style | |
2173 | as the comment sequence which began this comment | |
2174 | section. */ | |
2175 | break; | |
2176 | if (nesting > 0 | |
2177 | && code == Scomment | |
abf8a9ff | 2178 | && SYNTAX_FLAGS_COMMENT_NESTED (syntax) |
95ff8dfc RS |
2179 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style) |
2180 | /* we have encountered a nested comment of the same style | |
2181 | as the comment sequence which began this comment section */ | |
2182 | nesting++; | |
2183 | INC_BOTH (from, from_byte); | |
2184 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
7d0393cf | 2185 | |
95ff8dfc RS |
2186 | forw_incomment: |
2187 | if (from < stop && SYNTAX_FLAGS_COMEND_FIRST (syntax) | |
2188 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style | |
8f924df7 | 2189 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
abf8a9ff SM |
2190 | SYNTAX_COMEND_SECOND (c1)) |
2191 | && ((SYNTAX_FLAGS_COMMENT_NESTED (syntax) || | |
2192 | SYNTAX_COMMENT_NESTED (c1)) ? nesting > 0 : nesting < 0)) | |
4ffe723b GM |
2193 | { |
2194 | if (--nesting <= 0) | |
2195 | /* we have encountered a comment end of the same style | |
2196 | as the comment sequence which began this comment | |
2197 | section */ | |
2198 | break; | |
2199 | else | |
2200 | { | |
2201 | INC_BOTH (from, from_byte); | |
2202 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
2203 | } | |
2204 | } | |
95ff8dfc RS |
2205 | if (nesting > 0 |
2206 | && from < stop | |
2207 | && SYNTAX_FLAGS_COMSTART_FIRST (syntax) | |
8f924df7 | 2208 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
95ff8dfc | 2209 | SYNTAX_COMMENT_STYLE (c1) == style |
abf8a9ff SM |
2210 | && SYNTAX_COMSTART_SECOND (c1)) |
2211 | && (SYNTAX_FLAGS_COMMENT_NESTED (syntax) || | |
2212 | SYNTAX_COMMENT_NESTED (c1))) | |
95ff8dfc RS |
2213 | /* we have encountered a nested comment of the same style |
2214 | as the comment sequence which began this comment | |
2215 | section */ | |
2216 | { | |
2217 | INC_BOTH (from, from_byte); | |
2218 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
2219 | nesting++; | |
2220 | } | |
2221 | } | |
2222 | *charpos_ptr = from; | |
2223 | *bytepos_ptr = from_byte; | |
2224 | return 1; | |
2225 | } | |
2226 | ||
b3cfe0c8 | 2227 | DEFUN ("forward-comment", Fforward_comment, Sforward_comment, 1, 1, 0, |
177c0ea7 | 2228 | doc: /* |
ee648c11 | 2229 | Move forward across up to COUNT comments. If COUNT is negative, move backward. |
fdb82f93 PJ |
2230 | Stop scanning if we find something other than a comment or whitespace. |
2231 | Set point to where scanning stops. | |
ee648c11 | 2232 | If COUNT comments are found as expected, with nothing except whitespace |
fdb82f93 PJ |
2233 | between them, return t; otherwise return nil. */) |
2234 | (count) | |
840f481c | 2235 | Lisp_Object count; |
b3cfe0c8 | 2236 | { |
2312c580 SM |
2237 | register EMACS_INT from; |
2238 | EMACS_INT from_byte; | |
2239 | register EMACS_INT stop; | |
8ea151b2 | 2240 | register int c, c1; |
b3cfe0c8 RS |
2241 | register enum syntaxcode code; |
2242 | int comstyle = 0; /* style of comment encountered */ | |
95ff8dfc | 2243 | int comnested = 0; /* whether the comment is nestable or not */ |
be720845 | 2244 | int found; |
2312c580 SM |
2245 | EMACS_INT count1; |
2246 | EMACS_INT out_charpos, out_bytepos; | |
95ff8dfc | 2247 | int dummy; |
840f481c | 2248 | |
b7826503 | 2249 | CHECK_NUMBER (count); |
840f481c | 2250 | count1 = XINT (count); |
195d1361 | 2251 | stop = count1 > 0 ? ZV : BEGV; |
b3cfe0c8 RS |
2252 | |
2253 | immediate_quit = 1; | |
2254 | QUIT; | |
2255 | ||
2256 | from = PT; | |
6a140a74 | 2257 | from_byte = PT_BYTE; |
b3cfe0c8 | 2258 | |
195d1361 | 2259 | SETUP_SYNTAX_TABLE (from, count1); |
840f481c | 2260 | while (count1 > 0) |
b3cfe0c8 | 2261 | { |
04882296 | 2262 | do |
b3cfe0c8 | 2263 | { |
f902a008 RS |
2264 | int comstart_first; |
2265 | ||
04882296 KH |
2266 | if (from == stop) |
2267 | { | |
ef316cf0 | 2268 | SET_PT_BOTH (from, from_byte); |
b7e6e612 | 2269 | immediate_quit = 0; |
04882296 KH |
2270 | return Qnil; |
2271 | } | |
b7dbcc19 | 2272 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
b3cfe0c8 | 2273 | code = SYNTAX (c); |
f902a008 | 2274 | comstart_first = SYNTAX_COMSTART_FIRST (c); |
95ff8dfc | 2275 | comnested = SYNTAX_COMMENT_NESTED (c); |
e6365855 | 2276 | comstyle = SYNTAX_COMMENT_STYLE (c); |
6a140a74 | 2277 | INC_BOTH (from, from_byte); |
f902a008 | 2278 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
f902a008 | 2279 | if (from < stop && comstart_first |
b7dbcc19 | 2280 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
8ea151b2 | 2281 | SYNTAX_COMSTART_SECOND (c1))) |
b3cfe0c8 | 2282 | { |
7d0393cf | 2283 | /* We have encountered a comment start sequence and we |
7fc8191e | 2284 | are ignoring all text inside comments. We must record |
b3cfe0c8 RS |
2285 | the comment style this sequence begins so that later, |
2286 | only a comment end of the same style actually ends | |
7fc8191e | 2287 | the comment section. */ |
b3cfe0c8 | 2288 | code = Scomment; |
8ea151b2 | 2289 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
95ff8dfc | 2290 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); |
6a140a74 | 2291 | INC_BOTH (from, from_byte); |
95ff8dfc | 2292 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b3cfe0c8 | 2293 | } |
04882296 | 2294 | } |
abf8a9ff | 2295 | while (code == Swhitespace || (code == Sendcomment && c == '\n')); |
6a140a74 | 2296 | |
da71ab91 KH |
2297 | if (code == Scomment_fence) |
2298 | comstyle = ST_COMMENT_STYLE; | |
2299 | else if (code != Scomment) | |
04882296 KH |
2300 | { |
2301 | immediate_quit = 0; | |
6a140a74 | 2302 | DEC_BOTH (from, from_byte); |
ef316cf0 | 2303 | SET_PT_BOTH (from, from_byte); |
04882296 KH |
2304 | return Qnil; |
2305 | } | |
2306 | /* We're at the start of a comment. */ | |
95ff8dfc RS |
2307 | found = forw_comment (from, from_byte, stop, comnested, comstyle, 0, |
2308 | &out_charpos, &out_bytepos, &dummy); | |
2309 | from = out_charpos; from_byte = out_bytepos; | |
2310 | if (!found) | |
04882296 | 2311 | { |
95ff8dfc RS |
2312 | immediate_quit = 0; |
2313 | SET_PT_BOTH (from, from_byte); | |
2314 | return Qnil; | |
b3cfe0c8 | 2315 | } |
95ff8dfc RS |
2316 | INC_BOTH (from, from_byte); |
2317 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
04882296 | 2318 | /* We have skipped one comment. */ |
840f481c | 2319 | count1--; |
b3cfe0c8 RS |
2320 | } |
2321 | ||
840f481c | 2322 | while (count1 < 0) |
b3cfe0c8 | 2323 | { |
b9145dbb | 2324 | while (1) |
b3cfe0c8 | 2325 | { |
c65adb44 | 2326 | int quoted; |
f902a008 | 2327 | |
b9145dbb RS |
2328 | if (from <= stop) |
2329 | { | |
6a140a74 | 2330 | SET_PT_BOTH (BEGV, BEGV_BYTE); |
b9145dbb RS |
2331 | immediate_quit = 0; |
2332 | return Qnil; | |
2333 | } | |
b3cfe0c8 | 2334 | |
6a140a74 | 2335 | DEC_BOTH (from, from_byte); |
f902a008 | 2336 | /* char_quoted does UPDATE_SYNTAX_TABLE_BACKWARD (from). */ |
6a140a74 | 2337 | quoted = char_quoted (from, from_byte); |
b7dbcc19 | 2338 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
b3cfe0c8 RS |
2339 | code = SYNTAX (c); |
2340 | comstyle = 0; | |
95ff8dfc | 2341 | comnested = SYNTAX_COMMENT_NESTED (c); |
7fc8191e RS |
2342 | if (code == Sendcomment) |
2343 | comstyle = SYNTAX_COMMENT_STYLE (c); | |
b3cfe0c8 | 2344 | if (from > stop && SYNTAX_COMEND_SECOND (c) |
f902a008 | 2345 | && prev_char_comend_first (from, from_byte) |
89c6809a | 2346 | && !char_quoted (from - 1, dec_bytepos (from_byte))) |
b3cfe0c8 | 2347 | { |
7fc8191e | 2348 | /* We must record the comment style encountered so that |
b3cfe0c8 | 2349 | later, we can match only the proper comment begin |
7fc8191e | 2350 | sequence of the same style. */ |
6a140a74 | 2351 | DEC_BOTH (from, from_byte); |
f902a008 RS |
2352 | code = Sendcomment; |
2353 | /* Calling char_quoted, above, set up global syntax position | |
2354 | at the new value of FROM. */ | |
b7dbcc19 | 2355 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2356 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
2357 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
b3cfe0c8 RS |
2358 | } |
2359 | ||
195d1361 RS |
2360 | if (code == Scomment_fence) |
2361 | { | |
2362 | /* Skip until first preceding unquoted comment_fence. */ | |
6a140a74 | 2363 | int found = 0, ini = from, ini_byte = from_byte; |
7d0393cf | 2364 | |
6a140a74 | 2365 | while (1) |
195d1361 | 2366 | { |
6a140a74 | 2367 | DEC_BOTH (from, from_byte); |
195d1361 | 2368 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 2369 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
6a140a74 | 2370 | if (SYNTAX (c) == Scomment_fence |
7d0393cf | 2371 | && !char_quoted (from, from_byte)) |
195d1361 | 2372 | { |
7d0393cf | 2373 | found = 1; |
195d1361 RS |
2374 | break; |
2375 | } | |
fcdd4585 SM |
2376 | else if (from == stop) |
2377 | break; | |
195d1361 RS |
2378 | } |
2379 | if (found == 0) | |
2380 | { | |
2381 | from = ini; /* Set point to ini + 1. */ | |
6a140a74 | 2382 | from_byte = ini_byte; |
195d1361 RS |
2383 | goto leave; |
2384 | } | |
39c41ad4 SM |
2385 | else |
2386 | /* We have skipped one comment. */ | |
2387 | break; | |
195d1361 RS |
2388 | } |
2389 | else if (code == Sendcomment) | |
b3cfe0c8 | 2390 | { |
95ff8dfc | 2391 | found = back_comment (from, from_byte, stop, comnested, comstyle, |
6a140a74 | 2392 | &out_charpos, &out_bytepos); |
1aa963c8 SM |
2393 | if (found == -1) |
2394 | { | |
abf8a9ff SM |
2395 | if (c == '\n') |
2396 | /* This end-of-line is not an end-of-comment. | |
2397 | Treat it like a whitespace. | |
2398 | CC-mode (and maybe others) relies on this behavior. */ | |
2399 | ; | |
2400 | else | |
2401 | { | |
2402 | /* Failure: we should go back to the end of this | |
2403 | not-quite-endcomment. */ | |
2404 | if (SYNTAX(c) != code) | |
2405 | /* It was a two-char Sendcomment. */ | |
2406 | INC_BOTH (from, from_byte); | |
2407 | goto leave; | |
2408 | } | |
1aa963c8 | 2409 | } |
2b927d02 | 2410 | else |
abf8a9ff SM |
2411 | { |
2412 | /* We have skipped one comment. */ | |
2413 | from = out_charpos, from_byte = out_bytepos; | |
2414 | break; | |
2415 | } | |
b3cfe0c8 | 2416 | } |
bb0de084 | 2417 | else if (code != Swhitespace || quoted) |
b3cfe0c8 | 2418 | { |
195d1361 | 2419 | leave: |
b3cfe0c8 | 2420 | immediate_quit = 0; |
6a140a74 | 2421 | INC_BOTH (from, from_byte); |
ef316cf0 | 2422 | SET_PT_BOTH (from, from_byte); |
b3cfe0c8 RS |
2423 | return Qnil; |
2424 | } | |
2425 | } | |
2426 | ||
840f481c | 2427 | count1++; |
b3cfe0c8 RS |
2428 | } |
2429 | ||
ef316cf0 | 2430 | SET_PT_BOTH (from, from_byte); |
b3cfe0c8 RS |
2431 | immediate_quit = 0; |
2432 | return Qt; | |
2433 | } | |
2434 | \f | |
b7dbcc19 | 2435 | /* Return syntax code of character C if C is an ASCII character |
db9f48c3 | 2436 | or `multibyte_symbol_p' is zero. Otherwise, return Ssymbol. */ |
bd25db08 | 2437 | |
b7dbcc19 KH |
2438 | #define SYNTAX_WITH_MULTIBYTE_CHECK(c) \ |
2439 | ((ASCII_CHAR_P (c) || !multibyte_symbol_p) \ | |
bd25db08 KH |
2440 | ? SYNTAX (c) : Ssymbol) |
2441 | ||
6a140a74 | 2442 | static Lisp_Object |
8489eb67 | 2443 | scan_lists (from, count, depth, sexpflag) |
932e6895 SM |
2444 | register EMACS_INT from; |
2445 | EMACS_INT count, depth; | |
2446 | int sexpflag; | |
8489eb67 RS |
2447 | { |
2448 | Lisp_Object val; | |
932e6895 | 2449 | register EMACS_INT stop = count > 0 ? ZV : BEGV; |
93da5fff KH |
2450 | register int c, c1; |
2451 | int stringterm; | |
8489eb67 RS |
2452 | int quoted; |
2453 | int mathexit = 0; | |
93da5fff | 2454 | register enum syntaxcode code, temp_code; |
195d1361 | 2455 | int min_depth = depth; /* Err out if depth gets less than this. */ |
e5d4f4dc | 2456 | int comstyle = 0; /* style of comment encountered */ |
95ff8dfc | 2457 | int comnested = 0; /* whether the comment is nestable or not */ |
932e6895 SM |
2458 | EMACS_INT temp_pos; |
2459 | EMACS_INT last_good = from; | |
195d1361 | 2460 | int found; |
932e6895 SM |
2461 | EMACS_INT from_byte; |
2462 | EMACS_INT out_bytepos, out_charpos; | |
95ff8dfc | 2463 | int temp, dummy; |
bd25db08 | 2464 | int multibyte_symbol_p = sexpflag && multibyte_syntax_as_symbol; |
8489eb67 RS |
2465 | |
2466 | if (depth > 0) min_depth = 0; | |
2467 | ||
c3907a7e RS |
2468 | if (from > ZV) from = ZV; |
2469 | if (from < BEGV) from = BEGV; | |
2470 | ||
2471 | from_byte = CHAR_TO_BYTE (from); | |
2472 | ||
8489eb67 RS |
2473 | immediate_quit = 1; |
2474 | QUIT; | |
2475 | ||
195d1361 | 2476 | SETUP_SYNTAX_TABLE (from, count); |
8489eb67 RS |
2477 | while (count > 0) |
2478 | { | |
8489eb67 RS |
2479 | while (from < stop) |
2480 | { | |
f902a008 | 2481 | int comstart_first, prefix; |
195d1361 | 2482 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 2483 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2484 | code = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
f902a008 | 2485 | comstart_first = SYNTAX_COMSTART_FIRST (c); |
95ff8dfc | 2486 | comnested = SYNTAX_COMMENT_NESTED (c); |
e6365855 | 2487 | comstyle = SYNTAX_COMMENT_STYLE (c); |
f902a008 | 2488 | prefix = SYNTAX_PREFIX (c); |
7bf5e9e4 RS |
2489 | if (depth == min_depth) |
2490 | last_good = from; | |
6a140a74 | 2491 | INC_BOTH (from, from_byte); |
195d1361 | 2492 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
f902a008 | 2493 | if (from < stop && comstart_first |
327719ee MB |
2494 | && (c = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
2495 | SYNTAX_COMSTART_SECOND (c)) | |
8489eb67 | 2496 | && parse_sexp_ignore_comments) |
e5d4f4dc | 2497 | { |
7d0393cf | 2498 | /* we have encountered a comment start sequence and we |
195d1361 | 2499 | are ignoring all text inside comments. We must record |
e5d4f4dc RS |
2500 | the comment style this sequence begins so that later, |
2501 | only a comment end of the same style actually ends | |
2502 | the comment section */ | |
2503 | code = Scomment; | |
b7dbcc19 | 2504 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2505 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
2506 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
6a140a74 | 2507 | INC_BOTH (from, from_byte); |
f902a008 | 2508 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
e5d4f4dc | 2509 | } |
7d0393cf | 2510 | |
f902a008 | 2511 | if (prefix) |
8489eb67 RS |
2512 | continue; |
2513 | ||
0220c518 | 2514 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
2515 | { |
2516 | case Sescape: | |
2517 | case Scharquote: | |
8b8bf8e6 SM |
2518 | if (from == stop) |
2519 | goto lose; | |
6a140a74 | 2520 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2521 | /* treat following character as a word constituent */ |
2522 | case Sword: | |
2523 | case Ssymbol: | |
2524 | if (depth || !sexpflag) break; | |
195d1361 | 2525 | /* This word counts as a sexp; return at end of it. */ |
8489eb67 RS |
2526 | while (from < stop) |
2527 | { | |
195d1361 | 2528 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
7c7ca3d2 RS |
2529 | |
2530 | /* Some compilers can't handle this inside the switch. */ | |
b7dbcc19 | 2531 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2532 | temp = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7c7ca3d2 | 2533 | switch (temp) |
8489eb67 RS |
2534 | { |
2535 | case Scharquote: | |
2536 | case Sescape: | |
6a140a74 | 2537 | INC_BOTH (from, from_byte); |
8b8bf8e6 SM |
2538 | if (from == stop) |
2539 | goto lose; | |
8489eb67 RS |
2540 | break; |
2541 | case Sword: | |
2542 | case Ssymbol: | |
2543 | case Squote: | |
2544 | break; | |
2545 | default: | |
2546 | goto done; | |
2547 | } | |
6a140a74 | 2548 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2549 | } |
2550 | goto done; | |
2551 | ||
195d1361 | 2552 | case Scomment_fence: |
95ff8dfc RS |
2553 | comstyle = ST_COMMENT_STYLE; |
2554 | /* FALLTHROUGH */ | |
2555 | case Scomment: | |
8489eb67 | 2556 | if (!parse_sexp_ignore_comments) break; |
95ff8dfc RS |
2557 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
2558 | found = forw_comment (from, from_byte, stop, | |
2559 | comnested, comstyle, 0, | |
2560 | &out_charpos, &out_bytepos, &dummy); | |
2561 | from = out_charpos, from_byte = out_bytepos; | |
2562 | if (!found) | |
8489eb67 | 2563 | { |
95ff8dfc RS |
2564 | if (depth == 0) |
2565 | goto done; | |
2566 | goto lose; | |
8489eb67 | 2567 | } |
95ff8dfc RS |
2568 | INC_BOTH (from, from_byte); |
2569 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
8489eb67 RS |
2570 | break; |
2571 | ||
2572 | case Smath: | |
2573 | if (!sexpflag) | |
2574 | break; | |
b7dbcc19 | 2575 | if (from != stop && c == FETCH_CHAR_AS_MULTIBYTE (from_byte)) |
6a140a74 RS |
2576 | { |
2577 | INC_BOTH (from, from_byte); | |
2578 | } | |
8489eb67 RS |
2579 | if (mathexit) |
2580 | { | |
2581 | mathexit = 0; | |
2582 | goto close1; | |
2583 | } | |
2584 | mathexit = 1; | |
2585 | ||
2586 | case Sopen: | |
2587 | if (!++depth) goto done; | |
2588 | break; | |
2589 | ||
2590 | case Sclose: | |
2591 | close1: | |
2592 | if (!--depth) goto done; | |
2593 | if (depth < min_depth) | |
9471945b KS |
2594 | xsignal3 (Qscan_error, |
2595 | build_string ("Containing expression ends prematurely"), | |
2596 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2597 | break; |
2598 | ||
2599 | case Sstring: | |
195d1361 | 2600 | case Sstring_fence: |
ef316cf0 | 2601 | temp_pos = dec_bytepos (from_byte); |
b7dbcc19 | 2602 | stringterm = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
8489eb67 RS |
2603 | while (1) |
2604 | { | |
8b8bf8e6 SM |
2605 | if (from >= stop) |
2606 | goto lose; | |
195d1361 | 2607 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 2608 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
42ebfa31 SM |
2609 | if (code == Sstring |
2610 | ? (c == stringterm | |
2611 | && SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring) | |
bd25db08 | 2612 | : SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring_fence) |
195d1361 | 2613 | break; |
7c7ca3d2 RS |
2614 | |
2615 | /* Some compilers can't handle this inside the switch. */ | |
bd25db08 | 2616 | temp = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7c7ca3d2 | 2617 | switch (temp) |
8489eb67 RS |
2618 | { |
2619 | case Scharquote: | |
2620 | case Sescape: | |
6a140a74 | 2621 | INC_BOTH (from, from_byte); |
8489eb67 | 2622 | } |
6a140a74 | 2623 | INC_BOTH (from, from_byte); |
8489eb67 | 2624 | } |
6a140a74 | 2625 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2626 | if (!depth && sexpflag) goto done; |
2627 | break; | |
240c43e8 SM |
2628 | default: |
2629 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
2630 | break; | |
8489eb67 RS |
2631 | } |
2632 | } | |
2633 | ||
2634 | /* Reached end of buffer. Error if within object, return nil if between */ | |
8b8bf8e6 SM |
2635 | if (depth) |
2636 | goto lose; | |
8489eb67 RS |
2637 | |
2638 | immediate_quit = 0; | |
2639 | return Qnil; | |
2640 | ||
2641 | /* End of object reached */ | |
2642 | done: | |
2643 | count--; | |
2644 | } | |
2645 | ||
2646 | while (count < 0) | |
2647 | { | |
8489eb67 RS |
2648 | while (from > stop) |
2649 | { | |
6a140a74 | 2650 | DEC_BOTH (from, from_byte); |
195d1361 | 2651 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 2652 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2653 | code = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7bf5e9e4 RS |
2654 | if (depth == min_depth) |
2655 | last_good = from; | |
7fc8191e | 2656 | comstyle = 0; |
95ff8dfc | 2657 | comnested = SYNTAX_COMMENT_NESTED (c); |
7fc8191e RS |
2658 | if (code == Sendcomment) |
2659 | comstyle = SYNTAX_COMMENT_STYLE (c); | |
8489eb67 | 2660 | if (from > stop && SYNTAX_COMEND_SECOND (c) |
1674d9a2 | 2661 | && prev_char_comend_first (from, from_byte) |
8489eb67 | 2662 | && parse_sexp_ignore_comments) |
e5d4f4dc | 2663 | { |
f902a008 | 2664 | /* We must record the comment style encountered so that |
e5d4f4dc | 2665 | later, we can match only the proper comment begin |
f902a008 | 2666 | sequence of the same style. */ |
6a140a74 | 2667 | DEC_BOTH (from, from_byte); |
f902a008 RS |
2668 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
2669 | code = Sendcomment; | |
b7dbcc19 | 2670 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2671 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
2672 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
e5d4f4dc | 2673 | } |
7d0393cf | 2674 | |
9828a477 | 2675 | /* Quoting turns anything except a comment-ender |
e6365855 | 2676 | into a word character. Note that this cannot be true |
f902a008 | 2677 | if we decremented FROM in the if-statement above. */ |
9828a477 | 2678 | if (code != Sendcomment && char_quoted (from, from_byte)) |
240c43e8 SM |
2679 | { |
2680 | DEC_BOTH (from, from_byte); | |
2681 | code = Sword; | |
2682 | } | |
9828a477 | 2683 | else if (SYNTAX_PREFIX (c)) |
8489eb67 RS |
2684 | continue; |
2685 | ||
9828a477 | 2686 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
2687 | { |
2688 | case Sword: | |
2689 | case Ssymbol: | |
9828a477 KH |
2690 | case Sescape: |
2691 | case Scharquote: | |
8489eb67 | 2692 | if (depth || !sexpflag) break; |
195d1361 RS |
2693 | /* This word counts as a sexp; count object finished |
2694 | after passing it. */ | |
8489eb67 RS |
2695 | while (from > stop) |
2696 | { | |
6a140a74 | 2697 | temp_pos = from_byte; |
ef316cf0 RS |
2698 | if (! NILP (current_buffer->enable_multibyte_characters)) |
2699 | DEC_POS (temp_pos); | |
2700 | else | |
2701 | temp_pos--; | |
6a140a74 | 2702 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
b7dbcc19 | 2703 | c1 = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
bd25db08 | 2704 | temp_code = SYNTAX_WITH_MULTIBYTE_CHECK (c1); |
9828a477 KH |
2705 | /* Don't allow comment-end to be quoted. */ |
2706 | if (temp_code == Sendcomment) | |
2707 | goto done2; | |
6a140a74 | 2708 | quoted = char_quoted (from - 1, temp_pos); |
8489eb67 | 2709 | if (quoted) |
93da5fff | 2710 | { |
6a140a74 | 2711 | DEC_BOTH (from, from_byte); |
ef316cf0 | 2712 | temp_pos = dec_bytepos (temp_pos); |
6a140a74 | 2713 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
93da5fff | 2714 | } |
b7dbcc19 | 2715 | c1 = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
bd25db08 | 2716 | temp_code = SYNTAX_WITH_MULTIBYTE_CHECK (c1); |
93da5fff KH |
2717 | if (! (quoted || temp_code == Sword |
2718 | || temp_code == Ssymbol | |
2719 | || temp_code == Squote)) | |
8489eb67 | 2720 | goto done2; |
6a140a74 | 2721 | DEC_BOTH (from, from_byte); |
8489eb67 RS |
2722 | } |
2723 | goto done2; | |
2724 | ||
2725 | case Smath: | |
2726 | if (!sexpflag) | |
2727 | break; | |
ef316cf0 | 2728 | temp_pos = dec_bytepos (from_byte); |
6a140a74 | 2729 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
b7dbcc19 | 2730 | if (from != stop && c == FETCH_CHAR_AS_MULTIBYTE (temp_pos)) |
6a140a74 | 2731 | DEC_BOTH (from, from_byte); |
8489eb67 RS |
2732 | if (mathexit) |
2733 | { | |
2734 | mathexit = 0; | |
2735 | goto open2; | |
2736 | } | |
2737 | mathexit = 1; | |
2738 | ||
2739 | case Sclose: | |
2740 | if (!++depth) goto done2; | |
2741 | break; | |
2742 | ||
2743 | case Sopen: | |
2744 | open2: | |
2745 | if (!--depth) goto done2; | |
2746 | if (depth < min_depth) | |
9471945b KS |
2747 | xsignal3 (Qscan_error, |
2748 | build_string ("Containing expression ends prematurely"), | |
2749 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2750 | break; |
2751 | ||
2752 | case Sendcomment: | |
2753 | if (!parse_sexp_ignore_comments) | |
2754 | break; | |
95ff8dfc | 2755 | found = back_comment (from, from_byte, stop, comnested, comstyle, |
6a140a74 | 2756 | &out_charpos, &out_bytepos); |
1aa963c8 SM |
2757 | /* FIXME: if found == -1, then it really wasn't a comment-end. |
2758 | For single-char Sendcomment, we can't do much about it apart | |
2759 | from skipping the char. | |
2760 | For 2-char endcomments, we could try again, taking both | |
2761 | chars as separate entities, but it's a lot of trouble | |
2762 | for very little gain, so we don't bother either. -sm */ | |
6a140a74 RS |
2763 | if (found != -1) |
2764 | from = out_charpos, from_byte = out_bytepos; | |
8489eb67 RS |
2765 | break; |
2766 | ||
195d1361 RS |
2767 | case Scomment_fence: |
2768 | case Sstring_fence: | |
2769 | while (1) | |
2770 | { | |
8b8bf8e6 SM |
2771 | if (from == stop) |
2772 | goto lose; | |
6b61353c | 2773 | DEC_BOTH (from, from_byte); |
195d1361 | 2774 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
7d0393cf | 2775 | if (!char_quoted (from, from_byte) |
b7dbcc19 | 2776 | && (c = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
bd25db08 | 2777 | SYNTAX_WITH_MULTIBYTE_CHECK (c) == code)) |
195d1361 RS |
2778 | break; |
2779 | } | |
2780 | if (code == Sstring_fence && !depth && sexpflag) goto done2; | |
2781 | break; | |
7d0393cf | 2782 | |
8489eb67 | 2783 | case Sstring: |
b7dbcc19 | 2784 | stringterm = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8489eb67 RS |
2785 | while (1) |
2786 | { | |
8b8bf8e6 SM |
2787 | if (from == stop) |
2788 | goto lose; | |
6b61353c KH |
2789 | DEC_BOTH (from, from_byte); |
2790 | UPDATE_SYNTAX_TABLE_BACKWARD (from); | |
2791 | if (!char_quoted (from, from_byte) | |
2792 | && (stringterm | |
2793 | == (c = FETCH_CHAR_AS_MULTIBYTE (from_byte))) | |
42ebfa31 | 2794 | && SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring) |
8489eb67 | 2795 | break; |
8489eb67 | 2796 | } |
8489eb67 RS |
2797 | if (!depth && sexpflag) goto done2; |
2798 | break; | |
240c43e8 SM |
2799 | default: |
2800 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
2801 | break; | |
8489eb67 RS |
2802 | } |
2803 | } | |
2804 | ||
2805 | /* Reached start of buffer. Error if within object, return nil if between */ | |
8b8bf8e6 SM |
2806 | if (depth) |
2807 | goto lose; | |
8489eb67 RS |
2808 | |
2809 | immediate_quit = 0; | |
2810 | return Qnil; | |
2811 | ||
2812 | done2: | |
2813 | count++; | |
2814 | } | |
2815 | ||
2816 | ||
2817 | immediate_quit = 0; | |
1e142fb7 | 2818 | XSETFASTINT (val, from); |
8489eb67 RS |
2819 | return val; |
2820 | ||
2821 | lose: | |
9471945b KS |
2822 | xsignal3 (Qscan_error, |
2823 | build_string ("Unbalanced parentheses"), | |
2824 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2825 | } |
2826 | ||
8489eb67 | 2827 | DEFUN ("scan-lists", Fscan_lists, Sscan_lists, 3, 3, 0, |
fdb82f93 PJ |
2828 | doc: /* Scan from character number FROM by COUNT lists. |
2829 | Returns the character number of the position thus found. | |
2830 | ||
2831 | If DEPTH is nonzero, paren depth begins counting from that value, | |
2832 | only places where the depth in parentheses becomes zero | |
2833 | are candidates for stopping; COUNT such places are counted. | |
2834 | Thus, a positive value for DEPTH means go out levels. | |
2835 | ||
2836 | Comments are ignored if `parse-sexp-ignore-comments' is non-nil. | |
2837 | ||
2838 | If the beginning or end of (the accessible part of) the buffer is reached | |
2839 | and the depth is wrong, an error is signaled. | |
2840 | If the depth is right but the count is not used up, nil is returned. */) | |
2841 | (from, count, depth) | |
8489eb67 RS |
2842 | Lisp_Object from, count, depth; |
2843 | { | |
b7826503 PJ |
2844 | CHECK_NUMBER (from); |
2845 | CHECK_NUMBER (count); | |
2846 | CHECK_NUMBER (depth); | |
8489eb67 RS |
2847 | |
2848 | return scan_lists (XINT (from), XINT (count), XINT (depth), 0); | |
2849 | } | |
2850 | ||
2851 | DEFUN ("scan-sexps", Fscan_sexps, Sscan_sexps, 2, 2, 0, | |
fdb82f93 PJ |
2852 | doc: /* Scan from character number FROM by COUNT balanced expressions. |
2853 | If COUNT is negative, scan backwards. | |
2854 | Returns the character number of the position thus found. | |
2855 | ||
2856 | Comments are ignored if `parse-sexp-ignore-comments' is non-nil. | |
2857 | ||
2858 | If the beginning or end of (the accessible part of) the buffer is reached | |
2859 | in the middle of a parenthetical grouping, an error is signaled. | |
2860 | If the beginning or end is reached between groupings | |
2861 | but before count is used up, nil is returned. */) | |
2862 | (from, count) | |
8489eb67 RS |
2863 | Lisp_Object from, count; |
2864 | { | |
b7826503 PJ |
2865 | CHECK_NUMBER (from); |
2866 | CHECK_NUMBER (count); | |
8489eb67 RS |
2867 | |
2868 | return scan_lists (XINT (from), XINT (count), 0, 1); | |
2869 | } | |
2870 | ||
2871 | DEFUN ("backward-prefix-chars", Fbackward_prefix_chars, Sbackward_prefix_chars, | |
fdb82f93 PJ |
2872 | 0, 0, 0, |
2873 | doc: /* Move point backward over any number of chars with prefix syntax. | |
2874 | This includes chars with "quote" or "prefix" syntax (' or p). */) | |
2875 | () | |
8489eb67 RS |
2876 | { |
2877 | int beg = BEGV; | |
6a140a74 RS |
2878 | int opoint = PT; |
2879 | int opoint_byte = PT_BYTE; | |
6ec8bbd2 | 2880 | int pos = PT; |
6a140a74 | 2881 | int pos_byte = PT_BYTE; |
93da5fff | 2882 | int c; |
93da5fff | 2883 | |
7d0393cf | 2884 | if (pos <= beg) |
195d1361 | 2885 | { |
f902a008 RS |
2886 | SET_PT_BOTH (opoint, opoint_byte); |
2887 | ||
2888 | return Qnil; | |
195d1361 | 2889 | } |
8489eb67 | 2890 | |
f902a008 RS |
2891 | SETUP_SYNTAX_TABLE (pos, -1); |
2892 | ||
6a140a74 RS |
2893 | DEC_BOTH (pos, pos_byte); |
2894 | ||
1fd3172d | 2895 | while (!char_quoted (pos, pos_byte) |
195d1361 | 2896 | /* Previous statement updates syntax table. */ |
b7dbcc19 | 2897 | && ((c = FETCH_CHAR_AS_MULTIBYTE (pos_byte), SYNTAX (c) == Squote) |
93da5fff KH |
2898 | || SYNTAX_PREFIX (c))) |
2899 | { | |
1fd3172d RS |
2900 | opoint = pos; |
2901 | opoint_byte = pos_byte; | |
2902 | ||
2903 | if (pos + 1 > beg) | |
2904 | DEC_BOTH (pos, pos_byte); | |
93da5fff | 2905 | } |
8489eb67 | 2906 | |
6a140a74 | 2907 | SET_PT_BOTH (opoint, opoint_byte); |
8489eb67 RS |
2908 | |
2909 | return Qnil; | |
2910 | } | |
2911 | \f | |
6a140a74 | 2912 | /* Parse forward from FROM / FROM_BYTE to END, |
e5d4f4dc RS |
2913 | assuming that FROM has state OLDSTATE (nil means FROM is start of function), |
2914 | and return a description of the state of the parse at END. | |
c81a3712 | 2915 | If STOPBEFORE is nonzero, stop at the start of an atom. |
644ea4df RS |
2916 | If COMMENTSTOP is 1, stop at the start of a comment. |
2917 | If COMMENTSTOP is -1, stop at the start or end of a comment, | |
2918 | after the beginning of a string, or after the end of a string. */ | |
8489eb67 | 2919 | |
340f92b5 | 2920 | static void |
6a140a74 | 2921 | scan_sexps_forward (stateptr, from, from_byte, end, targetdepth, |
c81a3712 | 2922 | stopbefore, oldstate, commentstop) |
e5d4f4dc | 2923 | struct lisp_parse_state *stateptr; |
8489eb67 | 2924 | register int from; |
38355d47 | 2925 | int from_byte; |
8489eb67 RS |
2926 | int end, targetdepth, stopbefore; |
2927 | Lisp_Object oldstate; | |
c81a3712 | 2928 | int commentstop; |
8489eb67 RS |
2929 | { |
2930 | struct lisp_parse_state state; | |
2931 | ||
2932 | register enum syntaxcode code; | |
95ff8dfc RS |
2933 | int c1; |
2934 | int comnested; | |
8489eb67 RS |
2935 | struct level { int last, prev; }; |
2936 | struct level levelstart[100]; | |
2937 | register struct level *curlevel = levelstart; | |
2938 | struct level *endlevel = levelstart + 100; | |
8489eb67 RS |
2939 | register int depth; /* Paren depth of current scanning location. |
2940 | level - levelstart equals this except | |
2941 | when the depth becomes negative. */ | |
2942 | int mindepth; /* Lowest DEPTH value seen. */ | |
2943 | int start_quoted = 0; /* Nonzero means starting after a char quote */ | |
2944 | Lisp_Object tem; | |
93da5fff | 2945 | int prev_from; /* Keep one character before FROM. */ |
6a140a74 | 2946 | int prev_from_byte; |
1fd3172d | 2947 | int prev_from_syntax; |
195d1361 RS |
2948 | int boundary_stop = commentstop == -1; |
2949 | int nofence; | |
95ff8dfc | 2950 | int found; |
60c86a83 | 2951 | EMACS_INT out_bytepos, out_charpos; |
7c7ca3d2 | 2952 | int temp; |
93da5fff KH |
2953 | |
2954 | prev_from = from; | |
6a140a74 RS |
2955 | prev_from_byte = from_byte; |
2956 | if (from != BEGV) | |
2957 | DEC_BOTH (prev_from, prev_from_byte); | |
93da5fff KH |
2958 | |
2959 | /* Use this macro instead of `from++'. */ | |
6a140a74 RS |
2960 | #define INC_FROM \ |
2961 | do { prev_from = from; \ | |
2962 | prev_from_byte = from_byte; \ | |
327719ee | 2963 | temp = FETCH_CHAR_AS_MULTIBYTE (prev_from_byte); \ |
ab229fdd | 2964 | prev_from_syntax = SYNTAX_WITH_FLAGS (temp); \ |
ef316cf0 | 2965 | INC_BOTH (from, from_byte); \ |
3ceb4629 SM |
2966 | if (from < end) \ |
2967 | UPDATE_SYNTAX_TABLE_FORWARD (from); \ | |
6a140a74 | 2968 | } while (0) |
8489eb67 RS |
2969 | |
2970 | immediate_quit = 1; | |
2971 | QUIT; | |
2972 | ||
265a9e55 | 2973 | if (NILP (oldstate)) |
8489eb67 RS |
2974 | { |
2975 | depth = 0; | |
2976 | state.instring = -1; | |
2977 | state.incomment = 0; | |
195d1361 RS |
2978 | state.comstyle = 0; /* comment style a by default. */ |
2979 | state.comstr_start = -1; /* no comment/string seen. */ | |
8489eb67 RS |
2980 | } |
2981 | else | |
2982 | { | |
2983 | tem = Fcar (oldstate); | |
265a9e55 | 2984 | if (!NILP (tem)) |
8489eb67 RS |
2985 | depth = XINT (tem); |
2986 | else | |
2987 | depth = 0; | |
2988 | ||
2989 | oldstate = Fcdr (oldstate); | |
2990 | oldstate = Fcdr (oldstate); | |
2991 | oldstate = Fcdr (oldstate); | |
2992 | tem = Fcar (oldstate); | |
195d1361 | 2993 | /* Check whether we are inside string_fence-style string: */ |
7d0393cf JB |
2994 | state.instring = (!NILP (tem) |
2995 | ? (INTEGERP (tem) ? XINT (tem) : ST_STRING_STYLE) | |
e76f1c44 | 2996 | : -1); |
8489eb67 RS |
2997 | |
2998 | oldstate = Fcdr (oldstate); | |
2999 | tem = Fcar (oldstate); | |
e76f1c44 GM |
3000 | state.incomment = (!NILP (tem) |
3001 | ? (INTEGERP (tem) ? XINT (tem) : -1) | |
95ff8dfc | 3002 | : 0); |
8489eb67 RS |
3003 | |
3004 | oldstate = Fcdr (oldstate); | |
3005 | tem = Fcar (oldstate); | |
265a9e55 | 3006 | start_quoted = !NILP (tem); |
e5d4f4dc | 3007 | |
95ff8dfc | 3008 | /* if the eighth element of the list is nil, we are in comment |
195d1361 RS |
3009 | style a. If it is non-nil, we are in comment style b */ |
3010 | oldstate = Fcdr (oldstate); | |
e5d4f4dc | 3011 | oldstate = Fcdr (oldstate); |
195d1361 | 3012 | tem = Fcar (oldstate); |
7d0393cf | 3013 | state.comstyle = NILP (tem) ? 0 : (EQ (tem, Qsyntax_table) |
e76f1c44 | 3014 | ? ST_COMMENT_STYLE : 1); |
195d1361 | 3015 | |
e5d4f4dc | 3016 | oldstate = Fcdr (oldstate); |
e5d4f4dc | 3017 | tem = Fcar (oldstate); |
195d1361 | 3018 | state.comstr_start = NILP (tem) ? -1 : XINT (tem) ; |
1a102a58 RS |
3019 | oldstate = Fcdr (oldstate); |
3020 | tem = Fcar (oldstate); | |
3021 | while (!NILP (tem)) /* >= second enclosing sexps. */ | |
3022 | { | |
3023 | /* curlevel++->last ran into compiler bug on Apollo */ | |
3024 | curlevel->last = XINT (Fcar (tem)); | |
3025 | if (++curlevel == endlevel) | |
02010917 | 3026 | curlevel--; /* error ("Nesting too deep for parser"); */ |
1a102a58 RS |
3027 | curlevel->prev = -1; |
3028 | curlevel->last = -1; | |
3029 | tem = Fcdr (tem); | |
3030 | } | |
8489eb67 RS |
3031 | } |
3032 | state.quoted = 0; | |
3033 | mindepth = depth; | |
3034 | ||
3035 | curlevel->prev = -1; | |
3036 | curlevel->last = -1; | |
3037 | ||
326b283b | 3038 | SETUP_SYNTAX_TABLE (prev_from, 1); |
ab229fdd AS |
3039 | temp = FETCH_CHAR (prev_from_byte); |
3040 | prev_from_syntax = SYNTAX_WITH_FLAGS (temp); | |
3ceb4629 | 3041 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
326b283b | 3042 | |
195d1361 | 3043 | /* Enter the loop at a place appropriate for initial state. */ |
8489eb67 | 3044 | |
326b283b RS |
3045 | if (state.incomment) |
3046 | goto startincomment; | |
8489eb67 RS |
3047 | if (state.instring >= 0) |
3048 | { | |
195d1361 | 3049 | nofence = state.instring != ST_STRING_STYLE; |
326b283b RS |
3050 | if (start_quoted) |
3051 | goto startquotedinstring; | |
8489eb67 RS |
3052 | goto startinstring; |
3053 | } | |
326b283b RS |
3054 | else if (start_quoted) |
3055 | goto startquoted; | |
1fd3172d | 3056 | |
8489eb67 RS |
3057 | while (from < end) |
3058 | { | |
93da5fff | 3059 | INC_FROM; |
1fd3172d | 3060 | code = prev_from_syntax & 0xff; |
4c920633 | 3061 | |
c5d4b96f SM |
3062 | if (from < end |
3063 | && SYNTAX_FLAGS_COMSTART_FIRST (prev_from_syntax) | |
3064 | && (c1 = FETCH_CHAR (from_byte), | |
3065 | SYNTAX_COMSTART_SECOND (c1))) | |
3066 | /* Duplicate code to avoid a complex if-expression | |
3067 | which causes trouble for the SGI compiler. */ | |
95ff8dfc | 3068 | { |
c5d4b96f SM |
3069 | /* Record the comment style we have entered so that only |
3070 | the comment-end sequence of the same style actually | |
3071 | terminates the comment section. */ | |
3072 | state.comstyle = SYNTAX_COMMENT_STYLE (c1); | |
3073 | comnested = SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax); | |
3074 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
3075 | state.incomment = comnested ? 1 : -1; | |
95ff8dfc | 3076 | state.comstr_start = prev_from; |
c5d4b96f SM |
3077 | INC_FROM; |
3078 | code = Scomment; | |
95ff8dfc | 3079 | } |
4c920633 | 3080 | else if (code == Scomment_fence) |
e5d4f4dc RS |
3081 | { |
3082 | /* Record the comment style we have entered so that only | |
3083 | the comment-end sequence of the same style actually | |
3084 | terminates the comment section. */ | |
95ff8dfc RS |
3085 | state.comstyle = ST_COMMENT_STYLE; |
3086 | state.incomment = -1; | |
195d1361 | 3087 | state.comstr_start = prev_from; |
e5d4f4dc | 3088 | code = Scomment; |
e5d4f4dc | 3089 | } |
c5d4b96f SM |
3090 | else if (code == Scomment) |
3091 | { | |
3092 | state.comstyle = SYNTAX_FLAGS_COMMENT_STYLE (prev_from_syntax); | |
3093 | state.incomment = (SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax) ? | |
3094 | 1 : -1); | |
3095 | state.comstr_start = prev_from; | |
3096 | } | |
e5d4f4dc | 3097 | |
1fd3172d | 3098 | if (SYNTAX_FLAGS_PREFIX (prev_from_syntax)) |
8489eb67 | 3099 | continue; |
0220c518 | 3100 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
3101 | { |
3102 | case Sescape: | |
3103 | case Scharquote: | |
3104 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
93da5fff | 3105 | curlevel->last = prev_from; |
8489eb67 RS |
3106 | startquoted: |
3107 | if (from == end) goto endquoted; | |
93da5fff | 3108 | INC_FROM; |
8489eb67 RS |
3109 | goto symstarted; |
3110 | /* treat following character as a word constituent */ | |
3111 | case Sword: | |
3112 | case Ssymbol: | |
3113 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
93da5fff | 3114 | curlevel->last = prev_from; |
8489eb67 RS |
3115 | symstarted: |
3116 | while (from < end) | |
3117 | { | |
7c7ca3d2 | 3118 | /* Some compilers can't handle this inside the switch. */ |
327719ee | 3119 | temp = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
ab229fdd | 3120 | temp = SYNTAX (temp); |
7c7ca3d2 | 3121 | switch (temp) |
8489eb67 RS |
3122 | { |
3123 | case Scharquote: | |
3124 | case Sescape: | |
93da5fff | 3125 | INC_FROM; |
8489eb67 RS |
3126 | if (from == end) goto endquoted; |
3127 | break; | |
3128 | case Sword: | |
3129 | case Ssymbol: | |
3130 | case Squote: | |
3131 | break; | |
3132 | default: | |
3133 | goto symdone; | |
3134 | } | |
93da5fff | 3135 | INC_FROM; |
8489eb67 RS |
3136 | } |
3137 | symdone: | |
3138 | curlevel->prev = curlevel->last; | |
3139 | break; | |
3140 | ||
240c43e8 | 3141 | case Scomment_fence: /* Can't happen because it's handled above. */ |
8489eb67 | 3142 | case Scomment: |
195d1361 | 3143 | if (commentstop || boundary_stop) goto done; |
d355bd8a SM |
3144 | startincomment: |
3145 | /* The (from == BEGV) test was to enter the loop in the middle so | |
95ff8dfc | 3146 | that we find a 2-char comment ender even if we start in the |
e6365855 | 3147 | middle of it. We don't want to do that if we're just at the |
d355bd8a | 3148 | beginning of the comment (think of (*) ... (*)). */ |
95ff8dfc RS |
3149 | found = forw_comment (from, from_byte, end, |
3150 | state.incomment, state.comstyle, | |
e6365855 SM |
3151 | (from == BEGV || from < state.comstr_start + 3) |
3152 | ? 0 : prev_from_syntax, | |
95ff8dfc RS |
3153 | &out_charpos, &out_bytepos, &state.incomment); |
3154 | from = out_charpos; from_byte = out_bytepos; | |
3155 | /* Beware! prev_from and friends are invalid now. | |
3156 | Luckily, the `done' doesn't use them and the INC_FROM | |
3157 | sets them to a sane value without looking at them. */ | |
3158 | if (!found) goto done; | |
d355bd8a | 3159 | INC_FROM; |
8489eb67 | 3160 | state.incomment = 0; |
e5d4f4dc | 3161 | state.comstyle = 0; /* reset the comment style */ |
195d1361 | 3162 | if (boundary_stop) goto done; |
8489eb67 RS |
3163 | break; |
3164 | ||
3165 | case Sopen: | |
3166 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
3167 | depth++; | |
3168 | /* curlevel++->last ran into compiler bug on Apollo */ | |
93da5fff | 3169 | curlevel->last = prev_from; |
8489eb67 | 3170 | if (++curlevel == endlevel) |
02010917 | 3171 | curlevel--; /* error ("Nesting too deep for parser"); */ |
8489eb67 RS |
3172 | curlevel->prev = -1; |
3173 | curlevel->last = -1; | |
30844415 | 3174 | if (targetdepth == depth) goto done; |
8489eb67 RS |
3175 | break; |
3176 | ||
3177 | case Sclose: | |
3178 | depth--; | |
3179 | if (depth < mindepth) | |
3180 | mindepth = depth; | |
3181 | if (curlevel != levelstart) | |
3182 | curlevel--; | |
3183 | curlevel->prev = curlevel->last; | |
30844415 | 3184 | if (targetdepth == depth) goto done; |
8489eb67 RS |
3185 | break; |
3186 | ||
3187 | case Sstring: | |
195d1361 RS |
3188 | case Sstring_fence: |
3189 | state.comstr_start = from - 1; | |
8489eb67 | 3190 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ |
93da5fff | 3191 | curlevel->last = prev_from; |
7d0393cf | 3192 | state.instring = (code == Sstring |
b7dbcc19 | 3193 | ? (FETCH_CHAR_AS_MULTIBYTE (prev_from_byte)) |
195d1361 RS |
3194 | : ST_STRING_STYLE); |
3195 | if (boundary_stop) goto done; | |
8489eb67 | 3196 | startinstring: |
195d1361 | 3197 | { |
644ea4df | 3198 | nofence = state.instring != ST_STRING_STYLE; |
7d0393cf | 3199 | |
644ea4df RS |
3200 | while (1) |
3201 | { | |
3202 | int c; | |
195d1361 | 3203 | |
644ea4df | 3204 | if (from >= end) goto done; |
b7dbcc19 | 3205 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
7c7ca3d2 RS |
3206 | /* Some compilers can't handle this inside the switch. */ |
3207 | temp = SYNTAX (c); | |
ccf89641 KH |
3208 | |
3209 | /* Check TEMP here so that if the char has | |
3210 | a syntax-table property which says it is NOT | |
3211 | a string character, it does not end the string. */ | |
3212 | if (nofence && c == state.instring && temp == Sstring) | |
3213 | break; | |
3214 | ||
7c7ca3d2 | 3215 | switch (temp) |
644ea4df RS |
3216 | { |
3217 | case Sstring_fence: | |
3218 | if (!nofence) goto string_end; | |
3219 | break; | |
3220 | case Scharquote: | |
3221 | case Sescape: | |
3222 | INC_FROM; | |
3223 | startquotedinstring: | |
3224 | if (from >= end) goto endquoted; | |
195d1361 | 3225 | } |
644ea4df RS |
3226 | INC_FROM; |
3227 | } | |
195d1361 RS |
3228 | } |
3229 | string_end: | |
8489eb67 RS |
3230 | state.instring = -1; |
3231 | curlevel->prev = curlevel->last; | |
93da5fff | 3232 | INC_FROM; |
195d1361 | 3233 | if (boundary_stop) goto done; |
8489eb67 RS |
3234 | break; |
3235 | ||
3236 | case Smath: | |
240c43e8 SM |
3237 | /* FIXME: We should do something with it. */ |
3238 | break; | |
3239 | default: | |
3240 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
8489eb67 RS |
3241 | break; |
3242 | } | |
3243 | } | |
3244 | goto done; | |
3245 | ||
3246 | stop: /* Here if stopping before start of sexp. */ | |
93da5fff | 3247 | from = prev_from; /* We have just fetched the char that starts it; */ |
8489eb67 RS |
3248 | goto done; /* but return the position before it. */ |
3249 | ||
3250 | endquoted: | |
3251 | state.quoted = 1; | |
3252 | done: | |
3253 | state.depth = depth; | |
3254 | state.mindepth = mindepth; | |
3255 | state.thislevelstart = curlevel->prev; | |
3256 | state.prevlevelstart | |
3257 | = (curlevel == levelstart) ? -1 : (curlevel - 1)->last; | |
3258 | state.location = from; | |
1a102a58 RS |
3259 | state.levelstarts = Qnil; |
3260 | while (--curlevel >= levelstart) | |
3261 | state.levelstarts = Fcons (make_number (curlevel->last), | |
3262 | state.levelstarts); | |
8489eb67 RS |
3263 | immediate_quit = 0; |
3264 | ||
e5d4f4dc | 3265 | *stateptr = state; |
8489eb67 RS |
3266 | } |
3267 | ||
c81a3712 | 3268 | DEFUN ("parse-partial-sexp", Fparse_partial_sexp, Sparse_partial_sexp, 2, 6, 0, |
fdb82f93 PJ |
3269 | doc: /* Parse Lisp syntax starting at FROM until TO; return status of parse at TO. |
3270 | Parsing stops at TO or when certain criteria are met; | |
3271 | point is set to where parsing stops. | |
3272 | If fifth arg OLDSTATE is omitted or nil, | |
3273 | parsing assumes that FROM is the beginning of a function. | |
0e6228bc | 3274 | Value is a list of elements describing final state of parsing: |
fdb82f93 PJ |
3275 | 0. depth in parens. |
3276 | 1. character address of start of innermost containing list; nil if none. | |
3277 | 2. character address of start of last complete sexp terminated. | |
3278 | 3. non-nil if inside a string. | |
3279 | (it is the character that will terminate the string, | |
3280 | or t if the string should be terminated by a generic string delimiter.) | |
7d0393cf | 3281 | 4. nil if outside a comment, t if inside a non-nestable comment, |
fdb82f93 PJ |
3282 | else an integer (the current comment nesting). |
3283 | 5. t if following a quote character. | |
3284 | 6. the minimum paren-depth encountered during this scan. | |
3285 | 7. t if in a comment of style b; symbol `syntax-table' if the comment | |
3286 | should be terminated by a generic comment delimiter. | |
3287 | 8. character address of start of comment or string; nil if not in one. | |
3288 | 9. Intermediate data for continuation of parsing (subject to change). | |
3289 | If third arg TARGETDEPTH is non-nil, parsing stops if the depth | |
3290 | in parentheses becomes equal to TARGETDEPTH. | |
3291 | Fourth arg STOPBEFORE non-nil means stop when come to | |
3292 | any character that starts a sexp. | |
0e6228bc | 3293 | Fifth arg OLDSTATE is a list like what this function returns. |
fdb82f93 | 3294 | It is used to initialize the state of the parse. Elements number 1, 2, 6 |
43726c05 | 3295 | and 8 are ignored. |
fdb82f93 PJ |
3296 | Sixth arg COMMENTSTOP non-nil means stop at the start of a comment. |
3297 | If it is symbol `syntax-table', stop after the start of a comment or a | |
3298 | string, or after end of a comment or a string. */) | |
3299 | (from, to, targetdepth, stopbefore, oldstate, commentstop) | |
c81a3712 | 3300 | Lisp_Object from, to, targetdepth, stopbefore, oldstate, commentstop; |
8489eb67 RS |
3301 | { |
3302 | struct lisp_parse_state state; | |
3303 | int target; | |
3304 | ||
265a9e55 | 3305 | if (!NILP (targetdepth)) |
8489eb67 | 3306 | { |
b7826503 | 3307 | CHECK_NUMBER (targetdepth); |
8489eb67 RS |
3308 | target = XINT (targetdepth); |
3309 | } | |
3310 | else | |
3311 | target = -100000; /* We won't reach this depth */ | |
3312 | ||
3313 | validate_region (&from, &to); | |
6a140a74 RS |
3314 | scan_sexps_forward (&state, XINT (from), CHAR_TO_BYTE (XINT (from)), |
3315 | XINT (to), | |
c81a3712 | 3316 | target, !NILP (stopbefore), oldstate, |
7d0393cf | 3317 | (NILP (commentstop) |
195d1361 | 3318 | ? 0 : (EQ (commentstop, Qsyntax_table) ? -1 : 1))); |
8489eb67 RS |
3319 | |
3320 | SET_PT (state.location); | |
7d0393cf | 3321 | |
8489eb67 RS |
3322 | return Fcons (make_number (state.depth), |
3323 | Fcons (state.prevlevelstart < 0 ? Qnil : make_number (state.prevlevelstart), | |
3324 | Fcons (state.thislevelstart < 0 ? Qnil : make_number (state.thislevelstart), | |
7d0393cf JB |
3325 | Fcons (state.instring >= 0 |
3326 | ? (state.instring == ST_STRING_STYLE | |
195d1361 | 3327 | ? Qt : make_number (state.instring)) : Qnil, |
95ff8dfc RS |
3328 | Fcons (state.incomment < 0 ? Qt : |
3329 | (state.incomment == 0 ? Qnil : | |
3330 | make_number (state.incomment)), | |
8489eb67 | 3331 | Fcons (state.quoted ? Qt : Qnil, |
bff37d2a | 3332 | Fcons (make_number (state.mindepth), |
7d0393cf | 3333 | Fcons ((state.comstyle |
bff37d2a RS |
3334 | ? (state.comstyle == ST_COMMENT_STYLE |
3335 | ? Qsyntax_table : Qt) : | |
3336 | Qnil), | |
0beaf54f DL |
3337 | Fcons (((state.incomment |
3338 | || (state.instring >= 0)) | |
bff37d2a RS |
3339 | ? make_number (state.comstr_start) |
3340 | : Qnil), | |
1a102a58 | 3341 | Fcons (state.levelstarts, Qnil)))))))))); |
8489eb67 RS |
3342 | } |
3343 | \f | |
dfcf069d | 3344 | void |
8489eb67 RS |
3345 | init_syntax_once () |
3346 | { | |
78f9a1f7 | 3347 | register int i, c; |
8ea151b2 | 3348 | Lisp_Object temp; |
8489eb67 | 3349 | |
5ebaddf5 RS |
3350 | /* This has to be done here, before we call Fmake_char_table. */ |
3351 | Qsyntax_table = intern ("syntax-table"); | |
3352 | staticpro (&Qsyntax_table); | |
3353 | ||
3354 | /* Intern this now in case it isn't already done. | |
3355 | Setting this variable twice is harmless. | |
3356 | But don't staticpro it here--that is done in alloc.c. */ | |
3357 | Qchar_table_extra_slots = intern ("char-table-extra-slots"); | |
3358 | ||
93da5fff | 3359 | /* Create objects which can be shared among syntax tables. */ |
42ebfa31 | 3360 | Vsyntax_code_object = Fmake_vector (make_number (Smax), Qnil); |
93da5fff KH |
3361 | for (i = 0; i < XVECTOR (Vsyntax_code_object)->size; i++) |
3362 | XVECTOR (Vsyntax_code_object)->contents[i] | |
3363 | = Fcons (make_number (i), Qnil); | |
3364 | ||
5ebaddf5 RS |
3365 | /* Now we are ready to set up this property, so we can |
3366 | create syntax tables. */ | |
3367 | Fput (Qsyntax_table, Qchar_table_extra_slots, make_number (0)); | |
3368 | ||
93da5fff | 3369 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Swhitespace]; |
8489eb67 | 3370 | |
5ebaddf5 | 3371 | Vstandard_syntax_table = Fmake_char_table (Qsyntax_table, temp); |
8489eb67 | 3372 | |
aa7b08b4 RS |
3373 | /* Control characters should not be whitespace. */ |
3374 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Spunct]; | |
3375 | for (i = 0; i <= ' ' - 1; i++) | |
3376 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); | |
3377 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 0177, temp); | |
3378 | ||
3379 | /* Except that a few really are whitespace. */ | |
3380 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Swhitespace]; | |
3381 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ' ', temp); | |
3382 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\t', temp); | |
3383 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\n', temp); | |
3384 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 015, temp); | |
3385 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 014, temp); | |
3386 | ||
93da5fff | 3387 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Sword]; |
8489eb67 | 3388 | for (i = 'a'; i <= 'z'; i++) |
8ea151b2 | 3389 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
8489eb67 | 3390 | for (i = 'A'; i <= 'Z'; i++) |
8ea151b2 | 3391 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
8489eb67 | 3392 | for (i = '0'; i <= '9'; i++) |
8ea151b2 RS |
3393 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
3394 | ||
3395 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '$', temp); | |
3396 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '%', temp); | |
3397 | ||
3398 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '(', | |
3399 | Fcons (make_number (Sopen), make_number (')'))); | |
3400 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ')', | |
3401 | Fcons (make_number (Sclose), make_number ('('))); | |
3402 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '[', | |
3403 | Fcons (make_number (Sopen), make_number (']'))); | |
3404 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ']', | |
3405 | Fcons (make_number (Sclose), make_number ('['))); | |
3406 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '{', | |
3407 | Fcons (make_number (Sopen), make_number ('}'))); | |
3408 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '}', | |
3409 | Fcons (make_number (Sclose), make_number ('{'))); | |
3410 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '"', | |
3411 | Fcons (make_number ((int) Sstring), Qnil)); | |
3412 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\\', | |
3413 | Fcons (make_number ((int) Sescape), Qnil)); | |
3414 | ||
93da5fff | 3415 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Ssymbol]; |
8489eb67 | 3416 | for (i = 0; i < 10; i++) |
78f9a1f7 KH |
3417 | { |
3418 | c = "_-+*/&|<>="[i]; | |
3419 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, c, temp); | |
3420 | } | |
8489eb67 | 3421 | |
93da5fff | 3422 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Spunct]; |
8489eb67 | 3423 | for (i = 0; i < 12; i++) |
78f9a1f7 KH |
3424 | { |
3425 | c = ".,;:?!#@~^'`"[i]; | |
3426 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, c, temp); | |
3427 | } | |
bd25db08 KH |
3428 | |
3429 | /* All multibyte characters have syntax `word' by default. */ | |
3430 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Sword]; | |
5c7b02ab | 3431 | char_table_set_range (Vstandard_syntax_table, 0x80, MAX_CHAR, temp); |
8489eb67 RS |
3432 | } |
3433 | ||
dfcf069d | 3434 | void |
8489eb67 RS |
3435 | syms_of_syntax () |
3436 | { | |
3437 | Qsyntax_table_p = intern ("syntax-table-p"); | |
3438 | staticpro (&Qsyntax_table_p); | |
3439 | ||
93da5fff KH |
3440 | staticpro (&Vsyntax_code_object); |
3441 | ||
a2994c46 KS |
3442 | staticpro (&gl_state.object); |
3443 | staticpro (&gl_state.global_code); | |
3444 | staticpro (&gl_state.current_syntax_table); | |
3445 | staticpro (&gl_state.old_prop); | |
3446 | ||
3447 | /* Defined in regex.c */ | |
3448 | staticpro (&re_match_object); | |
3449 | ||
7bf5e9e4 RS |
3450 | Qscan_error = intern ("scan-error"); |
3451 | staticpro (&Qscan_error); | |
3452 | Fput (Qscan_error, Qerror_conditions, | |
f3be100f | 3453 | Fcons (Qscan_error, Fcons (Qerror, Qnil))); |
7bf5e9e4 RS |
3454 | Fput (Qscan_error, Qerror_message, |
3455 | build_string ("Scan error")); | |
3456 | ||
8489eb67 | 3457 | DEFVAR_BOOL ("parse-sexp-ignore-comments", &parse_sexp_ignore_comments, |
fdb82f93 | 3458 | doc: /* Non-nil means `forward-sexp', etc., should treat comments as whitespace. */); |
8489eb67 | 3459 | |
195d1361 | 3460 | DEFVAR_BOOL ("parse-sexp-lookup-properties", &parse_sexp_lookup_properties, |
fdb82f93 PJ |
3461 | doc: /* Non-nil means `forward-sexp', etc., obey `syntax-table' property. |
3462 | Otherwise, that text property is simply ignored. | |
3463 | See the info node `(elisp)Syntax Properties' for a description of the | |
3464 | `syntax-table' property. */); | |
195d1361 | 3465 | |
8489eb67 RS |
3466 | words_include_escapes = 0; |
3467 | DEFVAR_BOOL ("words-include-escapes", &words_include_escapes, | |
fdb82f93 | 3468 | doc: /* Non-nil means `forward-word', etc., should treat escape chars part of words. */); |
8489eb67 | 3469 | |
bd25db08 | 3470 | DEFVAR_BOOL ("multibyte-syntax-as-symbol", &multibyte_syntax_as_symbol, |
fdb82f93 | 3471 | doc: /* Non-nil means `scan-sexps' treats all multibyte characters as symbol. */); |
bd25db08 KH |
3472 | multibyte_syntax_as_symbol = 0; |
3473 | ||
f4ed767f GM |
3474 | DEFVAR_BOOL ("open-paren-in-column-0-is-defun-start", |
3475 | &open_paren_in_column_0_is_defun_start, | |
b222e415 | 3476 | doc: /* *Non-nil means an open paren in column 0 denotes the start of a defun. */); |
f4ed767f GM |
3477 | open_paren_in_column_0_is_defun_start = 1; |
3478 | ||
8f924df7 KH |
3479 | |
3480 | DEFVAR_LISP ("find-word-boundary-function-table", | |
3481 | &Vfind_word_boundary_function_table, | |
869bb237 | 3482 | doc: /* |
8f924df7 | 3483 | Char table of functions to search for the word boundary. |
869bb237 KH |
3484 | Each function is called with two arguments; POS and LIMIT. |
3485 | POS and LIMIT are character positions in the current buffer. | |
3486 | ||
3487 | If POS is less than LIMIT, POS is at the first character of a word, | |
3488 | and the return value of a function is a position after the last | |
3489 | character of that word. | |
3490 | ||
3491 | If POS is not less than LIMIT, POS is at the last character of a word, | |
3492 | and the return value of a function is a position at the first | |
3493 | character of that word. | |
3494 | ||
3495 | In both cases, LIMIT bounds the search. */); | |
8f924df7 | 3496 | Vfind_word_boundary_function_table = Fmake_char_table (Qnil, Qnil); |
869bb237 | 3497 | |
8489eb67 RS |
3498 | defsubr (&Ssyntax_table_p); |
3499 | defsubr (&Ssyntax_table); | |
3500 | defsubr (&Sstandard_syntax_table); | |
3501 | defsubr (&Scopy_syntax_table); | |
3502 | defsubr (&Sset_syntax_table); | |
3503 | defsubr (&Schar_syntax); | |
beefa22e | 3504 | defsubr (&Smatching_paren); |
c65adb44 | 3505 | defsubr (&Sstring_to_syntax); |
8489eb67 | 3506 | defsubr (&Smodify_syntax_entry); |
62abe9cb | 3507 | defsubr (&Sinternal_describe_syntax_value); |
8489eb67 RS |
3508 | |
3509 | defsubr (&Sforward_word); | |
3510 | ||
195d1361 RS |
3511 | defsubr (&Sskip_chars_forward); |
3512 | defsubr (&Sskip_chars_backward); | |
3513 | defsubr (&Sskip_syntax_forward); | |
3514 | defsubr (&Sskip_syntax_backward); | |
3515 | ||
b3cfe0c8 | 3516 | defsubr (&Sforward_comment); |
8489eb67 RS |
3517 | defsubr (&Sscan_lists); |
3518 | defsubr (&Sscan_sexps); | |
3519 | defsubr (&Sbackward_prefix_chars); | |
3520 | defsubr (&Sparse_partial_sexp); | |
3521 | } | |
839966f3 KH |
3522 | |
3523 | /* arch-tag: 3e297b9f-088e-4b64-8f4c-fb0b3443e412 | |
3524 | (do not change this comment) */ |