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, |
76b6f707 | 3 | 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009 |
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 | 23 | #include <ctype.h> |
d7306fe6 | 24 | #include <setjmp.h> |
8489eb67 RS |
25 | #include "lisp.h" |
26 | #include "commands.h" | |
27 | #include "buffer.h" | |
5c7b02ab | 28 | #include "character.h" |
e35f6ff7 | 29 | #include "keymap.h" |
a1bc88d4 | 30 | #include "regex.h" |
195d1361 RS |
31 | |
32 | /* Make syntax table lookup grant data in gl_state. */ | |
33 | #define SYNTAX_ENTRY_VIA_PROPERTY | |
34 | ||
8489eb67 | 35 | #include "syntax.h" |
195d1361 RS |
36 | #include "intervals.h" |
37 | ||
38 | /* We use these constants in place for comment-style and | |
39 | string-ender-char to distinguish comments/strings started by | |
40 | comment_fence and string_fence codes. */ | |
41 | ||
42 | #define ST_COMMENT_STYLE (256 + 1) | |
43 | #define ST_STRING_STYLE (256 + 2) | |
93da5fff | 44 | #include "category.h" |
8489eb67 | 45 | |
7bf5e9e4 | 46 | Lisp_Object Qsyntax_table_p, Qsyntax_table, Qscan_error; |
8489eb67 RS |
47 | |
48 | int words_include_escapes; | |
195d1361 | 49 | int parse_sexp_lookup_properties; |
8489eb67 | 50 | |
bd25db08 KH |
51 | /* Nonzero means `scan-sexps' treat all multibyte characters as symbol. */ |
52 | int multibyte_syntax_as_symbol; | |
53 | ||
8ea151b2 RS |
54 | /* Used as a temporary in SYNTAX_ENTRY and other macros in syntax.h, |
55 | if not compiled with GCC. No need to mark it, since it is used | |
56 | only very temporarily. */ | |
57 | Lisp_Object syntax_temp; | |
58 | ||
f4ed767f GM |
59 | /* Non-zero means an open parenthesis in column 0 is always considered |
60 | to be the start of a defun. Zero means an open parenthesis in | |
61 | column 0 has no special meaning. */ | |
62 | ||
63 | int open_paren_in_column_0_is_defun_start; | |
64 | ||
e5d4f4dc RS |
65 | /* This is the internal form of the parse state used in parse-partial-sexp. */ |
66 | ||
67 | struct lisp_parse_state | |
68 | { | |
6449674e SM |
69 | int depth; /* Depth at end of parsing. */ |
70 | int instring; /* -1 if not within string, else desired terminator. */ | |
71 | int incomment; /* -1 if in unnestable comment else comment nesting */ | |
72 | int comstyle; /* comment style a=0, or b=1, or ST_COMMENT_STYLE. */ | |
73 | int quoted; /* Nonzero if just after an escape char at end of parsing */ | |
74 | int mindepth; /* Minimum depth seen while scanning. */ | |
75 | /* Char number of most recent start-of-expression at current level */ | |
76 | EMACS_INT thislevelstart; | |
77 | /* Char number of start of containing expression */ | |
78 | EMACS_INT prevlevelstart; | |
79 | EMACS_INT location; /* Char number at which parsing stopped. */ | |
80 | EMACS_INT comstr_start; /* Position of last comment/string starter. */ | |
81 | Lisp_Object levelstarts; /* Char numbers of starts-of-expression | |
82 | of levels (starting from outermost). */ | |
e5d4f4dc RS |
83 | }; |
84 | \f | |
37bef230 RS |
85 | /* These variables are a cache for finding the start of a defun. |
86 | find_start_pos is the place for which the defun start was found. | |
87 | find_start_value is the defun start position found for it. | |
6a140a74 | 88 | find_start_value_byte is the corresponding byte position. |
37bef230 RS |
89 | find_start_buffer is the buffer it was found in. |
90 | find_start_begv is the BEGV value when it was found. | |
91 | find_start_modiff is the value of MODIFF when it was found. */ | |
92 | ||
8e2911c2 AS |
93 | static EMACS_INT find_start_pos; |
94 | static EMACS_INT find_start_value; | |
95 | static EMACS_INT find_start_value_byte; | |
37bef230 | 96 | static struct buffer *find_start_buffer; |
8e2911c2 | 97 | static EMACS_INT find_start_begv; |
37bef230 | 98 | static int find_start_modiff; |
6a140a74 RS |
99 | |
100 | ||
327719ee | 101 | static Lisp_Object skip_chars P_ ((int, Lisp_Object, Lisp_Object, int)); |
b7dbcc19 | 102 | static Lisp_Object skip_syntaxes P_ ((int, Lisp_Object, Lisp_Object)); |
932e6895 | 103 | static Lisp_Object scan_lists P_ ((EMACS_INT, EMACS_INT, EMACS_INT, int)); |
6a140a74 | 104 | static void scan_sexps_forward P_ ((struct lisp_parse_state *, |
8e2911c2 | 105 | EMACS_INT, EMACS_INT, EMACS_INT, int, |
6a140a74 | 106 | int, Lisp_Object, int)); |
a1bc88d4 | 107 | static int in_classes P_ ((int, Lisp_Object)); |
195d1361 RS |
108 | \f |
109 | ||
110 | struct gl_state_s gl_state; /* Global state of syntax parser. */ | |
111 | ||
112 | INTERVAL interval_of (); | |
113 | #define INTERVALS_AT_ONCE 10 /* 1 + max-number of intervals | |
114 | to scan to property-change. */ | |
115 | ||
6a140a74 RS |
116 | /* Update gl_state to an appropriate interval which contains CHARPOS. The |
117 | sign of COUNT give the relative position of CHARPOS wrt the previously | |
195d1361 | 118 | valid interval. If INIT, only [be]_property fields of gl_state are |
6a140a74 | 119 | valid at start, the rest is filled basing on OBJECT. |
195d1361 | 120 | |
6a140a74 | 121 | `gl_state.*_i' are the intervals, and CHARPOS is further in the search |
195d1361 RS |
122 | direction than the intervals - or in an interval. We update the |
123 | current syntax-table basing on the property of this interval, and | |
6a140a74 | 124 | update the interval to start further than CHARPOS - or be |
195d1361 | 125 | NULL_INTERVAL. We also update lim_property to be the next value of |
6a140a74 | 126 | charpos to call this subroutine again - or be before/after the |
195d1361 RS |
127 | start/end of OBJECT. */ |
128 | ||
129 | void | |
6a140a74 RS |
130 | update_syntax_table (charpos, count, init, object) |
131 | int charpos, count, init; | |
195d1361 RS |
132 | Lisp_Object object; |
133 | { | |
134 | Lisp_Object tmp_table; | |
4ffe723b | 135 | int cnt = 0, invalidate = 1; |
5a774522 | 136 | INTERVAL i; |
195d1361 RS |
137 | |
138 | if (init) | |
139 | { | |
bb0de084 | 140 | gl_state.old_prop = Qnil; |
195d1361 RS |
141 | gl_state.start = gl_state.b_property; |
142 | gl_state.stop = gl_state.e_property; | |
bb0de084 SM |
143 | i = interval_of (charpos, object); |
144 | gl_state.backward_i = gl_state.forward_i = i; | |
195d1361 RS |
145 | invalidate = 0; |
146 | if (NULL_INTERVAL_P (i)) | |
147 | return; | |
f902a008 | 148 | /* interval_of updates only ->position of the return value, so |
d80f4cc7 | 149 | update the parents manually to speed up update_interval. */ |
bb0de084 | 150 | while (!NULL_PARENT (i)) |
d80f4cc7 RS |
151 | { |
152 | if (AM_RIGHT_CHILD (i)) | |
439d5cb4 | 153 | INTERVAL_PARENT (i)->position = i->position |
d80f4cc7 | 154 | - LEFT_TOTAL_LENGTH (i) + TOTAL_LENGTH (i) /* right end */ |
439d5cb4 KR |
155 | - TOTAL_LENGTH (INTERVAL_PARENT (i)) |
156 | + LEFT_TOTAL_LENGTH (INTERVAL_PARENT (i)); | |
d80f4cc7 | 157 | else |
439d5cb4 | 158 | INTERVAL_PARENT (i)->position = i->position - LEFT_TOTAL_LENGTH (i) |
d80f4cc7 | 159 | + TOTAL_LENGTH (i); |
439d5cb4 | 160 | i = INTERVAL_PARENT (i); |
d80f4cc7 RS |
161 | } |
162 | i = gl_state.forward_i; | |
bb0de084 | 163 | gl_state.b_property = i->position - gl_state.offset; |
ee0cdb48 | 164 | gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset; |
195d1361 RS |
165 | goto update; |
166 | } | |
5a774522 | 167 | i = count > 0 ? gl_state.forward_i : gl_state.backward_i; |
195d1361 | 168 | |
423e705d | 169 | /* We are guaranteed to be called with CHARPOS either in i, |
6a140a74 | 170 | or further off. */ |
195d1361 RS |
171 | if (NULL_INTERVAL_P (i)) |
172 | error ("Error in syntax_table logic for to-the-end intervals"); | |
6a140a74 | 173 | else if (charpos < i->position) /* Move left. */ |
195d1361 RS |
174 | { |
175 | if (count > 0) | |
f902a008 | 176 | error ("Error in syntax_table logic for intervals <-"); |
195d1361 | 177 | /* Update the interval. */ |
6a140a74 | 178 | i = update_interval (i, charpos); |
bb0de084 | 179 | if (INTERVAL_LAST_POS (i) != gl_state.b_property) |
195d1361 RS |
180 | { |
181 | invalidate = 0; | |
195d1361 | 182 | gl_state.forward_i = i; |
ee0cdb48 | 183 | gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset; |
195d1361 | 184 | } |
423e705d | 185 | } |
6a140a74 | 186 | else if (charpos >= INTERVAL_LAST_POS (i)) /* Move right. */ |
195d1361 RS |
187 | { |
188 | if (count < 0) | |
f902a008 | 189 | error ("Error in syntax_table logic for intervals ->"); |
195d1361 | 190 | /* Update the interval. */ |
6a140a74 | 191 | i = update_interval (i, charpos); |
bb0de084 | 192 | if (i->position != gl_state.e_property) |
195d1361 RS |
193 | { |
194 | invalidate = 0; | |
195d1361 | 195 | gl_state.backward_i = i; |
bb0de084 | 196 | gl_state.b_property = i->position - gl_state.offset; |
195d1361 RS |
197 | } |
198 | } | |
195d1361 RS |
199 | |
200 | update: | |
201 | tmp_table = textget (i->plist, Qsyntax_table); | |
202 | ||
203 | if (invalidate) | |
204 | invalidate = !EQ (tmp_table, gl_state.old_prop); /* Need to invalidate? */ | |
7d0393cf | 205 | |
423e705d SM |
206 | if (invalidate) /* Did not get to adjacent interval. */ |
207 | { /* with the same table => */ | |
208 | /* invalidate the old range. */ | |
195d1361 RS |
209 | if (count > 0) |
210 | { | |
211 | gl_state.backward_i = i; | |
bb0de084 SM |
212 | gl_state.b_property = i->position - gl_state.offset; |
213 | } | |
214 | else | |
195d1361 | 215 | { |
bb0de084 | 216 | gl_state.forward_i = i; |
ee0cdb48 | 217 | gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset; |
195d1361 RS |
218 | } |
219 | } | |
37bef230 | 220 | |
bb0de084 | 221 | if (!EQ (tmp_table, gl_state.old_prop)) |
195d1361 | 222 | { |
bb0de084 SM |
223 | gl_state.current_syntax_table = tmp_table; |
224 | gl_state.old_prop = tmp_table; | |
225 | if (EQ (Fsyntax_table_p (tmp_table), Qt)) | |
226 | { | |
227 | gl_state.use_global = 0; | |
7d0393cf | 228 | } |
bb0de084 SM |
229 | else if (CONSP (tmp_table)) |
230 | { | |
231 | gl_state.use_global = 1; | |
232 | gl_state.global_code = tmp_table; | |
233 | } | |
7d0393cf | 234 | else |
bb0de084 SM |
235 | { |
236 | gl_state.use_global = 0; | |
237 | gl_state.current_syntax_table = current_buffer->syntax_table; | |
238 | } | |
195d1361 RS |
239 | } |
240 | ||
241 | while (!NULL_INTERVAL_P (i)) | |
242 | { | |
243 | if (cnt && !EQ (tmp_table, textget (i->plist, Qsyntax_table))) | |
244 | { | |
245 | if (count > 0) | |
bb0de084 SM |
246 | { |
247 | gl_state.e_property = i->position - gl_state.offset; | |
248 | gl_state.forward_i = i; | |
249 | } | |
250 | else | |
251 | { | |
5a774522 SM |
252 | gl_state.b_property |
253 | = i->position + LENGTH (i) - gl_state.offset; | |
bb0de084 SM |
254 | gl_state.backward_i = i; |
255 | } | |
256 | return; | |
195d1361 | 257 | } |
7d0393cf | 258 | else if (cnt == INTERVALS_AT_ONCE) |
195d1361 RS |
259 | { |
260 | if (count > 0) | |
bb0de084 | 261 | { |
5a774522 SM |
262 | gl_state.e_property |
263 | = i->position + LENGTH (i) - gl_state.offset | |
264 | /* e_property at EOB is not set to ZV but to ZV+1, so that | |
265 | we can do INC(from);UPDATE_SYNTAX_TABLE_FORWARD without | |
266 | having to check eob between the two. */ | |
267 | + (NULL_INTERVAL_P (next_interval (i)) ? 1 : 0); | |
bb0de084 SM |
268 | gl_state.forward_i = i; |
269 | } | |
270 | else | |
271 | { | |
272 | gl_state.b_property = i->position - gl_state.offset; | |
273 | gl_state.backward_i = i; | |
274 | } | |
275 | return; | |
195d1361 RS |
276 | } |
277 | cnt++; | |
278 | i = count > 0 ? next_interval (i) : previous_interval (i); | |
279 | } | |
bb0de084 SM |
280 | eassert (NULL_INTERVAL_P (i)); /* This property goes to the end. */ |
281 | if (count > 0) | |
282 | gl_state.e_property = gl_state.stop; | |
283 | else | |
284 | gl_state.b_property = gl_state.start; | |
195d1361 RS |
285 | } |
286 | \f | |
6a140a74 RS |
287 | /* Returns TRUE if char at CHARPOS is quoted. |
288 | Global syntax-table data should be set up already to be good at CHARPOS | |
289 | or after. On return global syntax data is good for lookup at CHARPOS. */ | |
195d1361 RS |
290 | |
291 | static int | |
6449674e | 292 | char_quoted (EMACS_INT charpos, EMACS_INT bytepos) |
195d1361 RS |
293 | { |
294 | register enum syntaxcode code; | |
6449674e | 295 | register EMACS_INT beg = BEGV; |
195d1361 | 296 | register int quoted = 0; |
6449674e | 297 | EMACS_INT orig = charpos; |
6a140a74 | 298 | |
02d8b017 | 299 | while (charpos > beg) |
195d1361 | 300 | { |
ab229fdd | 301 | int c; |
02d8b017 | 302 | DEC_BOTH (charpos, bytepos); |
ab229fdd | 303 | |
6a140a74 | 304 | UPDATE_SYNTAX_TABLE_BACKWARD (charpos); |
327719ee | 305 | c = FETCH_CHAR_AS_MULTIBYTE (bytepos); |
ab229fdd | 306 | code = SYNTAX (c); |
6a140a74 RS |
307 | if (! (code == Scharquote || code == Sescape)) |
308 | break; | |
309 | ||
6a140a74 | 310 | quoted = !quoted; |
195d1361 | 311 | } |
6a140a74 RS |
312 | |
313 | UPDATE_SYNTAX_TABLE (orig); | |
195d1361 RS |
314 | return quoted; |
315 | } | |
6a140a74 RS |
316 | |
317 | /* Return the bytepos one character after BYTEPOS. | |
318 | We assume that BYTEPOS is not at the end of the buffer. */ | |
319 | ||
6449674e | 320 | INLINE EMACS_INT |
6a140a74 | 321 | inc_bytepos (bytepos) |
6449674e | 322 | EMACS_INT bytepos; |
6a140a74 | 323 | { |
ef316cf0 RS |
324 | if (NILP (current_buffer->enable_multibyte_characters)) |
325 | return bytepos + 1; | |
326 | ||
6a140a74 RS |
327 | INC_POS (bytepos); |
328 | return bytepos; | |
329 | } | |
330 | ||
331 | /* Return the bytepos one character before BYTEPOS. | |
332 | We assume that BYTEPOS is not at the start of the buffer. */ | |
333 | ||
6449674e | 334 | INLINE EMACS_INT |
6a140a74 | 335 | dec_bytepos (bytepos) |
6449674e | 336 | EMACS_INT bytepos; |
6a140a74 | 337 | { |
ef316cf0 RS |
338 | if (NILP (current_buffer->enable_multibyte_characters)) |
339 | return bytepos - 1; | |
340 | ||
6a140a74 RS |
341 | DEC_POS (bytepos); |
342 | return bytepos; | |
343 | } | |
195d1361 | 344 | \f |
2b34df4e | 345 | /* Return a defun-start position before POS and not too far before. |
7cc80f0a RS |
346 | It should be the last one before POS, or nearly the last. |
347 | ||
348 | When open_paren_in_column_0_is_defun_start is nonzero, | |
1fd1cc2f | 349 | only the beginning of the buffer is treated as a defun-start. |
7cc80f0a RS |
350 | |
351 | We record the information about where the scan started | |
352 | and what its result was, so that another call in the same area | |
353 | can return the same value very quickly. | |
195d1361 RS |
354 | |
355 | There is no promise at which position the global syntax data is | |
356 | valid on return from the subroutine, so the caller should explicitly | |
357 | update the global data. */ | |
37bef230 | 358 | |
6449674e | 359 | static EMACS_INT |
6a140a74 | 360 | find_defun_start (pos, pos_byte) |
8e2911c2 | 361 | EMACS_INT pos, pos_byte; |
37bef230 | 362 | { |
8e2911c2 | 363 | EMACS_INT opoint = PT, opoint_byte = PT_BYTE; |
37bef230 | 364 | |
1fd1cc2f RS |
365 | if (!open_paren_in_column_0_is_defun_start) |
366 | { | |
367 | find_start_value_byte = BEGV_BYTE; | |
368 | return BEGV; | |
369 | } | |
370 | ||
37bef230 RS |
371 | /* Use previous finding, if it's valid and applies to this inquiry. */ |
372 | if (current_buffer == find_start_buffer | |
373 | /* Reuse the defun-start even if POS is a little farther on. | |
374 | POS might be in the next defun, but that's ok. | |
375 | Our value may not be the best possible, but will still be usable. */ | |
376 | && pos <= find_start_pos + 1000 | |
377 | && pos >= find_start_value | |
378 | && BEGV == find_start_begv | |
379 | && MODIFF == find_start_modiff) | |
380 | return find_start_value; | |
381 | ||
382 | /* Back up to start of line. */ | |
6a140a74 | 383 | scan_newline (pos, pos_byte, BEGV, BEGV_BYTE, -1, 1); |
37bef230 | 384 | |
195d1361 RS |
385 | /* We optimize syntax-table lookup for rare updates. Thus we accept |
386 | only those `^\s(' which are good in global _and_ text-property | |
387 | syntax-tables. */ | |
388 | gl_state.current_syntax_table = current_buffer->syntax_table; | |
389 | gl_state.use_global = 0; | |
1fd1cc2f | 390 | while (PT > BEGV) |
37bef230 | 391 | { |
ab229fdd AS |
392 | int c; |
393 | ||
1fd1cc2f RS |
394 | /* Open-paren at start of line means we may have found our |
395 | defun-start. */ | |
327719ee | 396 | c = FETCH_CHAR_AS_MULTIBYTE (PT_BYTE); |
ab229fdd | 397 | if (SYNTAX (c) == Sopen) |
195d1361 | 398 | { |
1fd1cc2f | 399 | SETUP_SYNTAX_TABLE (PT + 1, -1); /* Try again... */ |
327719ee | 400 | c = FETCH_CHAR_AS_MULTIBYTE (PT_BYTE); |
ab229fdd | 401 | if (SYNTAX (c) == Sopen) |
1fd1cc2f RS |
402 | break; |
403 | /* Now fallback to the default value. */ | |
404 | gl_state.current_syntax_table = current_buffer->syntax_table; | |
405 | gl_state.use_global = 0; | |
195d1361 | 406 | } |
1fd1cc2f RS |
407 | /* Move to beg of previous line. */ |
408 | scan_newline (PT, PT_BYTE, BEGV, BEGV_BYTE, -2, 1); | |
37bef230 RS |
409 | } |
410 | ||
411 | /* Record what we found, for the next try. */ | |
6a140a74 RS |
412 | find_start_value = PT; |
413 | find_start_value_byte = PT_BYTE; | |
37bef230 RS |
414 | find_start_buffer = current_buffer; |
415 | find_start_modiff = MODIFF; | |
416 | find_start_begv = BEGV; | |
417 | find_start_pos = pos; | |
418 | ||
6a140a74 RS |
419 | TEMP_SET_PT_BOTH (opoint, opoint_byte); |
420 | ||
37bef230 RS |
421 | return find_start_value; |
422 | } | |
423 | \f | |
f902a008 RS |
424 | /* Return the SYNTAX_COMEND_FIRST of the character before POS, POS_BYTE. */ |
425 | ||
426 | static int | |
427 | prev_char_comend_first (pos, pos_byte) | |
428 | int pos, pos_byte; | |
429 | { | |
430 | int c, val; | |
431 | ||
432 | DEC_BOTH (pos, pos_byte); | |
433 | UPDATE_SYNTAX_TABLE_BACKWARD (pos); | |
434 | c = FETCH_CHAR (pos_byte); | |
435 | val = SYNTAX_COMEND_FIRST (c); | |
436 | UPDATE_SYNTAX_TABLE_FORWARD (pos + 1); | |
437 | return val; | |
438 | } | |
439 | ||
440 | /* Return the SYNTAX_COMSTART_FIRST of the character before POS, POS_BYTE. */ | |
441 | ||
3f679f55 SM |
442 | /* static int |
443 | * prev_char_comstart_first (pos, pos_byte) | |
444 | * int pos, pos_byte; | |
445 | * { | |
446 | * int c, val; | |
7d0393cf | 447 | * |
3f679f55 SM |
448 | * DEC_BOTH (pos, pos_byte); |
449 | * UPDATE_SYNTAX_TABLE_BACKWARD (pos); | |
450 | * c = FETCH_CHAR (pos_byte); | |
451 | * val = SYNTAX_COMSTART_FIRST (c); | |
452 | * UPDATE_SYNTAX_TABLE_FORWARD (pos + 1); | |
453 | * return val; | |
454 | * } */ | |
f902a008 | 455 | |
6a140a74 RS |
456 | /* Checks whether charpos FROM is at the end of a comment. |
457 | FROM_BYTE is the bytepos corresponding to FROM. | |
458 | Do not move back before STOP. | |
459 | ||
460 | Return a positive value if we find a comment ending at FROM/FROM_BYTE; | |
461 | return -1 otherwise. | |
462 | ||
463 | If successful, store the charpos of the comment's beginning | |
464 | into *CHARPOS_PTR, and the bytepos into *BYTEPOS_PTR. | |
527a32d9 KH |
465 | |
466 | Global syntax data remains valid for backward search starting at | |
467 | the returned value (or at FROM, if the search was not successful). */ | |
195d1361 RS |
468 | |
469 | static int | |
95ff8dfc | 470 | back_comment (from, from_byte, stop, comnested, comstyle, charpos_ptr, bytepos_ptr) |
932e6895 | 471 | EMACS_INT from, from_byte, stop; |
95ff8dfc | 472 | int comnested, comstyle; |
932e6895 | 473 | EMACS_INT *charpos_ptr, *bytepos_ptr; |
195d1361 RS |
474 | { |
475 | /* Look back, counting the parity of string-quotes, | |
476 | and recording the comment-starters seen. | |
477 | When we reach a safe place, assume that's not in a string; | |
478 | then step the main scan to the earliest comment-starter seen | |
479 | an even number of string quotes away from the safe place. | |
480 | ||
481 | OFROM[I] is position of the earliest comment-starter seen | |
482 | which is I+2X quotes from the comment-end. | |
483 | PARITY is current parity of quotes from the comment end. */ | |
3ee5041c | 484 | int string_style = -1; /* Presumed outside of any string. */ |
195d1361 | 485 | int string_lossage = 0; |
f7c436c1 | 486 | /* Not a real lossage: indicates that we have passed a matching comment |
7d0393cf | 487 | starter plus a non-matching comment-ender, meaning that any matching |
f7c436c1 SM |
488 | comment-starter we might see later could be a false positive (hidden |
489 | inside another comment). | |
490 | Test case: { a (* b } c (* d *) */ | |
491 | int comment_lossage = 0; | |
8e2911c2 AS |
492 | EMACS_INT comment_end = from; |
493 | EMACS_INT comment_end_byte = from_byte; | |
494 | EMACS_INT comstart_pos = 0; | |
495 | EMACS_INT comstart_byte; | |
eb35b628 RS |
496 | /* Place where the containing defun starts, |
497 | or 0 if we didn't come across it yet. */ | |
8e2911c2 AS |
498 | EMACS_INT defun_start = 0; |
499 | EMACS_INT defun_start_byte = 0; | |
195d1361 | 500 | register enum syntaxcode code; |
95ff8dfc | 501 | int nesting = 1; /* current comment nesting */ |
ea315ed6 | 502 | int c; |
3f679f55 SM |
503 | int syntax = 0; |
504 | ||
505 | /* FIXME: A }} comment-ender style leads to incorrect behavior | |
506 | in the case of {{ c }}} because we ignore the last two chars which are | |
507 | assumed to be comment-enders although they aren't. */ | |
195d1361 RS |
508 | |
509 | /* At beginning of range to scan, we're outside of strings; | |
510 | that determines quote parity to the comment-end. */ | |
511 | while (from != stop) | |
512 | { | |
3f679f55 SM |
513 | int temp_byte, prev_syntax; |
514 | int com2start, com2end; | |
6a140a74 | 515 | |
195d1361 | 516 | /* Move back and examine a character. */ |
6a140a74 | 517 | DEC_BOTH (from, from_byte); |
195d1361 RS |
518 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
519 | ||
3f679f55 | 520 | prev_syntax = syntax; |
b7dbcc19 | 521 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
3f679f55 | 522 | syntax = SYNTAX_WITH_FLAGS (c); |
195d1361 RS |
523 | code = SYNTAX (c); |
524 | ||
3f679f55 SM |
525 | /* Check for 2-char comment markers. */ |
526 | com2start = (SYNTAX_FLAGS_COMSTART_FIRST (syntax) | |
527 | && SYNTAX_FLAGS_COMSTART_SECOND (prev_syntax) | |
528 | && comstyle == SYNTAX_FLAGS_COMMENT_STYLE (prev_syntax) | |
529 | && (SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax) | |
530 | || SYNTAX_FLAGS_COMMENT_NESTED (syntax)) == comnested); | |
531 | com2end = (SYNTAX_FLAGS_COMEND_FIRST (syntax) | |
532 | && SYNTAX_FLAGS_COMEND_SECOND (prev_syntax)); | |
533 | ||
534 | /* Nasty cases with overlapping 2-char comment markers: | |
535 | - snmp-mode: -- c -- foo -- c -- | |
536 | --- c -- | |
537 | ------ c -- | |
538 | - c-mode: *||* | |
539 | |* *|* *| | |
540 | |*| |* |*| | |
541 | /// */ | |
542 | ||
543 | /* If a 2-char comment sequence partly overlaps with another, | |
544 | we don't try to be clever. */ | |
545 | if (from > stop && (com2end || com2start)) | |
195d1361 | 546 | { |
3f679f55 SM |
547 | int next = from, next_byte = from_byte, next_c, next_syntax; |
548 | DEC_BOTH (next, next_byte); | |
549 | UPDATE_SYNTAX_TABLE_BACKWARD (next); | |
b7dbcc19 | 550 | next_c = FETCH_CHAR_AS_MULTIBYTE (next_byte); |
3f679f55 SM |
551 | next_syntax = SYNTAX_WITH_FLAGS (next_c); |
552 | if (((com2start || comnested) | |
553 | && SYNTAX_FLAGS_COMEND_SECOND (syntax) | |
554 | && SYNTAX_FLAGS_COMEND_FIRST (next_syntax)) | |
555 | || ((com2end || comnested) | |
556 | && SYNTAX_FLAGS_COMSTART_SECOND (syntax) | |
557 | && comstyle == SYNTAX_FLAGS_COMMENT_STYLE (syntax) | |
558 | && SYNTAX_FLAGS_COMSTART_FIRST (next_syntax))) | |
559 | goto lossage; | |
560 | /* UPDATE_SYNTAX_TABLE_FORWARD (next + 1); */ | |
f902a008 | 561 | } |
3f679f55 SM |
562 | |
563 | if (com2start && comstart_pos == 0) | |
564 | /* We're looking at a comment starter. But it might be a comment | |
565 | ender as well (see snmp-mode). The first time we see one, we | |
566 | need to consider it as a comment starter, | |
567 | and the subsequent times as a comment ender. */ | |
568 | com2end = 0; | |
569 | ||
570 | /* Turn a 2-char comment sequences into the appropriate syntax. */ | |
571 | if (com2end) | |
572 | code = Sendcomment; | |
573 | else if (com2start) | |
574 | code = Scomment; | |
575 | /* Ignore comment starters of a different style. */ | |
576 | else if (code == Scomment | |
577 | && (comstyle != SYNTAX_FLAGS_COMMENT_STYLE (syntax) | |
578 | || SYNTAX_FLAGS_COMMENT_NESTED (syntax) != comnested)) | |
1aa963c8 | 579 | continue; |
195d1361 | 580 | |
9828a477 KH |
581 | /* Ignore escaped characters, except comment-enders. */ |
582 | if (code != Sendcomment && char_quoted (from, from_byte)) | |
195d1361 RS |
583 | continue; |
584 | ||
3ee5041c | 585 | switch (code) |
195d1361 | 586 | { |
3ee5041c SM |
587 | case Sstring_fence: |
588 | case Scomment_fence: | |
589 | c = (code == Sstring_fence ? ST_STRING_STYLE : ST_COMMENT_STYLE); | |
590 | case Sstring: | |
591 | /* Track parity of quotes. */ | |
592 | if (string_style == -1) | |
593 | /* Entering a string. */ | |
594 | string_style = c; | |
595 | else if (string_style == c) | |
596 | /* Leaving the string. */ | |
597 | string_style = -1; | |
598 | else | |
599 | /* If we have two kinds of string delimiters. | |
600 | There's no way to grok this scanning backwards. */ | |
195d1361 | 601 | string_lossage = 1; |
3ee5041c | 602 | break; |
7d0393cf | 603 | |
3ee5041c SM |
604 | case Scomment: |
605 | /* We've already checked that it is the relevant comstyle. */ | |
f7c436c1 | 606 | if (string_style != -1 || comment_lossage || string_lossage) |
3ee5041c SM |
607 | /* There are odd string quotes involved, so let's be careful. |
608 | Test case in Pascal: " { " a { " } */ | |
609 | goto lossage; | |
610 | ||
f7c436c1 SM |
611 | if (!comnested) |
612 | { | |
613 | /* Record best comment-starter so far. */ | |
614 | comstart_pos = from; | |
615 | comstart_byte = from_byte; | |
616 | } | |
617 | else if (--nesting <= 0) | |
95ff8dfc RS |
618 | /* nested comments have to be balanced, so we don't need to |
619 | keep looking for earlier ones. We use here the same (slightly | |
620 | incorrect) reasoning as below: since it is followed by uniform | |
621 | paired string quotes, this comment-start has to be outside of | |
622 | strings, else the comment-end itself would be inside a string. */ | |
623 | goto done; | |
3ee5041c SM |
624 | break; |
625 | ||
02010917 | 626 | case Sendcomment: |
3f679f55 | 627 | if (SYNTAX_FLAGS_COMMENT_STYLE (syntax) == comstyle |
d070eb22 | 628 | && ((com2end && SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax)) |
3f679f55 | 629 | || SYNTAX_FLAGS_COMMENT_NESTED (syntax)) == comnested) |
02010917 SM |
630 | /* This is the same style of comment ender as ours. */ |
631 | { | |
632 | if (comnested) | |
633 | nesting++; | |
634 | else | |
635 | /* Anything before that can't count because it would match | |
636 | this comment-ender rather than ours. */ | |
637 | from = stop; /* Break out of the loop. */ | |
638 | } | |
f7c436c1 SM |
639 | else if (comstart_pos != 0 || c != '\n') |
640 | /* We're mixing comment styles here, so we'd better be careful. | |
641 | The (comstart_pos != 0 || c != '\n') check is not quite correct | |
642 | (we should just always set comment_lossage), but removing it | |
643 | would imply that any multiline comment in C would go through | |
644 | lossage, which seems overkill. | |
645 | The failure should only happen in the rare cases such as | |
646 | { (* } *) */ | |
647 | comment_lossage = 1; | |
02010917 | 648 | break; |
195d1361 | 649 | |
02010917 SM |
650 | case Sopen: |
651 | /* Assume a defun-start point is outside of strings. */ | |
652 | if (open_paren_in_column_0_is_defun_start | |
653 | && (from == stop | |
654 | || (temp_byte = dec_bytepos (from_byte), | |
655 | FETCH_CHAR (temp_byte) == '\n'))) | |
656 | { | |
657 | defun_start = from; | |
658 | defun_start_byte = from_byte; | |
659 | from = stop; /* Break out of the loop. */ | |
660 | } | |
eb35b628 | 661 | break; |
02010917 SM |
662 | |
663 | default: | |
f7c436c1 | 664 | break; |
eb35b628 | 665 | } |
195d1361 RS |
666 | } |
667 | ||
668 | if (comstart_pos == 0) | |
669 | { | |
670 | from = comment_end; | |
6a140a74 | 671 | from_byte = comment_end_byte; |
195d1361 RS |
672 | UPDATE_SYNTAX_TABLE_FORWARD (comment_end - 1); |
673 | } | |
f7c436c1 SM |
674 | /* If comstart_pos is set and we get here (ie. didn't jump to `lossage' |
675 | or `done'), then we've found the beginning of the non-nested comment. */ | |
676 | else if (1) /* !comnested */ | |
195d1361 RS |
677 | { |
678 | from = comstart_pos; | |
6a140a74 | 679 | from_byte = comstart_byte; |
789f3320 | 680 | UPDATE_SYNTAX_TABLE_FORWARD (from - 1); |
195d1361 RS |
681 | } |
682 | else | |
683 | { | |
3ee5041c SM |
684 | struct lisp_parse_state state; |
685 | lossage: | |
195d1361 RS |
686 | /* We had two kinds of string delimiters mixed up |
687 | together. Decode this going forwards. | |
02010917 | 688 | Scan fwd from a known safe place (beginning-of-defun) |
195d1361 RS |
689 | to the one in question; this records where we |
690 | last passed a comment starter. */ | |
eb35b628 RS |
691 | /* If we did not already find the defun start, find it now. */ |
692 | if (defun_start == 0) | |
693 | { | |
694 | defun_start = find_defun_start (comment_end, comment_end_byte); | |
695 | defun_start_byte = find_start_value_byte; | |
696 | } | |
02010917 | 697 | do |
195d1361 | 698 | { |
02010917 SM |
699 | scan_sexps_forward (&state, |
700 | defun_start, defun_start_byte, | |
701 | comment_end, -10000, 0, Qnil, 0); | |
702 | defun_start = comment_end; | |
703 | if (state.incomment == (comnested ? 1 : -1) | |
704 | && state.comstyle == comstyle) | |
705 | from = state.comstr_start; | |
706 | else | |
707 | { | |
708 | from = comment_end; | |
709 | if (state.incomment) | |
710 | /* If comment_end is inside some other comment, maybe ours | |
711 | is nested, so we need to try again from within the | |
712 | surrounding comment. Example: { a (* " *) */ | |
713 | { | |
714 | /* FIXME: We should advance by one or two chars. */ | |
715 | defun_start = state.comstr_start + 2; | |
716 | defun_start_byte = CHAR_TO_BYTE (defun_start); | |
717 | } | |
718 | } | |
719 | } while (defun_start < comment_end); | |
720 | ||
6a140a74 | 721 | from_byte = CHAR_TO_BYTE (from); |
195d1361 RS |
722 | UPDATE_SYNTAX_TABLE_FORWARD (from - 1); |
723 | } | |
7d0393cf | 724 | |
95ff8dfc | 725 | done: |
6a140a74 RS |
726 | *charpos_ptr = from; |
727 | *bytepos_ptr = from_byte; | |
728 | ||
1aa963c8 | 729 | return (from == comment_end) ? -1 : from; |
195d1361 RS |
730 | } |
731 | \f | |
8489eb67 | 732 | DEFUN ("syntax-table-p", Fsyntax_table_p, Ssyntax_table_p, 1, 1, 0, |
fdb82f93 PJ |
733 | doc: /* Return t if OBJECT is a syntax table. |
734 | Currently, any char-table counts as a syntax table. */) | |
735 | (object) | |
2203e1e8 | 736 | Lisp_Object object; |
8489eb67 | 737 | { |
2203e1e8 | 738 | if (CHAR_TABLE_P (object) |
e704cb4b | 739 | && EQ (XCHAR_TABLE (object)->purpose, Qsyntax_table)) |
8489eb67 RS |
740 | return Qt; |
741 | return Qnil; | |
742 | } | |
743 | ||
8ea151b2 | 744 | static void |
8489eb67 RS |
745 | check_syntax_table (obj) |
746 | Lisp_Object obj; | |
747 | { | |
47f5f6ae KS |
748 | CHECK_TYPE (CHAR_TABLE_P (obj) && EQ (XCHAR_TABLE (obj)->purpose, Qsyntax_table), |
749 | Qsyntax_table_p, obj); | |
7d0393cf | 750 | } |
8489eb67 | 751 | |
8489eb67 | 752 | DEFUN ("syntax-table", Fsyntax_table, Ssyntax_table, 0, 0, 0, |
fdb82f93 PJ |
753 | doc: /* Return the current syntax table. |
754 | This is the one specified by the current buffer. */) | |
755 | () | |
8489eb67 RS |
756 | { |
757 | return current_buffer->syntax_table; | |
758 | } | |
759 | ||
760 | DEFUN ("standard-syntax-table", Fstandard_syntax_table, | |
761 | Sstandard_syntax_table, 0, 0, 0, | |
fdb82f93 PJ |
762 | doc: /* Return the standard syntax table. |
763 | This is the one used for new buffers. */) | |
764 | () | |
8489eb67 RS |
765 | { |
766 | return Vstandard_syntax_table; | |
767 | } | |
768 | ||
769 | DEFUN ("copy-syntax-table", Fcopy_syntax_table, Scopy_syntax_table, 0, 1, 0, | |
fdb82f93 PJ |
770 | doc: /* Construct a new syntax table and return it. |
771 | It is a copy of the TABLE, which defaults to the standard syntax table. */) | |
772 | (table) | |
8489eb67 RS |
773 | Lisp_Object table; |
774 | { | |
8ea151b2 RS |
775 | Lisp_Object copy; |
776 | ||
265a9e55 | 777 | if (!NILP (table)) |
8ea151b2 RS |
778 | check_syntax_table (table); |
779 | else | |
780 | table = Vstandard_syntax_table; | |
781 | ||
782 | copy = Fcopy_sequence (table); | |
0f867324 RS |
783 | |
784 | /* Only the standard syntax table should have a default element. | |
785 | Other syntax tables should inherit from parents instead. */ | |
786 | XCHAR_TABLE (copy)->defalt = Qnil; | |
787 | ||
788 | /* Copied syntax tables should all have parents. | |
789 | If we copied one with no parent, such as the standard syntax table, | |
790 | use the standard syntax table as the copy's parent. */ | |
791 | if (NILP (XCHAR_TABLE (copy)->parent)) | |
792 | Fset_char_table_parent (copy, Vstandard_syntax_table); | |
8ea151b2 | 793 | return copy; |
8489eb67 RS |
794 | } |
795 | ||
796 | DEFUN ("set-syntax-table", Fset_syntax_table, Sset_syntax_table, 1, 1, 0, | |
fdb82f93 PJ |
797 | doc: /* Select a new syntax table for the current buffer. |
798 | One argument, a syntax table. */) | |
799 | (table) | |
8489eb67 RS |
800 | Lisp_Object table; |
801 | { | |
4e4dfa78 | 802 | int idx; |
8ea151b2 | 803 | check_syntax_table (table); |
8489eb67 RS |
804 | current_buffer->syntax_table = table; |
805 | /* Indicate that this buffer now has a specified syntax table. */ | |
f6cd0527 GM |
806 | idx = PER_BUFFER_VAR_IDX (syntax_table); |
807 | SET_PER_BUFFER_VALUE_P (current_buffer, idx, 1); | |
8489eb67 RS |
808 | return table; |
809 | } | |
810 | \f | |
811 | /* Convert a letter which signifies a syntax code | |
812 | into the code it signifies. | |
195d1361 | 813 | This is used by modify-syntax-entry, and other things. */ |
8489eb67 RS |
814 | |
815 | unsigned char syntax_spec_code[0400] = | |
816 | { 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
817 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
818 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
819 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
195d1361 | 820 | (char) Swhitespace, (char) Scomment_fence, (char) Sstring, 0377, |
8489eb67 RS |
821 | (char) Smath, 0377, 0377, (char) Squote, |
822 | (char) Sopen, (char) Sclose, 0377, 0377, | |
823 | 0377, (char) Swhitespace, (char) Spunct, (char) Scharquote, | |
824 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
825 | 0377, 0377, 0377, 0377, | |
826 | (char) Scomment, 0377, (char) Sendcomment, 0377, | |
6cb71bf6 | 827 | (char) Sinherit, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* @, A ... */ |
8489eb67 RS |
828 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, |
829 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword, | |
830 | 0377, 0377, 0377, 0377, (char) Sescape, 0377, 0377, (char) Ssymbol, | |
831 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* `, a, ... */ | |
832 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
833 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword, | |
195d1361 | 834 | 0377, 0377, 0377, 0377, (char) Sstring_fence, 0377, 0377, 0377 |
8489eb67 RS |
835 | }; |
836 | ||
195d1361 | 837 | /* Indexed by syntax code, give the letter that describes it. */ |
8489eb67 | 838 | |
195d1361 | 839 | char syntax_code_spec[16] = |
8489eb67 | 840 | { |
195d1361 RS |
841 | ' ', '.', 'w', '_', '(', ')', '\'', '\"', '$', '\\', '/', '<', '>', '@', |
842 | '!', '|' | |
8489eb67 | 843 | }; |
93da5fff KH |
844 | |
845 | /* Indexed by syntax code, give the object (cons of syntax code and | |
846 | nil) to be stored in syntax table. Since these objects can be | |
847 | shared among syntax tables, we generate them in advance. By | |
848 | sharing objects, the function `describe-syntax' can give a more | |
849 | compact listing. */ | |
850 | static Lisp_Object Vsyntax_code_object; | |
851 | ||
8489eb67 RS |
852 | \f |
853 | DEFUN ("char-syntax", Fchar_syntax, Schar_syntax, 1, 1, 0, | |
fdb82f93 | 854 | doc: /* Return the syntax code of CHARACTER, described by a character. |
e2b6daf4 CY |
855 | For example, if CHARACTER is a word constituent, the |
856 | character `w' (119) is returned. | |
fdb82f93 PJ |
857 | The characters that correspond to various syntax codes |
858 | are listed in the documentation of `modify-syntax-entry'. */) | |
859 | (character) | |
2203e1e8 | 860 | Lisp_Object character; |
8489eb67 | 861 | { |
8ea151b2 | 862 | int char_int; |
195d1361 RS |
863 | gl_state.current_syntax_table = current_buffer->syntax_table; |
864 | ||
865 | gl_state.use_global = 0; | |
b7826503 | 866 | CHECK_NUMBER (character); |
2203e1e8 | 867 | char_int = XINT (character); |
8ea151b2 | 868 | return make_number (syntax_code_spec[(int) SYNTAX (char_int)]); |
beefa22e RS |
869 | } |
870 | ||
871 | DEFUN ("matching-paren", Fmatching_paren, Smatching_paren, 1, 1, 0, | |
fdb82f93 PJ |
872 | doc: /* Return the matching parenthesis of CHARACTER, or nil if none. */) |
873 | (character) | |
2203e1e8 | 874 | Lisp_Object character; |
beefa22e | 875 | { |
8ea151b2 | 876 | int char_int, code; |
195d1361 RS |
877 | gl_state.current_syntax_table = current_buffer->syntax_table; |
878 | gl_state.use_global = 0; | |
b7826503 | 879 | CHECK_NUMBER (character); |
2203e1e8 | 880 | char_int = XINT (character); |
8ea151b2 | 881 | code = SYNTAX (char_int); |
a8bd7cd8 | 882 | if (code == Sopen || code == Sclose) |
2e34157c | 883 | return SYNTAX_MATCH (char_int); |
beefa22e | 884 | return Qnil; |
8489eb67 RS |
885 | } |
886 | ||
c65adb44 | 887 | DEFUN ("string-to-syntax", Fstring_to_syntax, Sstring_to_syntax, 1, 1, 0, |
fdb82f93 PJ |
888 | doc: /* Convert a syntax specification STRING into syntax cell form. |
889 | STRING should be a string as it is allowed as argument of | |
890 | `modify-syntax-entry'. Value is the equivalent cons cell | |
bce3dc36 | 891 | \(CODE . MATCHING-CHAR) that can be used as value of a `syntax-table' |
fdb82f93 PJ |
892 | text property. */) |
893 | (string) | |
c65adb44 | 894 | Lisp_Object string; |
8489eb67 | 895 | { |
2e567bd3 | 896 | register const unsigned char *p; |
8489eb67 | 897 | register enum syntaxcode code; |
247e20a8 | 898 | int val; |
8ea151b2 | 899 | Lisp_Object match; |
8489eb67 | 900 | |
b7826503 | 901 | CHECK_STRING (string); |
8489eb67 | 902 | |
d5db4077 | 903 | p = SDATA (string); |
8489eb67 RS |
904 | code = (enum syntaxcode) syntax_spec_code[*p++]; |
905 | if (((int) code & 0377) == 0377) | |
b764a653 | 906 | error ("Invalid syntax description letter: %c", p[-1]); |
8489eb67 | 907 | |
8ea151b2 | 908 | if (code == Sinherit) |
c65adb44 | 909 | return Qnil; |
8ea151b2 RS |
910 | |
911 | if (*p) | |
d1be9f0f | 912 | { |
93da5fff | 913 | int len; |
7c19e8e1 | 914 | int character = (STRING_CHAR_AND_LENGTH |
d5db4077 | 915 | (p, SBYTES (string) - 1, len)); |
93da5fff | 916 | XSETINT (match, character); |
d1be9f0f RS |
917 | if (XFASTINT (match) == ' ') |
918 | match = Qnil; | |
93da5fff | 919 | p += len; |
d1be9f0f RS |
920 | } |
921 | else | |
8ea151b2 | 922 | match = Qnil; |
8489eb67 | 923 | |
8ea151b2 | 924 | val = (int) code; |
8489eb67 RS |
925 | while (*p) |
926 | switch (*p++) | |
927 | { | |
928 | case '1': | |
247e20a8 | 929 | val |= 1 << 16; |
8489eb67 RS |
930 | break; |
931 | ||
932 | case '2': | |
247e20a8 | 933 | val |= 1 << 17; |
8489eb67 RS |
934 | break; |
935 | ||
936 | case '3': | |
247e20a8 | 937 | val |= 1 << 18; |
8489eb67 RS |
938 | break; |
939 | ||
940 | case '4': | |
247e20a8 | 941 | val |= 1 << 19; |
8489eb67 RS |
942 | break; |
943 | ||
944 | case 'p': | |
247e20a8 | 945 | val |= 1 << 20; |
8489eb67 | 946 | break; |
e5d4f4dc RS |
947 | |
948 | case 'b': | |
247e20a8 | 949 | val |= 1 << 21; |
e5d4f4dc | 950 | break; |
95ff8dfc RS |
951 | |
952 | case 'n': | |
953 | val |= 1 << 22; | |
954 | break; | |
8489eb67 | 955 | } |
7d0393cf | 956 | |
93da5fff | 957 | if (val < XVECTOR (Vsyntax_code_object)->size && NILP (match)) |
c65adb44 | 958 | return XVECTOR (Vsyntax_code_object)->contents[val]; |
93da5fff KH |
959 | else |
960 | /* Since we can't use a shared object, let's make a new one. */ | |
c65adb44 SM |
961 | return Fcons (make_number (val), match); |
962 | } | |
963 | ||
fdb82f93 | 964 | /* I really don't know why this is interactive |
2c9e1900 | 965 | help-form should at least be made useful whilst reading the second arg. */ |
7d0393cf | 966 | DEFUN ("modify-syntax-entry", Fmodify_syntax_entry, Smodify_syntax_entry, 2, 3, |
c65adb44 | 967 | "cSet syntax for character: \nsSet syntax for %s to: ", |
a50a10a0 | 968 | doc: /* Set syntax for character CHAR according to string NEWENTRY. |
d7ee9fab | 969 | The syntax is changed only for table SYNTAX-TABLE, which defaults to |
fdb82f93 | 970 | the current buffer's syntax table. |
5c7b02ab | 971 | CHAR may be a cons (MIN . MAX), in which case, syntaxes of all characters |
fedc6ab5 | 972 | in the range MIN to MAX are changed. |
fdb82f93 PJ |
973 | The first character of NEWENTRY should be one of the following: |
974 | Space or - whitespace syntax. w word constituent. | |
975 | _ symbol constituent. . punctuation. | |
976 | ( open-parenthesis. ) close-parenthesis. | |
977 | " string quote. \\ escape. | |
978 | $ paired delimiter. ' expression quote or prefix operator. | |
979 | < comment starter. > comment ender. | |
980 | / character-quote. @ inherit from `standard-syntax-table'. | |
981 | | generic string fence. ! generic comment fence. | |
982 | ||
983 | Only single-character comment start and end sequences are represented thus. | |
984 | Two-character sequences are represented as described below. | |
985 | The second character of NEWENTRY is the matching parenthesis, | |
986 | used only if the first character is `(' or `)'. | |
987 | Any additional characters are flags. | |
988 | Defined flags are the characters 1, 2, 3, 4, b, p, and n. | |
a50a10a0 PJ |
989 | 1 means CHAR is the start of a two-char comment start sequence. |
990 | 2 means CHAR is the second character of such a sequence. | |
991 | 3 means CHAR is the start of a two-char comment end sequence. | |
992 | 4 means CHAR is the second character of such a sequence. | |
fdb82f93 PJ |
993 | |
994 | There can be up to two orthogonal comment sequences. This is to support | |
995 | language modes such as C++. By default, all comment sequences are of style | |
996 | a, but you can set the comment sequence style to b (on the second character | |
997 | of a comment-start, or the first character of a comment-end sequence) using | |
998 | this flag: | |
a50a10a0 PJ |
999 | b means CHAR is part of comment sequence b. |
1000 | n means CHAR is part of a nestable comment sequence. | |
fdb82f93 | 1001 | |
a50a10a0 | 1002 | p means CHAR is a prefix character for `backward-prefix-chars'; |
fdb82f93 | 1003 | such characters are treated as whitespace when they occur |
a50a10a0 | 1004 | between expressions. |
fedc6ab5 | 1005 | usage: (modify-syntax-entry CHAR NEWENTRY &optional SYNTAX-TABLE) */) |
fdb82f93 | 1006 | (c, newentry, syntax_table) |
c65adb44 SM |
1007 | Lisp_Object c, newentry, syntax_table; |
1008 | { | |
5c7b02ab KH |
1009 | if (CONSP (c)) |
1010 | { | |
8f924df7 KH |
1011 | CHECK_CHARACTER_CAR (c); |
1012 | CHECK_CHARACTER_CDR (c); | |
5c7b02ab KH |
1013 | } |
1014 | else | |
1015 | CHECK_CHARACTER (c); | |
c65adb44 SM |
1016 | |
1017 | if (NILP (syntax_table)) | |
1018 | syntax_table = current_buffer->syntax_table; | |
1019 | else | |
1020 | check_syntax_table (syntax_table); | |
8489eb67 | 1021 | |
5c7b02ab KH |
1022 | newentry = Fstring_to_syntax (newentry); |
1023 | if (CONSP (c)) | |
1024 | SET_RAW_SYNTAX_ENTRY_RANGE (syntax_table, c, newentry); | |
1025 | else | |
1026 | SET_RAW_SYNTAX_ENTRY (syntax_table, XINT (c), newentry); | |
e5b94d44 CY |
1027 | |
1028 | /* We clear the regexp cache, since character classes can now have | |
1029 | different values from those in the compiled regexps.*/ | |
1030 | clear_regexp_cache (); | |
1031 | ||
8489eb67 RS |
1032 | return Qnil; |
1033 | } | |
1034 | \f | |
1035 | /* Dump syntax table to buffer in human-readable format */ | |
1036 | ||
62abe9cb SM |
1037 | DEFUN ("internal-describe-syntax-value", Finternal_describe_syntax_value, |
1038 | Sinternal_describe_syntax_value, 1, 1, 0, | |
1039 | doc: /* Insert a description of the internal syntax description SYNTAX at point. */) | |
1040 | (syntax) | |
1041 | Lisp_Object syntax; | |
8489eb67 RS |
1042 | { |
1043 | register enum syntaxcode code; | |
e6365855 | 1044 | char desc, start1, start2, end1, end2, prefix, comstyle, comnested; |
8489eb67 | 1045 | char str[2]; |
62abe9cb | 1046 | Lisp_Object first, match_lisp, value = syntax; |
8489eb67 | 1047 | |
8ea151b2 RS |
1048 | if (NILP (value)) |
1049 | { | |
62abe9cb SM |
1050 | insert_string ("default"); |
1051 | return syntax; | |
8ea151b2 RS |
1052 | } |
1053 | ||
908b7fea KH |
1054 | if (CHAR_TABLE_P (value)) |
1055 | { | |
62abe9cb SM |
1056 | insert_string ("deeper char-table ..."); |
1057 | return syntax; | |
908b7fea KH |
1058 | } |
1059 | ||
8ea151b2 RS |
1060 | if (!CONSP (value)) |
1061 | { | |
62abe9cb SM |
1062 | insert_string ("invalid"); |
1063 | return syntax; | |
8ea151b2 RS |
1064 | } |
1065 | ||
c1d497be KR |
1066 | first = XCAR (value); |
1067 | match_lisp = XCDR (value); | |
8ea151b2 RS |
1068 | |
1069 | if (!INTEGERP (first) || !(NILP (match_lisp) || INTEGERP (match_lisp))) | |
8489eb67 | 1070 | { |
62abe9cb SM |
1071 | insert_string ("invalid"); |
1072 | return syntax; | |
8489eb67 RS |
1073 | } |
1074 | ||
e704cb4b | 1075 | code = (enum syntaxcode) (XINT (first) & 0377); |
8ea151b2 RS |
1076 | start1 = (XINT (first) >> 16) & 1; |
1077 | start2 = (XINT (first) >> 17) & 1; | |
1078 | end1 = (XINT (first) >> 18) & 1; | |
1079 | end2 = (XINT (first) >> 19) & 1; | |
1080 | prefix = (XINT (first) >> 20) & 1; | |
1081 | comstyle = (XINT (first) >> 21) & 1; | |
e6365855 | 1082 | comnested = (XINT (first) >> 22) & 1; |
8489eb67 RS |
1083 | |
1084 | if ((int) code < 0 || (int) code >= (int) Smax) | |
1085 | { | |
1086 | insert_string ("invalid"); | |
62abe9cb | 1087 | return syntax; |
8489eb67 RS |
1088 | } |
1089 | desc = syntax_code_spec[(int) code]; | |
1090 | ||
1091 | str[0] = desc, str[1] = 0; | |
1092 | insert (str, 1); | |
1093 | ||
93da5fff KH |
1094 | if (NILP (match_lisp)) |
1095 | insert (" ", 1); | |
1096 | else | |
1097 | insert_char (XINT (match_lisp)); | |
8489eb67 | 1098 | |
8489eb67 RS |
1099 | if (start1) |
1100 | insert ("1", 1); | |
1101 | if (start2) | |
1102 | insert ("2", 1); | |
1103 | ||
1104 | if (end1) | |
1105 | insert ("3", 1); | |
1106 | if (end2) | |
1107 | insert ("4", 1); | |
1108 | ||
1109 | if (prefix) | |
1110 | insert ("p", 1); | |
e5d4f4dc RS |
1111 | if (comstyle) |
1112 | insert ("b", 1); | |
e6365855 SM |
1113 | if (comnested) |
1114 | insert ("n", 1); | |
8489eb67 RS |
1115 | |
1116 | insert_string ("\twhich means: "); | |
1117 | ||
0220c518 | 1118 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
1119 | { |
1120 | case Swhitespace: | |
1121 | insert_string ("whitespace"); break; | |
1122 | case Spunct: | |
1123 | insert_string ("punctuation"); break; | |
1124 | case Sword: | |
1125 | insert_string ("word"); break; | |
1126 | case Ssymbol: | |
1127 | insert_string ("symbol"); break; | |
1128 | case Sopen: | |
1129 | insert_string ("open"); break; | |
1130 | case Sclose: | |
1131 | insert_string ("close"); break; | |
1132 | case Squote: | |
d671382e | 1133 | insert_string ("prefix"); break; |
8489eb67 RS |
1134 | case Sstring: |
1135 | insert_string ("string"); break; | |
1136 | case Smath: | |
1137 | insert_string ("math"); break; | |
1138 | case Sescape: | |
1139 | insert_string ("escape"); break; | |
1140 | case Scharquote: | |
1141 | insert_string ("charquote"); break; | |
1142 | case Scomment: | |
1143 | insert_string ("comment"); break; | |
1144 | case Sendcomment: | |
1145 | insert_string ("endcomment"); break; | |
d671382e SM |
1146 | case Sinherit: |
1147 | insert_string ("inherit"); break; | |
1148 | case Scomment_fence: | |
1149 | insert_string ("comment fence"); break; | |
1150 | case Sstring_fence: | |
1151 | insert_string ("string fence"); break; | |
8489eb67 RS |
1152 | default: |
1153 | insert_string ("invalid"); | |
62abe9cb | 1154 | return syntax; |
8489eb67 RS |
1155 | } |
1156 | ||
8ea151b2 | 1157 | if (!NILP (match_lisp)) |
8489eb67 RS |
1158 | { |
1159 | insert_string (", matches "); | |
8ea151b2 | 1160 | insert_char (XINT (match_lisp)); |
8489eb67 RS |
1161 | } |
1162 | ||
1163 | if (start1) | |
1164 | insert_string (",\n\t is the first character of a comment-start sequence"); | |
1165 | if (start2) | |
1166 | insert_string (",\n\t is the second character of a comment-start sequence"); | |
1167 | ||
1168 | if (end1) | |
1169 | insert_string (",\n\t is the first character of a comment-end sequence"); | |
1170 | if (end2) | |
1171 | insert_string (",\n\t is the second character of a comment-end sequence"); | |
e5d4f4dc RS |
1172 | if (comstyle) |
1173 | insert_string (" (comment style b)"); | |
e6365855 SM |
1174 | if (comnested) |
1175 | insert_string (" (nestable)"); | |
e5d4f4dc | 1176 | |
8489eb67 RS |
1177 | if (prefix) |
1178 | insert_string (",\n\t is a prefix character for `backward-prefix-chars'"); | |
1179 | ||
62abe9cb SM |
1180 | return syntax; |
1181 | } | |
8489eb67 | 1182 | \f |
195d1361 RS |
1183 | int parse_sexp_ignore_comments; |
1184 | ||
8f924df7 KH |
1185 | /* Char-table of functions that find the next or previous word |
1186 | boundary. */ | |
1187 | Lisp_Object Vfind_word_boundary_function_table; | |
869bb237 | 1188 | |
8489eb67 RS |
1189 | /* Return the position across COUNT words from FROM. |
1190 | If that many words cannot be found before the end of the buffer, return 0. | |
1191 | COUNT negative means scan backward and stop at word beginning. */ | |
1192 | ||
dfcf069d | 1193 | int |
8489eb67 RS |
1194 | scan_words (from, count) |
1195 | register int from, count; | |
1196 | { | |
1197 | register int beg = BEGV; | |
1198 | register int end = ZV; | |
6a140a74 | 1199 | register int from_byte = CHAR_TO_BYTE (from); |
93da5fff KH |
1200 | register enum syntaxcode code; |
1201 | int ch0, ch1; | |
8f924df7 | 1202 | Lisp_Object func, script, pos; |
8489eb67 RS |
1203 | |
1204 | immediate_quit = 1; | |
1205 | QUIT; | |
1206 | ||
195d1361 RS |
1207 | SETUP_SYNTAX_TABLE (from, count); |
1208 | ||
8489eb67 RS |
1209 | while (count > 0) |
1210 | { | |
1211 | while (1) | |
1212 | { | |
1213 | if (from == end) | |
1214 | { | |
1215 | immediate_quit = 0; | |
1216 | return 0; | |
1217 | } | |
195d1361 | 1218 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 1219 | ch0 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
93da5fff | 1220 | code = SYNTAX (ch0); |
6a140a74 | 1221 | INC_BOTH (from, from_byte); |
8489eb67 RS |
1222 | if (words_include_escapes |
1223 | && (code == Sescape || code == Scharquote)) | |
1224 | break; | |
1225 | if (code == Sword) | |
1226 | break; | |
8489eb67 | 1227 | } |
93da5fff KH |
1228 | /* Now CH0 is a character which begins a word and FROM is the |
1229 | position of the next character. */ | |
8f924df7 | 1230 | func = CHAR_TABLE_REF (Vfind_word_boundary_function_table, ch0); |
869bb237 | 1231 | if (! NILP (Ffboundp (func))) |
8489eb67 | 1232 | { |
869bb237 | 1233 | pos = call2 (func, make_number (from - 1), make_number (end)); |
8f924df7 KH |
1234 | if (INTEGERP (pos) && XINT (pos) > from) |
1235 | { | |
1236 | from = XINT (pos); | |
1237 | from_byte = CHAR_TO_BYTE (from); | |
1238 | } | |
8489eb67 | 1239 | } |
869bb237 | 1240 | else |
8cb8232a | 1241 | { |
8cb8232a KH |
1242 | script = CHAR_TABLE_REF (Vchar_script_table, ch0); |
1243 | while (1) | |
1244 | { | |
1245 | if (from == end) break; | |
1246 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
b7dbcc19 | 1247 | ch1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8cb8232a KH |
1248 | code = SYNTAX (ch1); |
1249 | if ((code != Sword | |
1250 | && (! words_include_escapes | |
1251 | || (code != Sescape && code != Scharquote))) | |
2b70a11c | 1252 | || word_boundary_p (ch0, ch1)) |
869bb237 | 1253 | break; |
8cb8232a KH |
1254 | INC_BOTH (from, from_byte); |
1255 | ch0 = ch1; | |
1256 | } | |
8489eb67 RS |
1257 | } |
1258 | count--; | |
1259 | } | |
1260 | while (count < 0) | |
1261 | { | |
1262 | while (1) | |
1263 | { | |
1264 | if (from == beg) | |
1265 | { | |
1266 | immediate_quit = 0; | |
1267 | return 0; | |
1268 | } | |
6a140a74 | 1269 | DEC_BOTH (from, from_byte); |
195d1361 | 1270 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 1271 | ch1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
93da5fff | 1272 | code = SYNTAX (ch1); |
8489eb67 RS |
1273 | if (words_include_escapes |
1274 | && (code == Sescape || code == Scharquote)) | |
1275 | break; | |
1276 | if (code == Sword) | |
1277 | break; | |
8489eb67 | 1278 | } |
93da5fff KH |
1279 | /* Now CH1 is a character which ends a word and FROM is the |
1280 | position of it. */ | |
8f924df7 | 1281 | func = CHAR_TABLE_REF (Vfind_word_boundary_function_table, ch1); |
869bb237 | 1282 | if (! NILP (Ffboundp (func))) |
8f924df7 | 1283 | { |
869bb237 | 1284 | pos = call2 (func, make_number (from), make_number (beg)); |
8f924df7 KH |
1285 | if (INTEGERP (pos) && XINT (pos) < from) |
1286 | { | |
1287 | from = XINT (pos); | |
1288 | from_byte = CHAR_TO_BYTE (from); | |
1289 | } | |
8489eb67 | 1290 | } |
869bb237 | 1291 | else |
8489eb67 | 1292 | { |
8cb8232a KH |
1293 | script = CHAR_TABLE_REF (Vchar_script_table, ch1); |
1294 | while (1) | |
1295 | { | |
8cb8232a KH |
1296 | if (from == beg) |
1297 | break; | |
262be72a | 1298 | DEC_BOTH (from, from_byte); |
8cb8232a | 1299 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
262be72a | 1300 | ch0 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8cb8232a KH |
1301 | code = SYNTAX (ch0); |
1302 | if ((code != Sword | |
1303 | && (! words_include_escapes | |
1304 | || (code != Sescape && code != Scharquote))) | |
2b70a11c | 1305 | || word_boundary_p (ch0, ch1)) |
262be72a MB |
1306 | { |
1307 | INC_BOTH (from, from_byte); | |
1308 | break; | |
1309 | } | |
8cb8232a KH |
1310 | ch1 = ch0; |
1311 | } | |
8489eb67 RS |
1312 | } |
1313 | count++; | |
1314 | } | |
1315 | ||
1316 | immediate_quit = 0; | |
1317 | ||
1318 | return from; | |
1319 | } | |
1320 | ||
1e9dbb5f | 1321 | DEFUN ("forward-word", Fforward_word, Sforward_word, 0, 1, "^p", |
fdb82f93 PJ |
1322 | doc: /* Move point forward ARG words (backward if ARG is negative). |
1323 | Normally returns t. | |
1324 | If an edge of the buffer or a field boundary is reached, point is left there | |
1325 | and the function returns nil. Field boundaries are not noticed if | |
1326 | `inhibit-field-text-motion' is non-nil. */) | |
839966f3 KH |
1327 | (arg) |
1328 | Lisp_Object arg; | |
8489eb67 | 1329 | { |
c096ae4d | 1330 | Lisp_Object tmp; |
2c6ea900 | 1331 | int orig_val, val; |
8489eb67 | 1332 | |
839966f3 KH |
1333 | if (NILP (arg)) |
1334 | XSETFASTINT (arg, 1); | |
1335 | else | |
1336 | CHECK_NUMBER (arg); | |
1337 | ||
1338 | val = orig_val = scan_words (PT, XINT (arg)); | |
2c6ea900 | 1339 | if (! orig_val) |
839966f3 | 1340 | val = XINT (arg) > 0 ? ZV : BEGV; |
5878ee6f | 1341 | |
b8855607 | 1342 | /* Avoid jumping out of an input field. */ |
c096ae4d SM |
1343 | tmp = Fconstrain_to_field (make_number (val), make_number (PT), |
1344 | Qt, Qnil, Qnil); | |
1345 | val = XFASTINT (tmp); | |
7d0393cf | 1346 | |
8489eb67 | 1347 | SET_PT (val); |
e6d8341f | 1348 | return val == orig_val ? Qt : Qnil; |
8489eb67 RS |
1349 | } |
1350 | \f | |
195d1361 RS |
1351 | Lisp_Object skip_chars (); |
1352 | ||
1353 | DEFUN ("skip-chars-forward", Fskip_chars_forward, Sskip_chars_forward, 1, 2, 0, | |
fdb82f93 PJ |
1354 | doc: /* Move point forward, stopping before a char not in STRING, or at pos LIM. |
1355 | STRING is like the inside of a `[...]' in a regular expression | |
1356 | except that `]' is never special and `\\' quotes `^', `-' or `\\' | |
7087d5e9 | 1357 | (but not at the end of a range; quoting is never needed there). |
fdb82f93 PJ |
1358 | Thus, with arg "a-zA-Z", this skips letters stopping before first nonletter. |
1359 | With arg "^a-zA-Z", skips nonletters stopping before first letter. | |
a1bc88d4 RS |
1360 | Char classes, e.g. `[:alpha:]', are supported. |
1361 | ||
1362 | Returns the distance traveled, either zero or positive. */) | |
fdb82f93 | 1363 | (string, lim) |
195d1361 RS |
1364 | Lisp_Object string, lim; |
1365 | { | |
327719ee | 1366 | return skip_chars (1, string, lim, 1); |
195d1361 RS |
1367 | } |
1368 | ||
1369 | DEFUN ("skip-chars-backward", Fskip_chars_backward, Sskip_chars_backward, 1, 2, 0, | |
fdb82f93 PJ |
1370 | doc: /* Move point backward, stopping after a char not in STRING, or at pos LIM. |
1371 | See `skip-chars-forward' for details. | |
1372 | Returns the distance traveled, either zero or negative. */) | |
1373 | (string, lim) | |
195d1361 RS |
1374 | Lisp_Object string, lim; |
1375 | { | |
327719ee | 1376 | return skip_chars (0, string, lim, 1); |
195d1361 RS |
1377 | } |
1378 | ||
1379 | DEFUN ("skip-syntax-forward", Fskip_syntax_forward, Sskip_syntax_forward, 1, 2, 0, | |
fdb82f93 PJ |
1380 | doc: /* Move point forward across chars in specified syntax classes. |
1381 | SYNTAX is a string of syntax code characters. | |
1382 | Stop before a char whose syntax is not in SYNTAX, or at position LIM. | |
1383 | If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX. | |
1384 | This function returns the distance traveled, either zero or positive. */) | |
1385 | (syntax, lim) | |
195d1361 RS |
1386 | Lisp_Object syntax, lim; |
1387 | { | |
b7dbcc19 | 1388 | return skip_syntaxes (1, syntax, lim); |
195d1361 RS |
1389 | } |
1390 | ||
1391 | DEFUN ("skip-syntax-backward", Fskip_syntax_backward, Sskip_syntax_backward, 1, 2, 0, | |
fdb82f93 PJ |
1392 | doc: /* Move point backward across chars in specified syntax classes. |
1393 | SYNTAX is a string of syntax code characters. | |
1394 | Stop on reaching a char whose syntax is not in SYNTAX, or at position LIM. | |
1395 | If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX. | |
1396 | This function returns the distance traveled, either zero or negative. */) | |
1397 | (syntax, lim) | |
195d1361 RS |
1398 | Lisp_Object syntax, lim; |
1399 | { | |
b7dbcc19 | 1400 | return skip_syntaxes (0, syntax, lim); |
195d1361 RS |
1401 | } |
1402 | ||
6a140a74 | 1403 | static Lisp_Object |
327719ee | 1404 | skip_chars (forwardp, string, lim, handle_iso_classes) |
b7dbcc19 | 1405 | int forwardp; |
195d1361 | 1406 | Lisp_Object string, lim; |
a1bc88d4 | 1407 | int handle_iso_classes; |
195d1361 | 1408 | { |
195d1361 RS |
1409 | register unsigned int c; |
1410 | unsigned char fastmap[0400]; | |
b7dbcc19 | 1411 | /* Store the ranges of non-ASCII characters. */ |
5df4982e | 1412 | int *char_ranges; |
7e68b0ea | 1413 | int n_char_ranges = 0; |
195d1361 | 1414 | int negate = 0; |
4101e6fe | 1415 | register int i, i_byte; |
b7dbcc19 KH |
1416 | /* Set to 1 if the current buffer is multibyte and the region |
1417 | contains non-ASCII chars. */ | |
1418 | int multibyte; | |
1419 | /* Set to 1 if STRING is multibyte and it contains non-ASCII | |
1420 | chars. */ | |
1674d9a2 RS |
1421 | int string_multibyte; |
1422 | int size_byte; | |
2e567bd3 | 1423 | const unsigned char *str; |
82d497fc | 1424 | int len; |
a1bc88d4 | 1425 | Lisp_Object iso_classes; |
195d1361 | 1426 | |
b7826503 | 1427 | CHECK_STRING (string); |
a1bc88d4 | 1428 | iso_classes = Qnil; |
82d497fc | 1429 | |
195d1361 RS |
1430 | if (NILP (lim)) |
1431 | XSETINT (lim, forwardp ? ZV : BEGV); | |
1432 | else | |
b7826503 | 1433 | CHECK_NUMBER_COERCE_MARKER (lim); |
195d1361 RS |
1434 | |
1435 | /* In any case, don't allow scan outside bounds of buffer. */ | |
195d1361 RS |
1436 | if (XINT (lim) > ZV) |
1437 | XSETFASTINT (lim, ZV); | |
1438 | if (XINT (lim) < BEGV) | |
1439 | XSETFASTINT (lim, BEGV); | |
1440 | ||
b7dbcc19 | 1441 | multibyte = (!NILP (current_buffer->enable_multibyte_characters) |
92eaa22e | 1442 | && (XINT (lim) - PT != CHAR_TO_BYTE (XINT (lim)) - PT_BYTE)); |
8f924df7 | 1443 | string_multibyte = SBYTES (string) > SCHARS (string); |
b7dbcc19 | 1444 | |
195d1361 RS |
1445 | bzero (fastmap, sizeof fastmap); |
1446 | ||
8f924df7 KH |
1447 | str = SDATA (string); |
1448 | size_byte = SBYTES (string); | |
4101e6fe | 1449 | |
82d497fc | 1450 | i_byte = 0; |
13090112 | 1451 | if (i_byte < size_byte |
d5db4077 | 1452 | && SREF (string, 0) == '^') |
195d1361 | 1453 | { |
82d497fc | 1454 | negate = 1; i_byte++; |
195d1361 RS |
1455 | } |
1456 | ||
1457 | /* Find the characters specified and set their elements of fastmap. | |
b7dbcc19 | 1458 | Handle backslashes and ranges specially. |
195d1361 | 1459 | |
b7dbcc19 KH |
1460 | If STRING contains non-ASCII characters, setup char_ranges for |
1461 | them and use fastmap only for their leading codes. */ | |
4101e6fe | 1462 | |
b7dbcc19 | 1463 | if (! string_multibyte) |
195d1361 | 1464 | { |
b7dbcc19 | 1465 | int string_has_eight_bit = 0; |
4101e6fe | 1466 | |
b7dbcc19 KH |
1467 | /* At first setup fastmap. */ |
1468 | while (i_byte < size_byte) | |
1469 | { | |
1470 | c = str[i_byte++]; | |
1471 | ||
a1bc88d4 RS |
1472 | if (handle_iso_classes && c == '[' |
1473 | && i_byte < size_byte | |
327719ee | 1474 | && str[i_byte] == ':') |
a1bc88d4 RS |
1475 | { |
1476 | const unsigned char *class_beg = str + i_byte + 1; | |
1477 | const unsigned char *class_end = class_beg; | |
b3bda4fd | 1478 | const unsigned char *class_limit = str + size_byte - 2; |
db9cd97a | 1479 | /* Leave room for the null. */ |
a1bc88d4 RS |
1480 | unsigned char class_name[CHAR_CLASS_MAX_LENGTH + 1]; |
1481 | re_wctype_t cc; | |
1482 | ||
1483 | if (class_limit - class_beg > CHAR_CLASS_MAX_LENGTH) | |
1484 | class_limit = class_beg + CHAR_CLASS_MAX_LENGTH; | |
1485 | ||
b3bda4fd KS |
1486 | while (class_end < class_limit |
1487 | && *class_end >= 'a' && *class_end <= 'z') | |
a1bc88d4 RS |
1488 | class_end++; |
1489 | ||
b3bda4fd KS |
1490 | if (class_end == class_beg |
1491 | || *class_end != ':' || class_end[1] != ']') | |
1492 | goto not_a_class_name; | |
a1bc88d4 RS |
1493 | |
1494 | bcopy (class_beg, class_name, class_end - class_beg); | |
1495 | class_name[class_end - class_beg] = 0; | |
1496 | ||
1497 | cc = re_wctype (class_name); | |
1498 | if (cc == 0) | |
1499 | error ("Invalid ISO C character class"); | |
1500 | ||
1501 | iso_classes = Fcons (make_number (cc), iso_classes); | |
1502 | ||
1503 | i_byte = class_end + 2 - str; | |
1504 | continue; | |
1505 | } | |
1506 | ||
b3bda4fd | 1507 | not_a_class_name: |
b7dbcc19 KH |
1508 | if (c == '\\') |
1509 | { | |
13090112 | 1510 | if (i_byte == size_byte) |
4101e6fe RS |
1511 | break; |
1512 | ||
5c7b02ab | 1513 | c = str[i_byte++]; |
195d1361 | 1514 | } |
839966f3 KH |
1515 | /* Treat `-' as range character only if another character |
1516 | follows. */ | |
1517 | if (i_byte + 1 < size_byte | |
82d497fc | 1518 | && str[i_byte] == '-') |
195d1361 | 1519 | { |
9690d026 | 1520 | unsigned int c2; |
4101e6fe RS |
1521 | |
1522 | /* Skip over the dash. */ | |
82d497fc | 1523 | i_byte++; |
4101e6fe | 1524 | |
4101e6fe | 1525 | /* Get the end of the range. */ |
5c7b02ab | 1526 | c2 = str[i_byte++]; |
b7dbcc19 KH |
1527 | if (c2 == '\\' |
1528 | && i_byte < size_byte) | |
1529 | c2 = str[i_byte++]; | |
1530 | ||
8f924df7 KH |
1531 | if (c <= c2) |
1532 | { | |
1533 | while (c <= c2) | |
1534 | fastmap[c++] = 1; | |
1535 | if (! ASCII_CHAR_P (c2)) | |
1536 | string_has_eight_bit = 1; | |
1537 | } | |
195d1361 RS |
1538 | } |
1539 | else | |
b7dbcc19 KH |
1540 | { |
1541 | fastmap[c] = 1; | |
1542 | if (! ASCII_CHAR_P (c)) | |
1543 | string_has_eight_bit = 1; | |
1544 | } | |
1545 | } | |
1546 | ||
1547 | /* If the current range is multibyte and STRING contains | |
1548 | eight-bit chars, arrange fastmap and setup char_ranges for | |
1549 | the corresponding multibyte chars. */ | |
1550 | if (multibyte && string_has_eight_bit) | |
1551 | { | |
1552 | unsigned char fastmap2[0400]; | |
1553 | int range_start_byte, range_start_char; | |
1554 | ||
1555 | bcopy (fastmap2 + 0200, fastmap + 0200, 0200); | |
1556 | bzero (fastmap + 0200, 0200); | |
1557 | /* We are sure that this loop stops. */ | |
1558 | for (i = 0200; ! fastmap2[i]; i++); | |
4c0354d7 | 1559 | c = BYTE8_TO_CHAR (i); |
b7dbcc19 KH |
1560 | fastmap[CHAR_LEADING_CODE (c)] = 1; |
1561 | range_start_byte = i; | |
1562 | range_start_char = c; | |
f003ca54 | 1563 | char_ranges = (int *) alloca (sizeof (int) * 128 * 2); |
b7dbcc19 KH |
1564 | for (i = 129; i < 0400; i++) |
1565 | { | |
4c0354d7 | 1566 | c = BYTE8_TO_CHAR (i); |
b7dbcc19 KH |
1567 | fastmap[CHAR_LEADING_CODE (c)] = 1; |
1568 | if (i - range_start_byte != c - range_start_char) | |
1569 | { | |
1570 | char_ranges[n_char_ranges++] = range_start_char; | |
1571 | char_ranges[n_char_ranges++] = ((i - 1 - range_start_byte) | |
1572 | + range_start_char); | |
1573 | range_start_byte = i; | |
1574 | range_start_char = c; | |
8f924df7 | 1575 | } |
b7dbcc19 KH |
1576 | } |
1577 | char_ranges[n_char_ranges++] = range_start_char; | |
1578 | char_ranges[n_char_ranges++] = ((i - 1 - range_start_byte) | |
1579 | + range_start_char); | |
195d1361 RS |
1580 | } |
1581 | } | |
f003ca54 | 1582 | else /* STRING is multibyte */ |
b7dbcc19 | 1583 | { |
f003ca54 KH |
1584 | char_ranges = (int *) alloca (sizeof (int) * SCHARS (string) * 2); |
1585 | ||
b7dbcc19 | 1586 | while (i_byte < size_byte) |
195d1361 | 1587 | { |
b7dbcc19 | 1588 | unsigned char leading_code; |
5c7b02ab | 1589 | |
b7dbcc19 KH |
1590 | leading_code = str[i_byte]; |
1591 | c = STRING_CHAR_AND_LENGTH (str + i_byte, size_byte-i_byte, len); | |
1592 | i_byte += len; | |
5c7b02ab | 1593 | |
327719ee MB |
1594 | if (handle_iso_classes && c == '[' |
1595 | && i_byte < size_byte | |
1596 | && STRING_CHAR (str + i_byte, size_byte - i_byte) == ':') | |
1597 | { | |
1598 | const unsigned char *class_beg = str + i_byte + 1; | |
1599 | const unsigned char *class_end = class_beg; | |
1600 | const unsigned char *class_limit = str + size_byte - 2; | |
1601 | /* Leave room for the null. */ | |
1602 | unsigned char class_name[CHAR_CLASS_MAX_LENGTH + 1]; | |
1603 | re_wctype_t cc; | |
1604 | ||
1605 | if (class_limit - class_beg > CHAR_CLASS_MAX_LENGTH) | |
1606 | class_limit = class_beg + CHAR_CLASS_MAX_LENGTH; | |
1607 | ||
1608 | while (class_end < class_limit | |
1609 | && *class_end >= 'a' && *class_end <= 'z') | |
1610 | class_end++; | |
1611 | ||
1612 | if (class_end == class_beg | |
1613 | || *class_end != ':' || class_end[1] != ']') | |
1614 | goto not_a_class_name_multibyte; | |
1615 | ||
1616 | bcopy (class_beg, class_name, class_end - class_beg); | |
1617 | class_name[class_end - class_beg] = 0; | |
1618 | ||
1619 | cc = re_wctype (class_name); | |
1620 | if (cc == 0) | |
1621 | error ("Invalid ISO C character class"); | |
1622 | ||
1623 | iso_classes = Fcons (make_number (cc), iso_classes); | |
1624 | ||
1625 | i_byte = class_end + 2 - str; | |
1626 | continue; | |
1627 | } | |
1628 | ||
1629 | not_a_class_name_multibyte: | |
195d1361 RS |
1630 | if (c == '\\') |
1631 | { | |
13090112 | 1632 | if (i_byte == size_byte) |
4101e6fe RS |
1633 | break; |
1634 | ||
b7dbcc19 | 1635 | leading_code = str[i_byte]; |
839966f3 KH |
1636 | c = STRING_CHAR_AND_LENGTH (str + i_byte, |
1637 | size_byte - i_byte, len); | |
82d497fc | 1638 | i_byte += len; |
195d1361 | 1639 | } |
839966f3 KH |
1640 | /* Treat `-' as range character only if another character |
1641 | follows. */ | |
1642 | if (i_byte + 1 < size_byte | |
82d497fc | 1643 | && str[i_byte] == '-') |
195d1361 | 1644 | { |
9690d026 | 1645 | unsigned int c2; |
b7dbcc19 | 1646 | unsigned char leading_code2; |
4101e6fe RS |
1647 | |
1648 | /* Skip over the dash. */ | |
82d497fc | 1649 | i_byte++; |
4101e6fe | 1650 | |
4101e6fe | 1651 | /* Get the end of the range. */ |
b7dbcc19 | 1652 | leading_code2 = str[i_byte]; |
839966f3 KH |
1653 | c2 = STRING_CHAR_AND_LENGTH (str + i_byte, |
1654 | size_byte - i_byte, len); | |
82d497fc | 1655 | i_byte += len; |
7e68b0ea | 1656 | |
b7dbcc19 KH |
1657 | if (c2 == '\\' |
1658 | && i_byte < size_byte) | |
1659 | { | |
1660 | leading_code2 = str[i_byte]; | |
1661 | c2 =STRING_CHAR_AND_LENGTH (str + i_byte, size_byte-i_byte, len); | |
1662 | i_byte += len; | |
1663 | } | |
1664 | ||
f003ca54 KH |
1665 | if (c > c2) |
1666 | continue; | |
b7dbcc19 | 1667 | if (ASCII_CHAR_P (c)) |
06274af5 | 1668 | { |
5c7b02ab KH |
1669 | while (c <= c2 && c < 0x80) |
1670 | fastmap[c++] = 1; | |
b7dbcc19 KH |
1671 | leading_code = CHAR_LEADING_CODE (c); |
1672 | } | |
1673 | if (! ASCII_CHAR_P (c)) | |
1674 | { | |
1675 | while (leading_code <= leading_code2) | |
1676 | fastmap[leading_code++] = 1; | |
1677 | if (c <= c2) | |
e39091b6 | 1678 | { |
b7dbcc19 | 1679 | char_ranges[n_char_ranges++] = c; |
e39091b6 | 1680 | char_ranges[n_char_ranges++] = c2; |
06274af5 KH |
1681 | } |
1682 | } | |
195d1361 RS |
1683 | } |
1684 | else | |
7e68b0ea | 1685 | { |
b7dbcc19 | 1686 | if (ASCII_CHAR_P (c)) |
e39091b6 KH |
1687 | fastmap[c] = 1; |
1688 | else | |
7e68b0ea | 1689 | { |
b7dbcc19 | 1690 | fastmap[leading_code] = 1; |
4101e6fe RS |
1691 | char_ranges[n_char_ranges++] = c; |
1692 | char_ranges[n_char_ranges++] = c; | |
7e68b0ea RS |
1693 | } |
1694 | } | |
195d1361 | 1695 | } |
b7dbcc19 KH |
1696 | |
1697 | /* If the current range is unibyte and STRING contains non-ASCII | |
1698 | chars, arrange fastmap for the corresponding unibyte | |
1699 | chars. */ | |
1700 | ||
1701 | if (! multibyte && n_char_ranges > 0) | |
1702 | { | |
1703 | bzero (fastmap + 0200, 0200); | |
1704 | for (i = 0; i < n_char_ranges; i += 2) | |
1705 | { | |
1706 | int c1 = char_ranges[i]; | |
1707 | int c2 = char_ranges[i + 1]; | |
1708 | ||
1709 | for (; c1 <= c2; c1++) | |
2afc21f5 SM |
1710 | { |
1711 | int b = CHAR_TO_BYTE_SAFE (c1); | |
1712 | if (b >= 0) | |
1713 | fastmap[b] = 1; | |
1714 | } | |
b7dbcc19 KH |
1715 | } |
1716 | } | |
195d1361 RS |
1717 | } |
1718 | ||
9690d026 | 1719 | /* If ^ was the first character, complement the fastmap. */ |
195d1361 | 1720 | if (negate) |
b7dbcc19 KH |
1721 | { |
1722 | if (! multibyte) | |
1723 | for (i = 0; i < sizeof fastmap; i++) | |
1724 | fastmap[i] ^= 1; | |
1725 | else | |
1726 | { | |
1727 | for (i = 0; i < 0200; i++) | |
1728 | fastmap[i] ^= 1; | |
1729 | /* All non-ASCII chars possibly match. */ | |
1730 | for (; i < sizeof fastmap; i++) | |
1731 | fastmap[i] = 1; | |
1732 | } | |
1733 | } | |
195d1361 RS |
1734 | |
1735 | { | |
1736 | int start_point = PT; | |
1737 | int pos = PT; | |
6a140a74 | 1738 | int pos_byte = PT_BYTE; |
9af7511a KH |
1739 | unsigned char *p = PT_ADDR, *endp, *stop; |
1740 | ||
1741 | if (forwardp) | |
1742 | { | |
1743 | endp = (XINT (lim) == GPT) ? GPT_ADDR : CHAR_POS_ADDR (XINT (lim)); | |
d7ee9fab | 1744 | stop = (pos < GPT && GPT < XINT (lim)) ? GPT_ADDR : endp; |
9af7511a KH |
1745 | } |
1746 | else | |
1747 | { | |
1748 | endp = CHAR_POS_ADDR (XINT (lim)); | |
d7ee9fab | 1749 | stop = (pos >= GPT && GPT > XINT (lim)) ? GAP_END_ADDR : endp; |
9af7511a | 1750 | } |
195d1361 RS |
1751 | |
1752 | immediate_quit = 1; | |
b7dbcc19 | 1753 | if (forwardp) |
195d1361 | 1754 | { |
b7dbcc19 | 1755 | if (multibyte) |
8f924df7 | 1756 | while (1) |
b7dbcc19 | 1757 | { |
8f924df7 | 1758 | int nbytes; |
9af7511a | 1759 | |
8f924df7 | 1760 | if (p >= stop) |
9af7511a | 1761 | { |
8f924df7 | 1762 | if (p >= endp) |
9af7511a | 1763 | break; |
8f924df7 KH |
1764 | p = GAP_END_ADDR; |
1765 | stop = endp; | |
9af7511a | 1766 | } |
327719ee MB |
1767 | c = STRING_CHAR_AND_LENGTH (p, MAX_MULTIBYTE_LENGTH, nbytes); |
1768 | if (! NILP (iso_classes) && in_classes (c, iso_classes)) | |
9af7511a | 1769 | { |
327719ee | 1770 | if (negate) |
9af7511a | 1771 | break; |
327719ee MB |
1772 | else |
1773 | goto fwd_ok; | |
9af7511a | 1774 | } |
9af7511a | 1775 | |
8f924df7 | 1776 | if (! fastmap[*p]) |
b7dbcc19 KH |
1777 | break; |
1778 | if (! ASCII_CHAR_P (c)) | |
9af7511a | 1779 | { |
b7dbcc19 KH |
1780 | /* As we are looking at a multibyte character, we |
1781 | must look up the character in the table | |
1782 | CHAR_RANGES. If there's no data in the table, | |
1783 | that character is not what we want to skip. */ | |
1784 | ||
1785 | /* The following code do the right thing even if | |
1786 | n_char_ranges is zero (i.e. no data in | |
1787 | CHAR_RANGES). */ | |
1788 | for (i = 0; i < n_char_ranges; i += 2) | |
1789 | if (c >= char_ranges[i] && c <= char_ranges[i + 1]) | |
9af7511a | 1790 | break; |
b7dbcc19 KH |
1791 | if (!(negate ^ (i < n_char_ranges))) |
1792 | break; | |
9af7511a | 1793 | } |
327719ee | 1794 | fwd_ok: |
8f924df7 | 1795 | p += nbytes, pos++, pos_byte += nbytes; |
b7dbcc19 KH |
1796 | } |
1797 | else | |
8f924df7 KH |
1798 | while (1) |
1799 | { | |
1800 | if (p >= stop) | |
9af7511a | 1801 | { |
8f924df7 | 1802 | if (p >= endp) |
9af7511a | 1803 | break; |
8f924df7 KH |
1804 | p = GAP_END_ADDR; |
1805 | stop = endp; | |
9af7511a | 1806 | } |
327719ee MB |
1807 | |
1808 | if (!NILP (iso_classes) && in_classes (*p, iso_classes)) | |
1809 | { | |
1810 | if (negate) | |
1811 | break; | |
1812 | else | |
1813 | goto fwd_unibyte_ok; | |
1814 | } | |
1815 | ||
8f924df7 KH |
1816 | if (!fastmap[*p]) |
1817 | break; | |
327719ee | 1818 | fwd_unibyte_ok: |
8f924df7 KH |
1819 | p++, pos++, pos_byte++; |
1820 | } | |
195d1361 RS |
1821 | } |
1822 | else | |
1823 | { | |
b7dbcc19 | 1824 | if (multibyte) |
8f924df7 | 1825 | while (1) |
b7dbcc19 | 1826 | { |
8f924df7 | 1827 | unsigned char *prev_p; |
9af7511a | 1828 | |
8f924df7 | 1829 | if (p <= stop) |
9af7511a | 1830 | { |
8f924df7 | 1831 | if (p <= endp) |
9af7511a | 1832 | break; |
8f924df7 KH |
1833 | p = GPT_ADDR; |
1834 | stop = endp; | |
9af7511a | 1835 | } |
8f924df7 KH |
1836 | prev_p = p; |
1837 | while (--p >= stop && ! CHAR_HEAD_P (*p)); | |
327719ee | 1838 | c = STRING_CHAR (p, MAX_MULTIBYTE_LENGTH); |
a1bc88d4 | 1839 | |
327719ee MB |
1840 | if (! NILP (iso_classes) && in_classes (c, iso_classes)) |
1841 | { | |
1842 | if (negate) | |
1843 | break; | |
9690d026 | 1844 | else |
327719ee | 1845 | goto back_ok; |
7e68b0ea | 1846 | } |
a1bc88d4 | 1847 | |
8f924df7 | 1848 | if (! fastmap[*p]) |
b7dbcc19 KH |
1849 | break; |
1850 | if (! ASCII_CHAR_P (c)) | |
7e68b0ea | 1851 | { |
b7dbcc19 KH |
1852 | /* See the comment in the previous similar code. */ |
1853 | for (i = 0; i < n_char_ranges; i += 2) | |
1854 | if (c >= char_ranges[i] && c <= char_ranges[i + 1]) | |
1855 | break; | |
1856 | if (!(negate ^ (i < n_char_ranges))) | |
1857 | break; | |
7e68b0ea | 1858 | } |
327719ee | 1859 | back_ok: |
8f924df7 | 1860 | pos--, pos_byte -= prev_p - p; |
b7dbcc19 KH |
1861 | } |
1862 | else | |
8f924df7 KH |
1863 | while (1) |
1864 | { | |
1865 | if (p <= stop) | |
9af7511a | 1866 | { |
8f924df7 | 1867 | if (p <= endp) |
9af7511a | 1868 | break; |
8f924df7 KH |
1869 | p = GPT_ADDR; |
1870 | stop = endp; | |
9af7511a | 1871 | } |
e39091b6 | 1872 | |
327719ee MB |
1873 | if (! NILP (iso_classes) && in_classes (p[-1], iso_classes)) |
1874 | { | |
1875 | if (negate) | |
1876 | break; | |
9af7511a | 1877 | else |
327719ee MB |
1878 | goto back_unibyte_ok; |
1879 | } | |
a1bc88d4 | 1880 | |
8f924df7 KH |
1881 | if (!fastmap[p[-1]]) |
1882 | break; | |
327719ee | 1883 | back_unibyte_ok: |
8f924df7 KH |
1884 | p--, pos--, pos_byte--; |
1885 | } | |
195d1361 | 1886 | } |
7e68b0ea | 1887 | |
b7dbcc19 KH |
1888 | SET_PT_BOTH (pos, pos_byte); |
1889 | immediate_quit = 0; | |
1890 | ||
1891 | return make_number (PT - start_point); | |
1892 | } | |
1893 | } | |
7e68b0ea | 1894 | |
b7dbcc19 KH |
1895 | |
1896 | static Lisp_Object | |
1897 | skip_syntaxes (forwardp, string, lim) | |
1898 | int forwardp; | |
1899 | Lisp_Object string, lim; | |
1900 | { | |
1901 | register unsigned int c; | |
1902 | unsigned char fastmap[0400]; | |
1903 | int negate = 0; | |
1904 | register int i, i_byte; | |
1905 | int multibyte; | |
1906 | int size_byte; | |
1907 | unsigned char *str; | |
1908 | ||
1909 | CHECK_STRING (string); | |
1910 | ||
1911 | if (NILP (lim)) | |
1912 | XSETINT (lim, forwardp ? ZV : BEGV); | |
1913 | else | |
1914 | CHECK_NUMBER_COERCE_MARKER (lim); | |
1915 | ||
1916 | /* In any case, don't allow scan outside bounds of buffer. */ | |
1917 | if (XINT (lim) > ZV) | |
1918 | XSETFASTINT (lim, ZV); | |
1919 | if (XINT (lim) < BEGV) | |
1920 | XSETFASTINT (lim, BEGV); | |
1921 | ||
ade8ee9e | 1922 | if (forwardp ? (PT >= XFASTINT (lim)) : (PT <= XFASTINT (lim))) |
49feb1cd | 1923 | return make_number (0); |
8c6e735b | 1924 | |
b7dbcc19 | 1925 | multibyte = (!NILP (current_buffer->enable_multibyte_characters) |
92eaa22e | 1926 | && (XINT (lim) - PT != CHAR_TO_BYTE (XINT (lim)) - PT_BYTE)); |
b7dbcc19 KH |
1927 | |
1928 | bzero (fastmap, sizeof fastmap); | |
1929 | ||
8f924df7 KH |
1930 | if (SBYTES (string) > SCHARS (string)) |
1931 | /* As this is very rare case (syntax spec is ASCII only), don't | |
1932 | consider efficiency. */ | |
bc796a59 KH |
1933 | string = string_make_unibyte (string); |
1934 | ||
8f924df7 KH |
1935 | str = SDATA (string); |
1936 | size_byte = SBYTES (string); | |
bc796a59 | 1937 | |
b7dbcc19 KH |
1938 | i_byte = 0; |
1939 | if (i_byte < size_byte | |
8f924df7 | 1940 | && SREF (string, 0) == '^') |
b7dbcc19 KH |
1941 | { |
1942 | negate = 1; i_byte++; | |
1943 | } | |
1944 | ||
b7dbcc19 KH |
1945 | /* Find the syntaxes specified and set their elements of fastmap. */ |
1946 | ||
1947 | while (i_byte < size_byte) | |
1948 | { | |
1949 | c = str[i_byte++]; | |
8f924df7 | 1950 | fastmap[syntax_spec_code[c]] = 1; |
b7dbcc19 | 1951 | } |
195d1361 | 1952 | |
9690d026 | 1953 | /* If ^ was the first character, complement the fastmap. */ |
195d1361 RS |
1954 | if (negate) |
1955 | for (i = 0; i < sizeof fastmap; i++) | |
9690d026 | 1956 | fastmap[i] ^= 1; |
195d1361 RS |
1957 | |
1958 | { | |
1959 | int start_point = PT; | |
1960 | int pos = PT; | |
6a140a74 | 1961 | int pos_byte = PT_BYTE; |
8f924df7 KH |
1962 | unsigned char *p = PT_ADDR, *endp, *stop; |
1963 | ||
1964 | if (forwardp) | |
1965 | { | |
1966 | endp = (XINT (lim) == GPT) ? GPT_ADDR : CHAR_POS_ADDR (XINT (lim)); | |
1967 | stop = (pos < GPT && GPT < XINT (lim)) ? GPT_ADDR : endp; | |
1968 | } | |
1969 | else | |
1970 | { | |
1971 | endp = CHAR_POS_ADDR (XINT (lim)); | |
1972 | stop = (pos >= GPT && GPT > XINT (lim)) ? GAP_END_ADDR : endp; | |
1973 | } | |
195d1361 RS |
1974 | |
1975 | immediate_quit = 1; | |
b7dbcc19 KH |
1976 | SETUP_SYNTAX_TABLE (pos, forwardp ? 1 : -1); |
1977 | if (forwardp) | |
195d1361 | 1978 | { |
b7dbcc19 | 1979 | if (multibyte) |
195d1361 | 1980 | { |
8f924df7 | 1981 | while (1) |
8c6e735b | 1982 | { |
8f924df7 KH |
1983 | int nbytes; |
1984 | ||
1985 | if (p >= stop) | |
1986 | { | |
1987 | if (p >= endp) | |
1988 | break; | |
1989 | p = GAP_END_ADDR; | |
1990 | stop = endp; | |
1991 | } | |
1992 | c = STRING_CHAR_AND_LENGTH (p, MAX_MULTIBYTE_LENGTH, nbytes); | |
1993 | if (! fastmap[(int) SYNTAX (c)]) | |
8c6e735b | 1994 | break; |
8f924df7 | 1995 | p += nbytes, pos++, pos_byte += nbytes; |
8c6e735b KH |
1996 | UPDATE_SYNTAX_TABLE_FORWARD (pos); |
1997 | } | |
195d1361 RS |
1998 | } |
1999 | else | |
2000 | { | |
8c6e735b | 2001 | while (1) |
7e68b0ea | 2002 | { |
8f924df7 KH |
2003 | if (p >= stop) |
2004 | { | |
2005 | if (p >= endp) | |
2006 | break; | |
2007 | p = GAP_END_ADDR; | |
2008 | stop = endp; | |
2009 | } | |
2010 | if (! fastmap[(int) SYNTAX (*p)]) | |
8c6e735b | 2011 | break; |
8f924df7 | 2012 | p++, pos++, pos_byte++; |
b7dbcc19 | 2013 | UPDATE_SYNTAX_TABLE_FORWARD (pos); |
195d1361 RS |
2014 | } |
2015 | } | |
2016 | } | |
2017 | else | |
2018 | { | |
b7dbcc19 | 2019 | if (multibyte) |
195d1361 | 2020 | { |
8c6e735b | 2021 | while (1) |
b7dbcc19 | 2022 | { |
8f924df7 KH |
2023 | unsigned char *prev_p; |
2024 | ||
2025 | if (p <= stop) | |
b7dbcc19 | 2026 | { |
8f924df7 KH |
2027 | if (p <= endp) |
2028 | break; | |
2029 | p = GPT_ADDR; | |
2030 | stop = endp; | |
b7dbcc19 | 2031 | } |
262be72a | 2032 | UPDATE_SYNTAX_TABLE_BACKWARD (pos - 1); |
8f924df7 KH |
2033 | prev_p = p; |
2034 | while (--p >= stop && ! CHAR_HEAD_P (*p)); | |
2035 | c = STRING_CHAR (p, MAX_MULTIBYTE_LENGTH); | |
2036 | if (! fastmap[(int) SYNTAX (c)]) | |
2037 | break; | |
2038 | pos--, pos_byte -= prev_p - p; | |
b7dbcc19 | 2039 | } |
195d1361 RS |
2040 | } |
2041 | else | |
2042 | { | |
8c6e735b KH |
2043 | while (1) |
2044 | { | |
8f924df7 KH |
2045 | if (p <= stop) |
2046 | { | |
2047 | if (p <= endp) | |
2048 | break; | |
2049 | p = GPT_ADDR; | |
2050 | stop = endp; | |
2051 | } | |
102f6132 | 2052 | UPDATE_SYNTAX_TABLE_BACKWARD (pos - 1); |
8f924df7 | 2053 | if (! fastmap[(int) SYNTAX (p[-1])]) |
8c6e735b | 2054 | break; |
8f924df7 | 2055 | p--, pos--, pos_byte--; |
8c6e735b | 2056 | } |
195d1361 RS |
2057 | } |
2058 | } | |
6a140a74 RS |
2059 | |
2060 | SET_PT_BOTH (pos, pos_byte); | |
195d1361 RS |
2061 | immediate_quit = 0; |
2062 | ||
2063 | return make_number (PT - start_point); | |
2064 | } | |
2065 | } | |
a1bc88d4 RS |
2066 | |
2067 | /* Return 1 if character C belongs to one of the ISO classes | |
2068 | in the list ISO_CLASSES. Each class is represented by an | |
2069 | integer which is its type according to re_wctype. */ | |
2070 | ||
2071 | static int | |
2072 | in_classes (c, iso_classes) | |
2073 | int c; | |
2074 | Lisp_Object iso_classes; | |
2075 | { | |
2076 | int fits_class = 0; | |
2077 | ||
2078 | while (! NILP (iso_classes)) | |
2079 | { | |
2080 | Lisp_Object elt; | |
2081 | elt = XCAR (iso_classes); | |
2082 | iso_classes = XCDR (iso_classes); | |
2083 | ||
2084 | if (re_iswctype (c, XFASTINT (elt))) | |
2085 | fits_class = 1; | |
2086 | } | |
2087 | ||
2088 | return fits_class; | |
2089 | } | |
195d1361 | 2090 | \f |
95ff8dfc RS |
2091 | /* Jump over a comment, assuming we are at the beginning of one. |
2092 | FROM is the current position. | |
2093 | FROM_BYTE is the bytepos corresponding to FROM. | |
2094 | Do not move past STOP (a charpos). | |
2095 | The comment over which we have to jump is of style STYLE | |
2096 | (either SYNTAX_COMMENT_STYLE(foo) or ST_COMMENT_STYLE). | |
2097 | NESTING should be positive to indicate the nesting at the beginning | |
2098 | for nested comments and should be zero or negative else. | |
2099 | ST_COMMENT_STYLE cannot be nested. | |
2100 | PREV_SYNTAX is the SYNTAX_WITH_FLAGS of the previous character | |
2101 | (or 0 If the search cannot start in the middle of a two-character). | |
2102 | ||
2103 | If successful, return 1 and store the charpos of the comment's end | |
2104 | into *CHARPOS_PTR and the corresponding bytepos into *BYTEPOS_PTR. | |
2105 | Else, return 0 and store the charpos STOP into *CHARPOS_PTR, the | |
2106 | corresponding bytepos into *BYTEPOS_PTR and the current nesting | |
2107 | (as defined for state.incomment) in *INCOMMENT_PTR. | |
2108 | ||
2109 | The comment end is the last character of the comment rather than the | |
2110 | character just after the comment. | |
2111 | ||
2112 | Global syntax data is assumed to initially be valid for FROM and | |
2113 | remains valid for forward search starting at the returned position. */ | |
2114 | ||
2115 | static int | |
2116 | forw_comment (from, from_byte, stop, nesting, style, prev_syntax, | |
2117 | charpos_ptr, bytepos_ptr, incomment_ptr) | |
a0aa7fcf | 2118 | EMACS_INT from, from_byte, stop; |
95ff8dfc | 2119 | int nesting, style, prev_syntax; |
a0aa7fcf SM |
2120 | EMACS_INT *charpos_ptr, *bytepos_ptr; |
2121 | int *incomment_ptr; | |
95ff8dfc RS |
2122 | { |
2123 | register int c, c1; | |
2124 | register enum syntaxcode code; | |
2125 | register int syntax; | |
2126 | ||
2127 | if (nesting <= 0) nesting = -1; | |
2128 | ||
2129 | /* Enter the loop in the middle so that we find | |
2130 | a 2-char comment ender if we start in the middle of it. */ | |
2131 | syntax = prev_syntax; | |
2132 | if (syntax != 0) goto forw_incomment; | |
2133 | ||
2134 | while (1) | |
2135 | { | |
2136 | if (from == stop) | |
2137 | { | |
2138 | *incomment_ptr = nesting; | |
2139 | *charpos_ptr = from; | |
2140 | *bytepos_ptr = from_byte; | |
2141 | return 0; | |
2142 | } | |
8f924df7 | 2143 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2144 | syntax = SYNTAX_WITH_FLAGS (c); |
2145 | code = syntax & 0xff; | |
2146 | if (code == Sendcomment | |
2147 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style | |
abf8a9ff SM |
2148 | && (SYNTAX_FLAGS_COMMENT_NESTED (syntax) ? |
2149 | (nesting > 0 && --nesting == 0) : nesting < 0)) | |
95ff8dfc RS |
2150 | /* we have encountered a comment end of the same style |
2151 | as the comment sequence which began this comment | |
2152 | section */ | |
2153 | break; | |
2154 | if (code == Scomment_fence | |
2155 | && style == ST_COMMENT_STYLE) | |
2156 | /* we have encountered a comment end of the same style | |
2157 | as the comment sequence which began this comment | |
2158 | section. */ | |
2159 | break; | |
2160 | if (nesting > 0 | |
2161 | && code == Scomment | |
abf8a9ff | 2162 | && SYNTAX_FLAGS_COMMENT_NESTED (syntax) |
95ff8dfc RS |
2163 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style) |
2164 | /* we have encountered a nested comment of the same style | |
2165 | as the comment sequence which began this comment section */ | |
2166 | nesting++; | |
2167 | INC_BOTH (from, from_byte); | |
2168 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
7d0393cf | 2169 | |
95ff8dfc RS |
2170 | forw_incomment: |
2171 | if (from < stop && SYNTAX_FLAGS_COMEND_FIRST (syntax) | |
2172 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style | |
8f924df7 | 2173 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
abf8a9ff SM |
2174 | SYNTAX_COMEND_SECOND (c1)) |
2175 | && ((SYNTAX_FLAGS_COMMENT_NESTED (syntax) || | |
2176 | SYNTAX_COMMENT_NESTED (c1)) ? nesting > 0 : nesting < 0)) | |
4ffe723b GM |
2177 | { |
2178 | if (--nesting <= 0) | |
2179 | /* we have encountered a comment end of the same style | |
2180 | as the comment sequence which began this comment | |
2181 | section */ | |
2182 | break; | |
2183 | else | |
2184 | { | |
2185 | INC_BOTH (from, from_byte); | |
2186 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
2187 | } | |
2188 | } | |
95ff8dfc RS |
2189 | if (nesting > 0 |
2190 | && from < stop | |
2191 | && SYNTAX_FLAGS_COMSTART_FIRST (syntax) | |
8f924df7 | 2192 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
95ff8dfc | 2193 | SYNTAX_COMMENT_STYLE (c1) == style |
abf8a9ff SM |
2194 | && SYNTAX_COMSTART_SECOND (c1)) |
2195 | && (SYNTAX_FLAGS_COMMENT_NESTED (syntax) || | |
2196 | SYNTAX_COMMENT_NESTED (c1))) | |
95ff8dfc RS |
2197 | /* we have encountered a nested comment of the same style |
2198 | as the comment sequence which began this comment | |
2199 | section */ | |
2200 | { | |
2201 | INC_BOTH (from, from_byte); | |
2202 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
2203 | nesting++; | |
2204 | } | |
2205 | } | |
2206 | *charpos_ptr = from; | |
2207 | *bytepos_ptr = from_byte; | |
2208 | return 1; | |
2209 | } | |
2210 | ||
b3cfe0c8 | 2211 | DEFUN ("forward-comment", Fforward_comment, Sforward_comment, 1, 1, 0, |
177c0ea7 | 2212 | doc: /* |
ee648c11 | 2213 | Move forward across up to COUNT comments. If COUNT is negative, move backward. |
fdb82f93 PJ |
2214 | Stop scanning if we find something other than a comment or whitespace. |
2215 | Set point to where scanning stops. | |
ee648c11 | 2216 | If COUNT comments are found as expected, with nothing except whitespace |
fdb82f93 PJ |
2217 | between them, return t; otherwise return nil. */) |
2218 | (count) | |
840f481c | 2219 | Lisp_Object count; |
b3cfe0c8 | 2220 | { |
2312c580 SM |
2221 | register EMACS_INT from; |
2222 | EMACS_INT from_byte; | |
2223 | register EMACS_INT stop; | |
8ea151b2 | 2224 | register int c, c1; |
b3cfe0c8 RS |
2225 | register enum syntaxcode code; |
2226 | int comstyle = 0; /* style of comment encountered */ | |
95ff8dfc | 2227 | int comnested = 0; /* whether the comment is nestable or not */ |
be720845 | 2228 | int found; |
2312c580 SM |
2229 | EMACS_INT count1; |
2230 | EMACS_INT out_charpos, out_bytepos; | |
95ff8dfc | 2231 | int dummy; |
840f481c | 2232 | |
b7826503 | 2233 | CHECK_NUMBER (count); |
840f481c | 2234 | count1 = XINT (count); |
195d1361 | 2235 | stop = count1 > 0 ? ZV : BEGV; |
b3cfe0c8 RS |
2236 | |
2237 | immediate_quit = 1; | |
2238 | QUIT; | |
2239 | ||
2240 | from = PT; | |
6a140a74 | 2241 | from_byte = PT_BYTE; |
b3cfe0c8 | 2242 | |
195d1361 | 2243 | SETUP_SYNTAX_TABLE (from, count1); |
840f481c | 2244 | while (count1 > 0) |
b3cfe0c8 | 2245 | { |
04882296 | 2246 | do |
b3cfe0c8 | 2247 | { |
f902a008 RS |
2248 | int comstart_first; |
2249 | ||
04882296 KH |
2250 | if (from == stop) |
2251 | { | |
ef316cf0 | 2252 | SET_PT_BOTH (from, from_byte); |
b7e6e612 | 2253 | immediate_quit = 0; |
04882296 KH |
2254 | return Qnil; |
2255 | } | |
b7dbcc19 | 2256 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
b3cfe0c8 | 2257 | code = SYNTAX (c); |
f902a008 | 2258 | comstart_first = SYNTAX_COMSTART_FIRST (c); |
95ff8dfc | 2259 | comnested = SYNTAX_COMMENT_NESTED (c); |
e6365855 | 2260 | comstyle = SYNTAX_COMMENT_STYLE (c); |
6a140a74 | 2261 | INC_BOTH (from, from_byte); |
f902a008 | 2262 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
f902a008 | 2263 | if (from < stop && comstart_first |
b7dbcc19 | 2264 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
8ea151b2 | 2265 | SYNTAX_COMSTART_SECOND (c1))) |
b3cfe0c8 | 2266 | { |
7d0393cf | 2267 | /* We have encountered a comment start sequence and we |
7fc8191e | 2268 | are ignoring all text inside comments. We must record |
b3cfe0c8 RS |
2269 | the comment style this sequence begins so that later, |
2270 | only a comment end of the same style actually ends | |
7fc8191e | 2271 | the comment section. */ |
b3cfe0c8 | 2272 | code = Scomment; |
8ea151b2 | 2273 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
95ff8dfc | 2274 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); |
6a140a74 | 2275 | INC_BOTH (from, from_byte); |
95ff8dfc | 2276 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b3cfe0c8 | 2277 | } |
04882296 | 2278 | } |
abf8a9ff | 2279 | while (code == Swhitespace || (code == Sendcomment && c == '\n')); |
6a140a74 | 2280 | |
da71ab91 KH |
2281 | if (code == Scomment_fence) |
2282 | comstyle = ST_COMMENT_STYLE; | |
2283 | else if (code != Scomment) | |
04882296 KH |
2284 | { |
2285 | immediate_quit = 0; | |
6a140a74 | 2286 | DEC_BOTH (from, from_byte); |
ef316cf0 | 2287 | SET_PT_BOTH (from, from_byte); |
04882296 KH |
2288 | return Qnil; |
2289 | } | |
2290 | /* We're at the start of a comment. */ | |
95ff8dfc RS |
2291 | found = forw_comment (from, from_byte, stop, comnested, comstyle, 0, |
2292 | &out_charpos, &out_bytepos, &dummy); | |
2293 | from = out_charpos; from_byte = out_bytepos; | |
2294 | if (!found) | |
04882296 | 2295 | { |
95ff8dfc RS |
2296 | immediate_quit = 0; |
2297 | SET_PT_BOTH (from, from_byte); | |
2298 | return Qnil; | |
b3cfe0c8 | 2299 | } |
95ff8dfc RS |
2300 | INC_BOTH (from, from_byte); |
2301 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
04882296 | 2302 | /* We have skipped one comment. */ |
840f481c | 2303 | count1--; |
b3cfe0c8 RS |
2304 | } |
2305 | ||
840f481c | 2306 | while (count1 < 0) |
b3cfe0c8 | 2307 | { |
b9145dbb | 2308 | while (1) |
b3cfe0c8 | 2309 | { |
c65adb44 | 2310 | int quoted; |
f902a008 | 2311 | |
b9145dbb RS |
2312 | if (from <= stop) |
2313 | { | |
6a140a74 | 2314 | SET_PT_BOTH (BEGV, BEGV_BYTE); |
b9145dbb RS |
2315 | immediate_quit = 0; |
2316 | return Qnil; | |
2317 | } | |
b3cfe0c8 | 2318 | |
6a140a74 | 2319 | DEC_BOTH (from, from_byte); |
f902a008 | 2320 | /* char_quoted does UPDATE_SYNTAX_TABLE_BACKWARD (from). */ |
6a140a74 | 2321 | quoted = char_quoted (from, from_byte); |
b7dbcc19 | 2322 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
b3cfe0c8 RS |
2323 | code = SYNTAX (c); |
2324 | comstyle = 0; | |
95ff8dfc | 2325 | comnested = SYNTAX_COMMENT_NESTED (c); |
7fc8191e RS |
2326 | if (code == Sendcomment) |
2327 | comstyle = SYNTAX_COMMENT_STYLE (c); | |
b3cfe0c8 | 2328 | if (from > stop && SYNTAX_COMEND_SECOND (c) |
f902a008 | 2329 | && prev_char_comend_first (from, from_byte) |
89c6809a | 2330 | && !char_quoted (from - 1, dec_bytepos (from_byte))) |
b3cfe0c8 | 2331 | { |
7fc8191e | 2332 | /* We must record the comment style encountered so that |
b3cfe0c8 | 2333 | later, we can match only the proper comment begin |
7fc8191e | 2334 | sequence of the same style. */ |
6a140a74 | 2335 | DEC_BOTH (from, from_byte); |
f902a008 RS |
2336 | code = Sendcomment; |
2337 | /* Calling char_quoted, above, set up global syntax position | |
2338 | at the new value of FROM. */ | |
b7dbcc19 | 2339 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2340 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
2341 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
b3cfe0c8 RS |
2342 | } |
2343 | ||
195d1361 RS |
2344 | if (code == Scomment_fence) |
2345 | { | |
2346 | /* Skip until first preceding unquoted comment_fence. */ | |
6a140a74 | 2347 | int found = 0, ini = from, ini_byte = from_byte; |
7d0393cf | 2348 | |
6a140a74 | 2349 | while (1) |
195d1361 | 2350 | { |
6a140a74 | 2351 | DEC_BOTH (from, from_byte); |
195d1361 | 2352 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 2353 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
6a140a74 | 2354 | if (SYNTAX (c) == Scomment_fence |
7d0393cf | 2355 | && !char_quoted (from, from_byte)) |
195d1361 | 2356 | { |
7d0393cf | 2357 | found = 1; |
195d1361 RS |
2358 | break; |
2359 | } | |
fcdd4585 SM |
2360 | else if (from == stop) |
2361 | break; | |
195d1361 RS |
2362 | } |
2363 | if (found == 0) | |
2364 | { | |
2365 | from = ini; /* Set point to ini + 1. */ | |
6a140a74 | 2366 | from_byte = ini_byte; |
195d1361 RS |
2367 | goto leave; |
2368 | } | |
39c41ad4 SM |
2369 | else |
2370 | /* We have skipped one comment. */ | |
2371 | break; | |
195d1361 RS |
2372 | } |
2373 | else if (code == Sendcomment) | |
b3cfe0c8 | 2374 | { |
95ff8dfc | 2375 | found = back_comment (from, from_byte, stop, comnested, comstyle, |
6a140a74 | 2376 | &out_charpos, &out_bytepos); |
1aa963c8 SM |
2377 | if (found == -1) |
2378 | { | |
abf8a9ff SM |
2379 | if (c == '\n') |
2380 | /* This end-of-line is not an end-of-comment. | |
2381 | Treat it like a whitespace. | |
2382 | CC-mode (and maybe others) relies on this behavior. */ | |
2383 | ; | |
2384 | else | |
2385 | { | |
2386 | /* Failure: we should go back to the end of this | |
2387 | not-quite-endcomment. */ | |
2388 | if (SYNTAX(c) != code) | |
2389 | /* It was a two-char Sendcomment. */ | |
2390 | INC_BOTH (from, from_byte); | |
2391 | goto leave; | |
2392 | } | |
1aa963c8 | 2393 | } |
2b927d02 | 2394 | else |
abf8a9ff SM |
2395 | { |
2396 | /* We have skipped one comment. */ | |
2397 | from = out_charpos, from_byte = out_bytepos; | |
2398 | break; | |
2399 | } | |
b3cfe0c8 | 2400 | } |
bb0de084 | 2401 | else if (code != Swhitespace || quoted) |
b3cfe0c8 | 2402 | { |
195d1361 | 2403 | leave: |
b3cfe0c8 | 2404 | immediate_quit = 0; |
6a140a74 | 2405 | INC_BOTH (from, from_byte); |
ef316cf0 | 2406 | SET_PT_BOTH (from, from_byte); |
b3cfe0c8 RS |
2407 | return Qnil; |
2408 | } | |
2409 | } | |
2410 | ||
840f481c | 2411 | count1++; |
b3cfe0c8 RS |
2412 | } |
2413 | ||
ef316cf0 | 2414 | SET_PT_BOTH (from, from_byte); |
b3cfe0c8 RS |
2415 | immediate_quit = 0; |
2416 | return Qt; | |
2417 | } | |
2418 | \f | |
b7dbcc19 | 2419 | /* Return syntax code of character C if C is an ASCII character |
db9f48c3 | 2420 | or `multibyte_symbol_p' is zero. Otherwise, return Ssymbol. */ |
bd25db08 | 2421 | |
b7dbcc19 KH |
2422 | #define SYNTAX_WITH_MULTIBYTE_CHECK(c) \ |
2423 | ((ASCII_CHAR_P (c) || !multibyte_symbol_p) \ | |
bd25db08 KH |
2424 | ? SYNTAX (c) : Ssymbol) |
2425 | ||
6a140a74 | 2426 | static Lisp_Object |
8489eb67 | 2427 | scan_lists (from, count, depth, sexpflag) |
932e6895 SM |
2428 | register EMACS_INT from; |
2429 | EMACS_INT count, depth; | |
2430 | int sexpflag; | |
8489eb67 RS |
2431 | { |
2432 | Lisp_Object val; | |
932e6895 | 2433 | register EMACS_INT stop = count > 0 ? ZV : BEGV; |
93da5fff KH |
2434 | register int c, c1; |
2435 | int stringterm; | |
8489eb67 RS |
2436 | int quoted; |
2437 | int mathexit = 0; | |
93da5fff | 2438 | register enum syntaxcode code, temp_code; |
195d1361 | 2439 | int min_depth = depth; /* Err out if depth gets less than this. */ |
e5d4f4dc | 2440 | int comstyle = 0; /* style of comment encountered */ |
95ff8dfc | 2441 | int comnested = 0; /* whether the comment is nestable or not */ |
932e6895 SM |
2442 | EMACS_INT temp_pos; |
2443 | EMACS_INT last_good = from; | |
195d1361 | 2444 | int found; |
932e6895 SM |
2445 | EMACS_INT from_byte; |
2446 | EMACS_INT out_bytepos, out_charpos; | |
95ff8dfc | 2447 | int temp, dummy; |
bd25db08 | 2448 | int multibyte_symbol_p = sexpflag && multibyte_syntax_as_symbol; |
8489eb67 RS |
2449 | |
2450 | if (depth > 0) min_depth = 0; | |
2451 | ||
c3907a7e RS |
2452 | if (from > ZV) from = ZV; |
2453 | if (from < BEGV) from = BEGV; | |
2454 | ||
2455 | from_byte = CHAR_TO_BYTE (from); | |
2456 | ||
8489eb67 RS |
2457 | immediate_quit = 1; |
2458 | QUIT; | |
2459 | ||
195d1361 | 2460 | SETUP_SYNTAX_TABLE (from, count); |
8489eb67 RS |
2461 | while (count > 0) |
2462 | { | |
8489eb67 RS |
2463 | while (from < stop) |
2464 | { | |
f902a008 | 2465 | int comstart_first, prefix; |
195d1361 | 2466 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 2467 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2468 | code = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
f902a008 | 2469 | comstart_first = SYNTAX_COMSTART_FIRST (c); |
95ff8dfc | 2470 | comnested = SYNTAX_COMMENT_NESTED (c); |
e6365855 | 2471 | comstyle = SYNTAX_COMMENT_STYLE (c); |
f902a008 | 2472 | prefix = SYNTAX_PREFIX (c); |
7bf5e9e4 RS |
2473 | if (depth == min_depth) |
2474 | last_good = from; | |
6a140a74 | 2475 | INC_BOTH (from, from_byte); |
195d1361 | 2476 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
f902a008 | 2477 | if (from < stop && comstart_first |
327719ee MB |
2478 | && (c = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
2479 | SYNTAX_COMSTART_SECOND (c)) | |
8489eb67 | 2480 | && parse_sexp_ignore_comments) |
e5d4f4dc | 2481 | { |
7d0393cf | 2482 | /* we have encountered a comment start sequence and we |
195d1361 | 2483 | are ignoring all text inside comments. We must record |
e5d4f4dc RS |
2484 | the comment style this sequence begins so that later, |
2485 | only a comment end of the same style actually ends | |
2486 | the comment section */ | |
2487 | code = Scomment; | |
b7dbcc19 | 2488 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2489 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
2490 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
6a140a74 | 2491 | INC_BOTH (from, from_byte); |
f902a008 | 2492 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
e5d4f4dc | 2493 | } |
7d0393cf | 2494 | |
f902a008 | 2495 | if (prefix) |
8489eb67 RS |
2496 | continue; |
2497 | ||
0220c518 | 2498 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
2499 | { |
2500 | case Sescape: | |
2501 | case Scharquote: | |
8b8bf8e6 SM |
2502 | if (from == stop) |
2503 | goto lose; | |
6a140a74 | 2504 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2505 | /* treat following character as a word constituent */ |
2506 | case Sword: | |
2507 | case Ssymbol: | |
2508 | if (depth || !sexpflag) break; | |
195d1361 | 2509 | /* This word counts as a sexp; return at end of it. */ |
8489eb67 RS |
2510 | while (from < stop) |
2511 | { | |
195d1361 | 2512 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
7c7ca3d2 RS |
2513 | |
2514 | /* Some compilers can't handle this inside the switch. */ | |
b7dbcc19 | 2515 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2516 | temp = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7c7ca3d2 | 2517 | switch (temp) |
8489eb67 RS |
2518 | { |
2519 | case Scharquote: | |
2520 | case Sescape: | |
6a140a74 | 2521 | INC_BOTH (from, from_byte); |
8b8bf8e6 SM |
2522 | if (from == stop) |
2523 | goto lose; | |
8489eb67 RS |
2524 | break; |
2525 | case Sword: | |
2526 | case Ssymbol: | |
2527 | case Squote: | |
2528 | break; | |
2529 | default: | |
2530 | goto done; | |
2531 | } | |
6a140a74 | 2532 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2533 | } |
2534 | goto done; | |
2535 | ||
195d1361 | 2536 | case Scomment_fence: |
95ff8dfc RS |
2537 | comstyle = ST_COMMENT_STYLE; |
2538 | /* FALLTHROUGH */ | |
2539 | case Scomment: | |
8489eb67 | 2540 | if (!parse_sexp_ignore_comments) break; |
95ff8dfc RS |
2541 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
2542 | found = forw_comment (from, from_byte, stop, | |
2543 | comnested, comstyle, 0, | |
2544 | &out_charpos, &out_bytepos, &dummy); | |
2545 | from = out_charpos, from_byte = out_bytepos; | |
2546 | if (!found) | |
8489eb67 | 2547 | { |
95ff8dfc RS |
2548 | if (depth == 0) |
2549 | goto done; | |
2550 | goto lose; | |
8489eb67 | 2551 | } |
95ff8dfc RS |
2552 | INC_BOTH (from, from_byte); |
2553 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
8489eb67 RS |
2554 | break; |
2555 | ||
2556 | case Smath: | |
2557 | if (!sexpflag) | |
2558 | break; | |
b7dbcc19 | 2559 | if (from != stop && c == FETCH_CHAR_AS_MULTIBYTE (from_byte)) |
6a140a74 RS |
2560 | { |
2561 | INC_BOTH (from, from_byte); | |
2562 | } | |
8489eb67 RS |
2563 | if (mathexit) |
2564 | { | |
2565 | mathexit = 0; | |
2566 | goto close1; | |
2567 | } | |
2568 | mathexit = 1; | |
2569 | ||
2570 | case Sopen: | |
2571 | if (!++depth) goto done; | |
2572 | break; | |
2573 | ||
2574 | case Sclose: | |
2575 | close1: | |
2576 | if (!--depth) goto done; | |
2577 | if (depth < min_depth) | |
9471945b KS |
2578 | xsignal3 (Qscan_error, |
2579 | build_string ("Containing expression ends prematurely"), | |
2580 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2581 | break; |
2582 | ||
2583 | case Sstring: | |
195d1361 | 2584 | case Sstring_fence: |
ef316cf0 | 2585 | temp_pos = dec_bytepos (from_byte); |
b7dbcc19 | 2586 | stringterm = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
8489eb67 RS |
2587 | while (1) |
2588 | { | |
8b8bf8e6 SM |
2589 | if (from >= stop) |
2590 | goto lose; | |
195d1361 | 2591 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 2592 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
42ebfa31 SM |
2593 | if (code == Sstring |
2594 | ? (c == stringterm | |
2595 | && SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring) | |
bd25db08 | 2596 | : SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring_fence) |
195d1361 | 2597 | break; |
7c7ca3d2 RS |
2598 | |
2599 | /* Some compilers can't handle this inside the switch. */ | |
bd25db08 | 2600 | temp = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7c7ca3d2 | 2601 | switch (temp) |
8489eb67 RS |
2602 | { |
2603 | case Scharquote: | |
2604 | case Sescape: | |
6a140a74 | 2605 | INC_BOTH (from, from_byte); |
8489eb67 | 2606 | } |
6a140a74 | 2607 | INC_BOTH (from, from_byte); |
8489eb67 | 2608 | } |
6a140a74 | 2609 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2610 | if (!depth && sexpflag) goto done; |
2611 | break; | |
240c43e8 SM |
2612 | default: |
2613 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
2614 | break; | |
8489eb67 RS |
2615 | } |
2616 | } | |
2617 | ||
2618 | /* Reached end of buffer. Error if within object, return nil if between */ | |
8b8bf8e6 SM |
2619 | if (depth) |
2620 | goto lose; | |
8489eb67 RS |
2621 | |
2622 | immediate_quit = 0; | |
2623 | return Qnil; | |
2624 | ||
2625 | /* End of object reached */ | |
2626 | done: | |
2627 | count--; | |
2628 | } | |
2629 | ||
2630 | while (count < 0) | |
2631 | { | |
8489eb67 RS |
2632 | while (from > stop) |
2633 | { | |
6a140a74 | 2634 | DEC_BOTH (from, from_byte); |
195d1361 | 2635 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 2636 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2637 | code = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7bf5e9e4 RS |
2638 | if (depth == min_depth) |
2639 | last_good = from; | |
7fc8191e | 2640 | comstyle = 0; |
95ff8dfc | 2641 | comnested = SYNTAX_COMMENT_NESTED (c); |
7fc8191e RS |
2642 | if (code == Sendcomment) |
2643 | comstyle = SYNTAX_COMMENT_STYLE (c); | |
8489eb67 | 2644 | if (from > stop && SYNTAX_COMEND_SECOND (c) |
1674d9a2 | 2645 | && prev_char_comend_first (from, from_byte) |
8489eb67 | 2646 | && parse_sexp_ignore_comments) |
e5d4f4dc | 2647 | { |
f902a008 | 2648 | /* We must record the comment style encountered so that |
e5d4f4dc | 2649 | later, we can match only the proper comment begin |
f902a008 | 2650 | sequence of the same style. */ |
6a140a74 | 2651 | DEC_BOTH (from, from_byte); |
f902a008 RS |
2652 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
2653 | code = Sendcomment; | |
b7dbcc19 | 2654 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2655 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
2656 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
e5d4f4dc | 2657 | } |
7d0393cf | 2658 | |
9828a477 | 2659 | /* Quoting turns anything except a comment-ender |
e6365855 | 2660 | into a word character. Note that this cannot be true |
f902a008 | 2661 | if we decremented FROM in the if-statement above. */ |
9828a477 | 2662 | if (code != Sendcomment && char_quoted (from, from_byte)) |
240c43e8 SM |
2663 | { |
2664 | DEC_BOTH (from, from_byte); | |
2665 | code = Sword; | |
2666 | } | |
9828a477 | 2667 | else if (SYNTAX_PREFIX (c)) |
8489eb67 RS |
2668 | continue; |
2669 | ||
9828a477 | 2670 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
2671 | { |
2672 | case Sword: | |
2673 | case Ssymbol: | |
9828a477 KH |
2674 | case Sescape: |
2675 | case Scharquote: | |
8489eb67 | 2676 | if (depth || !sexpflag) break; |
195d1361 RS |
2677 | /* This word counts as a sexp; count object finished |
2678 | after passing it. */ | |
8489eb67 RS |
2679 | while (from > stop) |
2680 | { | |
6a140a74 | 2681 | temp_pos = from_byte; |
ef316cf0 RS |
2682 | if (! NILP (current_buffer->enable_multibyte_characters)) |
2683 | DEC_POS (temp_pos); | |
2684 | else | |
2685 | temp_pos--; | |
6a140a74 | 2686 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
b7dbcc19 | 2687 | c1 = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
bd25db08 | 2688 | temp_code = SYNTAX_WITH_MULTIBYTE_CHECK (c1); |
9828a477 KH |
2689 | /* Don't allow comment-end to be quoted. */ |
2690 | if (temp_code == Sendcomment) | |
2691 | goto done2; | |
6a140a74 | 2692 | quoted = char_quoted (from - 1, temp_pos); |
8489eb67 | 2693 | if (quoted) |
93da5fff | 2694 | { |
6a140a74 | 2695 | DEC_BOTH (from, from_byte); |
ef316cf0 | 2696 | temp_pos = dec_bytepos (temp_pos); |
6a140a74 | 2697 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
93da5fff | 2698 | } |
b7dbcc19 | 2699 | c1 = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
bd25db08 | 2700 | temp_code = SYNTAX_WITH_MULTIBYTE_CHECK (c1); |
93da5fff KH |
2701 | if (! (quoted || temp_code == Sword |
2702 | || temp_code == Ssymbol | |
2703 | || temp_code == Squote)) | |
8489eb67 | 2704 | goto done2; |
6a140a74 | 2705 | DEC_BOTH (from, from_byte); |
8489eb67 RS |
2706 | } |
2707 | goto done2; | |
2708 | ||
2709 | case Smath: | |
2710 | if (!sexpflag) | |
2711 | break; | |
ef316cf0 | 2712 | temp_pos = dec_bytepos (from_byte); |
6a140a74 | 2713 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
b7dbcc19 | 2714 | if (from != stop && c == FETCH_CHAR_AS_MULTIBYTE (temp_pos)) |
6a140a74 | 2715 | DEC_BOTH (from, from_byte); |
8489eb67 RS |
2716 | if (mathexit) |
2717 | { | |
2718 | mathexit = 0; | |
2719 | goto open2; | |
2720 | } | |
2721 | mathexit = 1; | |
2722 | ||
2723 | case Sclose: | |
2724 | if (!++depth) goto done2; | |
2725 | break; | |
2726 | ||
2727 | case Sopen: | |
2728 | open2: | |
2729 | if (!--depth) goto done2; | |
2730 | if (depth < min_depth) | |
9471945b KS |
2731 | xsignal3 (Qscan_error, |
2732 | build_string ("Containing expression ends prematurely"), | |
2733 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2734 | break; |
2735 | ||
2736 | case Sendcomment: | |
2737 | if (!parse_sexp_ignore_comments) | |
2738 | break; | |
95ff8dfc | 2739 | found = back_comment (from, from_byte, stop, comnested, comstyle, |
6a140a74 | 2740 | &out_charpos, &out_bytepos); |
1aa963c8 SM |
2741 | /* FIXME: if found == -1, then it really wasn't a comment-end. |
2742 | For single-char Sendcomment, we can't do much about it apart | |
2743 | from skipping the char. | |
2744 | For 2-char endcomments, we could try again, taking both | |
2745 | chars as separate entities, but it's a lot of trouble | |
2746 | for very little gain, so we don't bother either. -sm */ | |
6a140a74 RS |
2747 | if (found != -1) |
2748 | from = out_charpos, from_byte = out_bytepos; | |
8489eb67 RS |
2749 | break; |
2750 | ||
195d1361 RS |
2751 | case Scomment_fence: |
2752 | case Sstring_fence: | |
2753 | while (1) | |
2754 | { | |
8b8bf8e6 SM |
2755 | if (from == stop) |
2756 | goto lose; | |
6b61353c | 2757 | DEC_BOTH (from, from_byte); |
195d1361 | 2758 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
7d0393cf | 2759 | if (!char_quoted (from, from_byte) |
b7dbcc19 | 2760 | && (c = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
bd25db08 | 2761 | SYNTAX_WITH_MULTIBYTE_CHECK (c) == code)) |
195d1361 RS |
2762 | break; |
2763 | } | |
2764 | if (code == Sstring_fence && !depth && sexpflag) goto done2; | |
2765 | break; | |
7d0393cf | 2766 | |
8489eb67 | 2767 | case Sstring: |
b7dbcc19 | 2768 | stringterm = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8489eb67 RS |
2769 | while (1) |
2770 | { | |
8b8bf8e6 SM |
2771 | if (from == stop) |
2772 | goto lose; | |
6b61353c KH |
2773 | DEC_BOTH (from, from_byte); |
2774 | UPDATE_SYNTAX_TABLE_BACKWARD (from); | |
2775 | if (!char_quoted (from, from_byte) | |
2776 | && (stringterm | |
2777 | == (c = FETCH_CHAR_AS_MULTIBYTE (from_byte))) | |
42ebfa31 | 2778 | && SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring) |
8489eb67 | 2779 | break; |
8489eb67 | 2780 | } |
8489eb67 RS |
2781 | if (!depth && sexpflag) goto done2; |
2782 | break; | |
240c43e8 SM |
2783 | default: |
2784 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
2785 | break; | |
8489eb67 RS |
2786 | } |
2787 | } | |
2788 | ||
2789 | /* Reached start of buffer. Error if within object, return nil if between */ | |
8b8bf8e6 SM |
2790 | if (depth) |
2791 | goto lose; | |
8489eb67 RS |
2792 | |
2793 | immediate_quit = 0; | |
2794 | return Qnil; | |
2795 | ||
2796 | done2: | |
2797 | count++; | |
2798 | } | |
2799 | ||
2800 | ||
2801 | immediate_quit = 0; | |
1e142fb7 | 2802 | XSETFASTINT (val, from); |
8489eb67 RS |
2803 | return val; |
2804 | ||
2805 | lose: | |
9471945b KS |
2806 | xsignal3 (Qscan_error, |
2807 | build_string ("Unbalanced parentheses"), | |
2808 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2809 | } |
2810 | ||
8489eb67 | 2811 | DEFUN ("scan-lists", Fscan_lists, Sscan_lists, 3, 3, 0, |
fdb82f93 PJ |
2812 | doc: /* Scan from character number FROM by COUNT lists. |
2813 | Returns the character number of the position thus found. | |
2814 | ||
2815 | If DEPTH is nonzero, paren depth begins counting from that value, | |
2816 | only places where the depth in parentheses becomes zero | |
2817 | are candidates for stopping; COUNT such places are counted. | |
2818 | Thus, a positive value for DEPTH means go out levels. | |
2819 | ||
2820 | Comments are ignored if `parse-sexp-ignore-comments' is non-nil. | |
2821 | ||
2822 | If the beginning or end of (the accessible part of) the buffer is reached | |
2823 | and the depth is wrong, an error is signaled. | |
2824 | If the depth is right but the count is not used up, nil is returned. */) | |
2825 | (from, count, depth) | |
8489eb67 RS |
2826 | Lisp_Object from, count, depth; |
2827 | { | |
b7826503 PJ |
2828 | CHECK_NUMBER (from); |
2829 | CHECK_NUMBER (count); | |
2830 | CHECK_NUMBER (depth); | |
8489eb67 RS |
2831 | |
2832 | return scan_lists (XINT (from), XINT (count), XINT (depth), 0); | |
2833 | } | |
2834 | ||
2835 | DEFUN ("scan-sexps", Fscan_sexps, Sscan_sexps, 2, 2, 0, | |
fdb82f93 PJ |
2836 | doc: /* Scan from character number FROM by COUNT balanced expressions. |
2837 | If COUNT is negative, scan backwards. | |
2838 | Returns the character number of the position thus found. | |
2839 | ||
2840 | Comments are ignored if `parse-sexp-ignore-comments' is non-nil. | |
2841 | ||
2842 | If the beginning or end of (the accessible part of) the buffer is reached | |
2843 | in the middle of a parenthetical grouping, an error is signaled. | |
2844 | If the beginning or end is reached between groupings | |
2845 | but before count is used up, nil is returned. */) | |
2846 | (from, count) | |
8489eb67 RS |
2847 | Lisp_Object from, count; |
2848 | { | |
b7826503 PJ |
2849 | CHECK_NUMBER (from); |
2850 | CHECK_NUMBER (count); | |
8489eb67 RS |
2851 | |
2852 | return scan_lists (XINT (from), XINT (count), 0, 1); | |
2853 | } | |
2854 | ||
2855 | DEFUN ("backward-prefix-chars", Fbackward_prefix_chars, Sbackward_prefix_chars, | |
fdb82f93 PJ |
2856 | 0, 0, 0, |
2857 | doc: /* Move point backward over any number of chars with prefix syntax. | |
2858 | This includes chars with "quote" or "prefix" syntax (' or p). */) | |
2859 | () | |
8489eb67 RS |
2860 | { |
2861 | int beg = BEGV; | |
6a140a74 RS |
2862 | int opoint = PT; |
2863 | int opoint_byte = PT_BYTE; | |
6ec8bbd2 | 2864 | int pos = PT; |
6a140a74 | 2865 | int pos_byte = PT_BYTE; |
93da5fff | 2866 | int c; |
93da5fff | 2867 | |
7d0393cf | 2868 | if (pos <= beg) |
195d1361 | 2869 | { |
f902a008 RS |
2870 | SET_PT_BOTH (opoint, opoint_byte); |
2871 | ||
2872 | return Qnil; | |
195d1361 | 2873 | } |
8489eb67 | 2874 | |
f902a008 RS |
2875 | SETUP_SYNTAX_TABLE (pos, -1); |
2876 | ||
6a140a74 RS |
2877 | DEC_BOTH (pos, pos_byte); |
2878 | ||
1fd3172d | 2879 | while (!char_quoted (pos, pos_byte) |
195d1361 | 2880 | /* Previous statement updates syntax table. */ |
b7dbcc19 | 2881 | && ((c = FETCH_CHAR_AS_MULTIBYTE (pos_byte), SYNTAX (c) == Squote) |
93da5fff KH |
2882 | || SYNTAX_PREFIX (c))) |
2883 | { | |
1fd3172d RS |
2884 | opoint = pos; |
2885 | opoint_byte = pos_byte; | |
2886 | ||
2887 | if (pos + 1 > beg) | |
2888 | DEC_BOTH (pos, pos_byte); | |
93da5fff | 2889 | } |
8489eb67 | 2890 | |
6a140a74 | 2891 | SET_PT_BOTH (opoint, opoint_byte); |
8489eb67 RS |
2892 | |
2893 | return Qnil; | |
2894 | } | |
2895 | \f | |
6a140a74 | 2896 | /* Parse forward from FROM / FROM_BYTE to END, |
e5d4f4dc RS |
2897 | assuming that FROM has state OLDSTATE (nil means FROM is start of function), |
2898 | and return a description of the state of the parse at END. | |
c81a3712 | 2899 | If STOPBEFORE is nonzero, stop at the start of an atom. |
644ea4df RS |
2900 | If COMMENTSTOP is 1, stop at the start of a comment. |
2901 | If COMMENTSTOP is -1, stop at the start or end of a comment, | |
2902 | after the beginning of a string, or after the end of a string. */ | |
8489eb67 | 2903 | |
340f92b5 | 2904 | static void |
6a140a74 | 2905 | scan_sexps_forward (stateptr, from, from_byte, end, targetdepth, |
c81a3712 | 2906 | stopbefore, oldstate, commentstop) |
e5d4f4dc | 2907 | struct lisp_parse_state *stateptr; |
8e2911c2 AS |
2908 | register EMACS_INT from; |
2909 | EMACS_INT from_byte, end; | |
2910 | int targetdepth, stopbefore; | |
8489eb67 | 2911 | Lisp_Object oldstate; |
c81a3712 | 2912 | int commentstop; |
8489eb67 RS |
2913 | { |
2914 | struct lisp_parse_state state; | |
2915 | ||
2916 | register enum syntaxcode code; | |
95ff8dfc RS |
2917 | int c1; |
2918 | int comnested; | |
8489eb67 RS |
2919 | struct level { int last, prev; }; |
2920 | struct level levelstart[100]; | |
2921 | register struct level *curlevel = levelstart; | |
2922 | struct level *endlevel = levelstart + 100; | |
8489eb67 RS |
2923 | register int depth; /* Paren depth of current scanning location. |
2924 | level - levelstart equals this except | |
2925 | when the depth becomes negative. */ | |
2926 | int mindepth; /* Lowest DEPTH value seen. */ | |
2927 | int start_quoted = 0; /* Nonzero means starting after a char quote */ | |
2928 | Lisp_Object tem; | |
8e2911c2 AS |
2929 | EMACS_INT prev_from; /* Keep one character before FROM. */ |
2930 | EMACS_INT prev_from_byte; | |
1fd3172d | 2931 | int prev_from_syntax; |
195d1361 RS |
2932 | int boundary_stop = commentstop == -1; |
2933 | int nofence; | |
95ff8dfc | 2934 | int found; |
60c86a83 | 2935 | EMACS_INT out_bytepos, out_charpos; |
7c7ca3d2 | 2936 | int temp; |
93da5fff KH |
2937 | |
2938 | prev_from = from; | |
6a140a74 RS |
2939 | prev_from_byte = from_byte; |
2940 | if (from != BEGV) | |
2941 | DEC_BOTH (prev_from, prev_from_byte); | |
93da5fff KH |
2942 | |
2943 | /* Use this macro instead of `from++'. */ | |
6a140a74 RS |
2944 | #define INC_FROM \ |
2945 | do { prev_from = from; \ | |
2946 | prev_from_byte = from_byte; \ | |
327719ee | 2947 | temp = FETCH_CHAR_AS_MULTIBYTE (prev_from_byte); \ |
ab229fdd | 2948 | prev_from_syntax = SYNTAX_WITH_FLAGS (temp); \ |
ef316cf0 | 2949 | INC_BOTH (from, from_byte); \ |
3ceb4629 SM |
2950 | if (from < end) \ |
2951 | UPDATE_SYNTAX_TABLE_FORWARD (from); \ | |
6a140a74 | 2952 | } while (0) |
8489eb67 RS |
2953 | |
2954 | immediate_quit = 1; | |
2955 | QUIT; | |
2956 | ||
265a9e55 | 2957 | if (NILP (oldstate)) |
8489eb67 RS |
2958 | { |
2959 | depth = 0; | |
2960 | state.instring = -1; | |
2961 | state.incomment = 0; | |
195d1361 RS |
2962 | state.comstyle = 0; /* comment style a by default. */ |
2963 | state.comstr_start = -1; /* no comment/string seen. */ | |
8489eb67 RS |
2964 | } |
2965 | else | |
2966 | { | |
2967 | tem = Fcar (oldstate); | |
265a9e55 | 2968 | if (!NILP (tem)) |
8489eb67 RS |
2969 | depth = XINT (tem); |
2970 | else | |
2971 | depth = 0; | |
2972 | ||
2973 | oldstate = Fcdr (oldstate); | |
2974 | oldstate = Fcdr (oldstate); | |
2975 | oldstate = Fcdr (oldstate); | |
2976 | tem = Fcar (oldstate); | |
195d1361 | 2977 | /* Check whether we are inside string_fence-style string: */ |
7d0393cf JB |
2978 | state.instring = (!NILP (tem) |
2979 | ? (INTEGERP (tem) ? XINT (tem) : ST_STRING_STYLE) | |
e76f1c44 | 2980 | : -1); |
8489eb67 RS |
2981 | |
2982 | oldstate = Fcdr (oldstate); | |
2983 | tem = Fcar (oldstate); | |
e76f1c44 GM |
2984 | state.incomment = (!NILP (tem) |
2985 | ? (INTEGERP (tem) ? XINT (tem) : -1) | |
95ff8dfc | 2986 | : 0); |
8489eb67 RS |
2987 | |
2988 | oldstate = Fcdr (oldstate); | |
2989 | tem = Fcar (oldstate); | |
265a9e55 | 2990 | start_quoted = !NILP (tem); |
e5d4f4dc | 2991 | |
95ff8dfc | 2992 | /* if the eighth element of the list is nil, we are in comment |
195d1361 RS |
2993 | style a. If it is non-nil, we are in comment style b */ |
2994 | oldstate = Fcdr (oldstate); | |
e5d4f4dc | 2995 | oldstate = Fcdr (oldstate); |
195d1361 | 2996 | tem = Fcar (oldstate); |
7d0393cf | 2997 | state.comstyle = NILP (tem) ? 0 : (EQ (tem, Qsyntax_table) |
e76f1c44 | 2998 | ? ST_COMMENT_STYLE : 1); |
195d1361 | 2999 | |
e5d4f4dc | 3000 | oldstate = Fcdr (oldstate); |
e5d4f4dc | 3001 | tem = Fcar (oldstate); |
195d1361 | 3002 | state.comstr_start = NILP (tem) ? -1 : XINT (tem) ; |
1a102a58 RS |
3003 | oldstate = Fcdr (oldstate); |
3004 | tem = Fcar (oldstate); | |
3005 | while (!NILP (tem)) /* >= second enclosing sexps. */ | |
3006 | { | |
3007 | /* curlevel++->last ran into compiler bug on Apollo */ | |
3008 | curlevel->last = XINT (Fcar (tem)); | |
3009 | if (++curlevel == endlevel) | |
02010917 | 3010 | curlevel--; /* error ("Nesting too deep for parser"); */ |
1a102a58 RS |
3011 | curlevel->prev = -1; |
3012 | curlevel->last = -1; | |
3013 | tem = Fcdr (tem); | |
3014 | } | |
8489eb67 RS |
3015 | } |
3016 | state.quoted = 0; | |
3017 | mindepth = depth; | |
3018 | ||
3019 | curlevel->prev = -1; | |
3020 | curlevel->last = -1; | |
3021 | ||
326b283b | 3022 | SETUP_SYNTAX_TABLE (prev_from, 1); |
ab229fdd AS |
3023 | temp = FETCH_CHAR (prev_from_byte); |
3024 | prev_from_syntax = SYNTAX_WITH_FLAGS (temp); | |
3ceb4629 | 3025 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
326b283b | 3026 | |
195d1361 | 3027 | /* Enter the loop at a place appropriate for initial state. */ |
8489eb67 | 3028 | |
326b283b RS |
3029 | if (state.incomment) |
3030 | goto startincomment; | |
8489eb67 RS |
3031 | if (state.instring >= 0) |
3032 | { | |
195d1361 | 3033 | nofence = state.instring != ST_STRING_STYLE; |
326b283b RS |
3034 | if (start_quoted) |
3035 | goto startquotedinstring; | |
8489eb67 RS |
3036 | goto startinstring; |
3037 | } | |
326b283b RS |
3038 | else if (start_quoted) |
3039 | goto startquoted; | |
1fd3172d | 3040 | |
8489eb67 RS |
3041 | while (from < end) |
3042 | { | |
93da5fff | 3043 | INC_FROM; |
1fd3172d | 3044 | code = prev_from_syntax & 0xff; |
4c920633 | 3045 | |
c5d4b96f SM |
3046 | if (from < end |
3047 | && SYNTAX_FLAGS_COMSTART_FIRST (prev_from_syntax) | |
3048 | && (c1 = FETCH_CHAR (from_byte), | |
3049 | SYNTAX_COMSTART_SECOND (c1))) | |
3050 | /* Duplicate code to avoid a complex if-expression | |
3051 | which causes trouble for the SGI compiler. */ | |
95ff8dfc | 3052 | { |
c5d4b96f SM |
3053 | /* Record the comment style we have entered so that only |
3054 | the comment-end sequence of the same style actually | |
3055 | terminates the comment section. */ | |
3056 | state.comstyle = SYNTAX_COMMENT_STYLE (c1); | |
3057 | comnested = SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax); | |
3058 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
3059 | state.incomment = comnested ? 1 : -1; | |
95ff8dfc | 3060 | state.comstr_start = prev_from; |
c5d4b96f SM |
3061 | INC_FROM; |
3062 | code = Scomment; | |
95ff8dfc | 3063 | } |
4c920633 | 3064 | else if (code == Scomment_fence) |
e5d4f4dc RS |
3065 | { |
3066 | /* Record the comment style we have entered so that only | |
3067 | the comment-end sequence of the same style actually | |
3068 | terminates the comment section. */ | |
95ff8dfc RS |
3069 | state.comstyle = ST_COMMENT_STYLE; |
3070 | state.incomment = -1; | |
195d1361 | 3071 | state.comstr_start = prev_from; |
e5d4f4dc | 3072 | code = Scomment; |
e5d4f4dc | 3073 | } |
c5d4b96f SM |
3074 | else if (code == Scomment) |
3075 | { | |
3076 | state.comstyle = SYNTAX_FLAGS_COMMENT_STYLE (prev_from_syntax); | |
3077 | state.incomment = (SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax) ? | |
3078 | 1 : -1); | |
3079 | state.comstr_start = prev_from; | |
3080 | } | |
e5d4f4dc | 3081 | |
1fd3172d | 3082 | if (SYNTAX_FLAGS_PREFIX (prev_from_syntax)) |
8489eb67 | 3083 | continue; |
0220c518 | 3084 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
3085 | { |
3086 | case Sescape: | |
3087 | case Scharquote: | |
3088 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
93da5fff | 3089 | curlevel->last = prev_from; |
8489eb67 RS |
3090 | startquoted: |
3091 | if (from == end) goto endquoted; | |
93da5fff | 3092 | INC_FROM; |
8489eb67 RS |
3093 | goto symstarted; |
3094 | /* treat following character as a word constituent */ | |
3095 | case Sword: | |
3096 | case Ssymbol: | |
3097 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
93da5fff | 3098 | curlevel->last = prev_from; |
8489eb67 RS |
3099 | symstarted: |
3100 | while (from < end) | |
3101 | { | |
7c7ca3d2 | 3102 | /* Some compilers can't handle this inside the switch. */ |
327719ee | 3103 | temp = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
ab229fdd | 3104 | temp = SYNTAX (temp); |
7c7ca3d2 | 3105 | switch (temp) |
8489eb67 RS |
3106 | { |
3107 | case Scharquote: | |
3108 | case Sescape: | |
93da5fff | 3109 | INC_FROM; |
8489eb67 RS |
3110 | if (from == end) goto endquoted; |
3111 | break; | |
3112 | case Sword: | |
3113 | case Ssymbol: | |
3114 | case Squote: | |
3115 | break; | |
3116 | default: | |
3117 | goto symdone; | |
3118 | } | |
93da5fff | 3119 | INC_FROM; |
8489eb67 RS |
3120 | } |
3121 | symdone: | |
3122 | curlevel->prev = curlevel->last; | |
3123 | break; | |
3124 | ||
240c43e8 | 3125 | case Scomment_fence: /* Can't happen because it's handled above. */ |
8489eb67 | 3126 | case Scomment: |
195d1361 | 3127 | if (commentstop || boundary_stop) goto done; |
d355bd8a SM |
3128 | startincomment: |
3129 | /* The (from == BEGV) test was to enter the loop in the middle so | |
95ff8dfc | 3130 | that we find a 2-char comment ender even if we start in the |
e6365855 | 3131 | middle of it. We don't want to do that if we're just at the |
d355bd8a | 3132 | beginning of the comment (think of (*) ... (*)). */ |
95ff8dfc RS |
3133 | found = forw_comment (from, from_byte, end, |
3134 | state.incomment, state.comstyle, | |
e6365855 SM |
3135 | (from == BEGV || from < state.comstr_start + 3) |
3136 | ? 0 : prev_from_syntax, | |
95ff8dfc RS |
3137 | &out_charpos, &out_bytepos, &state.incomment); |
3138 | from = out_charpos; from_byte = out_bytepos; | |
3139 | /* Beware! prev_from and friends are invalid now. | |
3140 | Luckily, the `done' doesn't use them and the INC_FROM | |
3141 | sets them to a sane value without looking at them. */ | |
3142 | if (!found) goto done; | |
d355bd8a | 3143 | INC_FROM; |
8489eb67 | 3144 | state.incomment = 0; |
e5d4f4dc | 3145 | state.comstyle = 0; /* reset the comment style */ |
195d1361 | 3146 | if (boundary_stop) goto done; |
8489eb67 RS |
3147 | break; |
3148 | ||
3149 | case Sopen: | |
3150 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
3151 | depth++; | |
3152 | /* curlevel++->last ran into compiler bug on Apollo */ | |
93da5fff | 3153 | curlevel->last = prev_from; |
8489eb67 | 3154 | if (++curlevel == endlevel) |
02010917 | 3155 | curlevel--; /* error ("Nesting too deep for parser"); */ |
8489eb67 RS |
3156 | curlevel->prev = -1; |
3157 | curlevel->last = -1; | |
30844415 | 3158 | if (targetdepth == depth) goto done; |
8489eb67 RS |
3159 | break; |
3160 | ||
3161 | case Sclose: | |
3162 | depth--; | |
3163 | if (depth < mindepth) | |
3164 | mindepth = depth; | |
3165 | if (curlevel != levelstart) | |
3166 | curlevel--; | |
3167 | curlevel->prev = curlevel->last; | |
30844415 | 3168 | if (targetdepth == depth) goto done; |
8489eb67 RS |
3169 | break; |
3170 | ||
3171 | case Sstring: | |
195d1361 RS |
3172 | case Sstring_fence: |
3173 | state.comstr_start = from - 1; | |
8489eb67 | 3174 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ |
93da5fff | 3175 | curlevel->last = prev_from; |
7d0393cf | 3176 | state.instring = (code == Sstring |
b7dbcc19 | 3177 | ? (FETCH_CHAR_AS_MULTIBYTE (prev_from_byte)) |
195d1361 RS |
3178 | : ST_STRING_STYLE); |
3179 | if (boundary_stop) goto done; | |
8489eb67 | 3180 | startinstring: |
195d1361 | 3181 | { |
644ea4df | 3182 | nofence = state.instring != ST_STRING_STYLE; |
7d0393cf | 3183 | |
644ea4df RS |
3184 | while (1) |
3185 | { | |
3186 | int c; | |
195d1361 | 3187 | |
644ea4df | 3188 | if (from >= end) goto done; |
b7dbcc19 | 3189 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
7c7ca3d2 RS |
3190 | /* Some compilers can't handle this inside the switch. */ |
3191 | temp = SYNTAX (c); | |
ccf89641 KH |
3192 | |
3193 | /* Check TEMP here so that if the char has | |
3194 | a syntax-table property which says it is NOT | |
3195 | a string character, it does not end the string. */ | |
3196 | if (nofence && c == state.instring && temp == Sstring) | |
3197 | break; | |
3198 | ||
7c7ca3d2 | 3199 | switch (temp) |
644ea4df RS |
3200 | { |
3201 | case Sstring_fence: | |
3202 | if (!nofence) goto string_end; | |
3203 | break; | |
3204 | case Scharquote: | |
3205 | case Sescape: | |
3206 | INC_FROM; | |
3207 | startquotedinstring: | |
3208 | if (from >= end) goto endquoted; | |
195d1361 | 3209 | } |
644ea4df RS |
3210 | INC_FROM; |
3211 | } | |
195d1361 RS |
3212 | } |
3213 | string_end: | |
8489eb67 RS |
3214 | state.instring = -1; |
3215 | curlevel->prev = curlevel->last; | |
93da5fff | 3216 | INC_FROM; |
195d1361 | 3217 | if (boundary_stop) goto done; |
8489eb67 RS |
3218 | break; |
3219 | ||
3220 | case Smath: | |
240c43e8 SM |
3221 | /* FIXME: We should do something with it. */ |
3222 | break; | |
3223 | default: | |
3224 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
8489eb67 RS |
3225 | break; |
3226 | } | |
3227 | } | |
3228 | goto done; | |
3229 | ||
3230 | stop: /* Here if stopping before start of sexp. */ | |
93da5fff | 3231 | from = prev_from; /* We have just fetched the char that starts it; */ |
8489eb67 RS |
3232 | goto done; /* but return the position before it. */ |
3233 | ||
3234 | endquoted: | |
3235 | state.quoted = 1; | |
3236 | done: | |
3237 | state.depth = depth; | |
3238 | state.mindepth = mindepth; | |
3239 | state.thislevelstart = curlevel->prev; | |
3240 | state.prevlevelstart | |
3241 | = (curlevel == levelstart) ? -1 : (curlevel - 1)->last; | |
3242 | state.location = from; | |
1a102a58 RS |
3243 | state.levelstarts = Qnil; |
3244 | while (--curlevel >= levelstart) | |
3245 | state.levelstarts = Fcons (make_number (curlevel->last), | |
3246 | state.levelstarts); | |
8489eb67 RS |
3247 | immediate_quit = 0; |
3248 | ||
e5d4f4dc | 3249 | *stateptr = state; |
8489eb67 RS |
3250 | } |
3251 | ||
c81a3712 | 3252 | DEFUN ("parse-partial-sexp", Fparse_partial_sexp, Sparse_partial_sexp, 2, 6, 0, |
fdb82f93 PJ |
3253 | doc: /* Parse Lisp syntax starting at FROM until TO; return status of parse at TO. |
3254 | Parsing stops at TO or when certain criteria are met; | |
3255 | point is set to where parsing stops. | |
3256 | If fifth arg OLDSTATE is omitted or nil, | |
3257 | parsing assumes that FROM is the beginning of a function. | |
0e6228bc | 3258 | Value is a list of elements describing final state of parsing: |
fdb82f93 PJ |
3259 | 0. depth in parens. |
3260 | 1. character address of start of innermost containing list; nil if none. | |
3261 | 2. character address of start of last complete sexp terminated. | |
3262 | 3. non-nil if inside a string. | |
3263 | (it is the character that will terminate the string, | |
3264 | or t if the string should be terminated by a generic string delimiter.) | |
7d0393cf | 3265 | 4. nil if outside a comment, t if inside a non-nestable comment, |
fdb82f93 PJ |
3266 | else an integer (the current comment nesting). |
3267 | 5. t if following a quote character. | |
3268 | 6. the minimum paren-depth encountered during this scan. | |
3269 | 7. t if in a comment of style b; symbol `syntax-table' if the comment | |
3270 | should be terminated by a generic comment delimiter. | |
3271 | 8. character address of start of comment or string; nil if not in one. | |
3272 | 9. Intermediate data for continuation of parsing (subject to change). | |
3273 | If third arg TARGETDEPTH is non-nil, parsing stops if the depth | |
3274 | in parentheses becomes equal to TARGETDEPTH. | |
3275 | Fourth arg STOPBEFORE non-nil means stop when come to | |
3276 | any character that starts a sexp. | |
0e6228bc | 3277 | Fifth arg OLDSTATE is a list like what this function returns. |
fdb82f93 | 3278 | It is used to initialize the state of the parse. Elements number 1, 2, 6 |
43726c05 | 3279 | and 8 are ignored. |
fdb82f93 PJ |
3280 | Sixth arg COMMENTSTOP non-nil means stop at the start of a comment. |
3281 | If it is symbol `syntax-table', stop after the start of a comment or a | |
3282 | string, or after end of a comment or a string. */) | |
3283 | (from, to, targetdepth, stopbefore, oldstate, commentstop) | |
c81a3712 | 3284 | Lisp_Object from, to, targetdepth, stopbefore, oldstate, commentstop; |
8489eb67 RS |
3285 | { |
3286 | struct lisp_parse_state state; | |
3287 | int target; | |
3288 | ||
265a9e55 | 3289 | if (!NILP (targetdepth)) |
8489eb67 | 3290 | { |
b7826503 | 3291 | CHECK_NUMBER (targetdepth); |
8489eb67 RS |
3292 | target = XINT (targetdepth); |
3293 | } | |
3294 | else | |
3295 | target = -100000; /* We won't reach this depth */ | |
3296 | ||
3297 | validate_region (&from, &to); | |
6a140a74 RS |
3298 | scan_sexps_forward (&state, XINT (from), CHAR_TO_BYTE (XINT (from)), |
3299 | XINT (to), | |
c81a3712 | 3300 | target, !NILP (stopbefore), oldstate, |
7d0393cf | 3301 | (NILP (commentstop) |
195d1361 | 3302 | ? 0 : (EQ (commentstop, Qsyntax_table) ? -1 : 1))); |
8489eb67 RS |
3303 | |
3304 | SET_PT (state.location); | |
7d0393cf | 3305 | |
8489eb67 RS |
3306 | return Fcons (make_number (state.depth), |
3307 | Fcons (state.prevlevelstart < 0 ? Qnil : make_number (state.prevlevelstart), | |
3308 | Fcons (state.thislevelstart < 0 ? Qnil : make_number (state.thislevelstart), | |
7d0393cf JB |
3309 | Fcons (state.instring >= 0 |
3310 | ? (state.instring == ST_STRING_STYLE | |
195d1361 | 3311 | ? Qt : make_number (state.instring)) : Qnil, |
95ff8dfc RS |
3312 | Fcons (state.incomment < 0 ? Qt : |
3313 | (state.incomment == 0 ? Qnil : | |
3314 | make_number (state.incomment)), | |
8489eb67 | 3315 | Fcons (state.quoted ? Qt : Qnil, |
bff37d2a | 3316 | Fcons (make_number (state.mindepth), |
7d0393cf | 3317 | Fcons ((state.comstyle |
bff37d2a RS |
3318 | ? (state.comstyle == ST_COMMENT_STYLE |
3319 | ? Qsyntax_table : Qt) : | |
3320 | Qnil), | |
0beaf54f DL |
3321 | Fcons (((state.incomment |
3322 | || (state.instring >= 0)) | |
bff37d2a RS |
3323 | ? make_number (state.comstr_start) |
3324 | : Qnil), | |
1a102a58 | 3325 | Fcons (state.levelstarts, Qnil)))))))))); |
8489eb67 RS |
3326 | } |
3327 | \f | |
dfcf069d | 3328 | void |
8489eb67 RS |
3329 | init_syntax_once () |
3330 | { | |
78f9a1f7 | 3331 | register int i, c; |
8ea151b2 | 3332 | Lisp_Object temp; |
8489eb67 | 3333 | |
5ebaddf5 RS |
3334 | /* This has to be done here, before we call Fmake_char_table. */ |
3335 | Qsyntax_table = intern ("syntax-table"); | |
3336 | staticpro (&Qsyntax_table); | |
3337 | ||
3338 | /* Intern this now in case it isn't already done. | |
3339 | Setting this variable twice is harmless. | |
3340 | But don't staticpro it here--that is done in alloc.c. */ | |
3341 | Qchar_table_extra_slots = intern ("char-table-extra-slots"); | |
3342 | ||
93da5fff | 3343 | /* Create objects which can be shared among syntax tables. */ |
42ebfa31 | 3344 | Vsyntax_code_object = Fmake_vector (make_number (Smax), Qnil); |
93da5fff KH |
3345 | for (i = 0; i < XVECTOR (Vsyntax_code_object)->size; i++) |
3346 | XVECTOR (Vsyntax_code_object)->contents[i] | |
3347 | = Fcons (make_number (i), Qnil); | |
3348 | ||
5ebaddf5 RS |
3349 | /* Now we are ready to set up this property, so we can |
3350 | create syntax tables. */ | |
3351 | Fput (Qsyntax_table, Qchar_table_extra_slots, make_number (0)); | |
3352 | ||
93da5fff | 3353 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Swhitespace]; |
8489eb67 | 3354 | |
5ebaddf5 | 3355 | Vstandard_syntax_table = Fmake_char_table (Qsyntax_table, temp); |
8489eb67 | 3356 | |
aa7b08b4 RS |
3357 | /* Control characters should not be whitespace. */ |
3358 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Spunct]; | |
3359 | for (i = 0; i <= ' ' - 1; i++) | |
3360 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); | |
3361 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 0177, temp); | |
3362 | ||
3363 | /* Except that a few really are whitespace. */ | |
3364 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Swhitespace]; | |
3365 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ' ', temp); | |
3366 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\t', temp); | |
3367 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\n', temp); | |
3368 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 015, temp); | |
3369 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 014, temp); | |
3370 | ||
93da5fff | 3371 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Sword]; |
8489eb67 | 3372 | for (i = 'a'; i <= 'z'; i++) |
8ea151b2 | 3373 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
8489eb67 | 3374 | for (i = 'A'; i <= 'Z'; i++) |
8ea151b2 | 3375 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
8489eb67 | 3376 | for (i = '0'; i <= '9'; i++) |
8ea151b2 RS |
3377 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
3378 | ||
3379 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '$', temp); | |
3380 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '%', temp); | |
3381 | ||
3382 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '(', | |
3383 | Fcons (make_number (Sopen), make_number (')'))); | |
3384 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ')', | |
3385 | Fcons (make_number (Sclose), make_number ('('))); | |
3386 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '[', | |
3387 | Fcons (make_number (Sopen), make_number (']'))); | |
3388 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ']', | |
3389 | Fcons (make_number (Sclose), make_number ('['))); | |
3390 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '{', | |
3391 | Fcons (make_number (Sopen), make_number ('}'))); | |
3392 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '}', | |
3393 | Fcons (make_number (Sclose), make_number ('{'))); | |
3394 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '"', | |
3395 | Fcons (make_number ((int) Sstring), Qnil)); | |
3396 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\\', | |
3397 | Fcons (make_number ((int) Sescape), Qnil)); | |
3398 | ||
93da5fff | 3399 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Ssymbol]; |
8489eb67 | 3400 | for (i = 0; i < 10; i++) |
78f9a1f7 KH |
3401 | { |
3402 | c = "_-+*/&|<>="[i]; | |
3403 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, c, temp); | |
3404 | } | |
8489eb67 | 3405 | |
93da5fff | 3406 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Spunct]; |
8489eb67 | 3407 | for (i = 0; i < 12; i++) |
78f9a1f7 KH |
3408 | { |
3409 | c = ".,;:?!#@~^'`"[i]; | |
3410 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, c, temp); | |
3411 | } | |
bd25db08 KH |
3412 | |
3413 | /* All multibyte characters have syntax `word' by default. */ | |
3414 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Sword]; | |
5c7b02ab | 3415 | char_table_set_range (Vstandard_syntax_table, 0x80, MAX_CHAR, temp); |
8489eb67 RS |
3416 | } |
3417 | ||
dfcf069d | 3418 | void |
8489eb67 RS |
3419 | syms_of_syntax () |
3420 | { | |
3421 | Qsyntax_table_p = intern ("syntax-table-p"); | |
3422 | staticpro (&Qsyntax_table_p); | |
3423 | ||
93da5fff KH |
3424 | staticpro (&Vsyntax_code_object); |
3425 | ||
a2994c46 KS |
3426 | staticpro (&gl_state.object); |
3427 | staticpro (&gl_state.global_code); | |
3428 | staticpro (&gl_state.current_syntax_table); | |
3429 | staticpro (&gl_state.old_prop); | |
3430 | ||
3431 | /* Defined in regex.c */ | |
3432 | staticpro (&re_match_object); | |
3433 | ||
7bf5e9e4 RS |
3434 | Qscan_error = intern ("scan-error"); |
3435 | staticpro (&Qscan_error); | |
3436 | Fput (Qscan_error, Qerror_conditions, | |
f3be100f | 3437 | Fcons (Qscan_error, Fcons (Qerror, Qnil))); |
7bf5e9e4 RS |
3438 | Fput (Qscan_error, Qerror_message, |
3439 | build_string ("Scan error")); | |
3440 | ||
8489eb67 | 3441 | DEFVAR_BOOL ("parse-sexp-ignore-comments", &parse_sexp_ignore_comments, |
fdb82f93 | 3442 | doc: /* Non-nil means `forward-sexp', etc., should treat comments as whitespace. */); |
8489eb67 | 3443 | |
195d1361 | 3444 | DEFVAR_BOOL ("parse-sexp-lookup-properties", &parse_sexp_lookup_properties, |
fdb82f93 PJ |
3445 | doc: /* Non-nil means `forward-sexp', etc., obey `syntax-table' property. |
3446 | Otherwise, that text property is simply ignored. | |
3447 | See the info node `(elisp)Syntax Properties' for a description of the | |
3448 | `syntax-table' property. */); | |
195d1361 | 3449 | |
8489eb67 RS |
3450 | words_include_escapes = 0; |
3451 | DEFVAR_BOOL ("words-include-escapes", &words_include_escapes, | |
fdb82f93 | 3452 | doc: /* Non-nil means `forward-word', etc., should treat escape chars part of words. */); |
8489eb67 | 3453 | |
bd25db08 | 3454 | DEFVAR_BOOL ("multibyte-syntax-as-symbol", &multibyte_syntax_as_symbol, |
fdb82f93 | 3455 | doc: /* Non-nil means `scan-sexps' treats all multibyte characters as symbol. */); |
bd25db08 KH |
3456 | multibyte_syntax_as_symbol = 0; |
3457 | ||
f4ed767f GM |
3458 | DEFVAR_BOOL ("open-paren-in-column-0-is-defun-start", |
3459 | &open_paren_in_column_0_is_defun_start, | |
b222e415 | 3460 | doc: /* *Non-nil means an open paren in column 0 denotes the start of a defun. */); |
f4ed767f GM |
3461 | open_paren_in_column_0_is_defun_start = 1; |
3462 | ||
8f924df7 KH |
3463 | |
3464 | DEFVAR_LISP ("find-word-boundary-function-table", | |
3465 | &Vfind_word_boundary_function_table, | |
869bb237 | 3466 | doc: /* |
8f924df7 | 3467 | Char table of functions to search for the word boundary. |
869bb237 KH |
3468 | Each function is called with two arguments; POS and LIMIT. |
3469 | POS and LIMIT are character positions in the current buffer. | |
3470 | ||
3471 | If POS is less than LIMIT, POS is at the first character of a word, | |
3472 | and the return value of a function is a position after the last | |
3473 | character of that word. | |
3474 | ||
3475 | If POS is not less than LIMIT, POS is at the last character of a word, | |
3476 | and the return value of a function is a position at the first | |
3477 | character of that word. | |
3478 | ||
3479 | In both cases, LIMIT bounds the search. */); | |
8f924df7 | 3480 | Vfind_word_boundary_function_table = Fmake_char_table (Qnil, Qnil); |
869bb237 | 3481 | |
8489eb67 RS |
3482 | defsubr (&Ssyntax_table_p); |
3483 | defsubr (&Ssyntax_table); | |
3484 | defsubr (&Sstandard_syntax_table); | |
3485 | defsubr (&Scopy_syntax_table); | |
3486 | defsubr (&Sset_syntax_table); | |
3487 | defsubr (&Schar_syntax); | |
beefa22e | 3488 | defsubr (&Smatching_paren); |
c65adb44 | 3489 | defsubr (&Sstring_to_syntax); |
8489eb67 | 3490 | defsubr (&Smodify_syntax_entry); |
62abe9cb | 3491 | defsubr (&Sinternal_describe_syntax_value); |
8489eb67 RS |
3492 | |
3493 | defsubr (&Sforward_word); | |
3494 | ||
195d1361 RS |
3495 | defsubr (&Sskip_chars_forward); |
3496 | defsubr (&Sskip_chars_backward); | |
3497 | defsubr (&Sskip_syntax_forward); | |
3498 | defsubr (&Sskip_syntax_backward); | |
3499 | ||
b3cfe0c8 | 3500 | defsubr (&Sforward_comment); |
8489eb67 RS |
3501 | defsubr (&Sscan_lists); |
3502 | defsubr (&Sscan_sexps); | |
3503 | defsubr (&Sbackward_prefix_chars); | |
3504 | defsubr (&Sparse_partial_sexp); | |
3505 | } | |
839966f3 KH |
3506 | |
3507 | /* arch-tag: 3e297b9f-088e-4b64-8f4c-fb0b3443e412 | |
3508 | (do not change this comment) */ |