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, |
114f9c96 | 3 | 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010 |
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; |
62a6e103 | 914 | int character = STRING_CHAR_AND_LENGTH (p, len); |
93da5fff | 915 | XSETINT (match, character); |
d1be9f0f RS |
916 | if (XFASTINT (match) == ' ') |
917 | match = Qnil; | |
93da5fff | 918 | p += len; |
d1be9f0f RS |
919 | } |
920 | else | |
8ea151b2 | 921 | match = Qnil; |
8489eb67 | 922 | |
8ea151b2 | 923 | val = (int) code; |
8489eb67 RS |
924 | while (*p) |
925 | switch (*p++) | |
926 | { | |
927 | case '1': | |
247e20a8 | 928 | val |= 1 << 16; |
8489eb67 RS |
929 | break; |
930 | ||
931 | case '2': | |
247e20a8 | 932 | val |= 1 << 17; |
8489eb67 RS |
933 | break; |
934 | ||
935 | case '3': | |
247e20a8 | 936 | val |= 1 << 18; |
8489eb67 RS |
937 | break; |
938 | ||
939 | case '4': | |
247e20a8 | 940 | val |= 1 << 19; |
8489eb67 RS |
941 | break; |
942 | ||
943 | case 'p': | |
247e20a8 | 944 | val |= 1 << 20; |
8489eb67 | 945 | break; |
e5d4f4dc RS |
946 | |
947 | case 'b': | |
247e20a8 | 948 | val |= 1 << 21; |
e5d4f4dc | 949 | break; |
95ff8dfc RS |
950 | |
951 | case 'n': | |
952 | val |= 1 << 22; | |
953 | break; | |
8489eb67 | 954 | } |
7d0393cf | 955 | |
93da5fff | 956 | if (val < XVECTOR (Vsyntax_code_object)->size && NILP (match)) |
c65adb44 | 957 | return XVECTOR (Vsyntax_code_object)->contents[val]; |
93da5fff KH |
958 | else |
959 | /* Since we can't use a shared object, let's make a new one. */ | |
c65adb44 SM |
960 | return Fcons (make_number (val), match); |
961 | } | |
962 | ||
fdb82f93 | 963 | /* I really don't know why this is interactive |
2c9e1900 | 964 | help-form should at least be made useful whilst reading the second arg. */ |
7d0393cf | 965 | DEFUN ("modify-syntax-entry", Fmodify_syntax_entry, Smodify_syntax_entry, 2, 3, |
c65adb44 | 966 | "cSet syntax for character: \nsSet syntax for %s to: ", |
a50a10a0 | 967 | doc: /* Set syntax for character CHAR according to string NEWENTRY. |
d7ee9fab | 968 | The syntax is changed only for table SYNTAX-TABLE, which defaults to |
fdb82f93 | 969 | the current buffer's syntax table. |
5c7b02ab | 970 | CHAR may be a cons (MIN . MAX), in which case, syntaxes of all characters |
fedc6ab5 | 971 | in the range MIN to MAX are changed. |
fdb82f93 PJ |
972 | The first character of NEWENTRY should be one of the following: |
973 | Space or - whitespace syntax. w word constituent. | |
974 | _ symbol constituent. . punctuation. | |
975 | ( open-parenthesis. ) close-parenthesis. | |
976 | " string quote. \\ escape. | |
977 | $ paired delimiter. ' expression quote or prefix operator. | |
978 | < comment starter. > comment ender. | |
979 | / character-quote. @ inherit from `standard-syntax-table'. | |
980 | | generic string fence. ! generic comment fence. | |
981 | ||
982 | Only single-character comment start and end sequences are represented thus. | |
983 | Two-character sequences are represented as described below. | |
984 | The second character of NEWENTRY is the matching parenthesis, | |
985 | used only if the first character is `(' or `)'. | |
986 | Any additional characters are flags. | |
987 | Defined flags are the characters 1, 2, 3, 4, b, p, and n. | |
a50a10a0 PJ |
988 | 1 means CHAR is the start of a two-char comment start sequence. |
989 | 2 means CHAR is the second character of such a sequence. | |
990 | 3 means CHAR is the start of a two-char comment end sequence. | |
991 | 4 means CHAR is the second character of such a sequence. | |
fdb82f93 PJ |
992 | |
993 | There can be up to two orthogonal comment sequences. This is to support | |
994 | language modes such as C++. By default, all comment sequences are of style | |
995 | a, but you can set the comment sequence style to b (on the second character | |
996 | of a comment-start, or the first character of a comment-end sequence) using | |
997 | this flag: | |
a50a10a0 PJ |
998 | b means CHAR is part of comment sequence b. |
999 | n means CHAR is part of a nestable comment sequence. | |
fdb82f93 | 1000 | |
a50a10a0 | 1001 | p means CHAR is a prefix character for `backward-prefix-chars'; |
fdb82f93 | 1002 | such characters are treated as whitespace when they occur |
a50a10a0 | 1003 | between expressions. |
fedc6ab5 | 1004 | usage: (modify-syntax-entry CHAR NEWENTRY &optional SYNTAX-TABLE) */) |
fdb82f93 | 1005 | (c, newentry, syntax_table) |
c65adb44 SM |
1006 | Lisp_Object c, newentry, syntax_table; |
1007 | { | |
5c7b02ab KH |
1008 | if (CONSP (c)) |
1009 | { | |
8f924df7 KH |
1010 | CHECK_CHARACTER_CAR (c); |
1011 | CHECK_CHARACTER_CDR (c); | |
5c7b02ab KH |
1012 | } |
1013 | else | |
1014 | CHECK_CHARACTER (c); | |
c65adb44 SM |
1015 | |
1016 | if (NILP (syntax_table)) | |
1017 | syntax_table = current_buffer->syntax_table; | |
1018 | else | |
1019 | check_syntax_table (syntax_table); | |
8489eb67 | 1020 | |
5c7b02ab KH |
1021 | newentry = Fstring_to_syntax (newentry); |
1022 | if (CONSP (c)) | |
1023 | SET_RAW_SYNTAX_ENTRY_RANGE (syntax_table, c, newentry); | |
1024 | else | |
1025 | SET_RAW_SYNTAX_ENTRY (syntax_table, XINT (c), newentry); | |
e5b94d44 CY |
1026 | |
1027 | /* We clear the regexp cache, since character classes can now have | |
1028 | different values from those in the compiled regexps.*/ | |
1029 | clear_regexp_cache (); | |
1030 | ||
8489eb67 RS |
1031 | return Qnil; |
1032 | } | |
1033 | \f | |
1034 | /* Dump syntax table to buffer in human-readable format */ | |
1035 | ||
62abe9cb SM |
1036 | DEFUN ("internal-describe-syntax-value", Finternal_describe_syntax_value, |
1037 | Sinternal_describe_syntax_value, 1, 1, 0, | |
1038 | doc: /* Insert a description of the internal syntax description SYNTAX at point. */) | |
1039 | (syntax) | |
1040 | Lisp_Object syntax; | |
8489eb67 RS |
1041 | { |
1042 | register enum syntaxcode code; | |
e6365855 | 1043 | char desc, start1, start2, end1, end2, prefix, comstyle, comnested; |
8489eb67 | 1044 | char str[2]; |
62abe9cb | 1045 | Lisp_Object first, match_lisp, value = syntax; |
8489eb67 | 1046 | |
8ea151b2 RS |
1047 | if (NILP (value)) |
1048 | { | |
62abe9cb SM |
1049 | insert_string ("default"); |
1050 | return syntax; | |
8ea151b2 RS |
1051 | } |
1052 | ||
908b7fea KH |
1053 | if (CHAR_TABLE_P (value)) |
1054 | { | |
62abe9cb SM |
1055 | insert_string ("deeper char-table ..."); |
1056 | return syntax; | |
908b7fea KH |
1057 | } |
1058 | ||
8ea151b2 RS |
1059 | if (!CONSP (value)) |
1060 | { | |
62abe9cb SM |
1061 | insert_string ("invalid"); |
1062 | return syntax; | |
8ea151b2 RS |
1063 | } |
1064 | ||
c1d497be KR |
1065 | first = XCAR (value); |
1066 | match_lisp = XCDR (value); | |
8ea151b2 RS |
1067 | |
1068 | if (!INTEGERP (first) || !(NILP (match_lisp) || INTEGERP (match_lisp))) | |
8489eb67 | 1069 | { |
62abe9cb SM |
1070 | insert_string ("invalid"); |
1071 | return syntax; | |
8489eb67 RS |
1072 | } |
1073 | ||
e704cb4b | 1074 | code = (enum syntaxcode) (XINT (first) & 0377); |
8ea151b2 RS |
1075 | start1 = (XINT (first) >> 16) & 1; |
1076 | start2 = (XINT (first) >> 17) & 1; | |
1077 | end1 = (XINT (first) >> 18) & 1; | |
1078 | end2 = (XINT (first) >> 19) & 1; | |
1079 | prefix = (XINT (first) >> 20) & 1; | |
1080 | comstyle = (XINT (first) >> 21) & 1; | |
e6365855 | 1081 | comnested = (XINT (first) >> 22) & 1; |
8489eb67 RS |
1082 | |
1083 | if ((int) code < 0 || (int) code >= (int) Smax) | |
1084 | { | |
1085 | insert_string ("invalid"); | |
62abe9cb | 1086 | return syntax; |
8489eb67 RS |
1087 | } |
1088 | desc = syntax_code_spec[(int) code]; | |
1089 | ||
1090 | str[0] = desc, str[1] = 0; | |
1091 | insert (str, 1); | |
1092 | ||
93da5fff KH |
1093 | if (NILP (match_lisp)) |
1094 | insert (" ", 1); | |
1095 | else | |
1096 | insert_char (XINT (match_lisp)); | |
8489eb67 | 1097 | |
8489eb67 RS |
1098 | if (start1) |
1099 | insert ("1", 1); | |
1100 | if (start2) | |
1101 | insert ("2", 1); | |
1102 | ||
1103 | if (end1) | |
1104 | insert ("3", 1); | |
1105 | if (end2) | |
1106 | insert ("4", 1); | |
1107 | ||
1108 | if (prefix) | |
1109 | insert ("p", 1); | |
e5d4f4dc RS |
1110 | if (comstyle) |
1111 | insert ("b", 1); | |
e6365855 SM |
1112 | if (comnested) |
1113 | insert ("n", 1); | |
8489eb67 RS |
1114 | |
1115 | insert_string ("\twhich means: "); | |
1116 | ||
0220c518 | 1117 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
1118 | { |
1119 | case Swhitespace: | |
1120 | insert_string ("whitespace"); break; | |
1121 | case Spunct: | |
1122 | insert_string ("punctuation"); break; | |
1123 | case Sword: | |
1124 | insert_string ("word"); break; | |
1125 | case Ssymbol: | |
1126 | insert_string ("symbol"); break; | |
1127 | case Sopen: | |
1128 | insert_string ("open"); break; | |
1129 | case Sclose: | |
1130 | insert_string ("close"); break; | |
1131 | case Squote: | |
d671382e | 1132 | insert_string ("prefix"); break; |
8489eb67 RS |
1133 | case Sstring: |
1134 | insert_string ("string"); break; | |
1135 | case Smath: | |
1136 | insert_string ("math"); break; | |
1137 | case Sescape: | |
1138 | insert_string ("escape"); break; | |
1139 | case Scharquote: | |
1140 | insert_string ("charquote"); break; | |
1141 | case Scomment: | |
1142 | insert_string ("comment"); break; | |
1143 | case Sendcomment: | |
1144 | insert_string ("endcomment"); break; | |
d671382e SM |
1145 | case Sinherit: |
1146 | insert_string ("inherit"); break; | |
1147 | case Scomment_fence: | |
1148 | insert_string ("comment fence"); break; | |
1149 | case Sstring_fence: | |
1150 | insert_string ("string fence"); break; | |
8489eb67 RS |
1151 | default: |
1152 | insert_string ("invalid"); | |
62abe9cb | 1153 | return syntax; |
8489eb67 RS |
1154 | } |
1155 | ||
8ea151b2 | 1156 | if (!NILP (match_lisp)) |
8489eb67 RS |
1157 | { |
1158 | insert_string (", matches "); | |
8ea151b2 | 1159 | insert_char (XINT (match_lisp)); |
8489eb67 RS |
1160 | } |
1161 | ||
1162 | if (start1) | |
1163 | insert_string (",\n\t is the first character of a comment-start sequence"); | |
1164 | if (start2) | |
1165 | insert_string (",\n\t is the second character of a comment-start sequence"); | |
1166 | ||
1167 | if (end1) | |
1168 | insert_string (",\n\t is the first character of a comment-end sequence"); | |
1169 | if (end2) | |
1170 | insert_string (",\n\t is the second character of a comment-end sequence"); | |
e5d4f4dc RS |
1171 | if (comstyle) |
1172 | insert_string (" (comment style b)"); | |
e6365855 SM |
1173 | if (comnested) |
1174 | insert_string (" (nestable)"); | |
e5d4f4dc | 1175 | |
8489eb67 RS |
1176 | if (prefix) |
1177 | insert_string (",\n\t is a prefix character for `backward-prefix-chars'"); | |
1178 | ||
62abe9cb SM |
1179 | return syntax; |
1180 | } | |
8489eb67 | 1181 | \f |
195d1361 RS |
1182 | int parse_sexp_ignore_comments; |
1183 | ||
8f924df7 KH |
1184 | /* Char-table of functions that find the next or previous word |
1185 | boundary. */ | |
1186 | Lisp_Object Vfind_word_boundary_function_table; | |
869bb237 | 1187 | |
8489eb67 RS |
1188 | /* Return the position across COUNT words from FROM. |
1189 | If that many words cannot be found before the end of the buffer, return 0. | |
1190 | COUNT negative means scan backward and stop at word beginning. */ | |
1191 | ||
dfcf069d | 1192 | int |
8489eb67 RS |
1193 | scan_words (from, count) |
1194 | register int from, count; | |
1195 | { | |
1196 | register int beg = BEGV; | |
1197 | register int end = ZV; | |
6a140a74 | 1198 | register int from_byte = CHAR_TO_BYTE (from); |
93da5fff KH |
1199 | register enum syntaxcode code; |
1200 | int ch0, ch1; | |
8f924df7 | 1201 | Lisp_Object func, script, pos; |
8489eb67 RS |
1202 | |
1203 | immediate_quit = 1; | |
1204 | QUIT; | |
1205 | ||
195d1361 RS |
1206 | SETUP_SYNTAX_TABLE (from, count); |
1207 | ||
8489eb67 RS |
1208 | while (count > 0) |
1209 | { | |
1210 | while (1) | |
1211 | { | |
1212 | if (from == end) | |
1213 | { | |
1214 | immediate_quit = 0; | |
1215 | return 0; | |
1216 | } | |
195d1361 | 1217 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 1218 | ch0 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
93da5fff | 1219 | code = SYNTAX (ch0); |
6a140a74 | 1220 | INC_BOTH (from, from_byte); |
8489eb67 RS |
1221 | if (words_include_escapes |
1222 | && (code == Sescape || code == Scharquote)) | |
1223 | break; | |
1224 | if (code == Sword) | |
1225 | break; | |
8489eb67 | 1226 | } |
93da5fff KH |
1227 | /* Now CH0 is a character which begins a word and FROM is the |
1228 | position of the next character. */ | |
8f924df7 | 1229 | func = CHAR_TABLE_REF (Vfind_word_boundary_function_table, ch0); |
869bb237 | 1230 | if (! NILP (Ffboundp (func))) |
8489eb67 | 1231 | { |
869bb237 | 1232 | pos = call2 (func, make_number (from - 1), make_number (end)); |
8f924df7 KH |
1233 | if (INTEGERP (pos) && XINT (pos) > from) |
1234 | { | |
1235 | from = XINT (pos); | |
1236 | from_byte = CHAR_TO_BYTE (from); | |
1237 | } | |
8489eb67 | 1238 | } |
869bb237 | 1239 | else |
8cb8232a | 1240 | { |
8cb8232a KH |
1241 | script = CHAR_TABLE_REF (Vchar_script_table, ch0); |
1242 | while (1) | |
1243 | { | |
1244 | if (from == end) break; | |
1245 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
b7dbcc19 | 1246 | ch1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8cb8232a KH |
1247 | code = SYNTAX (ch1); |
1248 | if ((code != Sword | |
1249 | && (! words_include_escapes | |
1250 | || (code != Sescape && code != Scharquote))) | |
2b70a11c | 1251 | || word_boundary_p (ch0, ch1)) |
869bb237 | 1252 | break; |
8cb8232a KH |
1253 | INC_BOTH (from, from_byte); |
1254 | ch0 = ch1; | |
1255 | } | |
8489eb67 RS |
1256 | } |
1257 | count--; | |
1258 | } | |
1259 | while (count < 0) | |
1260 | { | |
1261 | while (1) | |
1262 | { | |
1263 | if (from == beg) | |
1264 | { | |
1265 | immediate_quit = 0; | |
1266 | return 0; | |
1267 | } | |
6a140a74 | 1268 | DEC_BOTH (from, from_byte); |
195d1361 | 1269 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 1270 | ch1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
93da5fff | 1271 | code = SYNTAX (ch1); |
8489eb67 RS |
1272 | if (words_include_escapes |
1273 | && (code == Sescape || code == Scharquote)) | |
1274 | break; | |
1275 | if (code == Sword) | |
1276 | break; | |
8489eb67 | 1277 | } |
93da5fff KH |
1278 | /* Now CH1 is a character which ends a word and FROM is the |
1279 | position of it. */ | |
8f924df7 | 1280 | func = CHAR_TABLE_REF (Vfind_word_boundary_function_table, ch1); |
869bb237 | 1281 | if (! NILP (Ffboundp (func))) |
8f924df7 | 1282 | { |
869bb237 | 1283 | pos = call2 (func, make_number (from), make_number (beg)); |
8f924df7 KH |
1284 | if (INTEGERP (pos) && XINT (pos) < from) |
1285 | { | |
1286 | from = XINT (pos); | |
1287 | from_byte = CHAR_TO_BYTE (from); | |
1288 | } | |
8489eb67 | 1289 | } |
869bb237 | 1290 | else |
8489eb67 | 1291 | { |
8cb8232a KH |
1292 | script = CHAR_TABLE_REF (Vchar_script_table, ch1); |
1293 | while (1) | |
1294 | { | |
8cb8232a KH |
1295 | if (from == beg) |
1296 | break; | |
262be72a | 1297 | DEC_BOTH (from, from_byte); |
8cb8232a | 1298 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
262be72a | 1299 | ch0 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8cb8232a KH |
1300 | code = SYNTAX (ch0); |
1301 | if ((code != Sword | |
1302 | && (! words_include_escapes | |
1303 | || (code != Sescape && code != Scharquote))) | |
2b70a11c | 1304 | || word_boundary_p (ch0, ch1)) |
262be72a MB |
1305 | { |
1306 | INC_BOTH (from, from_byte); | |
1307 | break; | |
1308 | } | |
8cb8232a KH |
1309 | ch1 = ch0; |
1310 | } | |
8489eb67 RS |
1311 | } |
1312 | count++; | |
1313 | } | |
1314 | ||
1315 | immediate_quit = 0; | |
1316 | ||
1317 | return from; | |
1318 | } | |
1319 | ||
1e9dbb5f | 1320 | DEFUN ("forward-word", Fforward_word, Sforward_word, 0, 1, "^p", |
fdb82f93 PJ |
1321 | doc: /* Move point forward ARG words (backward if ARG is negative). |
1322 | Normally returns t. | |
1323 | If an edge of the buffer or a field boundary is reached, point is left there | |
1324 | and the function returns nil. Field boundaries are not noticed if | |
1325 | `inhibit-field-text-motion' is non-nil. */) | |
839966f3 KH |
1326 | (arg) |
1327 | Lisp_Object arg; | |
8489eb67 | 1328 | { |
c096ae4d | 1329 | Lisp_Object tmp; |
2c6ea900 | 1330 | int orig_val, val; |
8489eb67 | 1331 | |
839966f3 KH |
1332 | if (NILP (arg)) |
1333 | XSETFASTINT (arg, 1); | |
1334 | else | |
1335 | CHECK_NUMBER (arg); | |
1336 | ||
1337 | val = orig_val = scan_words (PT, XINT (arg)); | |
2c6ea900 | 1338 | if (! orig_val) |
839966f3 | 1339 | val = XINT (arg) > 0 ? ZV : BEGV; |
5878ee6f | 1340 | |
b8855607 | 1341 | /* Avoid jumping out of an input field. */ |
c096ae4d SM |
1342 | tmp = Fconstrain_to_field (make_number (val), make_number (PT), |
1343 | Qt, Qnil, Qnil); | |
1344 | val = XFASTINT (tmp); | |
7d0393cf | 1345 | |
8489eb67 | 1346 | SET_PT (val); |
e6d8341f | 1347 | return val == orig_val ? Qt : Qnil; |
8489eb67 RS |
1348 | } |
1349 | \f | |
195d1361 RS |
1350 | Lisp_Object skip_chars (); |
1351 | ||
1352 | DEFUN ("skip-chars-forward", Fskip_chars_forward, Sskip_chars_forward, 1, 2, 0, | |
fdb82f93 PJ |
1353 | doc: /* Move point forward, stopping before a char not in STRING, or at pos LIM. |
1354 | STRING is like the inside of a `[...]' in a regular expression | |
1355 | except that `]' is never special and `\\' quotes `^', `-' or `\\' | |
7087d5e9 | 1356 | (but not at the end of a range; quoting is never needed there). |
fdb82f93 PJ |
1357 | Thus, with arg "a-zA-Z", this skips letters stopping before first nonletter. |
1358 | With arg "^a-zA-Z", skips nonletters stopping before first letter. | |
a1bc88d4 RS |
1359 | Char classes, e.g. `[:alpha:]', are supported. |
1360 | ||
1361 | Returns the distance traveled, either zero or positive. */) | |
fdb82f93 | 1362 | (string, lim) |
195d1361 RS |
1363 | Lisp_Object string, lim; |
1364 | { | |
327719ee | 1365 | return skip_chars (1, string, lim, 1); |
195d1361 RS |
1366 | } |
1367 | ||
1368 | DEFUN ("skip-chars-backward", Fskip_chars_backward, Sskip_chars_backward, 1, 2, 0, | |
fdb82f93 PJ |
1369 | doc: /* Move point backward, stopping after a char not in STRING, or at pos LIM. |
1370 | See `skip-chars-forward' for details. | |
1371 | Returns the distance traveled, either zero or negative. */) | |
1372 | (string, lim) | |
195d1361 RS |
1373 | Lisp_Object string, lim; |
1374 | { | |
327719ee | 1375 | return skip_chars (0, string, lim, 1); |
195d1361 RS |
1376 | } |
1377 | ||
1378 | DEFUN ("skip-syntax-forward", Fskip_syntax_forward, Sskip_syntax_forward, 1, 2, 0, | |
fdb82f93 PJ |
1379 | doc: /* Move point forward across chars in specified syntax classes. |
1380 | SYNTAX is a string of syntax code characters. | |
1381 | Stop before a char whose syntax is not in SYNTAX, or at position LIM. | |
1382 | If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX. | |
1383 | This function returns the distance traveled, either zero or positive. */) | |
1384 | (syntax, lim) | |
195d1361 RS |
1385 | Lisp_Object syntax, lim; |
1386 | { | |
b7dbcc19 | 1387 | return skip_syntaxes (1, syntax, lim); |
195d1361 RS |
1388 | } |
1389 | ||
1390 | DEFUN ("skip-syntax-backward", Fskip_syntax_backward, Sskip_syntax_backward, 1, 2, 0, | |
fdb82f93 PJ |
1391 | doc: /* Move point backward across chars in specified syntax classes. |
1392 | SYNTAX is a string of syntax code characters. | |
1393 | Stop on reaching a char whose syntax is not in SYNTAX, or at position LIM. | |
1394 | If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX. | |
1395 | This function returns the distance traveled, either zero or negative. */) | |
1396 | (syntax, lim) | |
195d1361 RS |
1397 | Lisp_Object syntax, lim; |
1398 | { | |
b7dbcc19 | 1399 | return skip_syntaxes (0, syntax, lim); |
195d1361 RS |
1400 | } |
1401 | ||
6a140a74 | 1402 | static Lisp_Object |
327719ee | 1403 | skip_chars (forwardp, string, lim, handle_iso_classes) |
b7dbcc19 | 1404 | int forwardp; |
195d1361 | 1405 | Lisp_Object string, lim; |
a1bc88d4 | 1406 | int handle_iso_classes; |
195d1361 | 1407 | { |
195d1361 RS |
1408 | register unsigned int c; |
1409 | unsigned char fastmap[0400]; | |
b7dbcc19 | 1410 | /* Store the ranges of non-ASCII characters. */ |
5df4982e | 1411 | int *char_ranges; |
7e68b0ea | 1412 | int n_char_ranges = 0; |
195d1361 | 1413 | int negate = 0; |
4101e6fe | 1414 | register int i, i_byte; |
b7dbcc19 KH |
1415 | /* Set to 1 if the current buffer is multibyte and the region |
1416 | contains non-ASCII chars. */ | |
1417 | int multibyte; | |
1418 | /* Set to 1 if STRING is multibyte and it contains non-ASCII | |
1419 | chars. */ | |
1674d9a2 RS |
1420 | int string_multibyte; |
1421 | int size_byte; | |
2e567bd3 | 1422 | const unsigned char *str; |
82d497fc | 1423 | int len; |
a1bc88d4 | 1424 | Lisp_Object iso_classes; |
195d1361 | 1425 | |
b7826503 | 1426 | CHECK_STRING (string); |
a1bc88d4 | 1427 | iso_classes = Qnil; |
82d497fc | 1428 | |
195d1361 RS |
1429 | if (NILP (lim)) |
1430 | XSETINT (lim, forwardp ? ZV : BEGV); | |
1431 | else | |
b7826503 | 1432 | CHECK_NUMBER_COERCE_MARKER (lim); |
195d1361 RS |
1433 | |
1434 | /* In any case, don't allow scan outside bounds of buffer. */ | |
195d1361 RS |
1435 | if (XINT (lim) > ZV) |
1436 | XSETFASTINT (lim, ZV); | |
1437 | if (XINT (lim) < BEGV) | |
1438 | XSETFASTINT (lim, BEGV); | |
1439 | ||
b7dbcc19 | 1440 | multibyte = (!NILP (current_buffer->enable_multibyte_characters) |
92eaa22e | 1441 | && (XINT (lim) - PT != CHAR_TO_BYTE (XINT (lim)) - PT_BYTE)); |
8f924df7 | 1442 | string_multibyte = SBYTES (string) > SCHARS (string); |
b7dbcc19 | 1443 | |
195d1361 RS |
1444 | bzero (fastmap, sizeof fastmap); |
1445 | ||
8f924df7 KH |
1446 | str = SDATA (string); |
1447 | size_byte = SBYTES (string); | |
4101e6fe | 1448 | |
82d497fc | 1449 | i_byte = 0; |
13090112 | 1450 | if (i_byte < size_byte |
d5db4077 | 1451 | && SREF (string, 0) == '^') |
195d1361 | 1452 | { |
82d497fc | 1453 | negate = 1; i_byte++; |
195d1361 RS |
1454 | } |
1455 | ||
1456 | /* Find the characters specified and set their elements of fastmap. | |
b7dbcc19 | 1457 | Handle backslashes and ranges specially. |
195d1361 | 1458 | |
b7dbcc19 KH |
1459 | If STRING contains non-ASCII characters, setup char_ranges for |
1460 | them and use fastmap only for their leading codes. */ | |
4101e6fe | 1461 | |
b7dbcc19 | 1462 | if (! string_multibyte) |
195d1361 | 1463 | { |
b7dbcc19 | 1464 | int string_has_eight_bit = 0; |
4101e6fe | 1465 | |
b7dbcc19 KH |
1466 | /* At first setup fastmap. */ |
1467 | while (i_byte < size_byte) | |
1468 | { | |
1469 | c = str[i_byte++]; | |
1470 | ||
a1bc88d4 RS |
1471 | if (handle_iso_classes && c == '[' |
1472 | && i_byte < size_byte | |
327719ee | 1473 | && str[i_byte] == ':') |
a1bc88d4 RS |
1474 | { |
1475 | const unsigned char *class_beg = str + i_byte + 1; | |
1476 | const unsigned char *class_end = class_beg; | |
b3bda4fd | 1477 | const unsigned char *class_limit = str + size_byte - 2; |
db9cd97a | 1478 | /* Leave room for the null. */ |
a1bc88d4 RS |
1479 | unsigned char class_name[CHAR_CLASS_MAX_LENGTH + 1]; |
1480 | re_wctype_t cc; | |
1481 | ||
1482 | if (class_limit - class_beg > CHAR_CLASS_MAX_LENGTH) | |
1483 | class_limit = class_beg + CHAR_CLASS_MAX_LENGTH; | |
1484 | ||
b3bda4fd KS |
1485 | while (class_end < class_limit |
1486 | && *class_end >= 'a' && *class_end <= 'z') | |
a1bc88d4 RS |
1487 | class_end++; |
1488 | ||
b3bda4fd KS |
1489 | if (class_end == class_beg |
1490 | || *class_end != ':' || class_end[1] != ']') | |
1491 | goto not_a_class_name; | |
a1bc88d4 RS |
1492 | |
1493 | bcopy (class_beg, class_name, class_end - class_beg); | |
1494 | class_name[class_end - class_beg] = 0; | |
1495 | ||
1496 | cc = re_wctype (class_name); | |
1497 | if (cc == 0) | |
1498 | error ("Invalid ISO C character class"); | |
1499 | ||
1500 | iso_classes = Fcons (make_number (cc), iso_classes); | |
1501 | ||
1502 | i_byte = class_end + 2 - str; | |
1503 | continue; | |
1504 | } | |
1505 | ||
b3bda4fd | 1506 | not_a_class_name: |
b7dbcc19 KH |
1507 | if (c == '\\') |
1508 | { | |
13090112 | 1509 | if (i_byte == size_byte) |
4101e6fe RS |
1510 | break; |
1511 | ||
5c7b02ab | 1512 | c = str[i_byte++]; |
195d1361 | 1513 | } |
839966f3 KH |
1514 | /* Treat `-' as range character only if another character |
1515 | follows. */ | |
1516 | if (i_byte + 1 < size_byte | |
82d497fc | 1517 | && str[i_byte] == '-') |
195d1361 | 1518 | { |
9690d026 | 1519 | unsigned int c2; |
4101e6fe RS |
1520 | |
1521 | /* Skip over the dash. */ | |
82d497fc | 1522 | i_byte++; |
4101e6fe | 1523 | |
4101e6fe | 1524 | /* Get the end of the range. */ |
5c7b02ab | 1525 | c2 = str[i_byte++]; |
b7dbcc19 KH |
1526 | if (c2 == '\\' |
1527 | && i_byte < size_byte) | |
1528 | c2 = str[i_byte++]; | |
1529 | ||
8f924df7 KH |
1530 | if (c <= c2) |
1531 | { | |
1532 | while (c <= c2) | |
1533 | fastmap[c++] = 1; | |
1534 | if (! ASCII_CHAR_P (c2)) | |
1535 | string_has_eight_bit = 1; | |
1536 | } | |
195d1361 RS |
1537 | } |
1538 | else | |
b7dbcc19 KH |
1539 | { |
1540 | fastmap[c] = 1; | |
1541 | if (! ASCII_CHAR_P (c)) | |
1542 | string_has_eight_bit = 1; | |
1543 | } | |
1544 | } | |
1545 | ||
1546 | /* If the current range is multibyte and STRING contains | |
1547 | eight-bit chars, arrange fastmap and setup char_ranges for | |
1548 | the corresponding multibyte chars. */ | |
1549 | if (multibyte && string_has_eight_bit) | |
1550 | { | |
1551 | unsigned char fastmap2[0400]; | |
1552 | int range_start_byte, range_start_char; | |
1553 | ||
1554 | bcopy (fastmap2 + 0200, fastmap + 0200, 0200); | |
1555 | bzero (fastmap + 0200, 0200); | |
1556 | /* We are sure that this loop stops. */ | |
1557 | for (i = 0200; ! fastmap2[i]; i++); | |
4c0354d7 | 1558 | c = BYTE8_TO_CHAR (i); |
b7dbcc19 KH |
1559 | fastmap[CHAR_LEADING_CODE (c)] = 1; |
1560 | range_start_byte = i; | |
1561 | range_start_char = c; | |
f003ca54 | 1562 | char_ranges = (int *) alloca (sizeof (int) * 128 * 2); |
b7dbcc19 KH |
1563 | for (i = 129; i < 0400; i++) |
1564 | { | |
4c0354d7 | 1565 | c = BYTE8_TO_CHAR (i); |
b7dbcc19 KH |
1566 | fastmap[CHAR_LEADING_CODE (c)] = 1; |
1567 | if (i - range_start_byte != c - range_start_char) | |
1568 | { | |
1569 | char_ranges[n_char_ranges++] = range_start_char; | |
1570 | char_ranges[n_char_ranges++] = ((i - 1 - range_start_byte) | |
1571 | + range_start_char); | |
1572 | range_start_byte = i; | |
1573 | range_start_char = c; | |
8f924df7 | 1574 | } |
b7dbcc19 KH |
1575 | } |
1576 | char_ranges[n_char_ranges++] = range_start_char; | |
1577 | char_ranges[n_char_ranges++] = ((i - 1 - range_start_byte) | |
1578 | + range_start_char); | |
195d1361 RS |
1579 | } |
1580 | } | |
f003ca54 | 1581 | else /* STRING is multibyte */ |
b7dbcc19 | 1582 | { |
f003ca54 KH |
1583 | char_ranges = (int *) alloca (sizeof (int) * SCHARS (string) * 2); |
1584 | ||
b7dbcc19 | 1585 | while (i_byte < size_byte) |
195d1361 | 1586 | { |
b7dbcc19 | 1587 | unsigned char leading_code; |
5c7b02ab | 1588 | |
b7dbcc19 | 1589 | leading_code = str[i_byte]; |
62a6e103 | 1590 | c = STRING_CHAR_AND_LENGTH (str + i_byte, len); |
b7dbcc19 | 1591 | i_byte += len; |
5c7b02ab | 1592 | |
327719ee MB |
1593 | if (handle_iso_classes && c == '[' |
1594 | && i_byte < size_byte | |
62a6e103 | 1595 | && STRING_CHAR (str + i_byte) == ':') |
327719ee MB |
1596 | { |
1597 | const unsigned char *class_beg = str + i_byte + 1; | |
1598 | const unsigned char *class_end = class_beg; | |
1599 | const unsigned char *class_limit = str + size_byte - 2; | |
1600 | /* Leave room for the null. */ | |
1601 | unsigned char class_name[CHAR_CLASS_MAX_LENGTH + 1]; | |
1602 | re_wctype_t cc; | |
1603 | ||
1604 | if (class_limit - class_beg > CHAR_CLASS_MAX_LENGTH) | |
1605 | class_limit = class_beg + CHAR_CLASS_MAX_LENGTH; | |
1606 | ||
1607 | while (class_end < class_limit | |
1608 | && *class_end >= 'a' && *class_end <= 'z') | |
1609 | class_end++; | |
1610 | ||
1611 | if (class_end == class_beg | |
1612 | || *class_end != ':' || class_end[1] != ']') | |
1613 | goto not_a_class_name_multibyte; | |
1614 | ||
1615 | bcopy (class_beg, class_name, class_end - class_beg); | |
1616 | class_name[class_end - class_beg] = 0; | |
1617 | ||
1618 | cc = re_wctype (class_name); | |
1619 | if (cc == 0) | |
1620 | error ("Invalid ISO C character class"); | |
1621 | ||
1622 | iso_classes = Fcons (make_number (cc), iso_classes); | |
1623 | ||
1624 | i_byte = class_end + 2 - str; | |
1625 | continue; | |
1626 | } | |
1627 | ||
1628 | not_a_class_name_multibyte: | |
195d1361 RS |
1629 | if (c == '\\') |
1630 | { | |
13090112 | 1631 | if (i_byte == size_byte) |
4101e6fe RS |
1632 | break; |
1633 | ||
b7dbcc19 | 1634 | leading_code = str[i_byte]; |
62a6e103 | 1635 | c = STRING_CHAR_AND_LENGTH (str + i_byte, len); |
82d497fc | 1636 | i_byte += len; |
195d1361 | 1637 | } |
839966f3 KH |
1638 | /* Treat `-' as range character only if another character |
1639 | follows. */ | |
1640 | if (i_byte + 1 < size_byte | |
82d497fc | 1641 | && str[i_byte] == '-') |
195d1361 | 1642 | { |
9690d026 | 1643 | unsigned int c2; |
b7dbcc19 | 1644 | unsigned char leading_code2; |
4101e6fe RS |
1645 | |
1646 | /* Skip over the dash. */ | |
82d497fc | 1647 | i_byte++; |
4101e6fe | 1648 | |
4101e6fe | 1649 | /* Get the end of the range. */ |
b7dbcc19 | 1650 | leading_code2 = str[i_byte]; |
62a6e103 | 1651 | c2 = STRING_CHAR_AND_LENGTH (str + i_byte, len); |
82d497fc | 1652 | i_byte += len; |
7e68b0ea | 1653 | |
b7dbcc19 KH |
1654 | if (c2 == '\\' |
1655 | && i_byte < size_byte) | |
1656 | { | |
1657 | leading_code2 = str[i_byte]; | |
62a6e103 | 1658 | c2 =STRING_CHAR_AND_LENGTH (str + i_byte, len); |
b7dbcc19 KH |
1659 | i_byte += len; |
1660 | } | |
1661 | ||
f003ca54 KH |
1662 | if (c > c2) |
1663 | continue; | |
b7dbcc19 | 1664 | if (ASCII_CHAR_P (c)) |
06274af5 | 1665 | { |
5c7b02ab KH |
1666 | while (c <= c2 && c < 0x80) |
1667 | fastmap[c++] = 1; | |
b7dbcc19 KH |
1668 | leading_code = CHAR_LEADING_CODE (c); |
1669 | } | |
1670 | if (! ASCII_CHAR_P (c)) | |
1671 | { | |
1672 | while (leading_code <= leading_code2) | |
1673 | fastmap[leading_code++] = 1; | |
1674 | if (c <= c2) | |
e39091b6 | 1675 | { |
b7dbcc19 | 1676 | char_ranges[n_char_ranges++] = c; |
e39091b6 | 1677 | char_ranges[n_char_ranges++] = c2; |
06274af5 KH |
1678 | } |
1679 | } | |
195d1361 RS |
1680 | } |
1681 | else | |
7e68b0ea | 1682 | { |
b7dbcc19 | 1683 | if (ASCII_CHAR_P (c)) |
e39091b6 KH |
1684 | fastmap[c] = 1; |
1685 | else | |
7e68b0ea | 1686 | { |
b7dbcc19 | 1687 | fastmap[leading_code] = 1; |
4101e6fe RS |
1688 | char_ranges[n_char_ranges++] = c; |
1689 | char_ranges[n_char_ranges++] = c; | |
7e68b0ea RS |
1690 | } |
1691 | } | |
195d1361 | 1692 | } |
b7dbcc19 KH |
1693 | |
1694 | /* If the current range is unibyte and STRING contains non-ASCII | |
1695 | chars, arrange fastmap for the corresponding unibyte | |
1696 | chars. */ | |
1697 | ||
1698 | if (! multibyte && n_char_ranges > 0) | |
1699 | { | |
1700 | bzero (fastmap + 0200, 0200); | |
1701 | for (i = 0; i < n_char_ranges; i += 2) | |
1702 | { | |
1703 | int c1 = char_ranges[i]; | |
1704 | int c2 = char_ranges[i + 1]; | |
1705 | ||
1706 | for (; c1 <= c2; c1++) | |
2afc21f5 SM |
1707 | { |
1708 | int b = CHAR_TO_BYTE_SAFE (c1); | |
1709 | if (b >= 0) | |
1710 | fastmap[b] = 1; | |
1711 | } | |
b7dbcc19 KH |
1712 | } |
1713 | } | |
195d1361 RS |
1714 | } |
1715 | ||
9690d026 | 1716 | /* If ^ was the first character, complement the fastmap. */ |
195d1361 | 1717 | if (negate) |
b7dbcc19 KH |
1718 | { |
1719 | if (! multibyte) | |
1720 | for (i = 0; i < sizeof fastmap; i++) | |
1721 | fastmap[i] ^= 1; | |
1722 | else | |
1723 | { | |
1724 | for (i = 0; i < 0200; i++) | |
1725 | fastmap[i] ^= 1; | |
1726 | /* All non-ASCII chars possibly match. */ | |
1727 | for (; i < sizeof fastmap; i++) | |
1728 | fastmap[i] = 1; | |
1729 | } | |
1730 | } | |
195d1361 RS |
1731 | |
1732 | { | |
1733 | int start_point = PT; | |
1734 | int pos = PT; | |
6a140a74 | 1735 | int pos_byte = PT_BYTE; |
9af7511a KH |
1736 | unsigned char *p = PT_ADDR, *endp, *stop; |
1737 | ||
1738 | if (forwardp) | |
1739 | { | |
1740 | endp = (XINT (lim) == GPT) ? GPT_ADDR : CHAR_POS_ADDR (XINT (lim)); | |
d7ee9fab | 1741 | stop = (pos < GPT && GPT < XINT (lim)) ? GPT_ADDR : endp; |
9af7511a KH |
1742 | } |
1743 | else | |
1744 | { | |
1745 | endp = CHAR_POS_ADDR (XINT (lim)); | |
d7ee9fab | 1746 | stop = (pos >= GPT && GPT > XINT (lim)) ? GAP_END_ADDR : endp; |
9af7511a | 1747 | } |
195d1361 RS |
1748 | |
1749 | immediate_quit = 1; | |
b7dbcc19 | 1750 | if (forwardp) |
195d1361 | 1751 | { |
b7dbcc19 | 1752 | if (multibyte) |
8f924df7 | 1753 | while (1) |
b7dbcc19 | 1754 | { |
8f924df7 | 1755 | int nbytes; |
9af7511a | 1756 | |
8f924df7 | 1757 | if (p >= stop) |
9af7511a | 1758 | { |
8f924df7 | 1759 | if (p >= endp) |
9af7511a | 1760 | break; |
8f924df7 KH |
1761 | p = GAP_END_ADDR; |
1762 | stop = endp; | |
9af7511a | 1763 | } |
62a6e103 | 1764 | c = STRING_CHAR_AND_LENGTH (p, nbytes); |
327719ee | 1765 | if (! NILP (iso_classes) && in_classes (c, iso_classes)) |
9af7511a | 1766 | { |
327719ee | 1767 | if (negate) |
9af7511a | 1768 | break; |
327719ee MB |
1769 | else |
1770 | goto fwd_ok; | |
9af7511a | 1771 | } |
9af7511a | 1772 | |
8f924df7 | 1773 | if (! fastmap[*p]) |
b7dbcc19 KH |
1774 | break; |
1775 | if (! ASCII_CHAR_P (c)) | |
9af7511a | 1776 | { |
b7dbcc19 KH |
1777 | /* As we are looking at a multibyte character, we |
1778 | must look up the character in the table | |
1779 | CHAR_RANGES. If there's no data in the table, | |
1780 | that character is not what we want to skip. */ | |
1781 | ||
1782 | /* The following code do the right thing even if | |
1783 | n_char_ranges is zero (i.e. no data in | |
1784 | CHAR_RANGES). */ | |
1785 | for (i = 0; i < n_char_ranges; i += 2) | |
1786 | if (c >= char_ranges[i] && c <= char_ranges[i + 1]) | |
9af7511a | 1787 | break; |
b7dbcc19 KH |
1788 | if (!(negate ^ (i < n_char_ranges))) |
1789 | break; | |
9af7511a | 1790 | } |
327719ee | 1791 | fwd_ok: |
8f924df7 | 1792 | p += nbytes, pos++, pos_byte += nbytes; |
b7dbcc19 KH |
1793 | } |
1794 | else | |
8f924df7 KH |
1795 | while (1) |
1796 | { | |
1797 | if (p >= stop) | |
9af7511a | 1798 | { |
8f924df7 | 1799 | if (p >= endp) |
9af7511a | 1800 | break; |
8f924df7 KH |
1801 | p = GAP_END_ADDR; |
1802 | stop = endp; | |
9af7511a | 1803 | } |
327719ee MB |
1804 | |
1805 | if (!NILP (iso_classes) && in_classes (*p, iso_classes)) | |
1806 | { | |
1807 | if (negate) | |
1808 | break; | |
1809 | else | |
1810 | goto fwd_unibyte_ok; | |
1811 | } | |
1812 | ||
8f924df7 KH |
1813 | if (!fastmap[*p]) |
1814 | break; | |
327719ee | 1815 | fwd_unibyte_ok: |
8f924df7 KH |
1816 | p++, pos++, pos_byte++; |
1817 | } | |
195d1361 RS |
1818 | } |
1819 | else | |
1820 | { | |
b7dbcc19 | 1821 | if (multibyte) |
8f924df7 | 1822 | while (1) |
b7dbcc19 | 1823 | { |
8f924df7 | 1824 | unsigned char *prev_p; |
9af7511a | 1825 | |
8f924df7 | 1826 | if (p <= stop) |
9af7511a | 1827 | { |
8f924df7 | 1828 | if (p <= endp) |
9af7511a | 1829 | break; |
8f924df7 KH |
1830 | p = GPT_ADDR; |
1831 | stop = endp; | |
9af7511a | 1832 | } |
8f924df7 KH |
1833 | prev_p = p; |
1834 | while (--p >= stop && ! CHAR_HEAD_P (*p)); | |
62a6e103 | 1835 | c = STRING_CHAR (p); |
a1bc88d4 | 1836 | |
327719ee MB |
1837 | if (! NILP (iso_classes) && in_classes (c, iso_classes)) |
1838 | { | |
1839 | if (negate) | |
1840 | break; | |
9690d026 | 1841 | else |
327719ee | 1842 | goto back_ok; |
7e68b0ea | 1843 | } |
a1bc88d4 | 1844 | |
8f924df7 | 1845 | if (! fastmap[*p]) |
b7dbcc19 KH |
1846 | break; |
1847 | if (! ASCII_CHAR_P (c)) | |
7e68b0ea | 1848 | { |
b7dbcc19 KH |
1849 | /* See the comment in the previous similar code. */ |
1850 | for (i = 0; i < n_char_ranges; i += 2) | |
1851 | if (c >= char_ranges[i] && c <= char_ranges[i + 1]) | |
1852 | break; | |
1853 | if (!(negate ^ (i < n_char_ranges))) | |
1854 | break; | |
7e68b0ea | 1855 | } |
327719ee | 1856 | back_ok: |
8f924df7 | 1857 | pos--, pos_byte -= prev_p - p; |
b7dbcc19 KH |
1858 | } |
1859 | else | |
8f924df7 KH |
1860 | while (1) |
1861 | { | |
1862 | if (p <= stop) | |
9af7511a | 1863 | { |
8f924df7 | 1864 | if (p <= endp) |
9af7511a | 1865 | break; |
8f924df7 KH |
1866 | p = GPT_ADDR; |
1867 | stop = endp; | |
9af7511a | 1868 | } |
e39091b6 | 1869 | |
327719ee MB |
1870 | if (! NILP (iso_classes) && in_classes (p[-1], iso_classes)) |
1871 | { | |
1872 | if (negate) | |
1873 | break; | |
9af7511a | 1874 | else |
327719ee MB |
1875 | goto back_unibyte_ok; |
1876 | } | |
a1bc88d4 | 1877 | |
8f924df7 KH |
1878 | if (!fastmap[p[-1]]) |
1879 | break; | |
327719ee | 1880 | back_unibyte_ok: |
8f924df7 KH |
1881 | p--, pos--, pos_byte--; |
1882 | } | |
195d1361 | 1883 | } |
7e68b0ea | 1884 | |
b7dbcc19 KH |
1885 | SET_PT_BOTH (pos, pos_byte); |
1886 | immediate_quit = 0; | |
1887 | ||
1888 | return make_number (PT - start_point); | |
1889 | } | |
1890 | } | |
7e68b0ea | 1891 | |
b7dbcc19 KH |
1892 | |
1893 | static Lisp_Object | |
1894 | skip_syntaxes (forwardp, string, lim) | |
1895 | int forwardp; | |
1896 | Lisp_Object string, lim; | |
1897 | { | |
1898 | register unsigned int c; | |
1899 | unsigned char fastmap[0400]; | |
1900 | int negate = 0; | |
1901 | register int i, i_byte; | |
1902 | int multibyte; | |
1903 | int size_byte; | |
1904 | unsigned char *str; | |
1905 | ||
1906 | CHECK_STRING (string); | |
1907 | ||
1908 | if (NILP (lim)) | |
1909 | XSETINT (lim, forwardp ? ZV : BEGV); | |
1910 | else | |
1911 | CHECK_NUMBER_COERCE_MARKER (lim); | |
1912 | ||
1913 | /* In any case, don't allow scan outside bounds of buffer. */ | |
1914 | if (XINT (lim) > ZV) | |
1915 | XSETFASTINT (lim, ZV); | |
1916 | if (XINT (lim) < BEGV) | |
1917 | XSETFASTINT (lim, BEGV); | |
1918 | ||
ade8ee9e | 1919 | if (forwardp ? (PT >= XFASTINT (lim)) : (PT <= XFASTINT (lim))) |
49feb1cd | 1920 | return make_number (0); |
8c6e735b | 1921 | |
b7dbcc19 | 1922 | multibyte = (!NILP (current_buffer->enable_multibyte_characters) |
92eaa22e | 1923 | && (XINT (lim) - PT != CHAR_TO_BYTE (XINT (lim)) - PT_BYTE)); |
b7dbcc19 KH |
1924 | |
1925 | bzero (fastmap, sizeof fastmap); | |
1926 | ||
8f924df7 KH |
1927 | if (SBYTES (string) > SCHARS (string)) |
1928 | /* As this is very rare case (syntax spec is ASCII only), don't | |
1929 | consider efficiency. */ | |
bc796a59 KH |
1930 | string = string_make_unibyte (string); |
1931 | ||
8f924df7 KH |
1932 | str = SDATA (string); |
1933 | size_byte = SBYTES (string); | |
bc796a59 | 1934 | |
b7dbcc19 KH |
1935 | i_byte = 0; |
1936 | if (i_byte < size_byte | |
8f924df7 | 1937 | && SREF (string, 0) == '^') |
b7dbcc19 KH |
1938 | { |
1939 | negate = 1; i_byte++; | |
1940 | } | |
1941 | ||
b7dbcc19 KH |
1942 | /* Find the syntaxes specified and set their elements of fastmap. */ |
1943 | ||
1944 | while (i_byte < size_byte) | |
1945 | { | |
1946 | c = str[i_byte++]; | |
8f924df7 | 1947 | fastmap[syntax_spec_code[c]] = 1; |
b7dbcc19 | 1948 | } |
195d1361 | 1949 | |
9690d026 | 1950 | /* If ^ was the first character, complement the fastmap. */ |
195d1361 RS |
1951 | if (negate) |
1952 | for (i = 0; i < sizeof fastmap; i++) | |
9690d026 | 1953 | fastmap[i] ^= 1; |
195d1361 RS |
1954 | |
1955 | { | |
1956 | int start_point = PT; | |
1957 | int pos = PT; | |
6a140a74 | 1958 | int pos_byte = PT_BYTE; |
8f924df7 KH |
1959 | unsigned char *p = PT_ADDR, *endp, *stop; |
1960 | ||
1961 | if (forwardp) | |
1962 | { | |
1963 | endp = (XINT (lim) == GPT) ? GPT_ADDR : CHAR_POS_ADDR (XINT (lim)); | |
1964 | stop = (pos < GPT && GPT < XINT (lim)) ? GPT_ADDR : endp; | |
1965 | } | |
1966 | else | |
1967 | { | |
1968 | endp = CHAR_POS_ADDR (XINT (lim)); | |
1969 | stop = (pos >= GPT && GPT > XINT (lim)) ? GAP_END_ADDR : endp; | |
1970 | } | |
195d1361 RS |
1971 | |
1972 | immediate_quit = 1; | |
b7dbcc19 KH |
1973 | SETUP_SYNTAX_TABLE (pos, forwardp ? 1 : -1); |
1974 | if (forwardp) | |
195d1361 | 1975 | { |
b7dbcc19 | 1976 | if (multibyte) |
195d1361 | 1977 | { |
8f924df7 | 1978 | while (1) |
8c6e735b | 1979 | { |
8f924df7 KH |
1980 | int nbytes; |
1981 | ||
1982 | if (p >= stop) | |
1983 | { | |
1984 | if (p >= endp) | |
1985 | break; | |
1986 | p = GAP_END_ADDR; | |
1987 | stop = endp; | |
1988 | } | |
62a6e103 | 1989 | c = STRING_CHAR_AND_LENGTH (p, nbytes); |
8f924df7 | 1990 | if (! fastmap[(int) SYNTAX (c)]) |
8c6e735b | 1991 | break; |
8f924df7 | 1992 | p += nbytes, pos++, pos_byte += nbytes; |
8c6e735b KH |
1993 | UPDATE_SYNTAX_TABLE_FORWARD (pos); |
1994 | } | |
195d1361 RS |
1995 | } |
1996 | else | |
1997 | { | |
8c6e735b | 1998 | while (1) |
7e68b0ea | 1999 | { |
8f924df7 KH |
2000 | if (p >= stop) |
2001 | { | |
2002 | if (p >= endp) | |
2003 | break; | |
2004 | p = GAP_END_ADDR; | |
2005 | stop = endp; | |
2006 | } | |
2007 | if (! fastmap[(int) SYNTAX (*p)]) | |
8c6e735b | 2008 | break; |
8f924df7 | 2009 | p++, pos++, pos_byte++; |
b7dbcc19 | 2010 | UPDATE_SYNTAX_TABLE_FORWARD (pos); |
195d1361 RS |
2011 | } |
2012 | } | |
2013 | } | |
2014 | else | |
2015 | { | |
b7dbcc19 | 2016 | if (multibyte) |
195d1361 | 2017 | { |
8c6e735b | 2018 | while (1) |
b7dbcc19 | 2019 | { |
8f924df7 KH |
2020 | unsigned char *prev_p; |
2021 | ||
2022 | if (p <= stop) | |
b7dbcc19 | 2023 | { |
8f924df7 KH |
2024 | if (p <= endp) |
2025 | break; | |
2026 | p = GPT_ADDR; | |
2027 | stop = endp; | |
b7dbcc19 | 2028 | } |
262be72a | 2029 | UPDATE_SYNTAX_TABLE_BACKWARD (pos - 1); |
8f924df7 KH |
2030 | prev_p = p; |
2031 | while (--p >= stop && ! CHAR_HEAD_P (*p)); | |
62a6e103 | 2032 | c = STRING_CHAR (p); |
8f924df7 KH |
2033 | if (! fastmap[(int) SYNTAX (c)]) |
2034 | break; | |
2035 | pos--, pos_byte -= prev_p - p; | |
b7dbcc19 | 2036 | } |
195d1361 RS |
2037 | } |
2038 | else | |
2039 | { | |
8c6e735b KH |
2040 | while (1) |
2041 | { | |
8f924df7 KH |
2042 | if (p <= stop) |
2043 | { | |
2044 | if (p <= endp) | |
2045 | break; | |
2046 | p = GPT_ADDR; | |
2047 | stop = endp; | |
2048 | } | |
102f6132 | 2049 | UPDATE_SYNTAX_TABLE_BACKWARD (pos - 1); |
8f924df7 | 2050 | if (! fastmap[(int) SYNTAX (p[-1])]) |
8c6e735b | 2051 | break; |
8f924df7 | 2052 | p--, pos--, pos_byte--; |
8c6e735b | 2053 | } |
195d1361 RS |
2054 | } |
2055 | } | |
6a140a74 RS |
2056 | |
2057 | SET_PT_BOTH (pos, pos_byte); | |
195d1361 RS |
2058 | immediate_quit = 0; |
2059 | ||
2060 | return make_number (PT - start_point); | |
2061 | } | |
2062 | } | |
a1bc88d4 RS |
2063 | |
2064 | /* Return 1 if character C belongs to one of the ISO classes | |
2065 | in the list ISO_CLASSES. Each class is represented by an | |
2066 | integer which is its type according to re_wctype. */ | |
2067 | ||
2068 | static int | |
2069 | in_classes (c, iso_classes) | |
2070 | int c; | |
2071 | Lisp_Object iso_classes; | |
2072 | { | |
2073 | int fits_class = 0; | |
2074 | ||
2075 | while (! NILP (iso_classes)) | |
2076 | { | |
2077 | Lisp_Object elt; | |
2078 | elt = XCAR (iso_classes); | |
2079 | iso_classes = XCDR (iso_classes); | |
2080 | ||
2081 | if (re_iswctype (c, XFASTINT (elt))) | |
2082 | fits_class = 1; | |
2083 | } | |
2084 | ||
2085 | return fits_class; | |
2086 | } | |
195d1361 | 2087 | \f |
95ff8dfc RS |
2088 | /* Jump over a comment, assuming we are at the beginning of one. |
2089 | FROM is the current position. | |
2090 | FROM_BYTE is the bytepos corresponding to FROM. | |
2091 | Do not move past STOP (a charpos). | |
2092 | The comment over which we have to jump is of style STYLE | |
2093 | (either SYNTAX_COMMENT_STYLE(foo) or ST_COMMENT_STYLE). | |
2094 | NESTING should be positive to indicate the nesting at the beginning | |
2095 | for nested comments and should be zero or negative else. | |
2096 | ST_COMMENT_STYLE cannot be nested. | |
2097 | PREV_SYNTAX is the SYNTAX_WITH_FLAGS of the previous character | |
2098 | (or 0 If the search cannot start in the middle of a two-character). | |
2099 | ||
2100 | If successful, return 1 and store the charpos of the comment's end | |
2101 | into *CHARPOS_PTR and the corresponding bytepos into *BYTEPOS_PTR. | |
2102 | Else, return 0 and store the charpos STOP into *CHARPOS_PTR, the | |
2103 | corresponding bytepos into *BYTEPOS_PTR and the current nesting | |
2104 | (as defined for state.incomment) in *INCOMMENT_PTR. | |
2105 | ||
2106 | The comment end is the last character of the comment rather than the | |
2107 | character just after the comment. | |
2108 | ||
2109 | Global syntax data is assumed to initially be valid for FROM and | |
2110 | remains valid for forward search starting at the returned position. */ | |
2111 | ||
2112 | static int | |
2113 | forw_comment (from, from_byte, stop, nesting, style, prev_syntax, | |
2114 | charpos_ptr, bytepos_ptr, incomment_ptr) | |
a0aa7fcf | 2115 | EMACS_INT from, from_byte, stop; |
95ff8dfc | 2116 | int nesting, style, prev_syntax; |
a0aa7fcf SM |
2117 | EMACS_INT *charpos_ptr, *bytepos_ptr; |
2118 | int *incomment_ptr; | |
95ff8dfc RS |
2119 | { |
2120 | register int c, c1; | |
2121 | register enum syntaxcode code; | |
2122 | register int syntax; | |
2123 | ||
2124 | if (nesting <= 0) nesting = -1; | |
2125 | ||
2126 | /* Enter the loop in the middle so that we find | |
2127 | a 2-char comment ender if we start in the middle of it. */ | |
2128 | syntax = prev_syntax; | |
2129 | if (syntax != 0) goto forw_incomment; | |
2130 | ||
2131 | while (1) | |
2132 | { | |
2133 | if (from == stop) | |
2134 | { | |
2135 | *incomment_ptr = nesting; | |
2136 | *charpos_ptr = from; | |
2137 | *bytepos_ptr = from_byte; | |
2138 | return 0; | |
2139 | } | |
8f924df7 | 2140 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2141 | syntax = SYNTAX_WITH_FLAGS (c); |
2142 | code = syntax & 0xff; | |
2143 | if (code == Sendcomment | |
2144 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style | |
abf8a9ff SM |
2145 | && (SYNTAX_FLAGS_COMMENT_NESTED (syntax) ? |
2146 | (nesting > 0 && --nesting == 0) : nesting < 0)) | |
95ff8dfc RS |
2147 | /* we have encountered a comment end of the same style |
2148 | as the comment sequence which began this comment | |
2149 | section */ | |
2150 | break; | |
2151 | if (code == Scomment_fence | |
2152 | && style == ST_COMMENT_STYLE) | |
2153 | /* we have encountered a comment end of the same style | |
2154 | as the comment sequence which began this comment | |
2155 | section. */ | |
2156 | break; | |
2157 | if (nesting > 0 | |
2158 | && code == Scomment | |
abf8a9ff | 2159 | && SYNTAX_FLAGS_COMMENT_NESTED (syntax) |
95ff8dfc RS |
2160 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style) |
2161 | /* we have encountered a nested comment of the same style | |
2162 | as the comment sequence which began this comment section */ | |
2163 | nesting++; | |
2164 | INC_BOTH (from, from_byte); | |
2165 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
7d0393cf | 2166 | |
95ff8dfc RS |
2167 | forw_incomment: |
2168 | if (from < stop && SYNTAX_FLAGS_COMEND_FIRST (syntax) | |
2169 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style | |
8f924df7 | 2170 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
abf8a9ff SM |
2171 | SYNTAX_COMEND_SECOND (c1)) |
2172 | && ((SYNTAX_FLAGS_COMMENT_NESTED (syntax) || | |
2173 | SYNTAX_COMMENT_NESTED (c1)) ? nesting > 0 : nesting < 0)) | |
4ffe723b GM |
2174 | { |
2175 | if (--nesting <= 0) | |
2176 | /* we have encountered a comment end of the same style | |
2177 | as the comment sequence which began this comment | |
2178 | section */ | |
2179 | break; | |
2180 | else | |
2181 | { | |
2182 | INC_BOTH (from, from_byte); | |
2183 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
2184 | } | |
2185 | } | |
95ff8dfc RS |
2186 | if (nesting > 0 |
2187 | && from < stop | |
2188 | && SYNTAX_FLAGS_COMSTART_FIRST (syntax) | |
8f924df7 | 2189 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
95ff8dfc | 2190 | SYNTAX_COMMENT_STYLE (c1) == style |
abf8a9ff SM |
2191 | && SYNTAX_COMSTART_SECOND (c1)) |
2192 | && (SYNTAX_FLAGS_COMMENT_NESTED (syntax) || | |
2193 | SYNTAX_COMMENT_NESTED (c1))) | |
95ff8dfc RS |
2194 | /* we have encountered a nested comment of the same style |
2195 | as the comment sequence which began this comment | |
2196 | section */ | |
2197 | { | |
2198 | INC_BOTH (from, from_byte); | |
2199 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
2200 | nesting++; | |
2201 | } | |
2202 | } | |
2203 | *charpos_ptr = from; | |
2204 | *bytepos_ptr = from_byte; | |
2205 | return 1; | |
2206 | } | |
2207 | ||
b3cfe0c8 | 2208 | DEFUN ("forward-comment", Fforward_comment, Sforward_comment, 1, 1, 0, |
177c0ea7 | 2209 | doc: /* |
ee648c11 | 2210 | Move forward across up to COUNT comments. If COUNT is negative, move backward. |
fdb82f93 PJ |
2211 | Stop scanning if we find something other than a comment or whitespace. |
2212 | Set point to where scanning stops. | |
ee648c11 | 2213 | If COUNT comments are found as expected, with nothing except whitespace |
fdb82f93 PJ |
2214 | between them, return t; otherwise return nil. */) |
2215 | (count) | |
840f481c | 2216 | Lisp_Object count; |
b3cfe0c8 | 2217 | { |
2312c580 SM |
2218 | register EMACS_INT from; |
2219 | EMACS_INT from_byte; | |
2220 | register EMACS_INT stop; | |
8ea151b2 | 2221 | register int c, c1; |
b3cfe0c8 RS |
2222 | register enum syntaxcode code; |
2223 | int comstyle = 0; /* style of comment encountered */ | |
95ff8dfc | 2224 | int comnested = 0; /* whether the comment is nestable or not */ |
be720845 | 2225 | int found; |
2312c580 SM |
2226 | EMACS_INT count1; |
2227 | EMACS_INT out_charpos, out_bytepos; | |
95ff8dfc | 2228 | int dummy; |
840f481c | 2229 | |
b7826503 | 2230 | CHECK_NUMBER (count); |
840f481c | 2231 | count1 = XINT (count); |
195d1361 | 2232 | stop = count1 > 0 ? ZV : BEGV; |
b3cfe0c8 RS |
2233 | |
2234 | immediate_quit = 1; | |
2235 | QUIT; | |
2236 | ||
2237 | from = PT; | |
6a140a74 | 2238 | from_byte = PT_BYTE; |
b3cfe0c8 | 2239 | |
195d1361 | 2240 | SETUP_SYNTAX_TABLE (from, count1); |
840f481c | 2241 | while (count1 > 0) |
b3cfe0c8 | 2242 | { |
04882296 | 2243 | do |
b3cfe0c8 | 2244 | { |
f902a008 RS |
2245 | int comstart_first; |
2246 | ||
04882296 KH |
2247 | if (from == stop) |
2248 | { | |
ef316cf0 | 2249 | SET_PT_BOTH (from, from_byte); |
b7e6e612 | 2250 | immediate_quit = 0; |
04882296 KH |
2251 | return Qnil; |
2252 | } | |
b7dbcc19 | 2253 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
b3cfe0c8 | 2254 | code = SYNTAX (c); |
f902a008 | 2255 | comstart_first = SYNTAX_COMSTART_FIRST (c); |
95ff8dfc | 2256 | comnested = SYNTAX_COMMENT_NESTED (c); |
e6365855 | 2257 | comstyle = SYNTAX_COMMENT_STYLE (c); |
6a140a74 | 2258 | INC_BOTH (from, from_byte); |
f902a008 | 2259 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
f902a008 | 2260 | if (from < stop && comstart_first |
b7dbcc19 | 2261 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
8ea151b2 | 2262 | SYNTAX_COMSTART_SECOND (c1))) |
b3cfe0c8 | 2263 | { |
7d0393cf | 2264 | /* We have encountered a comment start sequence and we |
7fc8191e | 2265 | are ignoring all text inside comments. We must record |
b3cfe0c8 RS |
2266 | the comment style this sequence begins so that later, |
2267 | only a comment end of the same style actually ends | |
7fc8191e | 2268 | the comment section. */ |
b3cfe0c8 | 2269 | code = Scomment; |
8ea151b2 | 2270 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
95ff8dfc | 2271 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); |
6a140a74 | 2272 | INC_BOTH (from, from_byte); |
95ff8dfc | 2273 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b3cfe0c8 | 2274 | } |
04882296 | 2275 | } |
abf8a9ff | 2276 | while (code == Swhitespace || (code == Sendcomment && c == '\n')); |
6a140a74 | 2277 | |
da71ab91 KH |
2278 | if (code == Scomment_fence) |
2279 | comstyle = ST_COMMENT_STYLE; | |
2280 | else if (code != Scomment) | |
04882296 KH |
2281 | { |
2282 | immediate_quit = 0; | |
6a140a74 | 2283 | DEC_BOTH (from, from_byte); |
ef316cf0 | 2284 | SET_PT_BOTH (from, from_byte); |
04882296 KH |
2285 | return Qnil; |
2286 | } | |
2287 | /* We're at the start of a comment. */ | |
95ff8dfc RS |
2288 | found = forw_comment (from, from_byte, stop, comnested, comstyle, 0, |
2289 | &out_charpos, &out_bytepos, &dummy); | |
2290 | from = out_charpos; from_byte = out_bytepos; | |
2291 | if (!found) | |
04882296 | 2292 | { |
95ff8dfc RS |
2293 | immediate_quit = 0; |
2294 | SET_PT_BOTH (from, from_byte); | |
2295 | return Qnil; | |
b3cfe0c8 | 2296 | } |
95ff8dfc RS |
2297 | INC_BOTH (from, from_byte); |
2298 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
04882296 | 2299 | /* We have skipped one comment. */ |
840f481c | 2300 | count1--; |
b3cfe0c8 RS |
2301 | } |
2302 | ||
840f481c | 2303 | while (count1 < 0) |
b3cfe0c8 | 2304 | { |
b9145dbb | 2305 | while (1) |
b3cfe0c8 | 2306 | { |
c65adb44 | 2307 | int quoted; |
f902a008 | 2308 | |
b9145dbb RS |
2309 | if (from <= stop) |
2310 | { | |
6a140a74 | 2311 | SET_PT_BOTH (BEGV, BEGV_BYTE); |
b9145dbb RS |
2312 | immediate_quit = 0; |
2313 | return Qnil; | |
2314 | } | |
b3cfe0c8 | 2315 | |
6a140a74 | 2316 | DEC_BOTH (from, from_byte); |
f902a008 | 2317 | /* char_quoted does UPDATE_SYNTAX_TABLE_BACKWARD (from). */ |
6a140a74 | 2318 | quoted = char_quoted (from, from_byte); |
b7dbcc19 | 2319 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
b3cfe0c8 RS |
2320 | code = SYNTAX (c); |
2321 | comstyle = 0; | |
95ff8dfc | 2322 | comnested = SYNTAX_COMMENT_NESTED (c); |
7fc8191e RS |
2323 | if (code == Sendcomment) |
2324 | comstyle = SYNTAX_COMMENT_STYLE (c); | |
b3cfe0c8 | 2325 | if (from > stop && SYNTAX_COMEND_SECOND (c) |
f902a008 | 2326 | && prev_char_comend_first (from, from_byte) |
89c6809a | 2327 | && !char_quoted (from - 1, dec_bytepos (from_byte))) |
b3cfe0c8 | 2328 | { |
7fc8191e | 2329 | /* We must record the comment style encountered so that |
b3cfe0c8 | 2330 | later, we can match only the proper comment begin |
7fc8191e | 2331 | sequence of the same style. */ |
6a140a74 | 2332 | DEC_BOTH (from, from_byte); |
f902a008 RS |
2333 | code = Sendcomment; |
2334 | /* Calling char_quoted, above, set up global syntax position | |
2335 | at the new value of FROM. */ | |
b7dbcc19 | 2336 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2337 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
2338 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
b3cfe0c8 RS |
2339 | } |
2340 | ||
195d1361 RS |
2341 | if (code == Scomment_fence) |
2342 | { | |
2343 | /* Skip until first preceding unquoted comment_fence. */ | |
6a140a74 | 2344 | int found = 0, ini = from, ini_byte = from_byte; |
7d0393cf | 2345 | |
6a140a74 | 2346 | while (1) |
195d1361 | 2347 | { |
6a140a74 | 2348 | DEC_BOTH (from, from_byte); |
195d1361 | 2349 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 2350 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
6a140a74 | 2351 | if (SYNTAX (c) == Scomment_fence |
7d0393cf | 2352 | && !char_quoted (from, from_byte)) |
195d1361 | 2353 | { |
7d0393cf | 2354 | found = 1; |
195d1361 RS |
2355 | break; |
2356 | } | |
fcdd4585 SM |
2357 | else if (from == stop) |
2358 | break; | |
195d1361 RS |
2359 | } |
2360 | if (found == 0) | |
2361 | { | |
2362 | from = ini; /* Set point to ini + 1. */ | |
6a140a74 | 2363 | from_byte = ini_byte; |
195d1361 RS |
2364 | goto leave; |
2365 | } | |
39c41ad4 SM |
2366 | else |
2367 | /* We have skipped one comment. */ | |
2368 | break; | |
195d1361 RS |
2369 | } |
2370 | else if (code == Sendcomment) | |
b3cfe0c8 | 2371 | { |
95ff8dfc | 2372 | found = back_comment (from, from_byte, stop, comnested, comstyle, |
6a140a74 | 2373 | &out_charpos, &out_bytepos); |
1aa963c8 SM |
2374 | if (found == -1) |
2375 | { | |
abf8a9ff SM |
2376 | if (c == '\n') |
2377 | /* This end-of-line is not an end-of-comment. | |
2378 | Treat it like a whitespace. | |
2379 | CC-mode (and maybe others) relies on this behavior. */ | |
2380 | ; | |
2381 | else | |
2382 | { | |
2383 | /* Failure: we should go back to the end of this | |
2384 | not-quite-endcomment. */ | |
2385 | if (SYNTAX(c) != code) | |
2386 | /* It was a two-char Sendcomment. */ | |
2387 | INC_BOTH (from, from_byte); | |
2388 | goto leave; | |
2389 | } | |
1aa963c8 | 2390 | } |
2b927d02 | 2391 | else |
abf8a9ff SM |
2392 | { |
2393 | /* We have skipped one comment. */ | |
2394 | from = out_charpos, from_byte = out_bytepos; | |
2395 | break; | |
2396 | } | |
b3cfe0c8 | 2397 | } |
bb0de084 | 2398 | else if (code != Swhitespace || quoted) |
b3cfe0c8 | 2399 | { |
195d1361 | 2400 | leave: |
b3cfe0c8 | 2401 | immediate_quit = 0; |
6a140a74 | 2402 | INC_BOTH (from, from_byte); |
ef316cf0 | 2403 | SET_PT_BOTH (from, from_byte); |
b3cfe0c8 RS |
2404 | return Qnil; |
2405 | } | |
2406 | } | |
2407 | ||
840f481c | 2408 | count1++; |
b3cfe0c8 RS |
2409 | } |
2410 | ||
ef316cf0 | 2411 | SET_PT_BOTH (from, from_byte); |
b3cfe0c8 RS |
2412 | immediate_quit = 0; |
2413 | return Qt; | |
2414 | } | |
2415 | \f | |
b7dbcc19 | 2416 | /* Return syntax code of character C if C is an ASCII character |
db9f48c3 | 2417 | or `multibyte_symbol_p' is zero. Otherwise, return Ssymbol. */ |
bd25db08 | 2418 | |
b7dbcc19 KH |
2419 | #define SYNTAX_WITH_MULTIBYTE_CHECK(c) \ |
2420 | ((ASCII_CHAR_P (c) || !multibyte_symbol_p) \ | |
bd25db08 KH |
2421 | ? SYNTAX (c) : Ssymbol) |
2422 | ||
6a140a74 | 2423 | static Lisp_Object |
8489eb67 | 2424 | scan_lists (from, count, depth, sexpflag) |
932e6895 SM |
2425 | register EMACS_INT from; |
2426 | EMACS_INT count, depth; | |
2427 | int sexpflag; | |
8489eb67 RS |
2428 | { |
2429 | Lisp_Object val; | |
932e6895 | 2430 | register EMACS_INT stop = count > 0 ? ZV : BEGV; |
93da5fff KH |
2431 | register int c, c1; |
2432 | int stringterm; | |
8489eb67 RS |
2433 | int quoted; |
2434 | int mathexit = 0; | |
93da5fff | 2435 | register enum syntaxcode code, temp_code; |
195d1361 | 2436 | int min_depth = depth; /* Err out if depth gets less than this. */ |
e5d4f4dc | 2437 | int comstyle = 0; /* style of comment encountered */ |
95ff8dfc | 2438 | int comnested = 0; /* whether the comment is nestable or not */ |
932e6895 SM |
2439 | EMACS_INT temp_pos; |
2440 | EMACS_INT last_good = from; | |
195d1361 | 2441 | int found; |
932e6895 SM |
2442 | EMACS_INT from_byte; |
2443 | EMACS_INT out_bytepos, out_charpos; | |
95ff8dfc | 2444 | int temp, dummy; |
bd25db08 | 2445 | int multibyte_symbol_p = sexpflag && multibyte_syntax_as_symbol; |
8489eb67 RS |
2446 | |
2447 | if (depth > 0) min_depth = 0; | |
2448 | ||
c3907a7e RS |
2449 | if (from > ZV) from = ZV; |
2450 | if (from < BEGV) from = BEGV; | |
2451 | ||
2452 | from_byte = CHAR_TO_BYTE (from); | |
2453 | ||
8489eb67 RS |
2454 | immediate_quit = 1; |
2455 | QUIT; | |
2456 | ||
195d1361 | 2457 | SETUP_SYNTAX_TABLE (from, count); |
8489eb67 RS |
2458 | while (count > 0) |
2459 | { | |
8489eb67 RS |
2460 | while (from < stop) |
2461 | { | |
f902a008 | 2462 | int comstart_first, prefix; |
195d1361 | 2463 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 2464 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2465 | code = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
f902a008 | 2466 | comstart_first = SYNTAX_COMSTART_FIRST (c); |
95ff8dfc | 2467 | comnested = SYNTAX_COMMENT_NESTED (c); |
e6365855 | 2468 | comstyle = SYNTAX_COMMENT_STYLE (c); |
f902a008 | 2469 | prefix = SYNTAX_PREFIX (c); |
7bf5e9e4 RS |
2470 | if (depth == min_depth) |
2471 | last_good = from; | |
6a140a74 | 2472 | INC_BOTH (from, from_byte); |
195d1361 | 2473 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
f902a008 | 2474 | if (from < stop && comstart_first |
327719ee MB |
2475 | && (c = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
2476 | SYNTAX_COMSTART_SECOND (c)) | |
8489eb67 | 2477 | && parse_sexp_ignore_comments) |
e5d4f4dc | 2478 | { |
7d0393cf | 2479 | /* we have encountered a comment start sequence and we |
195d1361 | 2480 | are ignoring all text inside comments. We must record |
e5d4f4dc RS |
2481 | the comment style this sequence begins so that later, |
2482 | only a comment end of the same style actually ends | |
2483 | the comment section */ | |
2484 | code = Scomment; | |
b7dbcc19 | 2485 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2486 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
2487 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
6a140a74 | 2488 | INC_BOTH (from, from_byte); |
f902a008 | 2489 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
e5d4f4dc | 2490 | } |
7d0393cf | 2491 | |
f902a008 | 2492 | if (prefix) |
8489eb67 RS |
2493 | continue; |
2494 | ||
0220c518 | 2495 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
2496 | { |
2497 | case Sescape: | |
2498 | case Scharquote: | |
8b8bf8e6 SM |
2499 | if (from == stop) |
2500 | goto lose; | |
6a140a74 | 2501 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2502 | /* treat following character as a word constituent */ |
2503 | case Sword: | |
2504 | case Ssymbol: | |
2505 | if (depth || !sexpflag) break; | |
195d1361 | 2506 | /* This word counts as a sexp; return at end of it. */ |
8489eb67 RS |
2507 | while (from < stop) |
2508 | { | |
195d1361 | 2509 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
7c7ca3d2 RS |
2510 | |
2511 | /* Some compilers can't handle this inside the switch. */ | |
b7dbcc19 | 2512 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2513 | temp = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7c7ca3d2 | 2514 | switch (temp) |
8489eb67 RS |
2515 | { |
2516 | case Scharquote: | |
2517 | case Sescape: | |
6a140a74 | 2518 | INC_BOTH (from, from_byte); |
8b8bf8e6 SM |
2519 | if (from == stop) |
2520 | goto lose; | |
8489eb67 RS |
2521 | break; |
2522 | case Sword: | |
2523 | case Ssymbol: | |
2524 | case Squote: | |
2525 | break; | |
2526 | default: | |
2527 | goto done; | |
2528 | } | |
6a140a74 | 2529 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2530 | } |
2531 | goto done; | |
2532 | ||
195d1361 | 2533 | case Scomment_fence: |
95ff8dfc RS |
2534 | comstyle = ST_COMMENT_STYLE; |
2535 | /* FALLTHROUGH */ | |
2536 | case Scomment: | |
8489eb67 | 2537 | if (!parse_sexp_ignore_comments) break; |
95ff8dfc RS |
2538 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
2539 | found = forw_comment (from, from_byte, stop, | |
2540 | comnested, comstyle, 0, | |
2541 | &out_charpos, &out_bytepos, &dummy); | |
2542 | from = out_charpos, from_byte = out_bytepos; | |
2543 | if (!found) | |
8489eb67 | 2544 | { |
95ff8dfc RS |
2545 | if (depth == 0) |
2546 | goto done; | |
2547 | goto lose; | |
8489eb67 | 2548 | } |
95ff8dfc RS |
2549 | INC_BOTH (from, from_byte); |
2550 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
8489eb67 RS |
2551 | break; |
2552 | ||
2553 | case Smath: | |
2554 | if (!sexpflag) | |
2555 | break; | |
b7dbcc19 | 2556 | if (from != stop && c == FETCH_CHAR_AS_MULTIBYTE (from_byte)) |
6a140a74 RS |
2557 | { |
2558 | INC_BOTH (from, from_byte); | |
2559 | } | |
8489eb67 RS |
2560 | if (mathexit) |
2561 | { | |
2562 | mathexit = 0; | |
2563 | goto close1; | |
2564 | } | |
2565 | mathexit = 1; | |
2566 | ||
2567 | case Sopen: | |
2568 | if (!++depth) goto done; | |
2569 | break; | |
2570 | ||
2571 | case Sclose: | |
2572 | close1: | |
2573 | if (!--depth) goto done; | |
2574 | if (depth < min_depth) | |
9471945b KS |
2575 | xsignal3 (Qscan_error, |
2576 | build_string ("Containing expression ends prematurely"), | |
2577 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2578 | break; |
2579 | ||
2580 | case Sstring: | |
195d1361 | 2581 | case Sstring_fence: |
ef316cf0 | 2582 | temp_pos = dec_bytepos (from_byte); |
b7dbcc19 | 2583 | stringterm = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
8489eb67 RS |
2584 | while (1) |
2585 | { | |
8b8bf8e6 SM |
2586 | if (from >= stop) |
2587 | goto lose; | |
195d1361 | 2588 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 2589 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
42ebfa31 SM |
2590 | if (code == Sstring |
2591 | ? (c == stringterm | |
2592 | && SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring) | |
bd25db08 | 2593 | : SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring_fence) |
195d1361 | 2594 | break; |
7c7ca3d2 RS |
2595 | |
2596 | /* Some compilers can't handle this inside the switch. */ | |
bd25db08 | 2597 | temp = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7c7ca3d2 | 2598 | switch (temp) |
8489eb67 RS |
2599 | { |
2600 | case Scharquote: | |
2601 | case Sescape: | |
6a140a74 | 2602 | INC_BOTH (from, from_byte); |
8489eb67 | 2603 | } |
6a140a74 | 2604 | INC_BOTH (from, from_byte); |
8489eb67 | 2605 | } |
6a140a74 | 2606 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2607 | if (!depth && sexpflag) goto done; |
2608 | break; | |
240c43e8 SM |
2609 | default: |
2610 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
2611 | break; | |
8489eb67 RS |
2612 | } |
2613 | } | |
2614 | ||
2615 | /* Reached end of buffer. Error if within object, return nil if between */ | |
8b8bf8e6 SM |
2616 | if (depth) |
2617 | goto lose; | |
8489eb67 RS |
2618 | |
2619 | immediate_quit = 0; | |
2620 | return Qnil; | |
2621 | ||
2622 | /* End of object reached */ | |
2623 | done: | |
2624 | count--; | |
2625 | } | |
2626 | ||
2627 | while (count < 0) | |
2628 | { | |
8489eb67 RS |
2629 | while (from > stop) |
2630 | { | |
6a140a74 | 2631 | DEC_BOTH (from, from_byte); |
195d1361 | 2632 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 2633 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2634 | code = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7bf5e9e4 RS |
2635 | if (depth == min_depth) |
2636 | last_good = from; | |
7fc8191e | 2637 | comstyle = 0; |
95ff8dfc | 2638 | comnested = SYNTAX_COMMENT_NESTED (c); |
7fc8191e RS |
2639 | if (code == Sendcomment) |
2640 | comstyle = SYNTAX_COMMENT_STYLE (c); | |
8489eb67 | 2641 | if (from > stop && SYNTAX_COMEND_SECOND (c) |
1674d9a2 | 2642 | && prev_char_comend_first (from, from_byte) |
8489eb67 | 2643 | && parse_sexp_ignore_comments) |
e5d4f4dc | 2644 | { |
f902a008 | 2645 | /* We must record the comment style encountered so that |
e5d4f4dc | 2646 | later, we can match only the proper comment begin |
f902a008 | 2647 | sequence of the same style. */ |
6a140a74 | 2648 | DEC_BOTH (from, from_byte); |
f902a008 RS |
2649 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
2650 | code = Sendcomment; | |
b7dbcc19 | 2651 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2652 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
2653 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
e5d4f4dc | 2654 | } |
7d0393cf | 2655 | |
9828a477 | 2656 | /* Quoting turns anything except a comment-ender |
e6365855 | 2657 | into a word character. Note that this cannot be true |
f902a008 | 2658 | if we decremented FROM in the if-statement above. */ |
9828a477 | 2659 | if (code != Sendcomment && char_quoted (from, from_byte)) |
240c43e8 SM |
2660 | { |
2661 | DEC_BOTH (from, from_byte); | |
2662 | code = Sword; | |
2663 | } | |
9828a477 | 2664 | else if (SYNTAX_PREFIX (c)) |
8489eb67 RS |
2665 | continue; |
2666 | ||
9828a477 | 2667 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
2668 | { |
2669 | case Sword: | |
2670 | case Ssymbol: | |
9828a477 KH |
2671 | case Sescape: |
2672 | case Scharquote: | |
8489eb67 | 2673 | if (depth || !sexpflag) break; |
195d1361 RS |
2674 | /* This word counts as a sexp; count object finished |
2675 | after passing it. */ | |
8489eb67 RS |
2676 | while (from > stop) |
2677 | { | |
6a140a74 | 2678 | temp_pos = from_byte; |
ef316cf0 RS |
2679 | if (! NILP (current_buffer->enable_multibyte_characters)) |
2680 | DEC_POS (temp_pos); | |
2681 | else | |
2682 | temp_pos--; | |
6a140a74 | 2683 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
b7dbcc19 | 2684 | c1 = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
bd25db08 | 2685 | temp_code = SYNTAX_WITH_MULTIBYTE_CHECK (c1); |
9828a477 KH |
2686 | /* Don't allow comment-end to be quoted. */ |
2687 | if (temp_code == Sendcomment) | |
2688 | goto done2; | |
6a140a74 | 2689 | quoted = char_quoted (from - 1, temp_pos); |
8489eb67 | 2690 | if (quoted) |
93da5fff | 2691 | { |
6a140a74 | 2692 | DEC_BOTH (from, from_byte); |
ef316cf0 | 2693 | temp_pos = dec_bytepos (temp_pos); |
6a140a74 | 2694 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
93da5fff | 2695 | } |
b7dbcc19 | 2696 | c1 = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
bd25db08 | 2697 | temp_code = SYNTAX_WITH_MULTIBYTE_CHECK (c1); |
93da5fff KH |
2698 | if (! (quoted || temp_code == Sword |
2699 | || temp_code == Ssymbol | |
2700 | || temp_code == Squote)) | |
8489eb67 | 2701 | goto done2; |
6a140a74 | 2702 | DEC_BOTH (from, from_byte); |
8489eb67 RS |
2703 | } |
2704 | goto done2; | |
2705 | ||
2706 | case Smath: | |
2707 | if (!sexpflag) | |
2708 | break; | |
ef316cf0 | 2709 | temp_pos = dec_bytepos (from_byte); |
6a140a74 | 2710 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
b7dbcc19 | 2711 | if (from != stop && c == FETCH_CHAR_AS_MULTIBYTE (temp_pos)) |
6a140a74 | 2712 | DEC_BOTH (from, from_byte); |
8489eb67 RS |
2713 | if (mathexit) |
2714 | { | |
2715 | mathexit = 0; | |
2716 | goto open2; | |
2717 | } | |
2718 | mathexit = 1; | |
2719 | ||
2720 | case Sclose: | |
2721 | if (!++depth) goto done2; | |
2722 | break; | |
2723 | ||
2724 | case Sopen: | |
2725 | open2: | |
2726 | if (!--depth) goto done2; | |
2727 | if (depth < min_depth) | |
9471945b KS |
2728 | xsignal3 (Qscan_error, |
2729 | build_string ("Containing expression ends prematurely"), | |
2730 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2731 | break; |
2732 | ||
2733 | case Sendcomment: | |
2734 | if (!parse_sexp_ignore_comments) | |
2735 | break; | |
95ff8dfc | 2736 | found = back_comment (from, from_byte, stop, comnested, comstyle, |
6a140a74 | 2737 | &out_charpos, &out_bytepos); |
1aa963c8 SM |
2738 | /* FIXME: if found == -1, then it really wasn't a comment-end. |
2739 | For single-char Sendcomment, we can't do much about it apart | |
2740 | from skipping the char. | |
2741 | For 2-char endcomments, we could try again, taking both | |
2742 | chars as separate entities, but it's a lot of trouble | |
2743 | for very little gain, so we don't bother either. -sm */ | |
6a140a74 RS |
2744 | if (found != -1) |
2745 | from = out_charpos, from_byte = out_bytepos; | |
8489eb67 RS |
2746 | break; |
2747 | ||
195d1361 RS |
2748 | case Scomment_fence: |
2749 | case Sstring_fence: | |
2750 | while (1) | |
2751 | { | |
8b8bf8e6 SM |
2752 | if (from == stop) |
2753 | goto lose; | |
6b61353c | 2754 | DEC_BOTH (from, from_byte); |
195d1361 | 2755 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
7d0393cf | 2756 | if (!char_quoted (from, from_byte) |
b7dbcc19 | 2757 | && (c = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
bd25db08 | 2758 | SYNTAX_WITH_MULTIBYTE_CHECK (c) == code)) |
195d1361 RS |
2759 | break; |
2760 | } | |
2761 | if (code == Sstring_fence && !depth && sexpflag) goto done2; | |
2762 | break; | |
7d0393cf | 2763 | |
8489eb67 | 2764 | case Sstring: |
b7dbcc19 | 2765 | stringterm = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8489eb67 RS |
2766 | while (1) |
2767 | { | |
8b8bf8e6 SM |
2768 | if (from == stop) |
2769 | goto lose; | |
6b61353c KH |
2770 | DEC_BOTH (from, from_byte); |
2771 | UPDATE_SYNTAX_TABLE_BACKWARD (from); | |
2772 | if (!char_quoted (from, from_byte) | |
2773 | && (stringterm | |
2774 | == (c = FETCH_CHAR_AS_MULTIBYTE (from_byte))) | |
42ebfa31 | 2775 | && SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring) |
8489eb67 | 2776 | break; |
8489eb67 | 2777 | } |
8489eb67 RS |
2778 | if (!depth && sexpflag) goto done2; |
2779 | break; | |
240c43e8 SM |
2780 | default: |
2781 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
2782 | break; | |
8489eb67 RS |
2783 | } |
2784 | } | |
2785 | ||
2786 | /* Reached start of buffer. Error if within object, return nil if between */ | |
8b8bf8e6 SM |
2787 | if (depth) |
2788 | goto lose; | |
8489eb67 RS |
2789 | |
2790 | immediate_quit = 0; | |
2791 | return Qnil; | |
2792 | ||
2793 | done2: | |
2794 | count++; | |
2795 | } | |
2796 | ||
2797 | ||
2798 | immediate_quit = 0; | |
1e142fb7 | 2799 | XSETFASTINT (val, from); |
8489eb67 RS |
2800 | return val; |
2801 | ||
2802 | lose: | |
9471945b KS |
2803 | xsignal3 (Qscan_error, |
2804 | build_string ("Unbalanced parentheses"), | |
2805 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2806 | } |
2807 | ||
8489eb67 | 2808 | DEFUN ("scan-lists", Fscan_lists, Sscan_lists, 3, 3, 0, |
fdb82f93 PJ |
2809 | doc: /* Scan from character number FROM by COUNT lists. |
2810 | Returns the character number of the position thus found. | |
2811 | ||
2812 | If DEPTH is nonzero, paren depth begins counting from that value, | |
2813 | only places where the depth in parentheses becomes zero | |
2814 | are candidates for stopping; COUNT such places are counted. | |
2815 | Thus, a positive value for DEPTH means go out levels. | |
2816 | ||
2817 | Comments are ignored if `parse-sexp-ignore-comments' is non-nil. | |
2818 | ||
2819 | If the beginning or end of (the accessible part of) the buffer is reached | |
2820 | and the depth is wrong, an error is signaled. | |
2821 | If the depth is right but the count is not used up, nil is returned. */) | |
2822 | (from, count, depth) | |
8489eb67 RS |
2823 | Lisp_Object from, count, depth; |
2824 | { | |
b7826503 PJ |
2825 | CHECK_NUMBER (from); |
2826 | CHECK_NUMBER (count); | |
2827 | CHECK_NUMBER (depth); | |
8489eb67 RS |
2828 | |
2829 | return scan_lists (XINT (from), XINT (count), XINT (depth), 0); | |
2830 | } | |
2831 | ||
2832 | DEFUN ("scan-sexps", Fscan_sexps, Sscan_sexps, 2, 2, 0, | |
fdb82f93 PJ |
2833 | doc: /* Scan from character number FROM by COUNT balanced expressions. |
2834 | If COUNT is negative, scan backwards. | |
2835 | Returns the character number of the position thus found. | |
2836 | ||
2837 | Comments are ignored if `parse-sexp-ignore-comments' is non-nil. | |
2838 | ||
2839 | If the beginning or end of (the accessible part of) the buffer is reached | |
2840 | in the middle of a parenthetical grouping, an error is signaled. | |
2841 | If the beginning or end is reached between groupings | |
2842 | but before count is used up, nil is returned. */) | |
2843 | (from, count) | |
8489eb67 RS |
2844 | Lisp_Object from, count; |
2845 | { | |
b7826503 PJ |
2846 | CHECK_NUMBER (from); |
2847 | CHECK_NUMBER (count); | |
8489eb67 RS |
2848 | |
2849 | return scan_lists (XINT (from), XINT (count), 0, 1); | |
2850 | } | |
2851 | ||
2852 | DEFUN ("backward-prefix-chars", Fbackward_prefix_chars, Sbackward_prefix_chars, | |
fdb82f93 PJ |
2853 | 0, 0, 0, |
2854 | doc: /* Move point backward over any number of chars with prefix syntax. | |
2855 | This includes chars with "quote" or "prefix" syntax (' or p). */) | |
2856 | () | |
8489eb67 RS |
2857 | { |
2858 | int beg = BEGV; | |
6a140a74 RS |
2859 | int opoint = PT; |
2860 | int opoint_byte = PT_BYTE; | |
6ec8bbd2 | 2861 | int pos = PT; |
6a140a74 | 2862 | int pos_byte = PT_BYTE; |
93da5fff | 2863 | int c; |
93da5fff | 2864 | |
7d0393cf | 2865 | if (pos <= beg) |
195d1361 | 2866 | { |
f902a008 RS |
2867 | SET_PT_BOTH (opoint, opoint_byte); |
2868 | ||
2869 | return Qnil; | |
195d1361 | 2870 | } |
8489eb67 | 2871 | |
f902a008 RS |
2872 | SETUP_SYNTAX_TABLE (pos, -1); |
2873 | ||
6a140a74 RS |
2874 | DEC_BOTH (pos, pos_byte); |
2875 | ||
1fd3172d | 2876 | while (!char_quoted (pos, pos_byte) |
195d1361 | 2877 | /* Previous statement updates syntax table. */ |
b7dbcc19 | 2878 | && ((c = FETCH_CHAR_AS_MULTIBYTE (pos_byte), SYNTAX (c) == Squote) |
93da5fff KH |
2879 | || SYNTAX_PREFIX (c))) |
2880 | { | |
1fd3172d RS |
2881 | opoint = pos; |
2882 | opoint_byte = pos_byte; | |
2883 | ||
2884 | if (pos + 1 > beg) | |
2885 | DEC_BOTH (pos, pos_byte); | |
93da5fff | 2886 | } |
8489eb67 | 2887 | |
6a140a74 | 2888 | SET_PT_BOTH (opoint, opoint_byte); |
8489eb67 RS |
2889 | |
2890 | return Qnil; | |
2891 | } | |
2892 | \f | |
6a140a74 | 2893 | /* Parse forward from FROM / FROM_BYTE to END, |
e5d4f4dc RS |
2894 | assuming that FROM has state OLDSTATE (nil means FROM is start of function), |
2895 | and return a description of the state of the parse at END. | |
c81a3712 | 2896 | If STOPBEFORE is nonzero, stop at the start of an atom. |
644ea4df RS |
2897 | If COMMENTSTOP is 1, stop at the start of a comment. |
2898 | If COMMENTSTOP is -1, stop at the start or end of a comment, | |
2899 | after the beginning of a string, or after the end of a string. */ | |
8489eb67 | 2900 | |
340f92b5 | 2901 | static void |
6a140a74 | 2902 | scan_sexps_forward (stateptr, from, from_byte, end, targetdepth, |
c81a3712 | 2903 | stopbefore, oldstate, commentstop) |
e5d4f4dc | 2904 | struct lisp_parse_state *stateptr; |
8e2911c2 AS |
2905 | register EMACS_INT from; |
2906 | EMACS_INT from_byte, end; | |
2907 | int targetdepth, stopbefore; | |
8489eb67 | 2908 | Lisp_Object oldstate; |
c81a3712 | 2909 | int commentstop; |
8489eb67 RS |
2910 | { |
2911 | struct lisp_parse_state state; | |
2912 | ||
2913 | register enum syntaxcode code; | |
95ff8dfc RS |
2914 | int c1; |
2915 | int comnested; | |
8489eb67 RS |
2916 | struct level { int last, prev; }; |
2917 | struct level levelstart[100]; | |
2918 | register struct level *curlevel = levelstart; | |
2919 | struct level *endlevel = levelstart + 100; | |
8489eb67 RS |
2920 | register int depth; /* Paren depth of current scanning location. |
2921 | level - levelstart equals this except | |
2922 | when the depth becomes negative. */ | |
2923 | int mindepth; /* Lowest DEPTH value seen. */ | |
2924 | int start_quoted = 0; /* Nonzero means starting after a char quote */ | |
2925 | Lisp_Object tem; | |
8e2911c2 AS |
2926 | EMACS_INT prev_from; /* Keep one character before FROM. */ |
2927 | EMACS_INT prev_from_byte; | |
1fd3172d | 2928 | int prev_from_syntax; |
195d1361 RS |
2929 | int boundary_stop = commentstop == -1; |
2930 | int nofence; | |
95ff8dfc | 2931 | int found; |
60c86a83 | 2932 | EMACS_INT out_bytepos, out_charpos; |
7c7ca3d2 | 2933 | int temp; |
93da5fff KH |
2934 | |
2935 | prev_from = from; | |
6a140a74 RS |
2936 | prev_from_byte = from_byte; |
2937 | if (from != BEGV) | |
2938 | DEC_BOTH (prev_from, prev_from_byte); | |
93da5fff KH |
2939 | |
2940 | /* Use this macro instead of `from++'. */ | |
6a140a74 RS |
2941 | #define INC_FROM \ |
2942 | do { prev_from = from; \ | |
2943 | prev_from_byte = from_byte; \ | |
327719ee | 2944 | temp = FETCH_CHAR_AS_MULTIBYTE (prev_from_byte); \ |
ab229fdd | 2945 | prev_from_syntax = SYNTAX_WITH_FLAGS (temp); \ |
ef316cf0 | 2946 | INC_BOTH (from, from_byte); \ |
3ceb4629 SM |
2947 | if (from < end) \ |
2948 | UPDATE_SYNTAX_TABLE_FORWARD (from); \ | |
6a140a74 | 2949 | } while (0) |
8489eb67 RS |
2950 | |
2951 | immediate_quit = 1; | |
2952 | QUIT; | |
2953 | ||
265a9e55 | 2954 | if (NILP (oldstate)) |
8489eb67 RS |
2955 | { |
2956 | depth = 0; | |
2957 | state.instring = -1; | |
2958 | state.incomment = 0; | |
195d1361 RS |
2959 | state.comstyle = 0; /* comment style a by default. */ |
2960 | state.comstr_start = -1; /* no comment/string seen. */ | |
8489eb67 RS |
2961 | } |
2962 | else | |
2963 | { | |
2964 | tem = Fcar (oldstate); | |
265a9e55 | 2965 | if (!NILP (tem)) |
8489eb67 RS |
2966 | depth = XINT (tem); |
2967 | else | |
2968 | depth = 0; | |
2969 | ||
2970 | oldstate = Fcdr (oldstate); | |
2971 | oldstate = Fcdr (oldstate); | |
2972 | oldstate = Fcdr (oldstate); | |
2973 | tem = Fcar (oldstate); | |
195d1361 | 2974 | /* Check whether we are inside string_fence-style string: */ |
7d0393cf JB |
2975 | state.instring = (!NILP (tem) |
2976 | ? (INTEGERP (tem) ? XINT (tem) : ST_STRING_STYLE) | |
e76f1c44 | 2977 | : -1); |
8489eb67 RS |
2978 | |
2979 | oldstate = Fcdr (oldstate); | |
2980 | tem = Fcar (oldstate); | |
e76f1c44 GM |
2981 | state.incomment = (!NILP (tem) |
2982 | ? (INTEGERP (tem) ? XINT (tem) : -1) | |
95ff8dfc | 2983 | : 0); |
8489eb67 RS |
2984 | |
2985 | oldstate = Fcdr (oldstate); | |
2986 | tem = Fcar (oldstate); | |
265a9e55 | 2987 | start_quoted = !NILP (tem); |
e5d4f4dc | 2988 | |
95ff8dfc | 2989 | /* if the eighth element of the list is nil, we are in comment |
195d1361 RS |
2990 | style a. If it is non-nil, we are in comment style b */ |
2991 | oldstate = Fcdr (oldstate); | |
e5d4f4dc | 2992 | oldstate = Fcdr (oldstate); |
195d1361 | 2993 | tem = Fcar (oldstate); |
7d0393cf | 2994 | state.comstyle = NILP (tem) ? 0 : (EQ (tem, Qsyntax_table) |
e76f1c44 | 2995 | ? ST_COMMENT_STYLE : 1); |
195d1361 | 2996 | |
e5d4f4dc | 2997 | oldstate = Fcdr (oldstate); |
e5d4f4dc | 2998 | tem = Fcar (oldstate); |
195d1361 | 2999 | state.comstr_start = NILP (tem) ? -1 : XINT (tem) ; |
1a102a58 RS |
3000 | oldstate = Fcdr (oldstate); |
3001 | tem = Fcar (oldstate); | |
3002 | while (!NILP (tem)) /* >= second enclosing sexps. */ | |
3003 | { | |
3004 | /* curlevel++->last ran into compiler bug on Apollo */ | |
3005 | curlevel->last = XINT (Fcar (tem)); | |
3006 | if (++curlevel == endlevel) | |
02010917 | 3007 | curlevel--; /* error ("Nesting too deep for parser"); */ |
1a102a58 RS |
3008 | curlevel->prev = -1; |
3009 | curlevel->last = -1; | |
3010 | tem = Fcdr (tem); | |
3011 | } | |
8489eb67 RS |
3012 | } |
3013 | state.quoted = 0; | |
3014 | mindepth = depth; | |
3015 | ||
3016 | curlevel->prev = -1; | |
3017 | curlevel->last = -1; | |
3018 | ||
326b283b | 3019 | SETUP_SYNTAX_TABLE (prev_from, 1); |
ab229fdd AS |
3020 | temp = FETCH_CHAR (prev_from_byte); |
3021 | prev_from_syntax = SYNTAX_WITH_FLAGS (temp); | |
3ceb4629 | 3022 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
326b283b | 3023 | |
195d1361 | 3024 | /* Enter the loop at a place appropriate for initial state. */ |
8489eb67 | 3025 | |
326b283b RS |
3026 | if (state.incomment) |
3027 | goto startincomment; | |
8489eb67 RS |
3028 | if (state.instring >= 0) |
3029 | { | |
195d1361 | 3030 | nofence = state.instring != ST_STRING_STYLE; |
326b283b RS |
3031 | if (start_quoted) |
3032 | goto startquotedinstring; | |
8489eb67 RS |
3033 | goto startinstring; |
3034 | } | |
326b283b RS |
3035 | else if (start_quoted) |
3036 | goto startquoted; | |
1fd3172d | 3037 | |
8489eb67 RS |
3038 | while (from < end) |
3039 | { | |
93da5fff | 3040 | INC_FROM; |
1fd3172d | 3041 | code = prev_from_syntax & 0xff; |
4c920633 | 3042 | |
c5d4b96f SM |
3043 | if (from < end |
3044 | && SYNTAX_FLAGS_COMSTART_FIRST (prev_from_syntax) | |
3045 | && (c1 = FETCH_CHAR (from_byte), | |
3046 | SYNTAX_COMSTART_SECOND (c1))) | |
3047 | /* Duplicate code to avoid a complex if-expression | |
3048 | which causes trouble for the SGI compiler. */ | |
95ff8dfc | 3049 | { |
c5d4b96f SM |
3050 | /* Record the comment style we have entered so that only |
3051 | the comment-end sequence of the same style actually | |
3052 | terminates the comment section. */ | |
3053 | state.comstyle = SYNTAX_COMMENT_STYLE (c1); | |
3054 | comnested = SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax); | |
3055 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
3056 | state.incomment = comnested ? 1 : -1; | |
95ff8dfc | 3057 | state.comstr_start = prev_from; |
c5d4b96f SM |
3058 | INC_FROM; |
3059 | code = Scomment; | |
95ff8dfc | 3060 | } |
4c920633 | 3061 | else if (code == Scomment_fence) |
e5d4f4dc RS |
3062 | { |
3063 | /* Record the comment style we have entered so that only | |
3064 | the comment-end sequence of the same style actually | |
3065 | terminates the comment section. */ | |
95ff8dfc RS |
3066 | state.comstyle = ST_COMMENT_STYLE; |
3067 | state.incomment = -1; | |
195d1361 | 3068 | state.comstr_start = prev_from; |
e5d4f4dc | 3069 | code = Scomment; |
e5d4f4dc | 3070 | } |
c5d4b96f SM |
3071 | else if (code == Scomment) |
3072 | { | |
3073 | state.comstyle = SYNTAX_FLAGS_COMMENT_STYLE (prev_from_syntax); | |
3074 | state.incomment = (SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax) ? | |
3075 | 1 : -1); | |
3076 | state.comstr_start = prev_from; | |
3077 | } | |
e5d4f4dc | 3078 | |
1fd3172d | 3079 | if (SYNTAX_FLAGS_PREFIX (prev_from_syntax)) |
8489eb67 | 3080 | continue; |
0220c518 | 3081 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
3082 | { |
3083 | case Sescape: | |
3084 | case Scharquote: | |
3085 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
93da5fff | 3086 | curlevel->last = prev_from; |
8489eb67 RS |
3087 | startquoted: |
3088 | if (from == end) goto endquoted; | |
93da5fff | 3089 | INC_FROM; |
8489eb67 RS |
3090 | goto symstarted; |
3091 | /* treat following character as a word constituent */ | |
3092 | case Sword: | |
3093 | case Ssymbol: | |
3094 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
93da5fff | 3095 | curlevel->last = prev_from; |
8489eb67 RS |
3096 | symstarted: |
3097 | while (from < end) | |
3098 | { | |
7c7ca3d2 | 3099 | /* Some compilers can't handle this inside the switch. */ |
327719ee | 3100 | temp = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
ab229fdd | 3101 | temp = SYNTAX (temp); |
7c7ca3d2 | 3102 | switch (temp) |
8489eb67 RS |
3103 | { |
3104 | case Scharquote: | |
3105 | case Sescape: | |
93da5fff | 3106 | INC_FROM; |
8489eb67 RS |
3107 | if (from == end) goto endquoted; |
3108 | break; | |
3109 | case Sword: | |
3110 | case Ssymbol: | |
3111 | case Squote: | |
3112 | break; | |
3113 | default: | |
3114 | goto symdone; | |
3115 | } | |
93da5fff | 3116 | INC_FROM; |
8489eb67 RS |
3117 | } |
3118 | symdone: | |
3119 | curlevel->prev = curlevel->last; | |
3120 | break; | |
3121 | ||
240c43e8 | 3122 | case Scomment_fence: /* Can't happen because it's handled above. */ |
8489eb67 | 3123 | case Scomment: |
195d1361 | 3124 | if (commentstop || boundary_stop) goto done; |
d355bd8a SM |
3125 | startincomment: |
3126 | /* The (from == BEGV) test was to enter the loop in the middle so | |
95ff8dfc | 3127 | that we find a 2-char comment ender even if we start in the |
e6365855 | 3128 | middle of it. We don't want to do that if we're just at the |
d355bd8a | 3129 | beginning of the comment (think of (*) ... (*)). */ |
95ff8dfc RS |
3130 | found = forw_comment (from, from_byte, end, |
3131 | state.incomment, state.comstyle, | |
e6365855 SM |
3132 | (from == BEGV || from < state.comstr_start + 3) |
3133 | ? 0 : prev_from_syntax, | |
95ff8dfc RS |
3134 | &out_charpos, &out_bytepos, &state.incomment); |
3135 | from = out_charpos; from_byte = out_bytepos; | |
3136 | /* Beware! prev_from and friends are invalid now. | |
3137 | Luckily, the `done' doesn't use them and the INC_FROM | |
3138 | sets them to a sane value without looking at them. */ | |
3139 | if (!found) goto done; | |
d355bd8a | 3140 | INC_FROM; |
8489eb67 | 3141 | state.incomment = 0; |
e5d4f4dc | 3142 | state.comstyle = 0; /* reset the comment style */ |
195d1361 | 3143 | if (boundary_stop) goto done; |
8489eb67 RS |
3144 | break; |
3145 | ||
3146 | case Sopen: | |
3147 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
3148 | depth++; | |
3149 | /* curlevel++->last ran into compiler bug on Apollo */ | |
93da5fff | 3150 | curlevel->last = prev_from; |
8489eb67 | 3151 | if (++curlevel == endlevel) |
02010917 | 3152 | curlevel--; /* error ("Nesting too deep for parser"); */ |
8489eb67 RS |
3153 | curlevel->prev = -1; |
3154 | curlevel->last = -1; | |
30844415 | 3155 | if (targetdepth == depth) goto done; |
8489eb67 RS |
3156 | break; |
3157 | ||
3158 | case Sclose: | |
3159 | depth--; | |
3160 | if (depth < mindepth) | |
3161 | mindepth = depth; | |
3162 | if (curlevel != levelstart) | |
3163 | curlevel--; | |
3164 | curlevel->prev = curlevel->last; | |
30844415 | 3165 | if (targetdepth == depth) goto done; |
8489eb67 RS |
3166 | break; |
3167 | ||
3168 | case Sstring: | |
195d1361 RS |
3169 | case Sstring_fence: |
3170 | state.comstr_start = from - 1; | |
8489eb67 | 3171 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ |
93da5fff | 3172 | curlevel->last = prev_from; |
7d0393cf | 3173 | state.instring = (code == Sstring |
b7dbcc19 | 3174 | ? (FETCH_CHAR_AS_MULTIBYTE (prev_from_byte)) |
195d1361 RS |
3175 | : ST_STRING_STYLE); |
3176 | if (boundary_stop) goto done; | |
8489eb67 | 3177 | startinstring: |
195d1361 | 3178 | { |
644ea4df | 3179 | nofence = state.instring != ST_STRING_STYLE; |
7d0393cf | 3180 | |
644ea4df RS |
3181 | while (1) |
3182 | { | |
3183 | int c; | |
195d1361 | 3184 | |
644ea4df | 3185 | if (from >= end) goto done; |
b7dbcc19 | 3186 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
7c7ca3d2 RS |
3187 | /* Some compilers can't handle this inside the switch. */ |
3188 | temp = SYNTAX (c); | |
ccf89641 KH |
3189 | |
3190 | /* Check TEMP here so that if the char has | |
3191 | a syntax-table property which says it is NOT | |
3192 | a string character, it does not end the string. */ | |
3193 | if (nofence && c == state.instring && temp == Sstring) | |
3194 | break; | |
3195 | ||
7c7ca3d2 | 3196 | switch (temp) |
644ea4df RS |
3197 | { |
3198 | case Sstring_fence: | |
3199 | if (!nofence) goto string_end; | |
3200 | break; | |
3201 | case Scharquote: | |
3202 | case Sescape: | |
3203 | INC_FROM; | |
3204 | startquotedinstring: | |
3205 | if (from >= end) goto endquoted; | |
195d1361 | 3206 | } |
644ea4df RS |
3207 | INC_FROM; |
3208 | } | |
195d1361 RS |
3209 | } |
3210 | string_end: | |
8489eb67 RS |
3211 | state.instring = -1; |
3212 | curlevel->prev = curlevel->last; | |
93da5fff | 3213 | INC_FROM; |
195d1361 | 3214 | if (boundary_stop) goto done; |
8489eb67 RS |
3215 | break; |
3216 | ||
3217 | case Smath: | |
240c43e8 SM |
3218 | /* FIXME: We should do something with it. */ |
3219 | break; | |
3220 | default: | |
3221 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
8489eb67 RS |
3222 | break; |
3223 | } | |
3224 | } | |
3225 | goto done; | |
3226 | ||
3227 | stop: /* Here if stopping before start of sexp. */ | |
93da5fff | 3228 | from = prev_from; /* We have just fetched the char that starts it; */ |
8489eb67 RS |
3229 | goto done; /* but return the position before it. */ |
3230 | ||
3231 | endquoted: | |
3232 | state.quoted = 1; | |
3233 | done: | |
3234 | state.depth = depth; | |
3235 | state.mindepth = mindepth; | |
3236 | state.thislevelstart = curlevel->prev; | |
3237 | state.prevlevelstart | |
3238 | = (curlevel == levelstart) ? -1 : (curlevel - 1)->last; | |
3239 | state.location = from; | |
1a102a58 RS |
3240 | state.levelstarts = Qnil; |
3241 | while (--curlevel >= levelstart) | |
3242 | state.levelstarts = Fcons (make_number (curlevel->last), | |
3243 | state.levelstarts); | |
8489eb67 RS |
3244 | immediate_quit = 0; |
3245 | ||
e5d4f4dc | 3246 | *stateptr = state; |
8489eb67 RS |
3247 | } |
3248 | ||
c81a3712 | 3249 | DEFUN ("parse-partial-sexp", Fparse_partial_sexp, Sparse_partial_sexp, 2, 6, 0, |
fdb82f93 PJ |
3250 | doc: /* Parse Lisp syntax starting at FROM until TO; return status of parse at TO. |
3251 | Parsing stops at TO or when certain criteria are met; | |
3252 | point is set to where parsing stops. | |
3253 | If fifth arg OLDSTATE is omitted or nil, | |
3254 | parsing assumes that FROM is the beginning of a function. | |
0e6228bc | 3255 | Value is a list of elements describing final state of parsing: |
fdb82f93 PJ |
3256 | 0. depth in parens. |
3257 | 1. character address of start of innermost containing list; nil if none. | |
3258 | 2. character address of start of last complete sexp terminated. | |
3259 | 3. non-nil if inside a string. | |
3260 | (it is the character that will terminate the string, | |
3261 | or t if the string should be terminated by a generic string delimiter.) | |
7d0393cf | 3262 | 4. nil if outside a comment, t if inside a non-nestable comment, |
fdb82f93 PJ |
3263 | else an integer (the current comment nesting). |
3264 | 5. t if following a quote character. | |
3265 | 6. the minimum paren-depth encountered during this scan. | |
3266 | 7. t if in a comment of style b; symbol `syntax-table' if the comment | |
3267 | should be terminated by a generic comment delimiter. | |
3268 | 8. character address of start of comment or string; nil if not in one. | |
3269 | 9. Intermediate data for continuation of parsing (subject to change). | |
3270 | If third arg TARGETDEPTH is non-nil, parsing stops if the depth | |
3271 | in parentheses becomes equal to TARGETDEPTH. | |
3272 | Fourth arg STOPBEFORE non-nil means stop when come to | |
3273 | any character that starts a sexp. | |
0e6228bc | 3274 | Fifth arg OLDSTATE is a list like what this function returns. |
fdb82f93 | 3275 | It is used to initialize the state of the parse. Elements number 1, 2, 6 |
43726c05 | 3276 | and 8 are ignored. |
fdb82f93 PJ |
3277 | Sixth arg COMMENTSTOP non-nil means stop at the start of a comment. |
3278 | If it is symbol `syntax-table', stop after the start of a comment or a | |
3279 | string, or after end of a comment or a string. */) | |
3280 | (from, to, targetdepth, stopbefore, oldstate, commentstop) | |
c81a3712 | 3281 | Lisp_Object from, to, targetdepth, stopbefore, oldstate, commentstop; |
8489eb67 RS |
3282 | { |
3283 | struct lisp_parse_state state; | |
3284 | int target; | |
3285 | ||
265a9e55 | 3286 | if (!NILP (targetdepth)) |
8489eb67 | 3287 | { |
b7826503 | 3288 | CHECK_NUMBER (targetdepth); |
8489eb67 RS |
3289 | target = XINT (targetdepth); |
3290 | } | |
3291 | else | |
3292 | target = -100000; /* We won't reach this depth */ | |
3293 | ||
3294 | validate_region (&from, &to); | |
6a140a74 RS |
3295 | scan_sexps_forward (&state, XINT (from), CHAR_TO_BYTE (XINT (from)), |
3296 | XINT (to), | |
c81a3712 | 3297 | target, !NILP (stopbefore), oldstate, |
7d0393cf | 3298 | (NILP (commentstop) |
195d1361 | 3299 | ? 0 : (EQ (commentstop, Qsyntax_table) ? -1 : 1))); |
8489eb67 RS |
3300 | |
3301 | SET_PT (state.location); | |
7d0393cf | 3302 | |
8489eb67 RS |
3303 | return Fcons (make_number (state.depth), |
3304 | Fcons (state.prevlevelstart < 0 ? Qnil : make_number (state.prevlevelstart), | |
3305 | Fcons (state.thislevelstart < 0 ? Qnil : make_number (state.thislevelstart), | |
7d0393cf JB |
3306 | Fcons (state.instring >= 0 |
3307 | ? (state.instring == ST_STRING_STYLE | |
195d1361 | 3308 | ? Qt : make_number (state.instring)) : Qnil, |
95ff8dfc RS |
3309 | Fcons (state.incomment < 0 ? Qt : |
3310 | (state.incomment == 0 ? Qnil : | |
3311 | make_number (state.incomment)), | |
8489eb67 | 3312 | Fcons (state.quoted ? Qt : Qnil, |
bff37d2a | 3313 | Fcons (make_number (state.mindepth), |
7d0393cf | 3314 | Fcons ((state.comstyle |
bff37d2a RS |
3315 | ? (state.comstyle == ST_COMMENT_STYLE |
3316 | ? Qsyntax_table : Qt) : | |
3317 | Qnil), | |
0beaf54f DL |
3318 | Fcons (((state.incomment |
3319 | || (state.instring >= 0)) | |
bff37d2a RS |
3320 | ? make_number (state.comstr_start) |
3321 | : Qnil), | |
1a102a58 | 3322 | Fcons (state.levelstarts, Qnil)))))))))); |
8489eb67 RS |
3323 | } |
3324 | \f | |
dfcf069d | 3325 | void |
8489eb67 RS |
3326 | init_syntax_once () |
3327 | { | |
78f9a1f7 | 3328 | register int i, c; |
8ea151b2 | 3329 | Lisp_Object temp; |
8489eb67 | 3330 | |
5ebaddf5 | 3331 | /* This has to be done here, before we call Fmake_char_table. */ |
d67b4f80 | 3332 | Qsyntax_table = intern_c_string ("syntax-table"); |
5ebaddf5 RS |
3333 | staticpro (&Qsyntax_table); |
3334 | ||
d67b4f80 | 3335 | /* Intern_C_String this now in case it isn't already done. |
5ebaddf5 RS |
3336 | Setting this variable twice is harmless. |
3337 | But don't staticpro it here--that is done in alloc.c. */ | |
d67b4f80 | 3338 | Qchar_table_extra_slots = intern_c_string ("char-table-extra-slots"); |
5ebaddf5 | 3339 | |
93da5fff | 3340 | /* Create objects which can be shared among syntax tables. */ |
42ebfa31 | 3341 | Vsyntax_code_object = Fmake_vector (make_number (Smax), Qnil); |
93da5fff KH |
3342 | for (i = 0; i < XVECTOR (Vsyntax_code_object)->size; i++) |
3343 | XVECTOR (Vsyntax_code_object)->contents[i] | |
3344 | = Fcons (make_number (i), Qnil); | |
3345 | ||
5ebaddf5 RS |
3346 | /* Now we are ready to set up this property, so we can |
3347 | create syntax tables. */ | |
3348 | Fput (Qsyntax_table, Qchar_table_extra_slots, make_number (0)); | |
3349 | ||
93da5fff | 3350 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Swhitespace]; |
8489eb67 | 3351 | |
5ebaddf5 | 3352 | Vstandard_syntax_table = Fmake_char_table (Qsyntax_table, temp); |
8489eb67 | 3353 | |
aa7b08b4 RS |
3354 | /* Control characters should not be whitespace. */ |
3355 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Spunct]; | |
3356 | for (i = 0; i <= ' ' - 1; i++) | |
3357 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); | |
3358 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 0177, temp); | |
3359 | ||
3360 | /* Except that a few really are whitespace. */ | |
3361 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Swhitespace]; | |
3362 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ' ', temp); | |
3363 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\t', temp); | |
3364 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\n', temp); | |
3365 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 015, temp); | |
3366 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 014, temp); | |
3367 | ||
93da5fff | 3368 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Sword]; |
8489eb67 | 3369 | for (i = 'a'; i <= 'z'; i++) |
8ea151b2 | 3370 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
8489eb67 | 3371 | for (i = 'A'; i <= 'Z'; i++) |
8ea151b2 | 3372 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
8489eb67 | 3373 | for (i = '0'; i <= '9'; i++) |
8ea151b2 RS |
3374 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
3375 | ||
3376 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '$', temp); | |
3377 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '%', temp); | |
3378 | ||
3379 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '(', | |
3380 | Fcons (make_number (Sopen), make_number (')'))); | |
3381 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ')', | |
3382 | Fcons (make_number (Sclose), make_number ('('))); | |
3383 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '[', | |
3384 | Fcons (make_number (Sopen), make_number (']'))); | |
3385 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ']', | |
3386 | Fcons (make_number (Sclose), make_number ('['))); | |
3387 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '{', | |
3388 | Fcons (make_number (Sopen), make_number ('}'))); | |
3389 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '}', | |
3390 | Fcons (make_number (Sclose), make_number ('{'))); | |
3391 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '"', | |
3392 | Fcons (make_number ((int) Sstring), Qnil)); | |
3393 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\\', | |
3394 | Fcons (make_number ((int) Sescape), Qnil)); | |
3395 | ||
93da5fff | 3396 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Ssymbol]; |
8489eb67 | 3397 | for (i = 0; i < 10; i++) |
78f9a1f7 KH |
3398 | { |
3399 | c = "_-+*/&|<>="[i]; | |
3400 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, c, temp); | |
3401 | } | |
8489eb67 | 3402 | |
93da5fff | 3403 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Spunct]; |
8489eb67 | 3404 | for (i = 0; i < 12; i++) |
78f9a1f7 KH |
3405 | { |
3406 | c = ".,;:?!#@~^'`"[i]; | |
3407 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, c, temp); | |
3408 | } | |
bd25db08 KH |
3409 | |
3410 | /* All multibyte characters have syntax `word' by default. */ | |
3411 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Sword]; | |
5c7b02ab | 3412 | char_table_set_range (Vstandard_syntax_table, 0x80, MAX_CHAR, temp); |
8489eb67 RS |
3413 | } |
3414 | ||
dfcf069d | 3415 | void |
8489eb67 RS |
3416 | syms_of_syntax () |
3417 | { | |
d67b4f80 | 3418 | Qsyntax_table_p = intern_c_string ("syntax-table-p"); |
8489eb67 RS |
3419 | staticpro (&Qsyntax_table_p); |
3420 | ||
93da5fff KH |
3421 | staticpro (&Vsyntax_code_object); |
3422 | ||
a2994c46 KS |
3423 | staticpro (&gl_state.object); |
3424 | staticpro (&gl_state.global_code); | |
3425 | staticpro (&gl_state.current_syntax_table); | |
3426 | staticpro (&gl_state.old_prop); | |
3427 | ||
3428 | /* Defined in regex.c */ | |
3429 | staticpro (&re_match_object); | |
3430 | ||
d67b4f80 | 3431 | Qscan_error = intern_c_string ("scan-error"); |
7bf5e9e4 RS |
3432 | staticpro (&Qscan_error); |
3433 | Fput (Qscan_error, Qerror_conditions, | |
d67b4f80 | 3434 | pure_cons (Qscan_error, pure_cons (Qerror, Qnil))); |
7bf5e9e4 | 3435 | Fput (Qscan_error, Qerror_message, |
d67b4f80 | 3436 | make_pure_c_string ("Scan error")); |
7bf5e9e4 | 3437 | |
8489eb67 | 3438 | DEFVAR_BOOL ("parse-sexp-ignore-comments", &parse_sexp_ignore_comments, |
fdb82f93 | 3439 | doc: /* Non-nil means `forward-sexp', etc., should treat comments as whitespace. */); |
8489eb67 | 3440 | |
195d1361 | 3441 | DEFVAR_BOOL ("parse-sexp-lookup-properties", &parse_sexp_lookup_properties, |
fdb82f93 PJ |
3442 | doc: /* Non-nil means `forward-sexp', etc., obey `syntax-table' property. |
3443 | Otherwise, that text property is simply ignored. | |
3444 | See the info node `(elisp)Syntax Properties' for a description of the | |
3445 | `syntax-table' property. */); | |
195d1361 | 3446 | |
8489eb67 RS |
3447 | words_include_escapes = 0; |
3448 | DEFVAR_BOOL ("words-include-escapes", &words_include_escapes, | |
fdb82f93 | 3449 | doc: /* Non-nil means `forward-word', etc., should treat escape chars part of words. */); |
8489eb67 | 3450 | |
bd25db08 | 3451 | DEFVAR_BOOL ("multibyte-syntax-as-symbol", &multibyte_syntax_as_symbol, |
fdb82f93 | 3452 | doc: /* Non-nil means `scan-sexps' treats all multibyte characters as symbol. */); |
bd25db08 KH |
3453 | multibyte_syntax_as_symbol = 0; |
3454 | ||
f4ed767f GM |
3455 | DEFVAR_BOOL ("open-paren-in-column-0-is-defun-start", |
3456 | &open_paren_in_column_0_is_defun_start, | |
b222e415 | 3457 | doc: /* *Non-nil means an open paren in column 0 denotes the start of a defun. */); |
f4ed767f GM |
3458 | open_paren_in_column_0_is_defun_start = 1; |
3459 | ||
8f924df7 KH |
3460 | |
3461 | DEFVAR_LISP ("find-word-boundary-function-table", | |
3462 | &Vfind_word_boundary_function_table, | |
869bb237 | 3463 | doc: /* |
8f924df7 | 3464 | Char table of functions to search for the word boundary. |
869bb237 KH |
3465 | Each function is called with two arguments; POS and LIMIT. |
3466 | POS and LIMIT are character positions in the current buffer. | |
3467 | ||
3468 | If POS is less than LIMIT, POS is at the first character of a word, | |
3469 | and the return value of a function is a position after the last | |
3470 | character of that word. | |
3471 | ||
3472 | If POS is not less than LIMIT, POS is at the last character of a word, | |
3473 | and the return value of a function is a position at the first | |
3474 | character of that word. | |
3475 | ||
3476 | In both cases, LIMIT bounds the search. */); | |
8f924df7 | 3477 | Vfind_word_boundary_function_table = Fmake_char_table (Qnil, Qnil); |
869bb237 | 3478 | |
8489eb67 RS |
3479 | defsubr (&Ssyntax_table_p); |
3480 | defsubr (&Ssyntax_table); | |
3481 | defsubr (&Sstandard_syntax_table); | |
3482 | defsubr (&Scopy_syntax_table); | |
3483 | defsubr (&Sset_syntax_table); | |
3484 | defsubr (&Schar_syntax); | |
beefa22e | 3485 | defsubr (&Smatching_paren); |
c65adb44 | 3486 | defsubr (&Sstring_to_syntax); |
8489eb67 | 3487 | defsubr (&Smodify_syntax_entry); |
62abe9cb | 3488 | defsubr (&Sinternal_describe_syntax_value); |
8489eb67 RS |
3489 | |
3490 | defsubr (&Sforward_word); | |
3491 | ||
195d1361 RS |
3492 | defsubr (&Sskip_chars_forward); |
3493 | defsubr (&Sskip_chars_backward); | |
3494 | defsubr (&Sskip_syntax_forward); | |
3495 | defsubr (&Sskip_syntax_backward); | |
3496 | ||
b3cfe0c8 | 3497 | defsubr (&Sforward_comment); |
8489eb67 RS |
3498 | defsubr (&Sscan_lists); |
3499 | defsubr (&Sscan_sexps); | |
3500 | defsubr (&Sbackward_prefix_chars); | |
3501 | defsubr (&Sparse_partial_sexp); | |
3502 | } | |
839966f3 KH |
3503 | |
3504 | /* arch-tag: 3e297b9f-088e-4b64-8f4c-fb0b3443e412 | |
3505 | (do not change this comment) */ |