Commit | Line | Data |
---|---|---|
8489eb67 | 1 | /* GNU Emacs routines to deal with syntax tables; also word and list parsing. |
429ab54e | 2 | Copyright (C) 1985, 1987, 1993, 1994, 1995, 1997, 1998, 1999, 2001, |
8cabe764 | 3 | 2002, 2003, 2004, 2005, 2006, 2007, 2008 |
429ab54e | 4 | Free Software Foundation, Inc. |
8489eb67 RS |
5 | |
6 | This file is part of GNU Emacs. | |
7 | ||
9ec0b715 | 8 | GNU Emacs is free software: you can redistribute it and/or modify |
8489eb67 | 9 | it under the terms of the GNU General Public License as published by |
9ec0b715 GM |
10 | the Free Software Foundation, either version 3 of the License, or |
11 | (at your option) any later version. | |
8489eb67 RS |
12 | |
13 | GNU Emacs is distributed in the hope that it will be useful, | |
14 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
15 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
16 | GNU General Public License for more details. | |
17 | ||
18 | You should have received a copy of the GNU General Public License | |
9ec0b715 | 19 | along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>. */ |
8489eb67 RS |
20 | |
21 | ||
18160b98 | 22 | #include <config.h> |
8489eb67 RS |
23 | #include <ctype.h> |
24 | #include "lisp.h" | |
25 | #include "commands.h" | |
26 | #include "buffer.h" | |
5c7b02ab | 27 | #include "character.h" |
e35f6ff7 | 28 | #include "keymap.h" |
a1bc88d4 | 29 | #include "regex.h" |
195d1361 RS |
30 | |
31 | /* Make syntax table lookup grant data in gl_state. */ | |
32 | #define SYNTAX_ENTRY_VIA_PROPERTY | |
33 | ||
8489eb67 | 34 | #include "syntax.h" |
195d1361 RS |
35 | #include "intervals.h" |
36 | ||
37 | /* We use these constants in place for comment-style and | |
38 | string-ender-char to distinguish comments/strings started by | |
39 | comment_fence and string_fence codes. */ | |
40 | ||
41 | #define ST_COMMENT_STYLE (256 + 1) | |
42 | #define ST_STRING_STYLE (256 + 2) | |
93da5fff | 43 | #include "category.h" |
8489eb67 | 44 | |
7bf5e9e4 | 45 | Lisp_Object Qsyntax_table_p, Qsyntax_table, Qscan_error; |
8489eb67 RS |
46 | |
47 | int words_include_escapes; | |
195d1361 | 48 | int parse_sexp_lookup_properties; |
8489eb67 | 49 | |
bd25db08 KH |
50 | /* Nonzero means `scan-sexps' treat all multibyte characters as symbol. */ |
51 | int multibyte_syntax_as_symbol; | |
52 | ||
8ea151b2 RS |
53 | /* Used as a temporary in SYNTAX_ENTRY and other macros in syntax.h, |
54 | if not compiled with GCC. No need to mark it, since it is used | |
55 | only very temporarily. */ | |
56 | Lisp_Object syntax_temp; | |
57 | ||
f4ed767f GM |
58 | /* Non-zero means an open parenthesis in column 0 is always considered |
59 | to be the start of a defun. Zero means an open parenthesis in | |
60 | column 0 has no special meaning. */ | |
61 | ||
62 | int open_paren_in_column_0_is_defun_start; | |
63 | ||
e5d4f4dc RS |
64 | /* This is the internal form of the parse state used in parse-partial-sexp. */ |
65 | ||
66 | struct lisp_parse_state | |
67 | { | |
6449674e SM |
68 | int depth; /* Depth at end of parsing. */ |
69 | int instring; /* -1 if not within string, else desired terminator. */ | |
70 | int incomment; /* -1 if in unnestable comment else comment nesting */ | |
71 | int comstyle; /* comment style a=0, or b=1, or ST_COMMENT_STYLE. */ | |
72 | int quoted; /* Nonzero if just after an escape char at end of parsing */ | |
73 | int mindepth; /* Minimum depth seen while scanning. */ | |
74 | /* Char number of most recent start-of-expression at current level */ | |
75 | EMACS_INT thislevelstart; | |
76 | /* Char number of start of containing expression */ | |
77 | EMACS_INT prevlevelstart; | |
78 | EMACS_INT location; /* Char number at which parsing stopped. */ | |
79 | EMACS_INT comstr_start; /* Position of last comment/string starter. */ | |
80 | Lisp_Object levelstarts; /* Char numbers of starts-of-expression | |
81 | of levels (starting from outermost). */ | |
e5d4f4dc RS |
82 | }; |
83 | \f | |
37bef230 RS |
84 | /* These variables are a cache for finding the start of a defun. |
85 | find_start_pos is the place for which the defun start was found. | |
86 | find_start_value is the defun start position found for it. | |
6a140a74 | 87 | find_start_value_byte is the corresponding byte position. |
37bef230 RS |
88 | find_start_buffer is the buffer it was found in. |
89 | find_start_begv is the BEGV value when it was found. | |
90 | find_start_modiff is the value of MODIFF when it was found. */ | |
91 | ||
8e2911c2 AS |
92 | static EMACS_INT find_start_pos; |
93 | static EMACS_INT find_start_value; | |
94 | static EMACS_INT find_start_value_byte; | |
37bef230 | 95 | static struct buffer *find_start_buffer; |
8e2911c2 | 96 | static EMACS_INT find_start_begv; |
37bef230 | 97 | static int find_start_modiff; |
6a140a74 RS |
98 | |
99 | ||
327719ee | 100 | static Lisp_Object skip_chars P_ ((int, Lisp_Object, Lisp_Object, int)); |
b7dbcc19 | 101 | static Lisp_Object skip_syntaxes P_ ((int, Lisp_Object, Lisp_Object)); |
932e6895 | 102 | static Lisp_Object scan_lists P_ ((EMACS_INT, EMACS_INT, EMACS_INT, int)); |
6a140a74 | 103 | static void scan_sexps_forward P_ ((struct lisp_parse_state *, |
8e2911c2 | 104 | EMACS_INT, EMACS_INT, EMACS_INT, int, |
6a140a74 | 105 | int, Lisp_Object, int)); |
a1bc88d4 | 106 | static int in_classes P_ ((int, Lisp_Object)); |
195d1361 RS |
107 | \f |
108 | ||
109 | struct gl_state_s gl_state; /* Global state of syntax parser. */ | |
110 | ||
111 | INTERVAL interval_of (); | |
112 | #define INTERVALS_AT_ONCE 10 /* 1 + max-number of intervals | |
113 | to scan to property-change. */ | |
114 | ||
6a140a74 RS |
115 | /* Update gl_state to an appropriate interval which contains CHARPOS. The |
116 | sign of COUNT give the relative position of CHARPOS wrt the previously | |
195d1361 | 117 | valid interval. If INIT, only [be]_property fields of gl_state are |
6a140a74 | 118 | valid at start, the rest is filled basing on OBJECT. |
195d1361 | 119 | |
6a140a74 | 120 | `gl_state.*_i' are the intervals, and CHARPOS is further in the search |
195d1361 RS |
121 | direction than the intervals - or in an interval. We update the |
122 | current syntax-table basing on the property of this interval, and | |
6a140a74 | 123 | update the interval to start further than CHARPOS - or be |
195d1361 | 124 | NULL_INTERVAL. We also update lim_property to be the next value of |
6a140a74 | 125 | charpos to call this subroutine again - or be before/after the |
195d1361 RS |
126 | start/end of OBJECT. */ |
127 | ||
128 | void | |
6a140a74 RS |
129 | update_syntax_table (charpos, count, init, object) |
130 | int charpos, count, init; | |
195d1361 RS |
131 | Lisp_Object object; |
132 | { | |
133 | Lisp_Object tmp_table; | |
4ffe723b | 134 | int cnt = 0, invalidate = 1; |
5a774522 | 135 | INTERVAL i; |
195d1361 RS |
136 | |
137 | if (init) | |
138 | { | |
bb0de084 | 139 | gl_state.old_prop = Qnil; |
195d1361 RS |
140 | gl_state.start = gl_state.b_property; |
141 | gl_state.stop = gl_state.e_property; | |
bb0de084 SM |
142 | i = interval_of (charpos, object); |
143 | gl_state.backward_i = gl_state.forward_i = i; | |
195d1361 RS |
144 | invalidate = 0; |
145 | if (NULL_INTERVAL_P (i)) | |
146 | return; | |
f902a008 | 147 | /* interval_of updates only ->position of the return value, so |
d80f4cc7 | 148 | update the parents manually to speed up update_interval. */ |
bb0de084 | 149 | while (!NULL_PARENT (i)) |
d80f4cc7 RS |
150 | { |
151 | if (AM_RIGHT_CHILD (i)) | |
439d5cb4 | 152 | INTERVAL_PARENT (i)->position = i->position |
d80f4cc7 | 153 | - LEFT_TOTAL_LENGTH (i) + TOTAL_LENGTH (i) /* right end */ |
439d5cb4 KR |
154 | - TOTAL_LENGTH (INTERVAL_PARENT (i)) |
155 | + LEFT_TOTAL_LENGTH (INTERVAL_PARENT (i)); | |
d80f4cc7 | 156 | else |
439d5cb4 | 157 | INTERVAL_PARENT (i)->position = i->position - LEFT_TOTAL_LENGTH (i) |
d80f4cc7 | 158 | + TOTAL_LENGTH (i); |
439d5cb4 | 159 | i = INTERVAL_PARENT (i); |
d80f4cc7 RS |
160 | } |
161 | i = gl_state.forward_i; | |
bb0de084 | 162 | gl_state.b_property = i->position - gl_state.offset; |
ee0cdb48 | 163 | gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset; |
195d1361 RS |
164 | goto update; |
165 | } | |
5a774522 | 166 | i = count > 0 ? gl_state.forward_i : gl_state.backward_i; |
195d1361 | 167 | |
423e705d | 168 | /* We are guaranteed to be called with CHARPOS either in i, |
6a140a74 | 169 | or further off. */ |
195d1361 RS |
170 | if (NULL_INTERVAL_P (i)) |
171 | error ("Error in syntax_table logic for to-the-end intervals"); | |
6a140a74 | 172 | else if (charpos < i->position) /* Move left. */ |
195d1361 RS |
173 | { |
174 | if (count > 0) | |
f902a008 | 175 | error ("Error in syntax_table logic for intervals <-"); |
195d1361 | 176 | /* Update the interval. */ |
6a140a74 | 177 | i = update_interval (i, charpos); |
bb0de084 | 178 | if (INTERVAL_LAST_POS (i) != gl_state.b_property) |
195d1361 RS |
179 | { |
180 | invalidate = 0; | |
195d1361 | 181 | gl_state.forward_i = i; |
ee0cdb48 | 182 | gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset; |
195d1361 | 183 | } |
423e705d | 184 | } |
6a140a74 | 185 | else if (charpos >= INTERVAL_LAST_POS (i)) /* Move right. */ |
195d1361 RS |
186 | { |
187 | if (count < 0) | |
f902a008 | 188 | error ("Error in syntax_table logic for intervals ->"); |
195d1361 | 189 | /* Update the interval. */ |
6a140a74 | 190 | i = update_interval (i, charpos); |
bb0de084 | 191 | if (i->position != gl_state.e_property) |
195d1361 RS |
192 | { |
193 | invalidate = 0; | |
195d1361 | 194 | gl_state.backward_i = i; |
bb0de084 | 195 | gl_state.b_property = i->position - gl_state.offset; |
195d1361 RS |
196 | } |
197 | } | |
195d1361 RS |
198 | |
199 | update: | |
200 | tmp_table = textget (i->plist, Qsyntax_table); | |
201 | ||
202 | if (invalidate) | |
203 | invalidate = !EQ (tmp_table, gl_state.old_prop); /* Need to invalidate? */ | |
7d0393cf | 204 | |
423e705d SM |
205 | if (invalidate) /* Did not get to adjacent interval. */ |
206 | { /* with the same table => */ | |
207 | /* invalidate the old range. */ | |
195d1361 RS |
208 | if (count > 0) |
209 | { | |
210 | gl_state.backward_i = i; | |
bb0de084 SM |
211 | gl_state.b_property = i->position - gl_state.offset; |
212 | } | |
213 | else | |
195d1361 | 214 | { |
bb0de084 | 215 | gl_state.forward_i = i; |
ee0cdb48 | 216 | gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset; |
195d1361 RS |
217 | } |
218 | } | |
37bef230 | 219 | |
bb0de084 | 220 | if (!EQ (tmp_table, gl_state.old_prop)) |
195d1361 | 221 | { |
bb0de084 SM |
222 | gl_state.current_syntax_table = tmp_table; |
223 | gl_state.old_prop = tmp_table; | |
224 | if (EQ (Fsyntax_table_p (tmp_table), Qt)) | |
225 | { | |
226 | gl_state.use_global = 0; | |
7d0393cf | 227 | } |
bb0de084 SM |
228 | else if (CONSP (tmp_table)) |
229 | { | |
230 | gl_state.use_global = 1; | |
231 | gl_state.global_code = tmp_table; | |
232 | } | |
7d0393cf | 233 | else |
bb0de084 SM |
234 | { |
235 | gl_state.use_global = 0; | |
236 | gl_state.current_syntax_table = current_buffer->syntax_table; | |
237 | } | |
195d1361 RS |
238 | } |
239 | ||
240 | while (!NULL_INTERVAL_P (i)) | |
241 | { | |
242 | if (cnt && !EQ (tmp_table, textget (i->plist, Qsyntax_table))) | |
243 | { | |
244 | if (count > 0) | |
bb0de084 SM |
245 | { |
246 | gl_state.e_property = i->position - gl_state.offset; | |
247 | gl_state.forward_i = i; | |
248 | } | |
249 | else | |
250 | { | |
5a774522 SM |
251 | gl_state.b_property |
252 | = i->position + LENGTH (i) - gl_state.offset; | |
bb0de084 SM |
253 | gl_state.backward_i = i; |
254 | } | |
255 | return; | |
195d1361 | 256 | } |
7d0393cf | 257 | else if (cnt == INTERVALS_AT_ONCE) |
195d1361 RS |
258 | { |
259 | if (count > 0) | |
bb0de084 | 260 | { |
5a774522 SM |
261 | gl_state.e_property |
262 | = i->position + LENGTH (i) - gl_state.offset | |
263 | /* e_property at EOB is not set to ZV but to ZV+1, so that | |
264 | we can do INC(from);UPDATE_SYNTAX_TABLE_FORWARD without | |
265 | having to check eob between the two. */ | |
266 | + (NULL_INTERVAL_P (next_interval (i)) ? 1 : 0); | |
bb0de084 SM |
267 | gl_state.forward_i = i; |
268 | } | |
269 | else | |
270 | { | |
271 | gl_state.b_property = i->position - gl_state.offset; | |
272 | gl_state.backward_i = i; | |
273 | } | |
274 | return; | |
195d1361 RS |
275 | } |
276 | cnt++; | |
277 | i = count > 0 ? next_interval (i) : previous_interval (i); | |
278 | } | |
bb0de084 SM |
279 | eassert (NULL_INTERVAL_P (i)); /* This property goes to the end. */ |
280 | if (count > 0) | |
281 | gl_state.e_property = gl_state.stop; | |
282 | else | |
283 | gl_state.b_property = gl_state.start; | |
195d1361 RS |
284 | } |
285 | \f | |
6a140a74 RS |
286 | /* Returns TRUE if char at CHARPOS is quoted. |
287 | Global syntax-table data should be set up already to be good at CHARPOS | |
288 | or after. On return global syntax data is good for lookup at CHARPOS. */ | |
195d1361 RS |
289 | |
290 | static int | |
6449674e | 291 | char_quoted (EMACS_INT charpos, EMACS_INT bytepos) |
195d1361 RS |
292 | { |
293 | register enum syntaxcode code; | |
6449674e | 294 | register EMACS_INT beg = BEGV; |
195d1361 | 295 | register int quoted = 0; |
6449674e | 296 | EMACS_INT orig = charpos; |
6a140a74 | 297 | |
02d8b017 | 298 | while (charpos > beg) |
195d1361 | 299 | { |
ab229fdd | 300 | int c; |
02d8b017 | 301 | DEC_BOTH (charpos, bytepos); |
ab229fdd | 302 | |
6a140a74 | 303 | UPDATE_SYNTAX_TABLE_BACKWARD (charpos); |
327719ee | 304 | c = FETCH_CHAR_AS_MULTIBYTE (bytepos); |
ab229fdd | 305 | code = SYNTAX (c); |
6a140a74 RS |
306 | if (! (code == Scharquote || code == Sescape)) |
307 | break; | |
308 | ||
6a140a74 | 309 | quoted = !quoted; |
195d1361 | 310 | } |
6a140a74 RS |
311 | |
312 | UPDATE_SYNTAX_TABLE (orig); | |
195d1361 RS |
313 | return quoted; |
314 | } | |
6a140a74 RS |
315 | |
316 | /* Return the bytepos one character after BYTEPOS. | |
317 | We assume that BYTEPOS is not at the end of the buffer. */ | |
318 | ||
6449674e | 319 | INLINE EMACS_INT |
6a140a74 | 320 | inc_bytepos (bytepos) |
6449674e | 321 | EMACS_INT bytepos; |
6a140a74 | 322 | { |
ef316cf0 RS |
323 | if (NILP (current_buffer->enable_multibyte_characters)) |
324 | return bytepos + 1; | |
325 | ||
6a140a74 RS |
326 | INC_POS (bytepos); |
327 | return bytepos; | |
328 | } | |
329 | ||
330 | /* Return the bytepos one character before BYTEPOS. | |
331 | We assume that BYTEPOS is not at the start of the buffer. */ | |
332 | ||
6449674e | 333 | INLINE EMACS_INT |
6a140a74 | 334 | dec_bytepos (bytepos) |
6449674e | 335 | EMACS_INT bytepos; |
6a140a74 | 336 | { |
ef316cf0 RS |
337 | if (NILP (current_buffer->enable_multibyte_characters)) |
338 | return bytepos - 1; | |
339 | ||
6a140a74 RS |
340 | DEC_POS (bytepos); |
341 | return bytepos; | |
342 | } | |
195d1361 | 343 | \f |
7cc80f0a RS |
344 | /* Return a defun-start position before before POS and not too far before. |
345 | It should be the last one before POS, or nearly the last. | |
346 | ||
347 | When open_paren_in_column_0_is_defun_start is nonzero, | |
1fd1cc2f | 348 | only the beginning of the buffer is treated as a defun-start. |
7cc80f0a RS |
349 | |
350 | We record the information about where the scan started | |
351 | and what its result was, so that another call in the same area | |
352 | can return the same value very quickly. | |
195d1361 RS |
353 | |
354 | There is no promise at which position the global syntax data is | |
355 | valid on return from the subroutine, so the caller should explicitly | |
356 | update the global data. */ | |
37bef230 | 357 | |
6449674e | 358 | static EMACS_INT |
6a140a74 | 359 | find_defun_start (pos, pos_byte) |
8e2911c2 | 360 | EMACS_INT pos, pos_byte; |
37bef230 | 361 | { |
8e2911c2 | 362 | EMACS_INT opoint = PT, opoint_byte = PT_BYTE; |
37bef230 | 363 | |
1fd1cc2f RS |
364 | if (!open_paren_in_column_0_is_defun_start) |
365 | { | |
366 | find_start_value_byte = BEGV_BYTE; | |
367 | return BEGV; | |
368 | } | |
369 | ||
37bef230 RS |
370 | /* Use previous finding, if it's valid and applies to this inquiry. */ |
371 | if (current_buffer == find_start_buffer | |
372 | /* Reuse the defun-start even if POS is a little farther on. | |
373 | POS might be in the next defun, but that's ok. | |
374 | Our value may not be the best possible, but will still be usable. */ | |
375 | && pos <= find_start_pos + 1000 | |
376 | && pos >= find_start_value | |
377 | && BEGV == find_start_begv | |
378 | && MODIFF == find_start_modiff) | |
379 | return find_start_value; | |
380 | ||
381 | /* Back up to start of line. */ | |
6a140a74 | 382 | scan_newline (pos, pos_byte, BEGV, BEGV_BYTE, -1, 1); |
37bef230 | 383 | |
195d1361 RS |
384 | /* We optimize syntax-table lookup for rare updates. Thus we accept |
385 | only those `^\s(' which are good in global _and_ text-property | |
386 | syntax-tables. */ | |
387 | gl_state.current_syntax_table = current_buffer->syntax_table; | |
388 | gl_state.use_global = 0; | |
1fd1cc2f | 389 | while (PT > BEGV) |
37bef230 | 390 | { |
ab229fdd AS |
391 | int c; |
392 | ||
1fd1cc2f RS |
393 | /* Open-paren at start of line means we may have found our |
394 | defun-start. */ | |
327719ee | 395 | c = FETCH_CHAR_AS_MULTIBYTE (PT_BYTE); |
ab229fdd | 396 | if (SYNTAX (c) == Sopen) |
195d1361 | 397 | { |
1fd1cc2f | 398 | SETUP_SYNTAX_TABLE (PT + 1, -1); /* Try again... */ |
327719ee | 399 | c = FETCH_CHAR_AS_MULTIBYTE (PT_BYTE); |
ab229fdd | 400 | if (SYNTAX (c) == Sopen) |
1fd1cc2f RS |
401 | break; |
402 | /* Now fallback to the default value. */ | |
403 | gl_state.current_syntax_table = current_buffer->syntax_table; | |
404 | gl_state.use_global = 0; | |
195d1361 | 405 | } |
1fd1cc2f RS |
406 | /* Move to beg of previous line. */ |
407 | scan_newline (PT, PT_BYTE, BEGV, BEGV_BYTE, -2, 1); | |
37bef230 RS |
408 | } |
409 | ||
410 | /* Record what we found, for the next try. */ | |
6a140a74 RS |
411 | find_start_value = PT; |
412 | find_start_value_byte = PT_BYTE; | |
37bef230 RS |
413 | find_start_buffer = current_buffer; |
414 | find_start_modiff = MODIFF; | |
415 | find_start_begv = BEGV; | |
416 | find_start_pos = pos; | |
417 | ||
6a140a74 RS |
418 | TEMP_SET_PT_BOTH (opoint, opoint_byte); |
419 | ||
37bef230 RS |
420 | return find_start_value; |
421 | } | |
422 | \f | |
f902a008 RS |
423 | /* Return the SYNTAX_COMEND_FIRST of the character before POS, POS_BYTE. */ |
424 | ||
425 | static int | |
426 | prev_char_comend_first (pos, pos_byte) | |
427 | int pos, pos_byte; | |
428 | { | |
429 | int c, val; | |
430 | ||
431 | DEC_BOTH (pos, pos_byte); | |
432 | UPDATE_SYNTAX_TABLE_BACKWARD (pos); | |
433 | c = FETCH_CHAR (pos_byte); | |
434 | val = SYNTAX_COMEND_FIRST (c); | |
435 | UPDATE_SYNTAX_TABLE_FORWARD (pos + 1); | |
436 | return val; | |
437 | } | |
438 | ||
439 | /* Return the SYNTAX_COMSTART_FIRST of the character before POS, POS_BYTE. */ | |
440 | ||
3f679f55 SM |
441 | /* static int |
442 | * prev_char_comstart_first (pos, pos_byte) | |
443 | * int pos, pos_byte; | |
444 | * { | |
445 | * int c, val; | |
7d0393cf | 446 | * |
3f679f55 SM |
447 | * DEC_BOTH (pos, pos_byte); |
448 | * UPDATE_SYNTAX_TABLE_BACKWARD (pos); | |
449 | * c = FETCH_CHAR (pos_byte); | |
450 | * val = SYNTAX_COMSTART_FIRST (c); | |
451 | * UPDATE_SYNTAX_TABLE_FORWARD (pos + 1); | |
452 | * return val; | |
453 | * } */ | |
f902a008 | 454 | |
6a140a74 RS |
455 | /* Checks whether charpos FROM is at the end of a comment. |
456 | FROM_BYTE is the bytepos corresponding to FROM. | |
457 | Do not move back before STOP. | |
458 | ||
459 | Return a positive value if we find a comment ending at FROM/FROM_BYTE; | |
460 | return -1 otherwise. | |
461 | ||
462 | If successful, store the charpos of the comment's beginning | |
463 | into *CHARPOS_PTR, and the bytepos into *BYTEPOS_PTR. | |
527a32d9 KH |
464 | |
465 | Global syntax data remains valid for backward search starting at | |
466 | the returned value (or at FROM, if the search was not successful). */ | |
195d1361 RS |
467 | |
468 | static int | |
95ff8dfc | 469 | back_comment (from, from_byte, stop, comnested, comstyle, charpos_ptr, bytepos_ptr) |
932e6895 | 470 | EMACS_INT from, from_byte, stop; |
95ff8dfc | 471 | int comnested, comstyle; |
932e6895 | 472 | EMACS_INT *charpos_ptr, *bytepos_ptr; |
195d1361 RS |
473 | { |
474 | /* Look back, counting the parity of string-quotes, | |
475 | and recording the comment-starters seen. | |
476 | When we reach a safe place, assume that's not in a string; | |
477 | then step the main scan to the earliest comment-starter seen | |
478 | an even number of string quotes away from the safe place. | |
479 | ||
480 | OFROM[I] is position of the earliest comment-starter seen | |
481 | which is I+2X quotes from the comment-end. | |
482 | PARITY is current parity of quotes from the comment end. */ | |
3ee5041c | 483 | int string_style = -1; /* Presumed outside of any string. */ |
195d1361 | 484 | int string_lossage = 0; |
f7c436c1 | 485 | /* Not a real lossage: indicates that we have passed a matching comment |
7d0393cf | 486 | starter plus a non-matching comment-ender, meaning that any matching |
f7c436c1 SM |
487 | comment-starter we might see later could be a false positive (hidden |
488 | inside another comment). | |
489 | Test case: { a (* b } c (* d *) */ | |
490 | int comment_lossage = 0; | |
8e2911c2 AS |
491 | EMACS_INT comment_end = from; |
492 | EMACS_INT comment_end_byte = from_byte; | |
493 | EMACS_INT comstart_pos = 0; | |
494 | EMACS_INT comstart_byte; | |
eb35b628 RS |
495 | /* Place where the containing defun starts, |
496 | or 0 if we didn't come across it yet. */ | |
8e2911c2 AS |
497 | EMACS_INT defun_start = 0; |
498 | EMACS_INT defun_start_byte = 0; | |
195d1361 | 499 | register enum syntaxcode code; |
95ff8dfc | 500 | int nesting = 1; /* current comment nesting */ |
ea315ed6 | 501 | int c; |
3f679f55 SM |
502 | int syntax = 0; |
503 | ||
504 | /* FIXME: A }} comment-ender style leads to incorrect behavior | |
505 | in the case of {{ c }}} because we ignore the last two chars which are | |
506 | assumed to be comment-enders although they aren't. */ | |
195d1361 RS |
507 | |
508 | /* At beginning of range to scan, we're outside of strings; | |
509 | that determines quote parity to the comment-end. */ | |
510 | while (from != stop) | |
511 | { | |
3f679f55 SM |
512 | int temp_byte, prev_syntax; |
513 | int com2start, com2end; | |
6a140a74 | 514 | |
195d1361 | 515 | /* Move back and examine a character. */ |
6a140a74 | 516 | DEC_BOTH (from, from_byte); |
195d1361 RS |
517 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
518 | ||
3f679f55 | 519 | prev_syntax = syntax; |
b7dbcc19 | 520 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
3f679f55 | 521 | syntax = SYNTAX_WITH_FLAGS (c); |
195d1361 RS |
522 | code = SYNTAX (c); |
523 | ||
3f679f55 SM |
524 | /* Check for 2-char comment markers. */ |
525 | com2start = (SYNTAX_FLAGS_COMSTART_FIRST (syntax) | |
526 | && SYNTAX_FLAGS_COMSTART_SECOND (prev_syntax) | |
527 | && comstyle == SYNTAX_FLAGS_COMMENT_STYLE (prev_syntax) | |
528 | && (SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax) | |
529 | || SYNTAX_FLAGS_COMMENT_NESTED (syntax)) == comnested); | |
530 | com2end = (SYNTAX_FLAGS_COMEND_FIRST (syntax) | |
531 | && SYNTAX_FLAGS_COMEND_SECOND (prev_syntax)); | |
532 | ||
533 | /* Nasty cases with overlapping 2-char comment markers: | |
534 | - snmp-mode: -- c -- foo -- c -- | |
535 | --- c -- | |
536 | ------ c -- | |
537 | - c-mode: *||* | |
538 | |* *|* *| | |
539 | |*| |* |*| | |
540 | /// */ | |
541 | ||
542 | /* If a 2-char comment sequence partly overlaps with another, | |
543 | we don't try to be clever. */ | |
544 | if (from > stop && (com2end || com2start)) | |
195d1361 | 545 | { |
3f679f55 SM |
546 | int next = from, next_byte = from_byte, next_c, next_syntax; |
547 | DEC_BOTH (next, next_byte); | |
548 | UPDATE_SYNTAX_TABLE_BACKWARD (next); | |
b7dbcc19 | 549 | next_c = FETCH_CHAR_AS_MULTIBYTE (next_byte); |
3f679f55 SM |
550 | next_syntax = SYNTAX_WITH_FLAGS (next_c); |
551 | if (((com2start || comnested) | |
552 | && SYNTAX_FLAGS_COMEND_SECOND (syntax) | |
553 | && SYNTAX_FLAGS_COMEND_FIRST (next_syntax)) | |
554 | || ((com2end || comnested) | |
555 | && SYNTAX_FLAGS_COMSTART_SECOND (syntax) | |
556 | && comstyle == SYNTAX_FLAGS_COMMENT_STYLE (syntax) | |
557 | && SYNTAX_FLAGS_COMSTART_FIRST (next_syntax))) | |
558 | goto lossage; | |
559 | /* UPDATE_SYNTAX_TABLE_FORWARD (next + 1); */ | |
f902a008 | 560 | } |
3f679f55 SM |
561 | |
562 | if (com2start && comstart_pos == 0) | |
563 | /* We're looking at a comment starter. But it might be a comment | |
564 | ender as well (see snmp-mode). The first time we see one, we | |
565 | need to consider it as a comment starter, | |
566 | and the subsequent times as a comment ender. */ | |
567 | com2end = 0; | |
568 | ||
569 | /* Turn a 2-char comment sequences into the appropriate syntax. */ | |
570 | if (com2end) | |
571 | code = Sendcomment; | |
572 | else if (com2start) | |
573 | code = Scomment; | |
574 | /* Ignore comment starters of a different style. */ | |
575 | else if (code == Scomment | |
576 | && (comstyle != SYNTAX_FLAGS_COMMENT_STYLE (syntax) | |
577 | || SYNTAX_FLAGS_COMMENT_NESTED (syntax) != comnested)) | |
1aa963c8 | 578 | continue; |
195d1361 | 579 | |
9828a477 KH |
580 | /* Ignore escaped characters, except comment-enders. */ |
581 | if (code != Sendcomment && char_quoted (from, from_byte)) | |
195d1361 RS |
582 | continue; |
583 | ||
3ee5041c | 584 | switch (code) |
195d1361 | 585 | { |
3ee5041c SM |
586 | case Sstring_fence: |
587 | case Scomment_fence: | |
588 | c = (code == Sstring_fence ? ST_STRING_STYLE : ST_COMMENT_STYLE); | |
589 | case Sstring: | |
590 | /* Track parity of quotes. */ | |
591 | if (string_style == -1) | |
592 | /* Entering a string. */ | |
593 | string_style = c; | |
594 | else if (string_style == c) | |
595 | /* Leaving the string. */ | |
596 | string_style = -1; | |
597 | else | |
598 | /* If we have two kinds of string delimiters. | |
599 | There's no way to grok this scanning backwards. */ | |
195d1361 | 600 | string_lossage = 1; |
3ee5041c | 601 | break; |
7d0393cf | 602 | |
3ee5041c SM |
603 | case Scomment: |
604 | /* We've already checked that it is the relevant comstyle. */ | |
f7c436c1 | 605 | if (string_style != -1 || comment_lossage || string_lossage) |
3ee5041c SM |
606 | /* There are odd string quotes involved, so let's be careful. |
607 | Test case in Pascal: " { " a { " } */ | |
608 | goto lossage; | |
609 | ||
f7c436c1 SM |
610 | if (!comnested) |
611 | { | |
612 | /* Record best comment-starter so far. */ | |
613 | comstart_pos = from; | |
614 | comstart_byte = from_byte; | |
615 | } | |
616 | else if (--nesting <= 0) | |
95ff8dfc RS |
617 | /* nested comments have to be balanced, so we don't need to |
618 | keep looking for earlier ones. We use here the same (slightly | |
619 | incorrect) reasoning as below: since it is followed by uniform | |
620 | paired string quotes, this comment-start has to be outside of | |
621 | strings, else the comment-end itself would be inside a string. */ | |
622 | goto done; | |
3ee5041c SM |
623 | break; |
624 | ||
02010917 | 625 | case Sendcomment: |
3f679f55 | 626 | if (SYNTAX_FLAGS_COMMENT_STYLE (syntax) == comstyle |
d070eb22 | 627 | && ((com2end && SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax)) |
3f679f55 | 628 | || SYNTAX_FLAGS_COMMENT_NESTED (syntax)) == comnested) |
02010917 SM |
629 | /* This is the same style of comment ender as ours. */ |
630 | { | |
631 | if (comnested) | |
632 | nesting++; | |
633 | else | |
634 | /* Anything before that can't count because it would match | |
635 | this comment-ender rather than ours. */ | |
636 | from = stop; /* Break out of the loop. */ | |
637 | } | |
f7c436c1 SM |
638 | else if (comstart_pos != 0 || c != '\n') |
639 | /* We're mixing comment styles here, so we'd better be careful. | |
640 | The (comstart_pos != 0 || c != '\n') check is not quite correct | |
641 | (we should just always set comment_lossage), but removing it | |
642 | would imply that any multiline comment in C would go through | |
643 | lossage, which seems overkill. | |
644 | The failure should only happen in the rare cases such as | |
645 | { (* } *) */ | |
646 | comment_lossage = 1; | |
02010917 | 647 | break; |
195d1361 | 648 | |
02010917 SM |
649 | case Sopen: |
650 | /* Assume a defun-start point is outside of strings. */ | |
651 | if (open_paren_in_column_0_is_defun_start | |
652 | && (from == stop | |
653 | || (temp_byte = dec_bytepos (from_byte), | |
654 | FETCH_CHAR (temp_byte) == '\n'))) | |
655 | { | |
656 | defun_start = from; | |
657 | defun_start_byte = from_byte; | |
658 | from = stop; /* Break out of the loop. */ | |
659 | } | |
eb35b628 | 660 | break; |
02010917 SM |
661 | |
662 | default: | |
f7c436c1 | 663 | break; |
eb35b628 | 664 | } |
195d1361 RS |
665 | } |
666 | ||
667 | if (comstart_pos == 0) | |
668 | { | |
669 | from = comment_end; | |
6a140a74 | 670 | from_byte = comment_end_byte; |
195d1361 RS |
671 | UPDATE_SYNTAX_TABLE_FORWARD (comment_end - 1); |
672 | } | |
f7c436c1 SM |
673 | /* If comstart_pos is set and we get here (ie. didn't jump to `lossage' |
674 | or `done'), then we've found the beginning of the non-nested comment. */ | |
675 | else if (1) /* !comnested */ | |
195d1361 RS |
676 | { |
677 | from = comstart_pos; | |
6a140a74 | 678 | from_byte = comstart_byte; |
789f3320 | 679 | UPDATE_SYNTAX_TABLE_FORWARD (from - 1); |
195d1361 RS |
680 | } |
681 | else | |
682 | { | |
3ee5041c SM |
683 | struct lisp_parse_state state; |
684 | lossage: | |
195d1361 RS |
685 | /* We had two kinds of string delimiters mixed up |
686 | together. Decode this going forwards. | |
02010917 | 687 | Scan fwd from a known safe place (beginning-of-defun) |
195d1361 RS |
688 | to the one in question; this records where we |
689 | last passed a comment starter. */ | |
eb35b628 RS |
690 | /* If we did not already find the defun start, find it now. */ |
691 | if (defun_start == 0) | |
692 | { | |
693 | defun_start = find_defun_start (comment_end, comment_end_byte); | |
694 | defun_start_byte = find_start_value_byte; | |
695 | } | |
02010917 | 696 | do |
195d1361 | 697 | { |
02010917 SM |
698 | scan_sexps_forward (&state, |
699 | defun_start, defun_start_byte, | |
700 | comment_end, -10000, 0, Qnil, 0); | |
701 | defun_start = comment_end; | |
702 | if (state.incomment == (comnested ? 1 : -1) | |
703 | && state.comstyle == comstyle) | |
704 | from = state.comstr_start; | |
705 | else | |
706 | { | |
707 | from = comment_end; | |
708 | if (state.incomment) | |
709 | /* If comment_end is inside some other comment, maybe ours | |
710 | is nested, so we need to try again from within the | |
711 | surrounding comment. Example: { a (* " *) */ | |
712 | { | |
713 | /* FIXME: We should advance by one or two chars. */ | |
714 | defun_start = state.comstr_start + 2; | |
715 | defun_start_byte = CHAR_TO_BYTE (defun_start); | |
716 | } | |
717 | } | |
718 | } while (defun_start < comment_end); | |
719 | ||
6a140a74 | 720 | from_byte = CHAR_TO_BYTE (from); |
195d1361 RS |
721 | UPDATE_SYNTAX_TABLE_FORWARD (from - 1); |
722 | } | |
7d0393cf | 723 | |
95ff8dfc | 724 | done: |
6a140a74 RS |
725 | *charpos_ptr = from; |
726 | *bytepos_ptr = from_byte; | |
727 | ||
1aa963c8 | 728 | return (from == comment_end) ? -1 : from; |
195d1361 RS |
729 | } |
730 | \f | |
8489eb67 | 731 | DEFUN ("syntax-table-p", Fsyntax_table_p, Ssyntax_table_p, 1, 1, 0, |
fdb82f93 PJ |
732 | doc: /* Return t if OBJECT is a syntax table. |
733 | Currently, any char-table counts as a syntax table. */) | |
734 | (object) | |
2203e1e8 | 735 | Lisp_Object object; |
8489eb67 | 736 | { |
2203e1e8 | 737 | if (CHAR_TABLE_P (object) |
e704cb4b | 738 | && EQ (XCHAR_TABLE (object)->purpose, Qsyntax_table)) |
8489eb67 RS |
739 | return Qt; |
740 | return Qnil; | |
741 | } | |
742 | ||
8ea151b2 | 743 | static void |
8489eb67 RS |
744 | check_syntax_table (obj) |
745 | Lisp_Object obj; | |
746 | { | |
47f5f6ae KS |
747 | CHECK_TYPE (CHAR_TABLE_P (obj) && EQ (XCHAR_TABLE (obj)->purpose, Qsyntax_table), |
748 | Qsyntax_table_p, obj); | |
7d0393cf | 749 | } |
8489eb67 | 750 | |
8489eb67 | 751 | DEFUN ("syntax-table", Fsyntax_table, Ssyntax_table, 0, 0, 0, |
fdb82f93 PJ |
752 | doc: /* Return the current syntax table. |
753 | This is the one specified by the current buffer. */) | |
754 | () | |
8489eb67 RS |
755 | { |
756 | return current_buffer->syntax_table; | |
757 | } | |
758 | ||
759 | DEFUN ("standard-syntax-table", Fstandard_syntax_table, | |
760 | Sstandard_syntax_table, 0, 0, 0, | |
fdb82f93 PJ |
761 | doc: /* Return the standard syntax table. |
762 | This is the one used for new buffers. */) | |
763 | () | |
8489eb67 RS |
764 | { |
765 | return Vstandard_syntax_table; | |
766 | } | |
767 | ||
768 | DEFUN ("copy-syntax-table", Fcopy_syntax_table, Scopy_syntax_table, 0, 1, 0, | |
fdb82f93 PJ |
769 | doc: /* Construct a new syntax table and return it. |
770 | It is a copy of the TABLE, which defaults to the standard syntax table. */) | |
771 | (table) | |
8489eb67 RS |
772 | Lisp_Object table; |
773 | { | |
8ea151b2 RS |
774 | Lisp_Object copy; |
775 | ||
265a9e55 | 776 | if (!NILP (table)) |
8ea151b2 RS |
777 | check_syntax_table (table); |
778 | else | |
779 | table = Vstandard_syntax_table; | |
780 | ||
781 | copy = Fcopy_sequence (table); | |
0f867324 RS |
782 | |
783 | /* Only the standard syntax table should have a default element. | |
784 | Other syntax tables should inherit from parents instead. */ | |
785 | XCHAR_TABLE (copy)->defalt = Qnil; | |
786 | ||
787 | /* Copied syntax tables should all have parents. | |
788 | If we copied one with no parent, such as the standard syntax table, | |
789 | use the standard syntax table as the copy's parent. */ | |
790 | if (NILP (XCHAR_TABLE (copy)->parent)) | |
791 | Fset_char_table_parent (copy, Vstandard_syntax_table); | |
8ea151b2 | 792 | return copy; |
8489eb67 RS |
793 | } |
794 | ||
795 | DEFUN ("set-syntax-table", Fset_syntax_table, Sset_syntax_table, 1, 1, 0, | |
fdb82f93 PJ |
796 | doc: /* Select a new syntax table for the current buffer. |
797 | One argument, a syntax table. */) | |
798 | (table) | |
8489eb67 RS |
799 | Lisp_Object table; |
800 | { | |
4e4dfa78 | 801 | int idx; |
8ea151b2 | 802 | check_syntax_table (table); |
8489eb67 RS |
803 | current_buffer->syntax_table = table; |
804 | /* Indicate that this buffer now has a specified syntax table. */ | |
f6cd0527 GM |
805 | idx = PER_BUFFER_VAR_IDX (syntax_table); |
806 | SET_PER_BUFFER_VALUE_P (current_buffer, idx, 1); | |
8489eb67 RS |
807 | return table; |
808 | } | |
809 | \f | |
810 | /* Convert a letter which signifies a syntax code | |
811 | into the code it signifies. | |
195d1361 | 812 | This is used by modify-syntax-entry, and other things. */ |
8489eb67 RS |
813 | |
814 | unsigned char syntax_spec_code[0400] = | |
815 | { 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
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, | |
195d1361 | 819 | (char) Swhitespace, (char) Scomment_fence, (char) Sstring, 0377, |
8489eb67 RS |
820 | (char) Smath, 0377, 0377, (char) Squote, |
821 | (char) Sopen, (char) Sclose, 0377, 0377, | |
822 | 0377, (char) Swhitespace, (char) Spunct, (char) Scharquote, | |
823 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
824 | 0377, 0377, 0377, 0377, | |
825 | (char) Scomment, 0377, (char) Sendcomment, 0377, | |
6cb71bf6 | 826 | (char) Sinherit, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* @, A ... */ |
8489eb67 RS |
827 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, |
828 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword, | |
829 | 0377, 0377, 0377, 0377, (char) Sescape, 0377, 0377, (char) Ssymbol, | |
830 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* `, a, ... */ | |
831 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
832 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword, | |
195d1361 | 833 | 0377, 0377, 0377, 0377, (char) Sstring_fence, 0377, 0377, 0377 |
8489eb67 RS |
834 | }; |
835 | ||
195d1361 | 836 | /* Indexed by syntax code, give the letter that describes it. */ |
8489eb67 | 837 | |
195d1361 | 838 | char syntax_code_spec[16] = |
8489eb67 | 839 | { |
195d1361 RS |
840 | ' ', '.', 'w', '_', '(', ')', '\'', '\"', '$', '\\', '/', '<', '>', '@', |
841 | '!', '|' | |
8489eb67 | 842 | }; |
93da5fff KH |
843 | |
844 | /* Indexed by syntax code, give the object (cons of syntax code and | |
845 | nil) to be stored in syntax table. Since these objects can be | |
846 | shared among syntax tables, we generate them in advance. By | |
847 | sharing objects, the function `describe-syntax' can give a more | |
848 | compact listing. */ | |
849 | static Lisp_Object Vsyntax_code_object; | |
850 | ||
8489eb67 RS |
851 | \f |
852 | DEFUN ("char-syntax", Fchar_syntax, Schar_syntax, 1, 1, 0, | |
fdb82f93 PJ |
853 | doc: /* Return the syntax code of CHARACTER, described by a character. |
854 | For example, if CHARACTER is a word constituent, | |
855 | the character `w' is returned. | |
856 | The characters that correspond to various syntax codes | |
857 | are listed in the documentation of `modify-syntax-entry'. */) | |
858 | (character) | |
2203e1e8 | 859 | Lisp_Object character; |
8489eb67 | 860 | { |
8ea151b2 | 861 | int char_int; |
195d1361 RS |
862 | gl_state.current_syntax_table = current_buffer->syntax_table; |
863 | ||
864 | gl_state.use_global = 0; | |
b7826503 | 865 | CHECK_NUMBER (character); |
2203e1e8 | 866 | char_int = XINT (character); |
8ea151b2 | 867 | return make_number (syntax_code_spec[(int) SYNTAX (char_int)]); |
beefa22e RS |
868 | } |
869 | ||
870 | DEFUN ("matching-paren", Fmatching_paren, Smatching_paren, 1, 1, 0, | |
fdb82f93 PJ |
871 | doc: /* Return the matching parenthesis of CHARACTER, or nil if none. */) |
872 | (character) | |
2203e1e8 | 873 | Lisp_Object character; |
beefa22e | 874 | { |
8ea151b2 | 875 | int char_int, code; |
195d1361 RS |
876 | gl_state.current_syntax_table = current_buffer->syntax_table; |
877 | gl_state.use_global = 0; | |
b7826503 | 878 | CHECK_NUMBER (character); |
2203e1e8 | 879 | char_int = XINT (character); |
8ea151b2 | 880 | code = SYNTAX (char_int); |
a8bd7cd8 | 881 | if (code == Sopen || code == Sclose) |
2e34157c | 882 | return SYNTAX_MATCH (char_int); |
beefa22e | 883 | return Qnil; |
8489eb67 RS |
884 | } |
885 | ||
c65adb44 | 886 | DEFUN ("string-to-syntax", Fstring_to_syntax, Sstring_to_syntax, 1, 1, 0, |
fdb82f93 PJ |
887 | doc: /* Convert a syntax specification STRING into syntax cell form. |
888 | STRING should be a string as it is allowed as argument of | |
889 | `modify-syntax-entry'. Value is the equivalent cons cell | |
bce3dc36 | 890 | \(CODE . MATCHING-CHAR) that can be used as value of a `syntax-table' |
fdb82f93 PJ |
891 | text property. */) |
892 | (string) | |
c65adb44 | 893 | Lisp_Object string; |
8489eb67 | 894 | { |
2e567bd3 | 895 | register const unsigned char *p; |
8489eb67 | 896 | register enum syntaxcode code; |
247e20a8 | 897 | int val; |
8ea151b2 | 898 | Lisp_Object match; |
8489eb67 | 899 | |
b7826503 | 900 | CHECK_STRING (string); |
8489eb67 | 901 | |
d5db4077 | 902 | p = SDATA (string); |
8489eb67 RS |
903 | code = (enum syntaxcode) syntax_spec_code[*p++]; |
904 | if (((int) code & 0377) == 0377) | |
b764a653 | 905 | error ("Invalid syntax description letter: %c", p[-1]); |
8489eb67 | 906 | |
8ea151b2 | 907 | if (code == Sinherit) |
c65adb44 | 908 | return Qnil; |
8ea151b2 RS |
909 | |
910 | if (*p) | |
d1be9f0f | 911 | { |
93da5fff | 912 | int len; |
7c19e8e1 | 913 | int character = (STRING_CHAR_AND_LENGTH |
d5db4077 | 914 | (p, SBYTES (string) - 1, 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 KH |
970 | CHAR may be a cons (MIN . MAX), in which case, syntaxes of all characters |
971 | in the range MIN and 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 PJ |
1003 | between expressions. |
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))) | |
1251 | || ! EQ (CHAR_TABLE_REF (Vchar_script_table, ch1), script)) | |
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))) | |
1304 | || ! EQ (CHAR_TABLE_REF (Vchar_script_table, ch0), script)) | |
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 `\\' | |
1356 | (but not as the end of a range; quoting is never needed there). | |
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++); | |
1558 | c = unibyte_char_to_multibyte (i); | |
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 | { | |
1565 | c = unibyte_char_to_multibyte (i); | |
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 KH |
1589 | leading_code = str[i_byte]; |
1590 | c = STRING_CHAR_AND_LENGTH (str + i_byte, size_byte-i_byte, len); | |
1591 | i_byte += len; | |
5c7b02ab | 1592 | |
327719ee MB |
1593 | if (handle_iso_classes && c == '[' |
1594 | && i_byte < size_byte | |
1595 | && STRING_CHAR (str + i_byte, size_byte - i_byte) == ':') | |
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]; |
839966f3 KH |
1635 | c = STRING_CHAR_AND_LENGTH (str + i_byte, |
1636 | size_byte - i_byte, len); | |
82d497fc | 1637 | i_byte += len; |
195d1361 | 1638 | } |
839966f3 KH |
1639 | /* Treat `-' as range character only if another character |
1640 | follows. */ | |
1641 | if (i_byte + 1 < size_byte | |
82d497fc | 1642 | && str[i_byte] == '-') |
195d1361 | 1643 | { |
9690d026 | 1644 | unsigned int c2; |
b7dbcc19 | 1645 | unsigned char leading_code2; |
4101e6fe RS |
1646 | |
1647 | /* Skip over the dash. */ | |
82d497fc | 1648 | i_byte++; |
4101e6fe | 1649 | |
4101e6fe | 1650 | /* Get the end of the range. */ |
b7dbcc19 | 1651 | leading_code2 = str[i_byte]; |
839966f3 KH |
1652 | c2 = STRING_CHAR_AND_LENGTH (str + i_byte, |
1653 | size_byte - i_byte, len); | |
82d497fc | 1654 | i_byte += len; |
7e68b0ea | 1655 | |
b7dbcc19 KH |
1656 | if (c2 == '\\' |
1657 | && i_byte < size_byte) | |
1658 | { | |
1659 | leading_code2 = str[i_byte]; | |
1660 | c2 =STRING_CHAR_AND_LENGTH (str + i_byte, size_byte-i_byte, len); | |
1661 | i_byte += len; | |
1662 | } | |
1663 | ||
f003ca54 KH |
1664 | if (c > c2) |
1665 | continue; | |
b7dbcc19 | 1666 | if (ASCII_CHAR_P (c)) |
06274af5 | 1667 | { |
5c7b02ab KH |
1668 | while (c <= c2 && c < 0x80) |
1669 | fastmap[c++] = 1; | |
b7dbcc19 KH |
1670 | leading_code = CHAR_LEADING_CODE (c); |
1671 | } | |
1672 | if (! ASCII_CHAR_P (c)) | |
1673 | { | |
1674 | while (leading_code <= leading_code2) | |
1675 | fastmap[leading_code++] = 1; | |
1676 | if (c <= c2) | |
e39091b6 | 1677 | { |
b7dbcc19 | 1678 | char_ranges[n_char_ranges++] = c; |
e39091b6 | 1679 | char_ranges[n_char_ranges++] = c2; |
06274af5 KH |
1680 | } |
1681 | } | |
195d1361 RS |
1682 | } |
1683 | else | |
7e68b0ea | 1684 | { |
b7dbcc19 | 1685 | if (ASCII_CHAR_P (c)) |
e39091b6 KH |
1686 | fastmap[c] = 1; |
1687 | else | |
7e68b0ea | 1688 | { |
b7dbcc19 | 1689 | fastmap[leading_code] = 1; |
4101e6fe RS |
1690 | char_ranges[n_char_ranges++] = c; |
1691 | char_ranges[n_char_ranges++] = c; | |
7e68b0ea RS |
1692 | } |
1693 | } | |
195d1361 | 1694 | } |
b7dbcc19 KH |
1695 | |
1696 | /* If the current range is unibyte and STRING contains non-ASCII | |
1697 | chars, arrange fastmap for the corresponding unibyte | |
1698 | chars. */ | |
1699 | ||
1700 | if (! multibyte && n_char_ranges > 0) | |
1701 | { | |
1702 | bzero (fastmap + 0200, 0200); | |
1703 | for (i = 0; i < n_char_ranges; i += 2) | |
1704 | { | |
1705 | int c1 = char_ranges[i]; | |
1706 | int c2 = char_ranges[i + 1]; | |
1707 | ||
1708 | for (; c1 <= c2; c1++) | |
2afc21f5 SM |
1709 | { |
1710 | int b = CHAR_TO_BYTE_SAFE (c1); | |
1711 | if (b >= 0) | |
1712 | fastmap[b] = 1; | |
1713 | } | |
b7dbcc19 KH |
1714 | } |
1715 | } | |
195d1361 RS |
1716 | } |
1717 | ||
9690d026 | 1718 | /* If ^ was the first character, complement the fastmap. */ |
195d1361 | 1719 | if (negate) |
b7dbcc19 KH |
1720 | { |
1721 | if (! multibyte) | |
1722 | for (i = 0; i < sizeof fastmap; i++) | |
1723 | fastmap[i] ^= 1; | |
1724 | else | |
1725 | { | |
1726 | for (i = 0; i < 0200; i++) | |
1727 | fastmap[i] ^= 1; | |
1728 | /* All non-ASCII chars possibly match. */ | |
1729 | for (; i < sizeof fastmap; i++) | |
1730 | fastmap[i] = 1; | |
1731 | } | |
1732 | } | |
195d1361 RS |
1733 | |
1734 | { | |
1735 | int start_point = PT; | |
1736 | int pos = PT; | |
6a140a74 | 1737 | int pos_byte = PT_BYTE; |
9af7511a KH |
1738 | unsigned char *p = PT_ADDR, *endp, *stop; |
1739 | ||
1740 | if (forwardp) | |
1741 | { | |
1742 | endp = (XINT (lim) == GPT) ? GPT_ADDR : CHAR_POS_ADDR (XINT (lim)); | |
d7ee9fab | 1743 | stop = (pos < GPT && GPT < XINT (lim)) ? GPT_ADDR : endp; |
9af7511a KH |
1744 | } |
1745 | else | |
1746 | { | |
1747 | endp = CHAR_POS_ADDR (XINT (lim)); | |
d7ee9fab | 1748 | stop = (pos >= GPT && GPT > XINT (lim)) ? GAP_END_ADDR : endp; |
9af7511a | 1749 | } |
195d1361 RS |
1750 | |
1751 | immediate_quit = 1; | |
b7dbcc19 | 1752 | if (forwardp) |
195d1361 | 1753 | { |
b7dbcc19 | 1754 | if (multibyte) |
8f924df7 | 1755 | while (1) |
b7dbcc19 | 1756 | { |
8f924df7 | 1757 | int nbytes; |
9af7511a | 1758 | |
8f924df7 | 1759 | if (p >= stop) |
9af7511a | 1760 | { |
8f924df7 | 1761 | if (p >= endp) |
9af7511a | 1762 | break; |
8f924df7 KH |
1763 | p = GAP_END_ADDR; |
1764 | stop = endp; | |
9af7511a | 1765 | } |
327719ee MB |
1766 | c = STRING_CHAR_AND_LENGTH (p, MAX_MULTIBYTE_LENGTH, nbytes); |
1767 | if (! NILP (iso_classes) && in_classes (c, iso_classes)) | |
9af7511a | 1768 | { |
327719ee | 1769 | if (negate) |
9af7511a | 1770 | break; |
327719ee MB |
1771 | else |
1772 | goto fwd_ok; | |
9af7511a | 1773 | } |
9af7511a | 1774 | |
8f924df7 | 1775 | if (! fastmap[*p]) |
b7dbcc19 KH |
1776 | break; |
1777 | if (! ASCII_CHAR_P (c)) | |
9af7511a | 1778 | { |
b7dbcc19 KH |
1779 | /* As we are looking at a multibyte character, we |
1780 | must look up the character in the table | |
1781 | CHAR_RANGES. If there's no data in the table, | |
1782 | that character is not what we want to skip. */ | |
1783 | ||
1784 | /* The following code do the right thing even if | |
1785 | n_char_ranges is zero (i.e. no data in | |
1786 | CHAR_RANGES). */ | |
1787 | for (i = 0; i < n_char_ranges; i += 2) | |
1788 | if (c >= char_ranges[i] && c <= char_ranges[i + 1]) | |
9af7511a | 1789 | break; |
b7dbcc19 KH |
1790 | if (!(negate ^ (i < n_char_ranges))) |
1791 | break; | |
9af7511a | 1792 | } |
327719ee | 1793 | fwd_ok: |
8f924df7 | 1794 | p += nbytes, pos++, pos_byte += nbytes; |
b7dbcc19 KH |
1795 | } |
1796 | else | |
8f924df7 KH |
1797 | while (1) |
1798 | { | |
1799 | if (p >= stop) | |
9af7511a | 1800 | { |
8f924df7 | 1801 | if (p >= endp) |
9af7511a | 1802 | break; |
8f924df7 KH |
1803 | p = GAP_END_ADDR; |
1804 | stop = endp; | |
9af7511a | 1805 | } |
327719ee MB |
1806 | |
1807 | if (!NILP (iso_classes) && in_classes (*p, iso_classes)) | |
1808 | { | |
1809 | if (negate) | |
1810 | break; | |
1811 | else | |
1812 | goto fwd_unibyte_ok; | |
1813 | } | |
1814 | ||
8f924df7 KH |
1815 | if (!fastmap[*p]) |
1816 | break; | |
327719ee | 1817 | fwd_unibyte_ok: |
8f924df7 KH |
1818 | p++, pos++, pos_byte++; |
1819 | } | |
195d1361 RS |
1820 | } |
1821 | else | |
1822 | { | |
b7dbcc19 | 1823 | if (multibyte) |
8f924df7 | 1824 | while (1) |
b7dbcc19 | 1825 | { |
8f924df7 | 1826 | unsigned char *prev_p; |
9af7511a | 1827 | |
8f924df7 | 1828 | if (p <= stop) |
9af7511a | 1829 | { |
8f924df7 | 1830 | if (p <= endp) |
9af7511a | 1831 | break; |
8f924df7 KH |
1832 | p = GPT_ADDR; |
1833 | stop = endp; | |
9af7511a | 1834 | } |
8f924df7 KH |
1835 | prev_p = p; |
1836 | while (--p >= stop && ! CHAR_HEAD_P (*p)); | |
327719ee | 1837 | c = STRING_CHAR (p, MAX_MULTIBYTE_LENGTH); |
a1bc88d4 | 1838 | |
327719ee MB |
1839 | if (! NILP (iso_classes) && in_classes (c, iso_classes)) |
1840 | { | |
1841 | if (negate) | |
1842 | break; | |
9690d026 | 1843 | else |
327719ee | 1844 | goto back_ok; |
7e68b0ea | 1845 | } |
a1bc88d4 | 1846 | |
8f924df7 | 1847 | if (! fastmap[*p]) |
b7dbcc19 KH |
1848 | break; |
1849 | if (! ASCII_CHAR_P (c)) | |
7e68b0ea | 1850 | { |
b7dbcc19 KH |
1851 | /* See the comment in the previous similar code. */ |
1852 | for (i = 0; i < n_char_ranges; i += 2) | |
1853 | if (c >= char_ranges[i] && c <= char_ranges[i + 1]) | |
1854 | break; | |
1855 | if (!(negate ^ (i < n_char_ranges))) | |
1856 | break; | |
7e68b0ea | 1857 | } |
327719ee | 1858 | back_ok: |
8f924df7 | 1859 | pos--, pos_byte -= prev_p - p; |
b7dbcc19 KH |
1860 | } |
1861 | else | |
8f924df7 KH |
1862 | while (1) |
1863 | { | |
1864 | if (p <= stop) | |
9af7511a | 1865 | { |
8f924df7 | 1866 | if (p <= endp) |
9af7511a | 1867 | break; |
8f924df7 KH |
1868 | p = GPT_ADDR; |
1869 | stop = endp; | |
9af7511a | 1870 | } |
e39091b6 | 1871 | |
327719ee MB |
1872 | if (! NILP (iso_classes) && in_classes (p[-1], iso_classes)) |
1873 | { | |
1874 | if (negate) | |
1875 | break; | |
9af7511a | 1876 | else |
327719ee MB |
1877 | goto back_unibyte_ok; |
1878 | } | |
a1bc88d4 | 1879 | |
8f924df7 KH |
1880 | if (!fastmap[p[-1]]) |
1881 | break; | |
327719ee | 1882 | back_unibyte_ok: |
8f924df7 KH |
1883 | p--, pos--, pos_byte--; |
1884 | } | |
195d1361 | 1885 | } |
7e68b0ea | 1886 | |
b7dbcc19 KH |
1887 | SET_PT_BOTH (pos, pos_byte); |
1888 | immediate_quit = 0; | |
1889 | ||
1890 | return make_number (PT - start_point); | |
1891 | } | |
1892 | } | |
7e68b0ea | 1893 | |
b7dbcc19 KH |
1894 | |
1895 | static Lisp_Object | |
1896 | skip_syntaxes (forwardp, string, lim) | |
1897 | int forwardp; | |
1898 | Lisp_Object string, lim; | |
1899 | { | |
1900 | register unsigned int c; | |
1901 | unsigned char fastmap[0400]; | |
1902 | int negate = 0; | |
1903 | register int i, i_byte; | |
1904 | int multibyte; | |
1905 | int size_byte; | |
1906 | unsigned char *str; | |
1907 | ||
1908 | CHECK_STRING (string); | |
1909 | ||
1910 | if (NILP (lim)) | |
1911 | XSETINT (lim, forwardp ? ZV : BEGV); | |
1912 | else | |
1913 | CHECK_NUMBER_COERCE_MARKER (lim); | |
1914 | ||
1915 | /* In any case, don't allow scan outside bounds of buffer. */ | |
1916 | if (XINT (lim) > ZV) | |
1917 | XSETFASTINT (lim, ZV); | |
1918 | if (XINT (lim) < BEGV) | |
1919 | XSETFASTINT (lim, BEGV); | |
1920 | ||
ade8ee9e | 1921 | if (forwardp ? (PT >= XFASTINT (lim)) : (PT <= XFASTINT (lim))) |
49feb1cd | 1922 | return make_number (0); |
8c6e735b | 1923 | |
b7dbcc19 | 1924 | multibyte = (!NILP (current_buffer->enable_multibyte_characters) |
92eaa22e | 1925 | && (XINT (lim) - PT != CHAR_TO_BYTE (XINT (lim)) - PT_BYTE)); |
b7dbcc19 KH |
1926 | |
1927 | bzero (fastmap, sizeof fastmap); | |
1928 | ||
8f924df7 KH |
1929 | if (SBYTES (string) > SCHARS (string)) |
1930 | /* As this is very rare case (syntax spec is ASCII only), don't | |
1931 | consider efficiency. */ | |
bc796a59 KH |
1932 | string = string_make_unibyte (string); |
1933 | ||
8f924df7 KH |
1934 | str = SDATA (string); |
1935 | size_byte = SBYTES (string); | |
bc796a59 | 1936 | |
b7dbcc19 KH |
1937 | i_byte = 0; |
1938 | if (i_byte < size_byte | |
8f924df7 | 1939 | && SREF (string, 0) == '^') |
b7dbcc19 KH |
1940 | { |
1941 | negate = 1; i_byte++; | |
1942 | } | |
1943 | ||
b7dbcc19 KH |
1944 | /* Find the syntaxes specified and set their elements of fastmap. */ |
1945 | ||
1946 | while (i_byte < size_byte) | |
1947 | { | |
1948 | c = str[i_byte++]; | |
8f924df7 | 1949 | fastmap[syntax_spec_code[c]] = 1; |
b7dbcc19 | 1950 | } |
195d1361 | 1951 | |
9690d026 | 1952 | /* If ^ was the first character, complement the fastmap. */ |
195d1361 RS |
1953 | if (negate) |
1954 | for (i = 0; i < sizeof fastmap; i++) | |
9690d026 | 1955 | fastmap[i] ^= 1; |
195d1361 RS |
1956 | |
1957 | { | |
1958 | int start_point = PT; | |
1959 | int pos = PT; | |
6a140a74 | 1960 | int pos_byte = PT_BYTE; |
8f924df7 KH |
1961 | unsigned char *p = PT_ADDR, *endp, *stop; |
1962 | ||
1963 | if (forwardp) | |
1964 | { | |
1965 | endp = (XINT (lim) == GPT) ? GPT_ADDR : CHAR_POS_ADDR (XINT (lim)); | |
1966 | stop = (pos < GPT && GPT < XINT (lim)) ? GPT_ADDR : endp; | |
1967 | } | |
1968 | else | |
1969 | { | |
1970 | endp = CHAR_POS_ADDR (XINT (lim)); | |
1971 | stop = (pos >= GPT && GPT > XINT (lim)) ? GAP_END_ADDR : endp; | |
1972 | } | |
195d1361 RS |
1973 | |
1974 | immediate_quit = 1; | |
b7dbcc19 KH |
1975 | SETUP_SYNTAX_TABLE (pos, forwardp ? 1 : -1); |
1976 | if (forwardp) | |
195d1361 | 1977 | { |
b7dbcc19 | 1978 | if (multibyte) |
195d1361 | 1979 | { |
8f924df7 | 1980 | while (1) |
8c6e735b | 1981 | { |
8f924df7 KH |
1982 | int nbytes; |
1983 | ||
1984 | if (p >= stop) | |
1985 | { | |
1986 | if (p >= endp) | |
1987 | break; | |
1988 | p = GAP_END_ADDR; | |
1989 | stop = endp; | |
1990 | } | |
1991 | c = STRING_CHAR_AND_LENGTH (p, MAX_MULTIBYTE_LENGTH, nbytes); | |
1992 | if (! fastmap[(int) SYNTAX (c)]) | |
8c6e735b | 1993 | break; |
8f924df7 | 1994 | p += nbytes, pos++, pos_byte += nbytes; |
8c6e735b KH |
1995 | UPDATE_SYNTAX_TABLE_FORWARD (pos); |
1996 | } | |
195d1361 RS |
1997 | } |
1998 | else | |
1999 | { | |
8c6e735b | 2000 | while (1) |
7e68b0ea | 2001 | { |
8f924df7 KH |
2002 | if (p >= stop) |
2003 | { | |
2004 | if (p >= endp) | |
2005 | break; | |
2006 | p = GAP_END_ADDR; | |
2007 | stop = endp; | |
2008 | } | |
2009 | if (! fastmap[(int) SYNTAX (*p)]) | |
8c6e735b | 2010 | break; |
8f924df7 | 2011 | p++, pos++, pos_byte++; |
b7dbcc19 | 2012 | UPDATE_SYNTAX_TABLE_FORWARD (pos); |
195d1361 RS |
2013 | } |
2014 | } | |
2015 | } | |
2016 | else | |
2017 | { | |
b7dbcc19 | 2018 | if (multibyte) |
195d1361 | 2019 | { |
8c6e735b | 2020 | while (1) |
b7dbcc19 | 2021 | { |
8f924df7 KH |
2022 | unsigned char *prev_p; |
2023 | ||
2024 | if (p <= stop) | |
b7dbcc19 | 2025 | { |
8f924df7 KH |
2026 | if (p <= endp) |
2027 | break; | |
2028 | p = GPT_ADDR; | |
2029 | stop = endp; | |
b7dbcc19 | 2030 | } |
262be72a | 2031 | UPDATE_SYNTAX_TABLE_BACKWARD (pos - 1); |
8f924df7 KH |
2032 | prev_p = p; |
2033 | while (--p >= stop && ! CHAR_HEAD_P (*p)); | |
2034 | c = STRING_CHAR (p, MAX_MULTIBYTE_LENGTH); | |
2035 | if (! fastmap[(int) SYNTAX (c)]) | |
2036 | break; | |
2037 | pos--, pos_byte -= prev_p - p; | |
b7dbcc19 | 2038 | } |
195d1361 RS |
2039 | } |
2040 | else | |
2041 | { | |
8c6e735b KH |
2042 | while (1) |
2043 | { | |
8f924df7 KH |
2044 | if (p <= stop) |
2045 | { | |
2046 | if (p <= endp) | |
2047 | break; | |
2048 | p = GPT_ADDR; | |
2049 | stop = endp; | |
2050 | } | |
102f6132 | 2051 | UPDATE_SYNTAX_TABLE_BACKWARD (pos - 1); |
8f924df7 | 2052 | if (! fastmap[(int) SYNTAX (p[-1])]) |
8c6e735b | 2053 | break; |
8f924df7 | 2054 | p--, pos--, pos_byte--; |
8c6e735b | 2055 | } |
195d1361 RS |
2056 | } |
2057 | } | |
6a140a74 RS |
2058 | |
2059 | SET_PT_BOTH (pos, pos_byte); | |
195d1361 RS |
2060 | immediate_quit = 0; |
2061 | ||
2062 | return make_number (PT - start_point); | |
2063 | } | |
2064 | } | |
a1bc88d4 RS |
2065 | |
2066 | /* Return 1 if character C belongs to one of the ISO classes | |
2067 | in the list ISO_CLASSES. Each class is represented by an | |
2068 | integer which is its type according to re_wctype. */ | |
2069 | ||
2070 | static int | |
2071 | in_classes (c, iso_classes) | |
2072 | int c; | |
2073 | Lisp_Object iso_classes; | |
2074 | { | |
2075 | int fits_class = 0; | |
2076 | ||
2077 | while (! NILP (iso_classes)) | |
2078 | { | |
2079 | Lisp_Object elt; | |
2080 | elt = XCAR (iso_classes); | |
2081 | iso_classes = XCDR (iso_classes); | |
2082 | ||
2083 | if (re_iswctype (c, XFASTINT (elt))) | |
2084 | fits_class = 1; | |
2085 | } | |
2086 | ||
2087 | return fits_class; | |
2088 | } | |
195d1361 | 2089 | \f |
95ff8dfc RS |
2090 | /* Jump over a comment, assuming we are at the beginning of one. |
2091 | FROM is the current position. | |
2092 | FROM_BYTE is the bytepos corresponding to FROM. | |
2093 | Do not move past STOP (a charpos). | |
2094 | The comment over which we have to jump is of style STYLE | |
2095 | (either SYNTAX_COMMENT_STYLE(foo) or ST_COMMENT_STYLE). | |
2096 | NESTING should be positive to indicate the nesting at the beginning | |
2097 | for nested comments and should be zero or negative else. | |
2098 | ST_COMMENT_STYLE cannot be nested. | |
2099 | PREV_SYNTAX is the SYNTAX_WITH_FLAGS of the previous character | |
2100 | (or 0 If the search cannot start in the middle of a two-character). | |
2101 | ||
2102 | If successful, return 1 and store the charpos of the comment's end | |
2103 | into *CHARPOS_PTR and the corresponding bytepos into *BYTEPOS_PTR. | |
2104 | Else, return 0 and store the charpos STOP into *CHARPOS_PTR, the | |
2105 | corresponding bytepos into *BYTEPOS_PTR and the current nesting | |
2106 | (as defined for state.incomment) in *INCOMMENT_PTR. | |
2107 | ||
2108 | The comment end is the last character of the comment rather than the | |
2109 | character just after the comment. | |
2110 | ||
2111 | Global syntax data is assumed to initially be valid for FROM and | |
2112 | remains valid for forward search starting at the returned position. */ | |
2113 | ||
2114 | static int | |
2115 | forw_comment (from, from_byte, stop, nesting, style, prev_syntax, | |
2116 | charpos_ptr, bytepos_ptr, incomment_ptr) | |
a0aa7fcf | 2117 | EMACS_INT from, from_byte, stop; |
95ff8dfc | 2118 | int nesting, style, prev_syntax; |
a0aa7fcf SM |
2119 | EMACS_INT *charpos_ptr, *bytepos_ptr; |
2120 | int *incomment_ptr; | |
95ff8dfc RS |
2121 | { |
2122 | register int c, c1; | |
2123 | register enum syntaxcode code; | |
2124 | register int syntax; | |
2125 | ||
2126 | if (nesting <= 0) nesting = -1; | |
2127 | ||
2128 | /* Enter the loop in the middle so that we find | |
2129 | a 2-char comment ender if we start in the middle of it. */ | |
2130 | syntax = prev_syntax; | |
2131 | if (syntax != 0) goto forw_incomment; | |
2132 | ||
2133 | while (1) | |
2134 | { | |
2135 | if (from == stop) | |
2136 | { | |
2137 | *incomment_ptr = nesting; | |
2138 | *charpos_ptr = from; | |
2139 | *bytepos_ptr = from_byte; | |
2140 | return 0; | |
2141 | } | |
8f924df7 | 2142 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2143 | syntax = SYNTAX_WITH_FLAGS (c); |
2144 | code = syntax & 0xff; | |
2145 | if (code == Sendcomment | |
2146 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style | |
abf8a9ff SM |
2147 | && (SYNTAX_FLAGS_COMMENT_NESTED (syntax) ? |
2148 | (nesting > 0 && --nesting == 0) : nesting < 0)) | |
95ff8dfc RS |
2149 | /* we have encountered a comment end of the same style |
2150 | as the comment sequence which began this comment | |
2151 | section */ | |
2152 | break; | |
2153 | if (code == Scomment_fence | |
2154 | && style == ST_COMMENT_STYLE) | |
2155 | /* we have encountered a comment end of the same style | |
2156 | as the comment sequence which began this comment | |
2157 | section. */ | |
2158 | break; | |
2159 | if (nesting > 0 | |
2160 | && code == Scomment | |
abf8a9ff | 2161 | && SYNTAX_FLAGS_COMMENT_NESTED (syntax) |
95ff8dfc RS |
2162 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style) |
2163 | /* we have encountered a nested comment of the same style | |
2164 | as the comment sequence which began this comment section */ | |
2165 | nesting++; | |
2166 | INC_BOTH (from, from_byte); | |
2167 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
7d0393cf | 2168 | |
95ff8dfc RS |
2169 | forw_incomment: |
2170 | if (from < stop && SYNTAX_FLAGS_COMEND_FIRST (syntax) | |
2171 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style | |
8f924df7 | 2172 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
abf8a9ff SM |
2173 | SYNTAX_COMEND_SECOND (c1)) |
2174 | && ((SYNTAX_FLAGS_COMMENT_NESTED (syntax) || | |
2175 | SYNTAX_COMMENT_NESTED (c1)) ? nesting > 0 : nesting < 0)) | |
4ffe723b GM |
2176 | { |
2177 | if (--nesting <= 0) | |
2178 | /* we have encountered a comment end of the same style | |
2179 | as the comment sequence which began this comment | |
2180 | section */ | |
2181 | break; | |
2182 | else | |
2183 | { | |
2184 | INC_BOTH (from, from_byte); | |
2185 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
2186 | } | |
2187 | } | |
95ff8dfc RS |
2188 | if (nesting > 0 |
2189 | && from < stop | |
2190 | && SYNTAX_FLAGS_COMSTART_FIRST (syntax) | |
8f924df7 | 2191 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
95ff8dfc | 2192 | SYNTAX_COMMENT_STYLE (c1) == style |
abf8a9ff SM |
2193 | && SYNTAX_COMSTART_SECOND (c1)) |
2194 | && (SYNTAX_FLAGS_COMMENT_NESTED (syntax) || | |
2195 | SYNTAX_COMMENT_NESTED (c1))) | |
95ff8dfc RS |
2196 | /* we have encountered a nested comment of the same style |
2197 | as the comment sequence which began this comment | |
2198 | section */ | |
2199 | { | |
2200 | INC_BOTH (from, from_byte); | |
2201 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
2202 | nesting++; | |
2203 | } | |
2204 | } | |
2205 | *charpos_ptr = from; | |
2206 | *bytepos_ptr = from_byte; | |
2207 | return 1; | |
2208 | } | |
2209 | ||
b3cfe0c8 | 2210 | DEFUN ("forward-comment", Fforward_comment, Sforward_comment, 1, 1, 0, |
177c0ea7 | 2211 | doc: /* |
ee648c11 | 2212 | Move forward across up to COUNT comments. If COUNT is negative, move backward. |
fdb82f93 PJ |
2213 | Stop scanning if we find something other than a comment or whitespace. |
2214 | Set point to where scanning stops. | |
ee648c11 | 2215 | If COUNT comments are found as expected, with nothing except whitespace |
fdb82f93 PJ |
2216 | between them, return t; otherwise return nil. */) |
2217 | (count) | |
840f481c | 2218 | Lisp_Object count; |
b3cfe0c8 | 2219 | { |
2312c580 SM |
2220 | register EMACS_INT from; |
2221 | EMACS_INT from_byte; | |
2222 | register EMACS_INT stop; | |
8ea151b2 | 2223 | register int c, c1; |
b3cfe0c8 RS |
2224 | register enum syntaxcode code; |
2225 | int comstyle = 0; /* style of comment encountered */ | |
95ff8dfc | 2226 | int comnested = 0; /* whether the comment is nestable or not */ |
be720845 | 2227 | int found; |
2312c580 SM |
2228 | EMACS_INT count1; |
2229 | EMACS_INT out_charpos, out_bytepos; | |
95ff8dfc | 2230 | int dummy; |
840f481c | 2231 | |
b7826503 | 2232 | CHECK_NUMBER (count); |
840f481c | 2233 | count1 = XINT (count); |
195d1361 | 2234 | stop = count1 > 0 ? ZV : BEGV; |
b3cfe0c8 RS |
2235 | |
2236 | immediate_quit = 1; | |
2237 | QUIT; | |
2238 | ||
2239 | from = PT; | |
6a140a74 | 2240 | from_byte = PT_BYTE; |
b3cfe0c8 | 2241 | |
195d1361 | 2242 | SETUP_SYNTAX_TABLE (from, count1); |
840f481c | 2243 | while (count1 > 0) |
b3cfe0c8 | 2244 | { |
04882296 | 2245 | do |
b3cfe0c8 | 2246 | { |
f902a008 RS |
2247 | int comstart_first; |
2248 | ||
04882296 KH |
2249 | if (from == stop) |
2250 | { | |
ef316cf0 | 2251 | SET_PT_BOTH (from, from_byte); |
b7e6e612 | 2252 | immediate_quit = 0; |
04882296 KH |
2253 | return Qnil; |
2254 | } | |
b7dbcc19 | 2255 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
b3cfe0c8 | 2256 | code = SYNTAX (c); |
f902a008 | 2257 | comstart_first = SYNTAX_COMSTART_FIRST (c); |
95ff8dfc | 2258 | comnested = SYNTAX_COMMENT_NESTED (c); |
e6365855 | 2259 | comstyle = SYNTAX_COMMENT_STYLE (c); |
6a140a74 | 2260 | INC_BOTH (from, from_byte); |
f902a008 | 2261 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
f902a008 | 2262 | if (from < stop && comstart_first |
b7dbcc19 | 2263 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
8ea151b2 | 2264 | SYNTAX_COMSTART_SECOND (c1))) |
b3cfe0c8 | 2265 | { |
7d0393cf | 2266 | /* We have encountered a comment start sequence and we |
7fc8191e | 2267 | are ignoring all text inside comments. We must record |
b3cfe0c8 RS |
2268 | the comment style this sequence begins so that later, |
2269 | only a comment end of the same style actually ends | |
7fc8191e | 2270 | the comment section. */ |
b3cfe0c8 | 2271 | code = Scomment; |
8ea151b2 | 2272 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
95ff8dfc | 2273 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); |
6a140a74 | 2274 | INC_BOTH (from, from_byte); |
95ff8dfc | 2275 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b3cfe0c8 | 2276 | } |
04882296 | 2277 | } |
abf8a9ff | 2278 | while (code == Swhitespace || (code == Sendcomment && c == '\n')); |
6a140a74 | 2279 | |
da71ab91 KH |
2280 | if (code == Scomment_fence) |
2281 | comstyle = ST_COMMENT_STYLE; | |
2282 | else if (code != Scomment) | |
04882296 KH |
2283 | { |
2284 | immediate_quit = 0; | |
6a140a74 | 2285 | DEC_BOTH (from, from_byte); |
ef316cf0 | 2286 | SET_PT_BOTH (from, from_byte); |
04882296 KH |
2287 | return Qnil; |
2288 | } | |
2289 | /* We're at the start of a comment. */ | |
95ff8dfc RS |
2290 | found = forw_comment (from, from_byte, stop, comnested, comstyle, 0, |
2291 | &out_charpos, &out_bytepos, &dummy); | |
2292 | from = out_charpos; from_byte = out_bytepos; | |
2293 | if (!found) | |
04882296 | 2294 | { |
95ff8dfc RS |
2295 | immediate_quit = 0; |
2296 | SET_PT_BOTH (from, from_byte); | |
2297 | return Qnil; | |
b3cfe0c8 | 2298 | } |
95ff8dfc RS |
2299 | INC_BOTH (from, from_byte); |
2300 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
04882296 | 2301 | /* We have skipped one comment. */ |
840f481c | 2302 | count1--; |
b3cfe0c8 RS |
2303 | } |
2304 | ||
840f481c | 2305 | while (count1 < 0) |
b3cfe0c8 | 2306 | { |
b9145dbb | 2307 | while (1) |
b3cfe0c8 | 2308 | { |
c65adb44 | 2309 | int quoted; |
f902a008 | 2310 | |
b9145dbb RS |
2311 | if (from <= stop) |
2312 | { | |
6a140a74 | 2313 | SET_PT_BOTH (BEGV, BEGV_BYTE); |
b9145dbb RS |
2314 | immediate_quit = 0; |
2315 | return Qnil; | |
2316 | } | |
b3cfe0c8 | 2317 | |
6a140a74 | 2318 | DEC_BOTH (from, from_byte); |
f902a008 | 2319 | /* char_quoted does UPDATE_SYNTAX_TABLE_BACKWARD (from). */ |
6a140a74 | 2320 | quoted = char_quoted (from, from_byte); |
b7dbcc19 | 2321 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
b3cfe0c8 RS |
2322 | code = SYNTAX (c); |
2323 | comstyle = 0; | |
95ff8dfc | 2324 | comnested = SYNTAX_COMMENT_NESTED (c); |
7fc8191e RS |
2325 | if (code == Sendcomment) |
2326 | comstyle = SYNTAX_COMMENT_STYLE (c); | |
b3cfe0c8 | 2327 | if (from > stop && SYNTAX_COMEND_SECOND (c) |
f902a008 | 2328 | && prev_char_comend_first (from, from_byte) |
89c6809a | 2329 | && !char_quoted (from - 1, dec_bytepos (from_byte))) |
b3cfe0c8 | 2330 | { |
7fc8191e | 2331 | /* We must record the comment style encountered so that |
b3cfe0c8 | 2332 | later, we can match only the proper comment begin |
7fc8191e | 2333 | sequence of the same style. */ |
6a140a74 | 2334 | DEC_BOTH (from, from_byte); |
f902a008 RS |
2335 | code = Sendcomment; |
2336 | /* Calling char_quoted, above, set up global syntax position | |
2337 | at the new value of FROM. */ | |
b7dbcc19 | 2338 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2339 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
2340 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
b3cfe0c8 RS |
2341 | } |
2342 | ||
195d1361 RS |
2343 | if (code == Scomment_fence) |
2344 | { | |
2345 | /* Skip until first preceding unquoted comment_fence. */ | |
6a140a74 | 2346 | int found = 0, ini = from, ini_byte = from_byte; |
7d0393cf | 2347 | |
6a140a74 | 2348 | while (1) |
195d1361 | 2349 | { |
6a140a74 | 2350 | DEC_BOTH (from, from_byte); |
195d1361 | 2351 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 2352 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
6a140a74 | 2353 | if (SYNTAX (c) == Scomment_fence |
7d0393cf | 2354 | && !char_quoted (from, from_byte)) |
195d1361 | 2355 | { |
7d0393cf | 2356 | found = 1; |
195d1361 RS |
2357 | break; |
2358 | } | |
fcdd4585 SM |
2359 | else if (from == stop) |
2360 | break; | |
195d1361 RS |
2361 | } |
2362 | if (found == 0) | |
2363 | { | |
2364 | from = ini; /* Set point to ini + 1. */ | |
6a140a74 | 2365 | from_byte = ini_byte; |
195d1361 RS |
2366 | goto leave; |
2367 | } | |
39c41ad4 SM |
2368 | else |
2369 | /* We have skipped one comment. */ | |
2370 | break; | |
195d1361 RS |
2371 | } |
2372 | else if (code == Sendcomment) | |
b3cfe0c8 | 2373 | { |
95ff8dfc | 2374 | found = back_comment (from, from_byte, stop, comnested, comstyle, |
6a140a74 | 2375 | &out_charpos, &out_bytepos); |
1aa963c8 SM |
2376 | if (found == -1) |
2377 | { | |
abf8a9ff SM |
2378 | if (c == '\n') |
2379 | /* This end-of-line is not an end-of-comment. | |
2380 | Treat it like a whitespace. | |
2381 | CC-mode (and maybe others) relies on this behavior. */ | |
2382 | ; | |
2383 | else | |
2384 | { | |
2385 | /* Failure: we should go back to the end of this | |
2386 | not-quite-endcomment. */ | |
2387 | if (SYNTAX(c) != code) | |
2388 | /* It was a two-char Sendcomment. */ | |
2389 | INC_BOTH (from, from_byte); | |
2390 | goto leave; | |
2391 | } | |
1aa963c8 | 2392 | } |
2b927d02 | 2393 | else |
abf8a9ff SM |
2394 | { |
2395 | /* We have skipped one comment. */ | |
2396 | from = out_charpos, from_byte = out_bytepos; | |
2397 | break; | |
2398 | } | |
b3cfe0c8 | 2399 | } |
bb0de084 | 2400 | else if (code != Swhitespace || quoted) |
b3cfe0c8 | 2401 | { |
195d1361 | 2402 | leave: |
b3cfe0c8 | 2403 | immediate_quit = 0; |
6a140a74 | 2404 | INC_BOTH (from, from_byte); |
ef316cf0 | 2405 | SET_PT_BOTH (from, from_byte); |
b3cfe0c8 RS |
2406 | return Qnil; |
2407 | } | |
2408 | } | |
2409 | ||
840f481c | 2410 | count1++; |
b3cfe0c8 RS |
2411 | } |
2412 | ||
ef316cf0 | 2413 | SET_PT_BOTH (from, from_byte); |
b3cfe0c8 RS |
2414 | immediate_quit = 0; |
2415 | return Qt; | |
2416 | } | |
2417 | \f | |
b7dbcc19 | 2418 | /* Return syntax code of character C if C is an ASCII character |
db9f48c3 | 2419 | or `multibyte_symbol_p' is zero. Otherwise, return Ssymbol. */ |
bd25db08 | 2420 | |
b7dbcc19 KH |
2421 | #define SYNTAX_WITH_MULTIBYTE_CHECK(c) \ |
2422 | ((ASCII_CHAR_P (c) || !multibyte_symbol_p) \ | |
bd25db08 KH |
2423 | ? SYNTAX (c) : Ssymbol) |
2424 | ||
6a140a74 | 2425 | static Lisp_Object |
8489eb67 | 2426 | scan_lists (from, count, depth, sexpflag) |
932e6895 SM |
2427 | register EMACS_INT from; |
2428 | EMACS_INT count, depth; | |
2429 | int sexpflag; | |
8489eb67 RS |
2430 | { |
2431 | Lisp_Object val; | |
932e6895 | 2432 | register EMACS_INT stop = count > 0 ? ZV : BEGV; |
93da5fff KH |
2433 | register int c, c1; |
2434 | int stringterm; | |
8489eb67 RS |
2435 | int quoted; |
2436 | int mathexit = 0; | |
93da5fff | 2437 | register enum syntaxcode code, temp_code; |
195d1361 | 2438 | int min_depth = depth; /* Err out if depth gets less than this. */ |
e5d4f4dc | 2439 | int comstyle = 0; /* style of comment encountered */ |
95ff8dfc | 2440 | int comnested = 0; /* whether the comment is nestable or not */ |
932e6895 SM |
2441 | EMACS_INT temp_pos; |
2442 | EMACS_INT last_good = from; | |
195d1361 | 2443 | int found; |
932e6895 SM |
2444 | EMACS_INT from_byte; |
2445 | EMACS_INT out_bytepos, out_charpos; | |
95ff8dfc | 2446 | int temp, dummy; |
bd25db08 | 2447 | int multibyte_symbol_p = sexpflag && multibyte_syntax_as_symbol; |
8489eb67 RS |
2448 | |
2449 | if (depth > 0) min_depth = 0; | |
2450 | ||
c3907a7e RS |
2451 | if (from > ZV) from = ZV; |
2452 | if (from < BEGV) from = BEGV; | |
2453 | ||
2454 | from_byte = CHAR_TO_BYTE (from); | |
2455 | ||
8489eb67 RS |
2456 | immediate_quit = 1; |
2457 | QUIT; | |
2458 | ||
195d1361 | 2459 | SETUP_SYNTAX_TABLE (from, count); |
8489eb67 RS |
2460 | while (count > 0) |
2461 | { | |
8489eb67 RS |
2462 | while (from < stop) |
2463 | { | |
f902a008 | 2464 | int comstart_first, prefix; |
195d1361 | 2465 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 2466 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2467 | code = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
f902a008 | 2468 | comstart_first = SYNTAX_COMSTART_FIRST (c); |
95ff8dfc | 2469 | comnested = SYNTAX_COMMENT_NESTED (c); |
e6365855 | 2470 | comstyle = SYNTAX_COMMENT_STYLE (c); |
f902a008 | 2471 | prefix = SYNTAX_PREFIX (c); |
7bf5e9e4 RS |
2472 | if (depth == min_depth) |
2473 | last_good = from; | |
6a140a74 | 2474 | INC_BOTH (from, from_byte); |
195d1361 | 2475 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
f902a008 | 2476 | if (from < stop && comstart_first |
327719ee MB |
2477 | && (c = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
2478 | SYNTAX_COMSTART_SECOND (c)) | |
8489eb67 | 2479 | && parse_sexp_ignore_comments) |
e5d4f4dc | 2480 | { |
7d0393cf | 2481 | /* we have encountered a comment start sequence and we |
195d1361 | 2482 | are ignoring all text inside comments. We must record |
e5d4f4dc RS |
2483 | the comment style this sequence begins so that later, |
2484 | only a comment end of the same style actually ends | |
2485 | the comment section */ | |
2486 | code = Scomment; | |
b7dbcc19 | 2487 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2488 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
2489 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
6a140a74 | 2490 | INC_BOTH (from, from_byte); |
f902a008 | 2491 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
e5d4f4dc | 2492 | } |
7d0393cf | 2493 | |
f902a008 | 2494 | if (prefix) |
8489eb67 RS |
2495 | continue; |
2496 | ||
0220c518 | 2497 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
2498 | { |
2499 | case Sescape: | |
2500 | case Scharquote: | |
8b8bf8e6 SM |
2501 | if (from == stop) |
2502 | goto lose; | |
6a140a74 | 2503 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2504 | /* treat following character as a word constituent */ |
2505 | case Sword: | |
2506 | case Ssymbol: | |
2507 | if (depth || !sexpflag) break; | |
195d1361 | 2508 | /* This word counts as a sexp; return at end of it. */ |
8489eb67 RS |
2509 | while (from < stop) |
2510 | { | |
195d1361 | 2511 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
7c7ca3d2 RS |
2512 | |
2513 | /* Some compilers can't handle this inside the switch. */ | |
b7dbcc19 | 2514 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2515 | temp = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7c7ca3d2 | 2516 | switch (temp) |
8489eb67 RS |
2517 | { |
2518 | case Scharquote: | |
2519 | case Sescape: | |
6a140a74 | 2520 | INC_BOTH (from, from_byte); |
8b8bf8e6 SM |
2521 | if (from == stop) |
2522 | goto lose; | |
8489eb67 RS |
2523 | break; |
2524 | case Sword: | |
2525 | case Ssymbol: | |
2526 | case Squote: | |
2527 | break; | |
2528 | default: | |
2529 | goto done; | |
2530 | } | |
6a140a74 | 2531 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2532 | } |
2533 | goto done; | |
2534 | ||
195d1361 | 2535 | case Scomment_fence: |
95ff8dfc RS |
2536 | comstyle = ST_COMMENT_STYLE; |
2537 | /* FALLTHROUGH */ | |
2538 | case Scomment: | |
8489eb67 | 2539 | if (!parse_sexp_ignore_comments) break; |
95ff8dfc RS |
2540 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
2541 | found = forw_comment (from, from_byte, stop, | |
2542 | comnested, comstyle, 0, | |
2543 | &out_charpos, &out_bytepos, &dummy); | |
2544 | from = out_charpos, from_byte = out_bytepos; | |
2545 | if (!found) | |
8489eb67 | 2546 | { |
95ff8dfc RS |
2547 | if (depth == 0) |
2548 | goto done; | |
2549 | goto lose; | |
8489eb67 | 2550 | } |
95ff8dfc RS |
2551 | INC_BOTH (from, from_byte); |
2552 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
8489eb67 RS |
2553 | break; |
2554 | ||
2555 | case Smath: | |
2556 | if (!sexpflag) | |
2557 | break; | |
b7dbcc19 | 2558 | if (from != stop && c == FETCH_CHAR_AS_MULTIBYTE (from_byte)) |
6a140a74 RS |
2559 | { |
2560 | INC_BOTH (from, from_byte); | |
2561 | } | |
8489eb67 RS |
2562 | if (mathexit) |
2563 | { | |
2564 | mathexit = 0; | |
2565 | goto close1; | |
2566 | } | |
2567 | mathexit = 1; | |
2568 | ||
2569 | case Sopen: | |
2570 | if (!++depth) goto done; | |
2571 | break; | |
2572 | ||
2573 | case Sclose: | |
2574 | close1: | |
2575 | if (!--depth) goto done; | |
2576 | if (depth < min_depth) | |
9471945b KS |
2577 | xsignal3 (Qscan_error, |
2578 | build_string ("Containing expression ends prematurely"), | |
2579 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2580 | break; |
2581 | ||
2582 | case Sstring: | |
195d1361 | 2583 | case Sstring_fence: |
ef316cf0 | 2584 | temp_pos = dec_bytepos (from_byte); |
b7dbcc19 | 2585 | stringterm = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
8489eb67 RS |
2586 | while (1) |
2587 | { | |
8b8bf8e6 SM |
2588 | if (from >= stop) |
2589 | goto lose; | |
195d1361 | 2590 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 2591 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
42ebfa31 SM |
2592 | if (code == Sstring |
2593 | ? (c == stringterm | |
2594 | && SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring) | |
bd25db08 | 2595 | : SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring_fence) |
195d1361 | 2596 | break; |
7c7ca3d2 RS |
2597 | |
2598 | /* Some compilers can't handle this inside the switch. */ | |
bd25db08 | 2599 | temp = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7c7ca3d2 | 2600 | switch (temp) |
8489eb67 RS |
2601 | { |
2602 | case Scharquote: | |
2603 | case Sescape: | |
6a140a74 | 2604 | INC_BOTH (from, from_byte); |
8489eb67 | 2605 | } |
6a140a74 | 2606 | INC_BOTH (from, from_byte); |
8489eb67 | 2607 | } |
6a140a74 | 2608 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2609 | if (!depth && sexpflag) goto done; |
2610 | break; | |
240c43e8 SM |
2611 | default: |
2612 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
2613 | break; | |
8489eb67 RS |
2614 | } |
2615 | } | |
2616 | ||
2617 | /* Reached end of buffer. Error if within object, return nil if between */ | |
8b8bf8e6 SM |
2618 | if (depth) |
2619 | goto lose; | |
8489eb67 RS |
2620 | |
2621 | immediate_quit = 0; | |
2622 | return Qnil; | |
2623 | ||
2624 | /* End of object reached */ | |
2625 | done: | |
2626 | count--; | |
2627 | } | |
2628 | ||
2629 | while (count < 0) | |
2630 | { | |
8489eb67 RS |
2631 | while (from > stop) |
2632 | { | |
6a140a74 | 2633 | DEC_BOTH (from, from_byte); |
195d1361 | 2634 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 2635 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2636 | code = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7bf5e9e4 RS |
2637 | if (depth == min_depth) |
2638 | last_good = from; | |
7fc8191e | 2639 | comstyle = 0; |
95ff8dfc | 2640 | comnested = SYNTAX_COMMENT_NESTED (c); |
7fc8191e RS |
2641 | if (code == Sendcomment) |
2642 | comstyle = SYNTAX_COMMENT_STYLE (c); | |
8489eb67 | 2643 | if (from > stop && SYNTAX_COMEND_SECOND (c) |
1674d9a2 | 2644 | && prev_char_comend_first (from, from_byte) |
8489eb67 | 2645 | && parse_sexp_ignore_comments) |
e5d4f4dc | 2646 | { |
f902a008 | 2647 | /* We must record the comment style encountered so that |
e5d4f4dc | 2648 | later, we can match only the proper comment begin |
f902a008 | 2649 | sequence of the same style. */ |
6a140a74 | 2650 | DEC_BOTH (from, from_byte); |
f902a008 RS |
2651 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
2652 | code = Sendcomment; | |
b7dbcc19 | 2653 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2654 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
2655 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
e5d4f4dc | 2656 | } |
7d0393cf | 2657 | |
9828a477 | 2658 | /* Quoting turns anything except a comment-ender |
e6365855 | 2659 | into a word character. Note that this cannot be true |
f902a008 | 2660 | if we decremented FROM in the if-statement above. */ |
9828a477 | 2661 | if (code != Sendcomment && char_quoted (from, from_byte)) |
240c43e8 SM |
2662 | { |
2663 | DEC_BOTH (from, from_byte); | |
2664 | code = Sword; | |
2665 | } | |
9828a477 | 2666 | else if (SYNTAX_PREFIX (c)) |
8489eb67 RS |
2667 | continue; |
2668 | ||
9828a477 | 2669 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
2670 | { |
2671 | case Sword: | |
2672 | case Ssymbol: | |
9828a477 KH |
2673 | case Sescape: |
2674 | case Scharquote: | |
8489eb67 | 2675 | if (depth || !sexpflag) break; |
195d1361 RS |
2676 | /* This word counts as a sexp; count object finished |
2677 | after passing it. */ | |
8489eb67 RS |
2678 | while (from > stop) |
2679 | { | |
6a140a74 | 2680 | temp_pos = from_byte; |
ef316cf0 RS |
2681 | if (! NILP (current_buffer->enable_multibyte_characters)) |
2682 | DEC_POS (temp_pos); | |
2683 | else | |
2684 | temp_pos--; | |
6a140a74 | 2685 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
b7dbcc19 | 2686 | c1 = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
bd25db08 | 2687 | temp_code = SYNTAX_WITH_MULTIBYTE_CHECK (c1); |
9828a477 KH |
2688 | /* Don't allow comment-end to be quoted. */ |
2689 | if (temp_code == Sendcomment) | |
2690 | goto done2; | |
6a140a74 | 2691 | quoted = char_quoted (from - 1, temp_pos); |
8489eb67 | 2692 | if (quoted) |
93da5fff | 2693 | { |
6a140a74 | 2694 | DEC_BOTH (from, from_byte); |
ef316cf0 | 2695 | temp_pos = dec_bytepos (temp_pos); |
6a140a74 | 2696 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
93da5fff | 2697 | } |
b7dbcc19 | 2698 | c1 = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
bd25db08 | 2699 | temp_code = SYNTAX_WITH_MULTIBYTE_CHECK (c1); |
93da5fff KH |
2700 | if (! (quoted || temp_code == Sword |
2701 | || temp_code == Ssymbol | |
2702 | || temp_code == Squote)) | |
8489eb67 | 2703 | goto done2; |
6a140a74 | 2704 | DEC_BOTH (from, from_byte); |
8489eb67 RS |
2705 | } |
2706 | goto done2; | |
2707 | ||
2708 | case Smath: | |
2709 | if (!sexpflag) | |
2710 | break; | |
ef316cf0 | 2711 | temp_pos = dec_bytepos (from_byte); |
6a140a74 | 2712 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
b7dbcc19 | 2713 | if (from != stop && c == FETCH_CHAR_AS_MULTIBYTE (temp_pos)) |
6a140a74 | 2714 | DEC_BOTH (from, from_byte); |
8489eb67 RS |
2715 | if (mathexit) |
2716 | { | |
2717 | mathexit = 0; | |
2718 | goto open2; | |
2719 | } | |
2720 | mathexit = 1; | |
2721 | ||
2722 | case Sclose: | |
2723 | if (!++depth) goto done2; | |
2724 | break; | |
2725 | ||
2726 | case Sopen: | |
2727 | open2: | |
2728 | if (!--depth) goto done2; | |
2729 | if (depth < min_depth) | |
9471945b KS |
2730 | xsignal3 (Qscan_error, |
2731 | build_string ("Containing expression ends prematurely"), | |
2732 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2733 | break; |
2734 | ||
2735 | case Sendcomment: | |
2736 | if (!parse_sexp_ignore_comments) | |
2737 | break; | |
95ff8dfc | 2738 | found = back_comment (from, from_byte, stop, comnested, comstyle, |
6a140a74 | 2739 | &out_charpos, &out_bytepos); |
1aa963c8 SM |
2740 | /* FIXME: if found == -1, then it really wasn't a comment-end. |
2741 | For single-char Sendcomment, we can't do much about it apart | |
2742 | from skipping the char. | |
2743 | For 2-char endcomments, we could try again, taking both | |
2744 | chars as separate entities, but it's a lot of trouble | |
2745 | for very little gain, so we don't bother either. -sm */ | |
6a140a74 RS |
2746 | if (found != -1) |
2747 | from = out_charpos, from_byte = out_bytepos; | |
8489eb67 RS |
2748 | break; |
2749 | ||
195d1361 RS |
2750 | case Scomment_fence: |
2751 | case Sstring_fence: | |
2752 | while (1) | |
2753 | { | |
8b8bf8e6 SM |
2754 | if (from == stop) |
2755 | goto lose; | |
6b61353c | 2756 | DEC_BOTH (from, from_byte); |
195d1361 | 2757 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
7d0393cf | 2758 | if (!char_quoted (from, from_byte) |
b7dbcc19 | 2759 | && (c = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
bd25db08 | 2760 | SYNTAX_WITH_MULTIBYTE_CHECK (c) == code)) |
195d1361 RS |
2761 | break; |
2762 | } | |
2763 | if (code == Sstring_fence && !depth && sexpflag) goto done2; | |
2764 | break; | |
7d0393cf | 2765 | |
8489eb67 | 2766 | case Sstring: |
b7dbcc19 | 2767 | stringterm = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8489eb67 RS |
2768 | while (1) |
2769 | { | |
8b8bf8e6 SM |
2770 | if (from == stop) |
2771 | goto lose; | |
6b61353c KH |
2772 | DEC_BOTH (from, from_byte); |
2773 | UPDATE_SYNTAX_TABLE_BACKWARD (from); | |
2774 | if (!char_quoted (from, from_byte) | |
2775 | && (stringterm | |
2776 | == (c = FETCH_CHAR_AS_MULTIBYTE (from_byte))) | |
42ebfa31 | 2777 | && SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring) |
8489eb67 | 2778 | break; |
8489eb67 | 2779 | } |
8489eb67 RS |
2780 | if (!depth && sexpflag) goto done2; |
2781 | break; | |
240c43e8 SM |
2782 | default: |
2783 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
2784 | break; | |
8489eb67 RS |
2785 | } |
2786 | } | |
2787 | ||
2788 | /* Reached start of buffer. Error if within object, return nil if between */ | |
8b8bf8e6 SM |
2789 | if (depth) |
2790 | goto lose; | |
8489eb67 RS |
2791 | |
2792 | immediate_quit = 0; | |
2793 | return Qnil; | |
2794 | ||
2795 | done2: | |
2796 | count++; | |
2797 | } | |
2798 | ||
2799 | ||
2800 | immediate_quit = 0; | |
1e142fb7 | 2801 | XSETFASTINT (val, from); |
8489eb67 RS |
2802 | return val; |
2803 | ||
2804 | lose: | |
9471945b KS |
2805 | xsignal3 (Qscan_error, |
2806 | build_string ("Unbalanced parentheses"), | |
2807 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2808 | } |
2809 | ||
8489eb67 | 2810 | DEFUN ("scan-lists", Fscan_lists, Sscan_lists, 3, 3, 0, |
fdb82f93 PJ |
2811 | doc: /* Scan from character number FROM by COUNT lists. |
2812 | Returns the character number of the position thus found. | |
2813 | ||
2814 | If DEPTH is nonzero, paren depth begins counting from that value, | |
2815 | only places where the depth in parentheses becomes zero | |
2816 | are candidates for stopping; COUNT such places are counted. | |
2817 | Thus, a positive value for DEPTH means go out levels. | |
2818 | ||
2819 | Comments are ignored if `parse-sexp-ignore-comments' is non-nil. | |
2820 | ||
2821 | If the beginning or end of (the accessible part of) the buffer is reached | |
2822 | and the depth is wrong, an error is signaled. | |
2823 | If the depth is right but the count is not used up, nil is returned. */) | |
2824 | (from, count, depth) | |
8489eb67 RS |
2825 | Lisp_Object from, count, depth; |
2826 | { | |
b7826503 PJ |
2827 | CHECK_NUMBER (from); |
2828 | CHECK_NUMBER (count); | |
2829 | CHECK_NUMBER (depth); | |
8489eb67 RS |
2830 | |
2831 | return scan_lists (XINT (from), XINT (count), XINT (depth), 0); | |
2832 | } | |
2833 | ||
2834 | DEFUN ("scan-sexps", Fscan_sexps, Sscan_sexps, 2, 2, 0, | |
fdb82f93 PJ |
2835 | doc: /* Scan from character number FROM by COUNT balanced expressions. |
2836 | If COUNT is negative, scan backwards. | |
2837 | Returns the character number of the position thus found. | |
2838 | ||
2839 | Comments are ignored if `parse-sexp-ignore-comments' is non-nil. | |
2840 | ||
2841 | If the beginning or end of (the accessible part of) the buffer is reached | |
2842 | in the middle of a parenthetical grouping, an error is signaled. | |
2843 | If the beginning or end is reached between groupings | |
2844 | but before count is used up, nil is returned. */) | |
2845 | (from, count) | |
8489eb67 RS |
2846 | Lisp_Object from, count; |
2847 | { | |
b7826503 PJ |
2848 | CHECK_NUMBER (from); |
2849 | CHECK_NUMBER (count); | |
8489eb67 RS |
2850 | |
2851 | return scan_lists (XINT (from), XINT (count), 0, 1); | |
2852 | } | |
2853 | ||
2854 | DEFUN ("backward-prefix-chars", Fbackward_prefix_chars, Sbackward_prefix_chars, | |
fdb82f93 PJ |
2855 | 0, 0, 0, |
2856 | doc: /* Move point backward over any number of chars with prefix syntax. | |
2857 | This includes chars with "quote" or "prefix" syntax (' or p). */) | |
2858 | () | |
8489eb67 RS |
2859 | { |
2860 | int beg = BEGV; | |
6a140a74 RS |
2861 | int opoint = PT; |
2862 | int opoint_byte = PT_BYTE; | |
6ec8bbd2 | 2863 | int pos = PT; |
6a140a74 | 2864 | int pos_byte = PT_BYTE; |
93da5fff | 2865 | int c; |
93da5fff | 2866 | |
7d0393cf | 2867 | if (pos <= beg) |
195d1361 | 2868 | { |
f902a008 RS |
2869 | SET_PT_BOTH (opoint, opoint_byte); |
2870 | ||
2871 | return Qnil; | |
195d1361 | 2872 | } |
8489eb67 | 2873 | |
f902a008 RS |
2874 | SETUP_SYNTAX_TABLE (pos, -1); |
2875 | ||
6a140a74 RS |
2876 | DEC_BOTH (pos, pos_byte); |
2877 | ||
1fd3172d | 2878 | while (!char_quoted (pos, pos_byte) |
195d1361 | 2879 | /* Previous statement updates syntax table. */ |
b7dbcc19 | 2880 | && ((c = FETCH_CHAR_AS_MULTIBYTE (pos_byte), SYNTAX (c) == Squote) |
93da5fff KH |
2881 | || SYNTAX_PREFIX (c))) |
2882 | { | |
1fd3172d RS |
2883 | opoint = pos; |
2884 | opoint_byte = pos_byte; | |
2885 | ||
2886 | if (pos + 1 > beg) | |
2887 | DEC_BOTH (pos, pos_byte); | |
93da5fff | 2888 | } |
8489eb67 | 2889 | |
6a140a74 | 2890 | SET_PT_BOTH (opoint, opoint_byte); |
8489eb67 RS |
2891 | |
2892 | return Qnil; | |
2893 | } | |
2894 | \f | |
6a140a74 | 2895 | /* Parse forward from FROM / FROM_BYTE to END, |
e5d4f4dc RS |
2896 | assuming that FROM has state OLDSTATE (nil means FROM is start of function), |
2897 | and return a description of the state of the parse at END. | |
c81a3712 | 2898 | If STOPBEFORE is nonzero, stop at the start of an atom. |
644ea4df RS |
2899 | If COMMENTSTOP is 1, stop at the start of a comment. |
2900 | If COMMENTSTOP is -1, stop at the start or end of a comment, | |
2901 | after the beginning of a string, or after the end of a string. */ | |
8489eb67 | 2902 | |
340f92b5 | 2903 | static void |
6a140a74 | 2904 | scan_sexps_forward (stateptr, from, from_byte, end, targetdepth, |
c81a3712 | 2905 | stopbefore, oldstate, commentstop) |
e5d4f4dc | 2906 | struct lisp_parse_state *stateptr; |
8e2911c2 AS |
2907 | register EMACS_INT from; |
2908 | EMACS_INT from_byte, end; | |
2909 | int targetdepth, stopbefore; | |
8489eb67 | 2910 | Lisp_Object oldstate; |
c81a3712 | 2911 | int commentstop; |
8489eb67 RS |
2912 | { |
2913 | struct lisp_parse_state state; | |
2914 | ||
2915 | register enum syntaxcode code; | |
95ff8dfc RS |
2916 | int c1; |
2917 | int comnested; | |
8489eb67 RS |
2918 | struct level { int last, prev; }; |
2919 | struct level levelstart[100]; | |
2920 | register struct level *curlevel = levelstart; | |
2921 | struct level *endlevel = levelstart + 100; | |
8489eb67 RS |
2922 | register int depth; /* Paren depth of current scanning location. |
2923 | level - levelstart equals this except | |
2924 | when the depth becomes negative. */ | |
2925 | int mindepth; /* Lowest DEPTH value seen. */ | |
2926 | int start_quoted = 0; /* Nonzero means starting after a char quote */ | |
2927 | Lisp_Object tem; | |
8e2911c2 AS |
2928 | EMACS_INT prev_from; /* Keep one character before FROM. */ |
2929 | EMACS_INT prev_from_byte; | |
1fd3172d | 2930 | int prev_from_syntax; |
195d1361 RS |
2931 | int boundary_stop = commentstop == -1; |
2932 | int nofence; | |
95ff8dfc | 2933 | int found; |
60c86a83 | 2934 | EMACS_INT out_bytepos, out_charpos; |
7c7ca3d2 | 2935 | int temp; |
93da5fff KH |
2936 | |
2937 | prev_from = from; | |
6a140a74 RS |
2938 | prev_from_byte = from_byte; |
2939 | if (from != BEGV) | |
2940 | DEC_BOTH (prev_from, prev_from_byte); | |
93da5fff KH |
2941 | |
2942 | /* Use this macro instead of `from++'. */ | |
6a140a74 RS |
2943 | #define INC_FROM \ |
2944 | do { prev_from = from; \ | |
2945 | prev_from_byte = from_byte; \ | |
327719ee | 2946 | temp = FETCH_CHAR_AS_MULTIBYTE (prev_from_byte); \ |
ab229fdd | 2947 | prev_from_syntax = SYNTAX_WITH_FLAGS (temp); \ |
ef316cf0 | 2948 | INC_BOTH (from, from_byte); \ |
3ceb4629 SM |
2949 | if (from < end) \ |
2950 | UPDATE_SYNTAX_TABLE_FORWARD (from); \ | |
6a140a74 | 2951 | } while (0) |
8489eb67 RS |
2952 | |
2953 | immediate_quit = 1; | |
2954 | QUIT; | |
2955 | ||
265a9e55 | 2956 | if (NILP (oldstate)) |
8489eb67 RS |
2957 | { |
2958 | depth = 0; | |
2959 | state.instring = -1; | |
2960 | state.incomment = 0; | |
195d1361 RS |
2961 | state.comstyle = 0; /* comment style a by default. */ |
2962 | state.comstr_start = -1; /* no comment/string seen. */ | |
8489eb67 RS |
2963 | } |
2964 | else | |
2965 | { | |
2966 | tem = Fcar (oldstate); | |
265a9e55 | 2967 | if (!NILP (tem)) |
8489eb67 RS |
2968 | depth = XINT (tem); |
2969 | else | |
2970 | depth = 0; | |
2971 | ||
2972 | oldstate = Fcdr (oldstate); | |
2973 | oldstate = Fcdr (oldstate); | |
2974 | oldstate = Fcdr (oldstate); | |
2975 | tem = Fcar (oldstate); | |
195d1361 | 2976 | /* Check whether we are inside string_fence-style string: */ |
7d0393cf JB |
2977 | state.instring = (!NILP (tem) |
2978 | ? (INTEGERP (tem) ? XINT (tem) : ST_STRING_STYLE) | |
e76f1c44 | 2979 | : -1); |
8489eb67 RS |
2980 | |
2981 | oldstate = Fcdr (oldstate); | |
2982 | tem = Fcar (oldstate); | |
e76f1c44 GM |
2983 | state.incomment = (!NILP (tem) |
2984 | ? (INTEGERP (tem) ? XINT (tem) : -1) | |
95ff8dfc | 2985 | : 0); |
8489eb67 RS |
2986 | |
2987 | oldstate = Fcdr (oldstate); | |
2988 | tem = Fcar (oldstate); | |
265a9e55 | 2989 | start_quoted = !NILP (tem); |
e5d4f4dc | 2990 | |
95ff8dfc | 2991 | /* if the eighth element of the list is nil, we are in comment |
195d1361 RS |
2992 | style a. If it is non-nil, we are in comment style b */ |
2993 | oldstate = Fcdr (oldstate); | |
e5d4f4dc | 2994 | oldstate = Fcdr (oldstate); |
195d1361 | 2995 | tem = Fcar (oldstate); |
7d0393cf | 2996 | state.comstyle = NILP (tem) ? 0 : (EQ (tem, Qsyntax_table) |
e76f1c44 | 2997 | ? ST_COMMENT_STYLE : 1); |
195d1361 | 2998 | |
e5d4f4dc | 2999 | oldstate = Fcdr (oldstate); |
e5d4f4dc | 3000 | tem = Fcar (oldstate); |
195d1361 | 3001 | state.comstr_start = NILP (tem) ? -1 : XINT (tem) ; |
1a102a58 RS |
3002 | oldstate = Fcdr (oldstate); |
3003 | tem = Fcar (oldstate); | |
3004 | while (!NILP (tem)) /* >= second enclosing sexps. */ | |
3005 | { | |
3006 | /* curlevel++->last ran into compiler bug on Apollo */ | |
3007 | curlevel->last = XINT (Fcar (tem)); | |
3008 | if (++curlevel == endlevel) | |
02010917 | 3009 | curlevel--; /* error ("Nesting too deep for parser"); */ |
1a102a58 RS |
3010 | curlevel->prev = -1; |
3011 | curlevel->last = -1; | |
3012 | tem = Fcdr (tem); | |
3013 | } | |
8489eb67 RS |
3014 | } |
3015 | state.quoted = 0; | |
3016 | mindepth = depth; | |
3017 | ||
3018 | curlevel->prev = -1; | |
3019 | curlevel->last = -1; | |
3020 | ||
326b283b | 3021 | SETUP_SYNTAX_TABLE (prev_from, 1); |
ab229fdd AS |
3022 | temp = FETCH_CHAR (prev_from_byte); |
3023 | prev_from_syntax = SYNTAX_WITH_FLAGS (temp); | |
3ceb4629 | 3024 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
326b283b | 3025 | |
195d1361 | 3026 | /* Enter the loop at a place appropriate for initial state. */ |
8489eb67 | 3027 | |
326b283b RS |
3028 | if (state.incomment) |
3029 | goto startincomment; | |
8489eb67 RS |
3030 | if (state.instring >= 0) |
3031 | { | |
195d1361 | 3032 | nofence = state.instring != ST_STRING_STYLE; |
326b283b RS |
3033 | if (start_quoted) |
3034 | goto startquotedinstring; | |
8489eb67 RS |
3035 | goto startinstring; |
3036 | } | |
326b283b RS |
3037 | else if (start_quoted) |
3038 | goto startquoted; | |
1fd3172d | 3039 | |
8489eb67 RS |
3040 | while (from < end) |
3041 | { | |
93da5fff | 3042 | INC_FROM; |
1fd3172d | 3043 | code = prev_from_syntax & 0xff; |
4c920633 | 3044 | |
c5d4b96f SM |
3045 | if (from < end |
3046 | && SYNTAX_FLAGS_COMSTART_FIRST (prev_from_syntax) | |
3047 | && (c1 = FETCH_CHAR (from_byte), | |
3048 | SYNTAX_COMSTART_SECOND (c1))) | |
3049 | /* Duplicate code to avoid a complex if-expression | |
3050 | which causes trouble for the SGI compiler. */ | |
95ff8dfc | 3051 | { |
c5d4b96f SM |
3052 | /* Record the comment style we have entered so that only |
3053 | the comment-end sequence of the same style actually | |
3054 | terminates the comment section. */ | |
3055 | state.comstyle = SYNTAX_COMMENT_STYLE (c1); | |
3056 | comnested = SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax); | |
3057 | comnested = comnested || SYNTAX_COMMENT_NESTED (c1); | |
3058 | state.incomment = comnested ? 1 : -1; | |
95ff8dfc | 3059 | state.comstr_start = prev_from; |
c5d4b96f SM |
3060 | INC_FROM; |
3061 | code = Scomment; | |
95ff8dfc | 3062 | } |
4c920633 | 3063 | else if (code == Scomment_fence) |
e5d4f4dc RS |
3064 | { |
3065 | /* Record the comment style we have entered so that only | |
3066 | the comment-end sequence of the same style actually | |
3067 | terminates the comment section. */ | |
95ff8dfc RS |
3068 | state.comstyle = ST_COMMENT_STYLE; |
3069 | state.incomment = -1; | |
195d1361 | 3070 | state.comstr_start = prev_from; |
e5d4f4dc | 3071 | code = Scomment; |
e5d4f4dc | 3072 | } |
c5d4b96f SM |
3073 | else if (code == Scomment) |
3074 | { | |
3075 | state.comstyle = SYNTAX_FLAGS_COMMENT_STYLE (prev_from_syntax); | |
3076 | state.incomment = (SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax) ? | |
3077 | 1 : -1); | |
3078 | state.comstr_start = prev_from; | |
3079 | } | |
e5d4f4dc | 3080 | |
1fd3172d | 3081 | if (SYNTAX_FLAGS_PREFIX (prev_from_syntax)) |
8489eb67 | 3082 | continue; |
0220c518 | 3083 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
3084 | { |
3085 | case Sescape: | |
3086 | case Scharquote: | |
3087 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
93da5fff | 3088 | curlevel->last = prev_from; |
8489eb67 RS |
3089 | startquoted: |
3090 | if (from == end) goto endquoted; | |
93da5fff | 3091 | INC_FROM; |
8489eb67 RS |
3092 | goto symstarted; |
3093 | /* treat following character as a word constituent */ | |
3094 | case Sword: | |
3095 | case Ssymbol: | |
3096 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
93da5fff | 3097 | curlevel->last = prev_from; |
8489eb67 RS |
3098 | symstarted: |
3099 | while (from < end) | |
3100 | { | |
7c7ca3d2 | 3101 | /* Some compilers can't handle this inside the switch. */ |
327719ee | 3102 | temp = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
ab229fdd | 3103 | temp = SYNTAX (temp); |
7c7ca3d2 | 3104 | switch (temp) |
8489eb67 RS |
3105 | { |
3106 | case Scharquote: | |
3107 | case Sescape: | |
93da5fff | 3108 | INC_FROM; |
8489eb67 RS |
3109 | if (from == end) goto endquoted; |
3110 | break; | |
3111 | case Sword: | |
3112 | case Ssymbol: | |
3113 | case Squote: | |
3114 | break; | |
3115 | default: | |
3116 | goto symdone; | |
3117 | } | |
93da5fff | 3118 | INC_FROM; |
8489eb67 RS |
3119 | } |
3120 | symdone: | |
3121 | curlevel->prev = curlevel->last; | |
3122 | break; | |
3123 | ||
240c43e8 | 3124 | case Scomment_fence: /* Can't happen because it's handled above. */ |
8489eb67 | 3125 | case Scomment: |
195d1361 | 3126 | if (commentstop || boundary_stop) goto done; |
d355bd8a SM |
3127 | startincomment: |
3128 | /* The (from == BEGV) test was to enter the loop in the middle so | |
95ff8dfc | 3129 | that we find a 2-char comment ender even if we start in the |
e6365855 | 3130 | middle of it. We don't want to do that if we're just at the |
d355bd8a | 3131 | beginning of the comment (think of (*) ... (*)). */ |
95ff8dfc RS |
3132 | found = forw_comment (from, from_byte, end, |
3133 | state.incomment, state.comstyle, | |
e6365855 SM |
3134 | (from == BEGV || from < state.comstr_start + 3) |
3135 | ? 0 : prev_from_syntax, | |
95ff8dfc RS |
3136 | &out_charpos, &out_bytepos, &state.incomment); |
3137 | from = out_charpos; from_byte = out_bytepos; | |
3138 | /* Beware! prev_from and friends are invalid now. | |
3139 | Luckily, the `done' doesn't use them and the INC_FROM | |
3140 | sets them to a sane value without looking at them. */ | |
3141 | if (!found) goto done; | |
d355bd8a | 3142 | INC_FROM; |
8489eb67 | 3143 | state.incomment = 0; |
e5d4f4dc | 3144 | state.comstyle = 0; /* reset the comment style */ |
195d1361 | 3145 | if (boundary_stop) goto done; |
8489eb67 RS |
3146 | break; |
3147 | ||
3148 | case Sopen: | |
3149 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
3150 | depth++; | |
3151 | /* curlevel++->last ran into compiler bug on Apollo */ | |
93da5fff | 3152 | curlevel->last = prev_from; |
8489eb67 | 3153 | if (++curlevel == endlevel) |
02010917 | 3154 | curlevel--; /* error ("Nesting too deep for parser"); */ |
8489eb67 RS |
3155 | curlevel->prev = -1; |
3156 | curlevel->last = -1; | |
30844415 | 3157 | if (targetdepth == depth) goto done; |
8489eb67 RS |
3158 | break; |
3159 | ||
3160 | case Sclose: | |
3161 | depth--; | |
3162 | if (depth < mindepth) | |
3163 | mindepth = depth; | |
3164 | if (curlevel != levelstart) | |
3165 | curlevel--; | |
3166 | curlevel->prev = curlevel->last; | |
30844415 | 3167 | if (targetdepth == depth) goto done; |
8489eb67 RS |
3168 | break; |
3169 | ||
3170 | case Sstring: | |
195d1361 RS |
3171 | case Sstring_fence: |
3172 | state.comstr_start = from - 1; | |
8489eb67 | 3173 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ |
93da5fff | 3174 | curlevel->last = prev_from; |
7d0393cf | 3175 | state.instring = (code == Sstring |
b7dbcc19 | 3176 | ? (FETCH_CHAR_AS_MULTIBYTE (prev_from_byte)) |
195d1361 RS |
3177 | : ST_STRING_STYLE); |
3178 | if (boundary_stop) goto done; | |
8489eb67 | 3179 | startinstring: |
195d1361 | 3180 | { |
644ea4df | 3181 | nofence = state.instring != ST_STRING_STYLE; |
7d0393cf | 3182 | |
644ea4df RS |
3183 | while (1) |
3184 | { | |
3185 | int c; | |
195d1361 | 3186 | |
644ea4df | 3187 | if (from >= end) goto done; |
b7dbcc19 | 3188 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
7c7ca3d2 RS |
3189 | /* Some compilers can't handle this inside the switch. */ |
3190 | temp = SYNTAX (c); | |
ccf89641 KH |
3191 | |
3192 | /* Check TEMP here so that if the char has | |
3193 | a syntax-table property which says it is NOT | |
3194 | a string character, it does not end the string. */ | |
3195 | if (nofence && c == state.instring && temp == Sstring) | |
3196 | break; | |
3197 | ||
7c7ca3d2 | 3198 | switch (temp) |
644ea4df RS |
3199 | { |
3200 | case Sstring_fence: | |
3201 | if (!nofence) goto string_end; | |
3202 | break; | |
3203 | case Scharquote: | |
3204 | case Sescape: | |
3205 | INC_FROM; | |
3206 | startquotedinstring: | |
3207 | if (from >= end) goto endquoted; | |
195d1361 | 3208 | } |
644ea4df RS |
3209 | INC_FROM; |
3210 | } | |
195d1361 RS |
3211 | } |
3212 | string_end: | |
8489eb67 RS |
3213 | state.instring = -1; |
3214 | curlevel->prev = curlevel->last; | |
93da5fff | 3215 | INC_FROM; |
195d1361 | 3216 | if (boundary_stop) goto done; |
8489eb67 RS |
3217 | break; |
3218 | ||
3219 | case Smath: | |
240c43e8 SM |
3220 | /* FIXME: We should do something with it. */ |
3221 | break; | |
3222 | default: | |
3223 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
8489eb67 RS |
3224 | break; |
3225 | } | |
3226 | } | |
3227 | goto done; | |
3228 | ||
3229 | stop: /* Here if stopping before start of sexp. */ | |
93da5fff | 3230 | from = prev_from; /* We have just fetched the char that starts it; */ |
8489eb67 RS |
3231 | goto done; /* but return the position before it. */ |
3232 | ||
3233 | endquoted: | |
3234 | state.quoted = 1; | |
3235 | done: | |
3236 | state.depth = depth; | |
3237 | state.mindepth = mindepth; | |
3238 | state.thislevelstart = curlevel->prev; | |
3239 | state.prevlevelstart | |
3240 | = (curlevel == levelstart) ? -1 : (curlevel - 1)->last; | |
3241 | state.location = from; | |
1a102a58 RS |
3242 | state.levelstarts = Qnil; |
3243 | while (--curlevel >= levelstart) | |
3244 | state.levelstarts = Fcons (make_number (curlevel->last), | |
3245 | state.levelstarts); | |
8489eb67 RS |
3246 | immediate_quit = 0; |
3247 | ||
e5d4f4dc | 3248 | *stateptr = state; |
8489eb67 RS |
3249 | } |
3250 | ||
c81a3712 | 3251 | DEFUN ("parse-partial-sexp", Fparse_partial_sexp, Sparse_partial_sexp, 2, 6, 0, |
fdb82f93 PJ |
3252 | doc: /* Parse Lisp syntax starting at FROM until TO; return status of parse at TO. |
3253 | Parsing stops at TO or when certain criteria are met; | |
3254 | point is set to where parsing stops. | |
3255 | If fifth arg OLDSTATE is omitted or nil, | |
3256 | parsing assumes that FROM is the beginning of a function. | |
0e6228bc | 3257 | Value is a list of elements describing final state of parsing: |
fdb82f93 PJ |
3258 | 0. depth in parens. |
3259 | 1. character address of start of innermost containing list; nil if none. | |
3260 | 2. character address of start of last complete sexp terminated. | |
3261 | 3. non-nil if inside a string. | |
3262 | (it is the character that will terminate the string, | |
3263 | or t if the string should be terminated by a generic string delimiter.) | |
7d0393cf | 3264 | 4. nil if outside a comment, t if inside a non-nestable comment, |
fdb82f93 PJ |
3265 | else an integer (the current comment nesting). |
3266 | 5. t if following a quote character. | |
3267 | 6. the minimum paren-depth encountered during this scan. | |
3268 | 7. t if in a comment of style b; symbol `syntax-table' if the comment | |
3269 | should be terminated by a generic comment delimiter. | |
3270 | 8. character address of start of comment or string; nil if not in one. | |
3271 | 9. Intermediate data for continuation of parsing (subject to change). | |
3272 | If third arg TARGETDEPTH is non-nil, parsing stops if the depth | |
3273 | in parentheses becomes equal to TARGETDEPTH. | |
3274 | Fourth arg STOPBEFORE non-nil means stop when come to | |
3275 | any character that starts a sexp. | |
0e6228bc | 3276 | Fifth arg OLDSTATE is a list like what this function returns. |
fdb82f93 | 3277 | It is used to initialize the state of the parse. Elements number 1, 2, 6 |
43726c05 | 3278 | and 8 are ignored. |
fdb82f93 PJ |
3279 | Sixth arg COMMENTSTOP non-nil means stop at the start of a comment. |
3280 | If it is symbol `syntax-table', stop after the start of a comment or a | |
3281 | string, or after end of a comment or a string. */) | |
3282 | (from, to, targetdepth, stopbefore, oldstate, commentstop) | |
c81a3712 | 3283 | Lisp_Object from, to, targetdepth, stopbefore, oldstate, commentstop; |
8489eb67 RS |
3284 | { |
3285 | struct lisp_parse_state state; | |
3286 | int target; | |
3287 | ||
265a9e55 | 3288 | if (!NILP (targetdepth)) |
8489eb67 | 3289 | { |
b7826503 | 3290 | CHECK_NUMBER (targetdepth); |
8489eb67 RS |
3291 | target = XINT (targetdepth); |
3292 | } | |
3293 | else | |
3294 | target = -100000; /* We won't reach this depth */ | |
3295 | ||
3296 | validate_region (&from, &to); | |
6a140a74 RS |
3297 | scan_sexps_forward (&state, XINT (from), CHAR_TO_BYTE (XINT (from)), |
3298 | XINT (to), | |
c81a3712 | 3299 | target, !NILP (stopbefore), oldstate, |
7d0393cf | 3300 | (NILP (commentstop) |
195d1361 | 3301 | ? 0 : (EQ (commentstop, Qsyntax_table) ? -1 : 1))); |
8489eb67 RS |
3302 | |
3303 | SET_PT (state.location); | |
7d0393cf | 3304 | |
8489eb67 RS |
3305 | return Fcons (make_number (state.depth), |
3306 | Fcons (state.prevlevelstart < 0 ? Qnil : make_number (state.prevlevelstart), | |
3307 | Fcons (state.thislevelstart < 0 ? Qnil : make_number (state.thislevelstart), | |
7d0393cf JB |
3308 | Fcons (state.instring >= 0 |
3309 | ? (state.instring == ST_STRING_STYLE | |
195d1361 | 3310 | ? Qt : make_number (state.instring)) : Qnil, |
95ff8dfc RS |
3311 | Fcons (state.incomment < 0 ? Qt : |
3312 | (state.incomment == 0 ? Qnil : | |
3313 | make_number (state.incomment)), | |
8489eb67 | 3314 | Fcons (state.quoted ? Qt : Qnil, |
bff37d2a | 3315 | Fcons (make_number (state.mindepth), |
7d0393cf | 3316 | Fcons ((state.comstyle |
bff37d2a RS |
3317 | ? (state.comstyle == ST_COMMENT_STYLE |
3318 | ? Qsyntax_table : Qt) : | |
3319 | Qnil), | |
0beaf54f DL |
3320 | Fcons (((state.incomment |
3321 | || (state.instring >= 0)) | |
bff37d2a RS |
3322 | ? make_number (state.comstr_start) |
3323 | : Qnil), | |
1a102a58 | 3324 | Fcons (state.levelstarts, Qnil)))))))))); |
8489eb67 RS |
3325 | } |
3326 | \f | |
dfcf069d | 3327 | void |
8489eb67 RS |
3328 | init_syntax_once () |
3329 | { | |
78f9a1f7 | 3330 | register int i, c; |
8ea151b2 | 3331 | Lisp_Object temp; |
8489eb67 | 3332 | |
5ebaddf5 RS |
3333 | /* This has to be done here, before we call Fmake_char_table. */ |
3334 | Qsyntax_table = intern ("syntax-table"); | |
3335 | staticpro (&Qsyntax_table); | |
3336 | ||
3337 | /* Intern this now in case it isn't already done. | |
3338 | Setting this variable twice is harmless. | |
3339 | But don't staticpro it here--that is done in alloc.c. */ | |
3340 | Qchar_table_extra_slots = intern ("char-table-extra-slots"); | |
3341 | ||
93da5fff | 3342 | /* Create objects which can be shared among syntax tables. */ |
42ebfa31 | 3343 | Vsyntax_code_object = Fmake_vector (make_number (Smax), Qnil); |
93da5fff KH |
3344 | for (i = 0; i < XVECTOR (Vsyntax_code_object)->size; i++) |
3345 | XVECTOR (Vsyntax_code_object)->contents[i] | |
3346 | = Fcons (make_number (i), Qnil); | |
3347 | ||
5ebaddf5 RS |
3348 | /* Now we are ready to set up this property, so we can |
3349 | create syntax tables. */ | |
3350 | Fput (Qsyntax_table, Qchar_table_extra_slots, make_number (0)); | |
3351 | ||
93da5fff | 3352 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Swhitespace]; |
8489eb67 | 3353 | |
5ebaddf5 | 3354 | Vstandard_syntax_table = Fmake_char_table (Qsyntax_table, temp); |
8489eb67 | 3355 | |
aa7b08b4 RS |
3356 | /* Control characters should not be whitespace. */ |
3357 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Spunct]; | |
3358 | for (i = 0; i <= ' ' - 1; i++) | |
3359 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); | |
3360 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 0177, temp); | |
3361 | ||
3362 | /* Except that a few really are whitespace. */ | |
3363 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Swhitespace]; | |
3364 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ' ', temp); | |
3365 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\t', temp); | |
3366 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\n', temp); | |
3367 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 015, temp); | |
3368 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 014, temp); | |
3369 | ||
93da5fff | 3370 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Sword]; |
8489eb67 | 3371 | for (i = 'a'; i <= 'z'; i++) |
8ea151b2 | 3372 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
8489eb67 | 3373 | for (i = 'A'; i <= 'Z'; i++) |
8ea151b2 | 3374 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
8489eb67 | 3375 | for (i = '0'; i <= '9'; i++) |
8ea151b2 RS |
3376 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
3377 | ||
3378 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '$', temp); | |
3379 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '%', temp); | |
3380 | ||
3381 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '(', | |
3382 | Fcons (make_number (Sopen), make_number (')'))); | |
3383 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ')', | |
3384 | Fcons (make_number (Sclose), make_number ('('))); | |
3385 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '[', | |
3386 | Fcons (make_number (Sopen), make_number (']'))); | |
3387 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ']', | |
3388 | Fcons (make_number (Sclose), make_number ('['))); | |
3389 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '{', | |
3390 | Fcons (make_number (Sopen), make_number ('}'))); | |
3391 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '}', | |
3392 | Fcons (make_number (Sclose), make_number ('{'))); | |
3393 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '"', | |
3394 | Fcons (make_number ((int) Sstring), Qnil)); | |
3395 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\\', | |
3396 | Fcons (make_number ((int) Sescape), Qnil)); | |
3397 | ||
93da5fff | 3398 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Ssymbol]; |
8489eb67 | 3399 | for (i = 0; i < 10; i++) |
78f9a1f7 KH |
3400 | { |
3401 | c = "_-+*/&|<>="[i]; | |
3402 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, c, temp); | |
3403 | } | |
8489eb67 | 3404 | |
93da5fff | 3405 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Spunct]; |
8489eb67 | 3406 | for (i = 0; i < 12; i++) |
78f9a1f7 KH |
3407 | { |
3408 | c = ".,;:?!#@~^'`"[i]; | |
3409 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, c, temp); | |
3410 | } | |
bd25db08 KH |
3411 | |
3412 | /* All multibyte characters have syntax `word' by default. */ | |
3413 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Sword]; | |
5c7b02ab | 3414 | char_table_set_range (Vstandard_syntax_table, 0x80, MAX_CHAR, temp); |
8489eb67 RS |
3415 | } |
3416 | ||
dfcf069d | 3417 | void |
8489eb67 RS |
3418 | syms_of_syntax () |
3419 | { | |
3420 | Qsyntax_table_p = intern ("syntax-table-p"); | |
3421 | staticpro (&Qsyntax_table_p); | |
3422 | ||
93da5fff KH |
3423 | staticpro (&Vsyntax_code_object); |
3424 | ||
a2994c46 KS |
3425 | staticpro (&gl_state.object); |
3426 | staticpro (&gl_state.global_code); | |
3427 | staticpro (&gl_state.current_syntax_table); | |
3428 | staticpro (&gl_state.old_prop); | |
3429 | ||
3430 | /* Defined in regex.c */ | |
3431 | staticpro (&re_match_object); | |
3432 | ||
7bf5e9e4 RS |
3433 | Qscan_error = intern ("scan-error"); |
3434 | staticpro (&Qscan_error); | |
3435 | Fput (Qscan_error, Qerror_conditions, | |
f3be100f | 3436 | Fcons (Qscan_error, Fcons (Qerror, Qnil))); |
7bf5e9e4 RS |
3437 | Fput (Qscan_error, Qerror_message, |
3438 | build_string ("Scan error")); | |
3439 | ||
8489eb67 | 3440 | DEFVAR_BOOL ("parse-sexp-ignore-comments", &parse_sexp_ignore_comments, |
fdb82f93 | 3441 | doc: /* Non-nil means `forward-sexp', etc., should treat comments as whitespace. */); |
8489eb67 | 3442 | |
195d1361 | 3443 | DEFVAR_BOOL ("parse-sexp-lookup-properties", &parse_sexp_lookup_properties, |
fdb82f93 PJ |
3444 | doc: /* Non-nil means `forward-sexp', etc., obey `syntax-table' property. |
3445 | Otherwise, that text property is simply ignored. | |
3446 | See the info node `(elisp)Syntax Properties' for a description of the | |
3447 | `syntax-table' property. */); | |
195d1361 | 3448 | |
8489eb67 RS |
3449 | words_include_escapes = 0; |
3450 | DEFVAR_BOOL ("words-include-escapes", &words_include_escapes, | |
fdb82f93 | 3451 | doc: /* Non-nil means `forward-word', etc., should treat escape chars part of words. */); |
8489eb67 | 3452 | |
bd25db08 | 3453 | DEFVAR_BOOL ("multibyte-syntax-as-symbol", &multibyte_syntax_as_symbol, |
fdb82f93 | 3454 | doc: /* Non-nil means `scan-sexps' treats all multibyte characters as symbol. */); |
bd25db08 KH |
3455 | multibyte_syntax_as_symbol = 0; |
3456 | ||
f4ed767f GM |
3457 | DEFVAR_BOOL ("open-paren-in-column-0-is-defun-start", |
3458 | &open_paren_in_column_0_is_defun_start, | |
b222e415 | 3459 | doc: /* *Non-nil means an open paren in column 0 denotes the start of a defun. */); |
f4ed767f GM |
3460 | open_paren_in_column_0_is_defun_start = 1; |
3461 | ||
8f924df7 KH |
3462 | |
3463 | DEFVAR_LISP ("find-word-boundary-function-table", | |
3464 | &Vfind_word_boundary_function_table, | |
869bb237 | 3465 | doc: /* |
8f924df7 | 3466 | Char table of functions to search for the word boundary. |
869bb237 KH |
3467 | Each function is called with two arguments; POS and LIMIT. |
3468 | POS and LIMIT are character positions in the current buffer. | |
3469 | ||
3470 | If POS is less than LIMIT, POS is at the first character of a word, | |
3471 | and the return value of a function is a position after the last | |
3472 | character of that word. | |
3473 | ||
3474 | If POS is not less than LIMIT, POS is at the last character of a word, | |
3475 | and the return value of a function is a position at the first | |
3476 | character of that word. | |
3477 | ||
3478 | In both cases, LIMIT bounds the search. */); | |
8f924df7 | 3479 | Vfind_word_boundary_function_table = Fmake_char_table (Qnil, Qnil); |
869bb237 | 3480 | |
8489eb67 RS |
3481 | defsubr (&Ssyntax_table_p); |
3482 | defsubr (&Ssyntax_table); | |
3483 | defsubr (&Sstandard_syntax_table); | |
3484 | defsubr (&Scopy_syntax_table); | |
3485 | defsubr (&Sset_syntax_table); | |
3486 | defsubr (&Schar_syntax); | |
beefa22e | 3487 | defsubr (&Smatching_paren); |
c65adb44 | 3488 | defsubr (&Sstring_to_syntax); |
8489eb67 | 3489 | defsubr (&Smodify_syntax_entry); |
62abe9cb | 3490 | defsubr (&Sinternal_describe_syntax_value); |
8489eb67 RS |
3491 | |
3492 | defsubr (&Sforward_word); | |
3493 | ||
195d1361 RS |
3494 | defsubr (&Sskip_chars_forward); |
3495 | defsubr (&Sskip_chars_backward); | |
3496 | defsubr (&Sskip_syntax_forward); | |
3497 | defsubr (&Sskip_syntax_backward); | |
3498 | ||
b3cfe0c8 | 3499 | defsubr (&Sforward_comment); |
8489eb67 RS |
3500 | defsubr (&Sscan_lists); |
3501 | defsubr (&Sscan_sexps); | |
3502 | defsubr (&Sbackward_prefix_chars); | |
3503 | defsubr (&Sparse_partial_sexp); | |
3504 | } | |
839966f3 KH |
3505 | |
3506 | /* arch-tag: 3e297b9f-088e-4b64-8f4c-fb0b3443e412 | |
3507 | (do not change this comment) */ |