Commit | Line | Data |
---|---|---|
8489eb67 | 1 | /* GNU Emacs routines to deal with syntax tables; also word and list parsing. |
73b0cd50 | 2 | Copyright (C) 1985, 1987, 1993-1995, 1997-1999, 2001-2011 |
429ab54e | 3 | Free Software Foundation, Inc. |
8489eb67 RS |
4 | |
5 | This file is part of GNU Emacs. | |
6 | ||
9ec0b715 | 7 | GNU Emacs is free software: you can redistribute it and/or modify |
8489eb67 | 8 | it under the terms of the GNU General Public License as published by |
9ec0b715 GM |
9 | the Free Software Foundation, either version 3 of the License, or |
10 | (at your option) any later version. | |
8489eb67 RS |
11 | |
12 | GNU Emacs is distributed in the hope that it will be useful, | |
13 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
14 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
15 | GNU General Public License for more details. | |
16 | ||
17 | You should have received a copy of the GNU General Public License | |
9ec0b715 | 18 | along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>. */ |
8489eb67 RS |
19 | |
20 | ||
18160b98 | 21 | #include <config.h> |
1b2de274 | 22 | |
8489eb67 | 23 | #include <ctype.h> |
1b2de274 | 24 | #include <sys/types.h> |
d7306fe6 | 25 | #include <setjmp.h> |
8489eb67 RS |
26 | #include "lisp.h" |
27 | #include "commands.h" | |
28 | #include "buffer.h" | |
5c7b02ab | 29 | #include "character.h" |
e35f6ff7 | 30 | #include "keymap.h" |
a1bc88d4 | 31 | #include "regex.h" |
195d1361 RS |
32 | |
33 | /* Make syntax table lookup grant data in gl_state. */ | |
34 | #define SYNTAX_ENTRY_VIA_PROPERTY | |
35 | ||
8489eb67 | 36 | #include "syntax.h" |
195d1361 | 37 | #include "intervals.h" |
c5683ceb SM |
38 | #include "category.h" |
39 | ||
40 | /* Then there are seven single-bit flags that have the following meanings: | |
41 | 1. This character is the first of a two-character comment-start sequence. | |
42 | 2. This character is the second of a two-character comment-start sequence. | |
43 | 3. This character is the first of a two-character comment-end sequence. | |
44 | 4. This character is the second of a two-character comment-end sequence. | |
45 | 5. This character is a prefix, for backward-prefix-chars. | |
46 | 6. The char is part of a delimiter for comments of style "b". | |
47 | 7. This character is part of a nestable comment sequence. | |
48 | 8. The char is part of a delimiter for comments of style "c". | |
49 | Note that any two-character sequence whose first character has flag 1 | |
50 | and whose second character has flag 2 will be interpreted as a comment start. | |
51 | ||
52 | bit 6 and 8 are used to discriminate between different comment styles. | |
53 | Languages such as C++ allow two orthogonal syntax start/end pairs | |
54 | and bit 6 is used to determine whether a comment-end or Scommentend | |
55 | ends style a or b. Comment markers can start style a, b, c, or bc. | |
56 | Style a is always the default. | |
57 | For 2-char comment markers, the style b flag is only looked up on the second | |
58 | char of the comment marker and on the first char of the comment ender. | |
59 | For style c (like to for the nested flag), the flag can be placed on any | |
60 | one of the chars. | |
61 | */ | |
62 | ||
63 | /* These macros extract specific flags from an integer | |
64 | that holds the syntax code and the flags. */ | |
65 | ||
66 | #define SYNTAX_FLAGS_COMSTART_FIRST(flags) (((flags) >> 16) & 1) | |
67 | ||
68 | #define SYNTAX_FLAGS_COMSTART_SECOND(flags) (((flags) >> 17) & 1) | |
69 | ||
70 | #define SYNTAX_FLAGS_COMEND_FIRST(flags) (((flags) >> 18) & 1) | |
71 | ||
72 | #define SYNTAX_FLAGS_COMEND_SECOND(flags) (((flags) >> 19) & 1) | |
73 | ||
74 | #define SYNTAX_FLAGS_PREFIX(flags) (((flags) >> 20) & 1) | |
75 | ||
76 | #define SYNTAX_FLAGS_COMMENT_STYLEB(flags) (((flags) >> 21) & 1) | |
77 | #define SYNTAX_FLAGS_COMMENT_STYLEC(flags) (((flags) >> 22) & 2) | |
78 | /* FLAGS should be the flags of the main char of the comment marker, e.g. | |
79 | the second for comstart and the first for comend. */ | |
80 | #define SYNTAX_FLAGS_COMMENT_STYLE(flags, other_flags) \ | |
81 | (SYNTAX_FLAGS_COMMENT_STYLEB (flags) \ | |
82 | | SYNTAX_FLAGS_COMMENT_STYLEC (flags) \ | |
83 | | SYNTAX_FLAGS_COMMENT_STYLEC (other_flags)) | |
84 | ||
85 | #define SYNTAX_FLAGS_COMMENT_NESTED(flags) (((flags) >> 22) & 1) | |
86 | ||
87 | /* These macros extract a particular flag for a given character. */ | |
88 | ||
89 | #define SYNTAX_COMEND_FIRST(c) \ | |
90 | (SYNTAX_FLAGS_COMEND_FIRST (SYNTAX_WITH_FLAGS (c))) | |
91 | #define SYNTAX_PREFIX(c) (SYNTAX_FLAGS_PREFIX (SYNTAX_WITH_FLAGS (c))) | |
195d1361 RS |
92 | |
93 | /* We use these constants in place for comment-style and | |
94 | string-ender-char to distinguish comments/strings started by | |
95 | comment_fence and string_fence codes. */ | |
96 | ||
97 | #define ST_COMMENT_STYLE (256 + 1) | |
98 | #define ST_STRING_STYLE (256 + 2) | |
8489eb67 | 99 | |
955cbe7b PE |
100 | static Lisp_Object Qsyntax_table_p; |
101 | static Lisp_Object Qsyntax_table, Qscan_error; | |
8489eb67 | 102 | |
d4b43b22 | 103 | #ifndef __GNUC__ |
8ea151b2 RS |
104 | /* Used as a temporary in SYNTAX_ENTRY and other macros in syntax.h, |
105 | if not compiled with GCC. No need to mark it, since it is used | |
106 | only very temporarily. */ | |
107 | Lisp_Object syntax_temp; | |
d4b43b22 | 108 | #endif |
8ea151b2 | 109 | |
e5d4f4dc RS |
110 | /* This is the internal form of the parse state used in parse-partial-sexp. */ |
111 | ||
112 | struct lisp_parse_state | |
113 | { | |
6449674e SM |
114 | int depth; /* Depth at end of parsing. */ |
115 | int instring; /* -1 if not within string, else desired terminator. */ | |
116 | int incomment; /* -1 if in unnestable comment else comment nesting */ | |
117 | int comstyle; /* comment style a=0, or b=1, or ST_COMMENT_STYLE. */ | |
118 | int quoted; /* Nonzero if just after an escape char at end of parsing */ | |
119 | int mindepth; /* Minimum depth seen while scanning. */ | |
120 | /* Char number of most recent start-of-expression at current level */ | |
121 | EMACS_INT thislevelstart; | |
122 | /* Char number of start of containing expression */ | |
123 | EMACS_INT prevlevelstart; | |
124 | EMACS_INT location; /* Char number at which parsing stopped. */ | |
125 | EMACS_INT comstr_start; /* Position of last comment/string starter. */ | |
126 | Lisp_Object levelstarts; /* Char numbers of starts-of-expression | |
127 | of levels (starting from outermost). */ | |
e5d4f4dc RS |
128 | }; |
129 | \f | |
37bef230 RS |
130 | /* These variables are a cache for finding the start of a defun. |
131 | find_start_pos is the place for which the defun start was found. | |
132 | find_start_value is the defun start position found for it. | |
6a140a74 | 133 | find_start_value_byte is the corresponding byte position. |
37bef230 RS |
134 | find_start_buffer is the buffer it was found in. |
135 | find_start_begv is the BEGV value when it was found. | |
136 | find_start_modiff is the value of MODIFF when it was found. */ | |
137 | ||
8e2911c2 AS |
138 | static EMACS_INT find_start_pos; |
139 | static EMACS_INT find_start_value; | |
140 | static EMACS_INT find_start_value_byte; | |
37bef230 | 141 | static struct buffer *find_start_buffer; |
8e2911c2 | 142 | static EMACS_INT find_start_begv; |
37bef230 | 143 | static int find_start_modiff; |
6a140a74 RS |
144 | |
145 | ||
cd64ea1d | 146 | static Lisp_Object Fsyntax_table_p (Lisp_Object); |
f57e2426 J |
147 | static Lisp_Object skip_chars (int, Lisp_Object, Lisp_Object, int); |
148 | static Lisp_Object skip_syntaxes (int, Lisp_Object, Lisp_Object); | |
149 | static Lisp_Object scan_lists (EMACS_INT, EMACS_INT, EMACS_INT, int); | |
150 | static void scan_sexps_forward (struct lisp_parse_state *, | |
151 | EMACS_INT, EMACS_INT, EMACS_INT, int, | |
152 | int, Lisp_Object, int); | |
153 | static int in_classes (int, Lisp_Object); | |
195d1361 | 154 | \f |
c5683ceb SM |
155 | /* Whether the syntax of the character C has the prefix flag set. */ |
156 | int syntax_prefix_flag_p (int c) | |
157 | { | |
158 | return SYNTAX_PREFIX (c); | |
159 | } | |
195d1361 RS |
160 | |
161 | struct gl_state_s gl_state; /* Global state of syntax parser. */ | |
162 | ||
195d1361 RS |
163 | #define INTERVALS_AT_ONCE 10 /* 1 + max-number of intervals |
164 | to scan to property-change. */ | |
165 | ||
6a140a74 RS |
166 | /* Update gl_state to an appropriate interval which contains CHARPOS. The |
167 | sign of COUNT give the relative position of CHARPOS wrt the previously | |
195d1361 | 168 | valid interval. If INIT, only [be]_property fields of gl_state are |
6a140a74 | 169 | valid at start, the rest is filled basing on OBJECT. |
195d1361 | 170 | |
6a140a74 | 171 | `gl_state.*_i' are the intervals, and CHARPOS is further in the search |
195d1361 RS |
172 | direction than the intervals - or in an interval. We update the |
173 | current syntax-table basing on the property of this interval, and | |
6a140a74 | 174 | update the interval to start further than CHARPOS - or be |
195d1361 | 175 | NULL_INTERVAL. We also update lim_property to be the next value of |
6a140a74 | 176 | charpos to call this subroutine again - or be before/after the |
195d1361 RS |
177 | start/end of OBJECT. */ |
178 | ||
179 | void | |
d1dfb56c | 180 | update_syntax_table (EMACS_INT charpos, EMACS_INT count, int init, |
4f3a2f8d | 181 | Lisp_Object object) |
195d1361 RS |
182 | { |
183 | Lisp_Object tmp_table; | |
12cbf13f PE |
184 | unsigned cnt = 0; |
185 | int invalidate = 1; | |
5a774522 | 186 | INTERVAL i; |
195d1361 RS |
187 | |
188 | if (init) | |
189 | { | |
bb0de084 | 190 | gl_state.old_prop = Qnil; |
195d1361 RS |
191 | gl_state.start = gl_state.b_property; |
192 | gl_state.stop = gl_state.e_property; | |
bb0de084 SM |
193 | i = interval_of (charpos, object); |
194 | gl_state.backward_i = gl_state.forward_i = i; | |
195d1361 RS |
195 | invalidate = 0; |
196 | if (NULL_INTERVAL_P (i)) | |
197 | return; | |
f902a008 | 198 | /* interval_of updates only ->position of the return value, so |
d80f4cc7 | 199 | update the parents manually to speed up update_interval. */ |
bb0de084 | 200 | while (!NULL_PARENT (i)) |
d80f4cc7 RS |
201 | { |
202 | if (AM_RIGHT_CHILD (i)) | |
439d5cb4 | 203 | INTERVAL_PARENT (i)->position = i->position |
d80f4cc7 | 204 | - LEFT_TOTAL_LENGTH (i) + TOTAL_LENGTH (i) /* right end */ |
439d5cb4 KR |
205 | - TOTAL_LENGTH (INTERVAL_PARENT (i)) |
206 | + LEFT_TOTAL_LENGTH (INTERVAL_PARENT (i)); | |
d80f4cc7 | 207 | else |
439d5cb4 | 208 | INTERVAL_PARENT (i)->position = i->position - LEFT_TOTAL_LENGTH (i) |
d80f4cc7 | 209 | + TOTAL_LENGTH (i); |
439d5cb4 | 210 | i = INTERVAL_PARENT (i); |
d80f4cc7 RS |
211 | } |
212 | i = gl_state.forward_i; | |
bb0de084 | 213 | gl_state.b_property = i->position - gl_state.offset; |
ee0cdb48 | 214 | gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset; |
195d1361 RS |
215 | goto update; |
216 | } | |
5a774522 | 217 | i = count > 0 ? gl_state.forward_i : gl_state.backward_i; |
195d1361 | 218 | |
423e705d | 219 | /* We are guaranteed to be called with CHARPOS either in i, |
6a140a74 | 220 | or further off. */ |
195d1361 RS |
221 | if (NULL_INTERVAL_P (i)) |
222 | error ("Error in syntax_table logic for to-the-end intervals"); | |
6a140a74 | 223 | else if (charpos < i->position) /* Move left. */ |
195d1361 RS |
224 | { |
225 | if (count > 0) | |
f902a008 | 226 | error ("Error in syntax_table logic for intervals <-"); |
195d1361 | 227 | /* Update the interval. */ |
6a140a74 | 228 | i = update_interval (i, charpos); |
bb0de084 | 229 | if (INTERVAL_LAST_POS (i) != gl_state.b_property) |
195d1361 RS |
230 | { |
231 | invalidate = 0; | |
195d1361 | 232 | gl_state.forward_i = i; |
ee0cdb48 | 233 | gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset; |
195d1361 | 234 | } |
423e705d | 235 | } |
6a140a74 | 236 | else if (charpos >= INTERVAL_LAST_POS (i)) /* Move right. */ |
195d1361 RS |
237 | { |
238 | if (count < 0) | |
f902a008 | 239 | error ("Error in syntax_table logic for intervals ->"); |
195d1361 | 240 | /* Update the interval. */ |
6a140a74 | 241 | i = update_interval (i, charpos); |
bb0de084 | 242 | if (i->position != gl_state.e_property) |
195d1361 RS |
243 | { |
244 | invalidate = 0; | |
195d1361 | 245 | gl_state.backward_i = i; |
bb0de084 | 246 | gl_state.b_property = i->position - gl_state.offset; |
195d1361 RS |
247 | } |
248 | } | |
195d1361 RS |
249 | |
250 | update: | |
251 | tmp_table = textget (i->plist, Qsyntax_table); | |
252 | ||
253 | if (invalidate) | |
254 | invalidate = !EQ (tmp_table, gl_state.old_prop); /* Need to invalidate? */ | |
7d0393cf | 255 | |
423e705d SM |
256 | if (invalidate) /* Did not get to adjacent interval. */ |
257 | { /* with the same table => */ | |
258 | /* invalidate the old range. */ | |
195d1361 RS |
259 | if (count > 0) |
260 | { | |
261 | gl_state.backward_i = i; | |
bb0de084 SM |
262 | gl_state.b_property = i->position - gl_state.offset; |
263 | } | |
264 | else | |
195d1361 | 265 | { |
bb0de084 | 266 | gl_state.forward_i = i; |
ee0cdb48 | 267 | gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset; |
195d1361 RS |
268 | } |
269 | } | |
37bef230 | 270 | |
bb0de084 | 271 | if (!EQ (tmp_table, gl_state.old_prop)) |
195d1361 | 272 | { |
bb0de084 SM |
273 | gl_state.current_syntax_table = tmp_table; |
274 | gl_state.old_prop = tmp_table; | |
275 | if (EQ (Fsyntax_table_p (tmp_table), Qt)) | |
276 | { | |
277 | gl_state.use_global = 0; | |
7d0393cf | 278 | } |
bb0de084 SM |
279 | else if (CONSP (tmp_table)) |
280 | { | |
281 | gl_state.use_global = 1; | |
282 | gl_state.global_code = tmp_table; | |
283 | } | |
7d0393cf | 284 | else |
bb0de084 SM |
285 | { |
286 | gl_state.use_global = 0; | |
4b4deea2 | 287 | gl_state.current_syntax_table = BVAR (current_buffer, syntax_table); |
bb0de084 | 288 | } |
195d1361 RS |
289 | } |
290 | ||
291 | while (!NULL_INTERVAL_P (i)) | |
292 | { | |
293 | if (cnt && !EQ (tmp_table, textget (i->plist, Qsyntax_table))) | |
294 | { | |
295 | if (count > 0) | |
bb0de084 SM |
296 | { |
297 | gl_state.e_property = i->position - gl_state.offset; | |
298 | gl_state.forward_i = i; | |
299 | } | |
300 | else | |
301 | { | |
5a774522 SM |
302 | gl_state.b_property |
303 | = i->position + LENGTH (i) - gl_state.offset; | |
bb0de084 SM |
304 | gl_state.backward_i = i; |
305 | } | |
306 | return; | |
195d1361 | 307 | } |
7d0393cf | 308 | else if (cnt == INTERVALS_AT_ONCE) |
195d1361 RS |
309 | { |
310 | if (count > 0) | |
bb0de084 | 311 | { |
5a774522 SM |
312 | gl_state.e_property |
313 | = i->position + LENGTH (i) - gl_state.offset | |
314 | /* e_property at EOB is not set to ZV but to ZV+1, so that | |
315 | we can do INC(from);UPDATE_SYNTAX_TABLE_FORWARD without | |
316 | having to check eob between the two. */ | |
317 | + (NULL_INTERVAL_P (next_interval (i)) ? 1 : 0); | |
bb0de084 SM |
318 | gl_state.forward_i = i; |
319 | } | |
320 | else | |
321 | { | |
322 | gl_state.b_property = i->position - gl_state.offset; | |
323 | gl_state.backward_i = i; | |
324 | } | |
325 | return; | |
195d1361 RS |
326 | } |
327 | cnt++; | |
328 | i = count > 0 ? next_interval (i) : previous_interval (i); | |
329 | } | |
bb0de084 SM |
330 | eassert (NULL_INTERVAL_P (i)); /* This property goes to the end. */ |
331 | if (count > 0) | |
332 | gl_state.e_property = gl_state.stop; | |
333 | else | |
334 | gl_state.b_property = gl_state.start; | |
195d1361 RS |
335 | } |
336 | \f | |
6a140a74 RS |
337 | /* Returns TRUE if char at CHARPOS is quoted. |
338 | Global syntax-table data should be set up already to be good at CHARPOS | |
339 | or after. On return global syntax data is good for lookup at CHARPOS. */ | |
195d1361 RS |
340 | |
341 | static int | |
6449674e | 342 | char_quoted (EMACS_INT charpos, EMACS_INT bytepos) |
195d1361 RS |
343 | { |
344 | register enum syntaxcode code; | |
6449674e | 345 | register EMACS_INT beg = BEGV; |
195d1361 | 346 | register int quoted = 0; |
6449674e | 347 | EMACS_INT orig = charpos; |
6a140a74 | 348 | |
02d8b017 | 349 | while (charpos > beg) |
195d1361 | 350 | { |
ab229fdd | 351 | int c; |
02d8b017 | 352 | DEC_BOTH (charpos, bytepos); |
ab229fdd | 353 | |
6a140a74 | 354 | UPDATE_SYNTAX_TABLE_BACKWARD (charpos); |
327719ee | 355 | c = FETCH_CHAR_AS_MULTIBYTE (bytepos); |
ab229fdd | 356 | code = SYNTAX (c); |
6a140a74 RS |
357 | if (! (code == Scharquote || code == Sescape)) |
358 | break; | |
359 | ||
6a140a74 | 360 | quoted = !quoted; |
195d1361 | 361 | } |
6a140a74 RS |
362 | |
363 | UPDATE_SYNTAX_TABLE (orig); | |
195d1361 RS |
364 | return quoted; |
365 | } | |
6a140a74 | 366 | |
6a140a74 RS |
367 | /* Return the bytepos one character before BYTEPOS. |
368 | We assume that BYTEPOS is not at the start of the buffer. */ | |
369 | ||
55d4c1b2 | 370 | static inline EMACS_INT |
971de7fb | 371 | dec_bytepos (EMACS_INT bytepos) |
6a140a74 | 372 | { |
4b4deea2 | 373 | if (NILP (BVAR (current_buffer, enable_multibyte_characters))) |
ef316cf0 RS |
374 | return bytepos - 1; |
375 | ||
6a140a74 RS |
376 | DEC_POS (bytepos); |
377 | return bytepos; | |
378 | } | |
195d1361 | 379 | \f |
2b34df4e | 380 | /* Return a defun-start position before POS and not too far before. |
7cc80f0a RS |
381 | It should be the last one before POS, or nearly the last. |
382 | ||
383 | When open_paren_in_column_0_is_defun_start is nonzero, | |
1fd1cc2f | 384 | only the beginning of the buffer is treated as a defun-start. |
7cc80f0a RS |
385 | |
386 | We record the information about where the scan started | |
387 | and what its result was, so that another call in the same area | |
388 | can return the same value very quickly. | |
195d1361 RS |
389 | |
390 | There is no promise at which position the global syntax data is | |
391 | valid on return from the subroutine, so the caller should explicitly | |
392 | update the global data. */ | |
37bef230 | 393 | |
6449674e | 394 | static EMACS_INT |
971de7fb | 395 | find_defun_start (EMACS_INT pos, EMACS_INT pos_byte) |
37bef230 | 396 | { |
8e2911c2 | 397 | EMACS_INT opoint = PT, opoint_byte = PT_BYTE; |
37bef230 | 398 | |
1fd1cc2f RS |
399 | if (!open_paren_in_column_0_is_defun_start) |
400 | { | |
401 | find_start_value_byte = BEGV_BYTE; | |
402 | return BEGV; | |
403 | } | |
404 | ||
37bef230 RS |
405 | /* Use previous finding, if it's valid and applies to this inquiry. */ |
406 | if (current_buffer == find_start_buffer | |
407 | /* Reuse the defun-start even if POS is a little farther on. | |
408 | POS might be in the next defun, but that's ok. | |
409 | Our value may not be the best possible, but will still be usable. */ | |
410 | && pos <= find_start_pos + 1000 | |
411 | && pos >= find_start_value | |
412 | && BEGV == find_start_begv | |
413 | && MODIFF == find_start_modiff) | |
414 | return find_start_value; | |
415 | ||
416 | /* Back up to start of line. */ | |
6a140a74 | 417 | scan_newline (pos, pos_byte, BEGV, BEGV_BYTE, -1, 1); |
37bef230 | 418 | |
195d1361 RS |
419 | /* We optimize syntax-table lookup for rare updates. Thus we accept |
420 | only those `^\s(' which are good in global _and_ text-property | |
421 | syntax-tables. */ | |
d48cd3f4 | 422 | SETUP_BUFFER_SYNTAX_TABLE (); |
1fd1cc2f | 423 | while (PT > BEGV) |
37bef230 | 424 | { |
ab229fdd AS |
425 | int c; |
426 | ||
1fd1cc2f RS |
427 | /* Open-paren at start of line means we may have found our |
428 | defun-start. */ | |
327719ee | 429 | c = FETCH_CHAR_AS_MULTIBYTE (PT_BYTE); |
ab229fdd | 430 | if (SYNTAX (c) == Sopen) |
195d1361 | 431 | { |
1fd1cc2f | 432 | SETUP_SYNTAX_TABLE (PT + 1, -1); /* Try again... */ |
327719ee | 433 | c = FETCH_CHAR_AS_MULTIBYTE (PT_BYTE); |
ab229fdd | 434 | if (SYNTAX (c) == Sopen) |
1fd1cc2f RS |
435 | break; |
436 | /* Now fallback to the default value. */ | |
d48cd3f4 | 437 | SETUP_BUFFER_SYNTAX_TABLE (); |
195d1361 | 438 | } |
1fd1cc2f RS |
439 | /* Move to beg of previous line. */ |
440 | scan_newline (PT, PT_BYTE, BEGV, BEGV_BYTE, -2, 1); | |
37bef230 RS |
441 | } |
442 | ||
443 | /* Record what we found, for the next try. */ | |
6a140a74 RS |
444 | find_start_value = PT; |
445 | find_start_value_byte = PT_BYTE; | |
37bef230 RS |
446 | find_start_buffer = current_buffer; |
447 | find_start_modiff = MODIFF; | |
448 | find_start_begv = BEGV; | |
449 | find_start_pos = pos; | |
450 | ||
6a140a74 RS |
451 | TEMP_SET_PT_BOTH (opoint, opoint_byte); |
452 | ||
37bef230 RS |
453 | return find_start_value; |
454 | } | |
455 | \f | |
f902a008 RS |
456 | /* Return the SYNTAX_COMEND_FIRST of the character before POS, POS_BYTE. */ |
457 | ||
458 | static int | |
4f3a2f8d | 459 | prev_char_comend_first (EMACS_INT pos, EMACS_INT pos_byte) |
f902a008 RS |
460 | { |
461 | int c, val; | |
462 | ||
463 | DEC_BOTH (pos, pos_byte); | |
464 | UPDATE_SYNTAX_TABLE_BACKWARD (pos); | |
465 | c = FETCH_CHAR (pos_byte); | |
466 | val = SYNTAX_COMEND_FIRST (c); | |
467 | UPDATE_SYNTAX_TABLE_FORWARD (pos + 1); | |
468 | return val; | |
469 | } | |
470 | ||
471 | /* Return the SYNTAX_COMSTART_FIRST of the character before POS, POS_BYTE. */ | |
472 | ||
3f679f55 SM |
473 | /* static int |
474 | * prev_char_comstart_first (pos, pos_byte) | |
475 | * int pos, pos_byte; | |
476 | * { | |
477 | * int c, val; | |
7d0393cf | 478 | * |
3f679f55 SM |
479 | * DEC_BOTH (pos, pos_byte); |
480 | * UPDATE_SYNTAX_TABLE_BACKWARD (pos); | |
481 | * c = FETCH_CHAR (pos_byte); | |
482 | * val = SYNTAX_COMSTART_FIRST (c); | |
483 | * UPDATE_SYNTAX_TABLE_FORWARD (pos + 1); | |
484 | * return val; | |
485 | * } */ | |
f902a008 | 486 | |
6a140a74 RS |
487 | /* Checks whether charpos FROM is at the end of a comment. |
488 | FROM_BYTE is the bytepos corresponding to FROM. | |
489 | Do not move back before STOP. | |
490 | ||
491 | Return a positive value if we find a comment ending at FROM/FROM_BYTE; | |
492 | return -1 otherwise. | |
493 | ||
494 | If successful, store the charpos of the comment's beginning | |
495 | into *CHARPOS_PTR, and the bytepos into *BYTEPOS_PTR. | |
527a32d9 KH |
496 | |
497 | Global syntax data remains valid for backward search starting at | |
498 | the returned value (or at FROM, if the search was not successful). */ | |
195d1361 RS |
499 | |
500 | static int | |
971de7fb | 501 | back_comment (EMACS_INT from, EMACS_INT from_byte, EMACS_INT stop, int comnested, int comstyle, EMACS_INT *charpos_ptr, EMACS_INT *bytepos_ptr) |
195d1361 RS |
502 | { |
503 | /* Look back, counting the parity of string-quotes, | |
504 | and recording the comment-starters seen. | |
505 | When we reach a safe place, assume that's not in a string; | |
506 | then step the main scan to the earliest comment-starter seen | |
507 | an even number of string quotes away from the safe place. | |
508 | ||
509 | OFROM[I] is position of the earliest comment-starter seen | |
510 | which is I+2X quotes from the comment-end. | |
511 | PARITY is current parity of quotes from the comment end. */ | |
3ee5041c | 512 | int string_style = -1; /* Presumed outside of any string. */ |
195d1361 | 513 | int string_lossage = 0; |
f7c436c1 | 514 | /* Not a real lossage: indicates that we have passed a matching comment |
7d0393cf | 515 | starter plus a non-matching comment-ender, meaning that any matching |
f7c436c1 SM |
516 | comment-starter we might see later could be a false positive (hidden |
517 | inside another comment). | |
518 | Test case: { a (* b } c (* d *) */ | |
519 | int comment_lossage = 0; | |
8e2911c2 AS |
520 | EMACS_INT comment_end = from; |
521 | EMACS_INT comment_end_byte = from_byte; | |
522 | EMACS_INT comstart_pos = 0; | |
4f63c6bb | 523 | EMACS_INT comstart_byte IF_LINT (= 0); |
eb35b628 RS |
524 | /* Place where the containing defun starts, |
525 | or 0 if we didn't come across it yet. */ | |
8e2911c2 AS |
526 | EMACS_INT defun_start = 0; |
527 | EMACS_INT defun_start_byte = 0; | |
195d1361 | 528 | register enum syntaxcode code; |
95ff8dfc | 529 | int nesting = 1; /* current comment nesting */ |
ea315ed6 | 530 | int c; |
3f679f55 SM |
531 | int syntax = 0; |
532 | ||
533 | /* FIXME: A }} comment-ender style leads to incorrect behavior | |
534 | in the case of {{ c }}} because we ignore the last two chars which are | |
535 | assumed to be comment-enders although they aren't. */ | |
195d1361 RS |
536 | |
537 | /* At beginning of range to scan, we're outside of strings; | |
538 | that determines quote parity to the comment-end. */ | |
539 | while (from != stop) | |
540 | { | |
4f3a2f8d EZ |
541 | EMACS_INT temp_byte; |
542 | int prev_syntax, com2start, com2end; | |
fbb3da77 | 543 | int comstart; |
6a140a74 | 544 | |
195d1361 | 545 | /* Move back and examine a character. */ |
6a140a74 | 546 | DEC_BOTH (from, from_byte); |
195d1361 RS |
547 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
548 | ||
3f679f55 | 549 | prev_syntax = syntax; |
b7dbcc19 | 550 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
3f679f55 | 551 | syntax = SYNTAX_WITH_FLAGS (c); |
195d1361 RS |
552 | code = SYNTAX (c); |
553 | ||
3f679f55 SM |
554 | /* Check for 2-char comment markers. */ |
555 | com2start = (SYNTAX_FLAGS_COMSTART_FIRST (syntax) | |
556 | && SYNTAX_FLAGS_COMSTART_SECOND (prev_syntax) | |
c5683ceb SM |
557 | && (comstyle |
558 | == SYNTAX_FLAGS_COMMENT_STYLE (prev_syntax, syntax)) | |
3f679f55 SM |
559 | && (SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax) |
560 | || SYNTAX_FLAGS_COMMENT_NESTED (syntax)) == comnested); | |
561 | com2end = (SYNTAX_FLAGS_COMEND_FIRST (syntax) | |
562 | && SYNTAX_FLAGS_COMEND_SECOND (prev_syntax)); | |
fbb3da77 | 563 | comstart = (com2start || code == Scomment); |
01f44d5a | 564 | |
3f679f55 SM |
565 | /* Nasty cases with overlapping 2-char comment markers: |
566 | - snmp-mode: -- c -- foo -- c -- | |
567 | --- c -- | |
568 | ------ c -- | |
569 | - c-mode: *||* | |
570 | |* *|* *| | |
571 | |*| |* |*| | |
572 | /// */ | |
573 | ||
574 | /* If a 2-char comment sequence partly overlaps with another, | |
fbb3da77 SM |
575 | we don't try to be clever. E.g. |*| in C, or }% in modes that |
576 | have %..\n and %{..}%. */ | |
577 | if (from > stop && (com2end || comstart)) | |
195d1361 | 578 | { |
4f3a2f8d EZ |
579 | EMACS_INT next = from, next_byte = from_byte; |
580 | int next_c, next_syntax; | |
3f679f55 SM |
581 | DEC_BOTH (next, next_byte); |
582 | UPDATE_SYNTAX_TABLE_BACKWARD (next); | |
b7dbcc19 | 583 | next_c = FETCH_CHAR_AS_MULTIBYTE (next_byte); |
3f679f55 | 584 | next_syntax = SYNTAX_WITH_FLAGS (next_c); |
fbb3da77 | 585 | if (((comstart || comnested) |
3f679f55 SM |
586 | && SYNTAX_FLAGS_COMEND_SECOND (syntax) |
587 | && SYNTAX_FLAGS_COMEND_FIRST (next_syntax)) | |
588 | || ((com2end || comnested) | |
589 | && SYNTAX_FLAGS_COMSTART_SECOND (syntax) | |
c5683ceb SM |
590 | && (comstyle |
591 | == SYNTAX_FLAGS_COMMENT_STYLE (syntax, prev_syntax)) | |
3f679f55 SM |
592 | && SYNTAX_FLAGS_COMSTART_FIRST (next_syntax))) |
593 | goto lossage; | |
594 | /* UPDATE_SYNTAX_TABLE_FORWARD (next + 1); */ | |
f902a008 | 595 | } |
3f679f55 SM |
596 | |
597 | if (com2start && comstart_pos == 0) | |
598 | /* We're looking at a comment starter. But it might be a comment | |
599 | ender as well (see snmp-mode). The first time we see one, we | |
600 | need to consider it as a comment starter, | |
601 | and the subsequent times as a comment ender. */ | |
602 | com2end = 0; | |
603 | ||
604 | /* Turn a 2-char comment sequences into the appropriate syntax. */ | |
605 | if (com2end) | |
606 | code = Sendcomment; | |
607 | else if (com2start) | |
608 | code = Scomment; | |
609 | /* Ignore comment starters of a different style. */ | |
610 | else if (code == Scomment | |
c5683ceb | 611 | && (comstyle != SYNTAX_FLAGS_COMMENT_STYLE (syntax, 0) |
3f679f55 | 612 | || SYNTAX_FLAGS_COMMENT_NESTED (syntax) != comnested)) |
1aa963c8 | 613 | continue; |
195d1361 | 614 | |
9828a477 KH |
615 | /* Ignore escaped characters, except comment-enders. */ |
616 | if (code != Sendcomment && char_quoted (from, from_byte)) | |
195d1361 RS |
617 | continue; |
618 | ||
3ee5041c | 619 | switch (code) |
195d1361 | 620 | { |
3ee5041c SM |
621 | case Sstring_fence: |
622 | case Scomment_fence: | |
623 | c = (code == Sstring_fence ? ST_STRING_STYLE : ST_COMMENT_STYLE); | |
624 | case Sstring: | |
625 | /* Track parity of quotes. */ | |
626 | if (string_style == -1) | |
627 | /* Entering a string. */ | |
628 | string_style = c; | |
629 | else if (string_style == c) | |
630 | /* Leaving the string. */ | |
631 | string_style = -1; | |
632 | else | |
633 | /* If we have two kinds of string delimiters. | |
634 | There's no way to grok this scanning backwards. */ | |
195d1361 | 635 | string_lossage = 1; |
3ee5041c | 636 | break; |
7d0393cf | 637 | |
3ee5041c SM |
638 | case Scomment: |
639 | /* We've already checked that it is the relevant comstyle. */ | |
f7c436c1 | 640 | if (string_style != -1 || comment_lossage || string_lossage) |
3ee5041c SM |
641 | /* There are odd string quotes involved, so let's be careful. |
642 | Test case in Pascal: " { " a { " } */ | |
643 | goto lossage; | |
644 | ||
f7c436c1 SM |
645 | if (!comnested) |
646 | { | |
647 | /* Record best comment-starter so far. */ | |
648 | comstart_pos = from; | |
649 | comstart_byte = from_byte; | |
650 | } | |
651 | else if (--nesting <= 0) | |
95ff8dfc RS |
652 | /* nested comments have to be balanced, so we don't need to |
653 | keep looking for earlier ones. We use here the same (slightly | |
654 | incorrect) reasoning as below: since it is followed by uniform | |
655 | paired string quotes, this comment-start has to be outside of | |
656 | strings, else the comment-end itself would be inside a string. */ | |
657 | goto done; | |
3ee5041c SM |
658 | break; |
659 | ||
02010917 | 660 | case Sendcomment: |
c5683ceb | 661 | if (SYNTAX_FLAGS_COMMENT_STYLE (syntax, 0) == comstyle |
d070eb22 | 662 | && ((com2end && SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax)) |
3f679f55 | 663 | || SYNTAX_FLAGS_COMMENT_NESTED (syntax)) == comnested) |
02010917 SM |
664 | /* This is the same style of comment ender as ours. */ |
665 | { | |
666 | if (comnested) | |
667 | nesting++; | |
668 | else | |
669 | /* Anything before that can't count because it would match | |
670 | this comment-ender rather than ours. */ | |
671 | from = stop; /* Break out of the loop. */ | |
672 | } | |
f7c436c1 SM |
673 | else if (comstart_pos != 0 || c != '\n') |
674 | /* We're mixing comment styles here, so we'd better be careful. | |
675 | The (comstart_pos != 0 || c != '\n') check is not quite correct | |
676 | (we should just always set comment_lossage), but removing it | |
677 | would imply that any multiline comment in C would go through | |
678 | lossage, which seems overkill. | |
679 | The failure should only happen in the rare cases such as | |
680 | { (* } *) */ | |
681 | comment_lossage = 1; | |
02010917 | 682 | break; |
195d1361 | 683 | |
02010917 SM |
684 | case Sopen: |
685 | /* Assume a defun-start point is outside of strings. */ | |
686 | if (open_paren_in_column_0_is_defun_start | |
687 | && (from == stop | |
688 | || (temp_byte = dec_bytepos (from_byte), | |
689 | FETCH_CHAR (temp_byte) == '\n'))) | |
690 | { | |
691 | defun_start = from; | |
692 | defun_start_byte = from_byte; | |
693 | from = stop; /* Break out of the loop. */ | |
694 | } | |
eb35b628 | 695 | break; |
02010917 SM |
696 | |
697 | default: | |
f7c436c1 | 698 | break; |
eb35b628 | 699 | } |
195d1361 RS |
700 | } |
701 | ||
702 | if (comstart_pos == 0) | |
703 | { | |
704 | from = comment_end; | |
6a140a74 | 705 | from_byte = comment_end_byte; |
195d1361 RS |
706 | UPDATE_SYNTAX_TABLE_FORWARD (comment_end - 1); |
707 | } | |
f7c436c1 SM |
708 | /* If comstart_pos is set and we get here (ie. didn't jump to `lossage' |
709 | or `done'), then we've found the beginning of the non-nested comment. */ | |
710 | else if (1) /* !comnested */ | |
195d1361 RS |
711 | { |
712 | from = comstart_pos; | |
6a140a74 | 713 | from_byte = comstart_byte; |
789f3320 | 714 | UPDATE_SYNTAX_TABLE_FORWARD (from - 1); |
195d1361 RS |
715 | } |
716 | else | |
717 | { | |
3ee5041c SM |
718 | struct lisp_parse_state state; |
719 | lossage: | |
195d1361 RS |
720 | /* We had two kinds of string delimiters mixed up |
721 | together. Decode this going forwards. | |
02010917 | 722 | Scan fwd from a known safe place (beginning-of-defun) |
195d1361 RS |
723 | to the one in question; this records where we |
724 | last passed a comment starter. */ | |
eb35b628 RS |
725 | /* If we did not already find the defun start, find it now. */ |
726 | if (defun_start == 0) | |
727 | { | |
728 | defun_start = find_defun_start (comment_end, comment_end_byte); | |
729 | defun_start_byte = find_start_value_byte; | |
730 | } | |
02010917 | 731 | do |
195d1361 | 732 | { |
02010917 SM |
733 | scan_sexps_forward (&state, |
734 | defun_start, defun_start_byte, | |
735 | comment_end, -10000, 0, Qnil, 0); | |
736 | defun_start = comment_end; | |
737 | if (state.incomment == (comnested ? 1 : -1) | |
738 | && state.comstyle == comstyle) | |
739 | from = state.comstr_start; | |
740 | else | |
741 | { | |
742 | from = comment_end; | |
743 | if (state.incomment) | |
744 | /* If comment_end is inside some other comment, maybe ours | |
745 | is nested, so we need to try again from within the | |
746 | surrounding comment. Example: { a (* " *) */ | |
747 | { | |
748 | /* FIXME: We should advance by one or two chars. */ | |
749 | defun_start = state.comstr_start + 2; | |
750 | defun_start_byte = CHAR_TO_BYTE (defun_start); | |
751 | } | |
752 | } | |
753 | } while (defun_start < comment_end); | |
754 | ||
6a140a74 | 755 | from_byte = CHAR_TO_BYTE (from); |
195d1361 RS |
756 | UPDATE_SYNTAX_TABLE_FORWARD (from - 1); |
757 | } | |
7d0393cf | 758 | |
95ff8dfc | 759 | done: |
6a140a74 RS |
760 | *charpos_ptr = from; |
761 | *bytepos_ptr = from_byte; | |
762 | ||
1aa963c8 | 763 | return (from == comment_end) ? -1 : from; |
195d1361 RS |
764 | } |
765 | \f | |
8489eb67 | 766 | DEFUN ("syntax-table-p", Fsyntax_table_p, Ssyntax_table_p, 1, 1, 0, |
fdb82f93 PJ |
767 | doc: /* Return t if OBJECT is a syntax table. |
768 | Currently, any char-table counts as a syntax table. */) | |
5842a27b | 769 | (Lisp_Object object) |
8489eb67 | 770 | { |
2203e1e8 | 771 | if (CHAR_TABLE_P (object) |
e704cb4b | 772 | && EQ (XCHAR_TABLE (object)->purpose, Qsyntax_table)) |
8489eb67 RS |
773 | return Qt; |
774 | return Qnil; | |
775 | } | |
776 | ||
8ea151b2 | 777 | static void |
971de7fb | 778 | check_syntax_table (Lisp_Object obj) |
8489eb67 | 779 | { |
47f5f6ae KS |
780 | CHECK_TYPE (CHAR_TABLE_P (obj) && EQ (XCHAR_TABLE (obj)->purpose, Qsyntax_table), |
781 | Qsyntax_table_p, obj); | |
7d0393cf | 782 | } |
8489eb67 | 783 | |
8489eb67 | 784 | DEFUN ("syntax-table", Fsyntax_table, Ssyntax_table, 0, 0, 0, |
fdb82f93 PJ |
785 | doc: /* Return the current syntax table. |
786 | This is the one specified by the current buffer. */) | |
5842a27b | 787 | (void) |
8489eb67 | 788 | { |
4b4deea2 | 789 | return BVAR (current_buffer, syntax_table); |
8489eb67 RS |
790 | } |
791 | ||
792 | DEFUN ("standard-syntax-table", Fstandard_syntax_table, | |
793 | Sstandard_syntax_table, 0, 0, 0, | |
fdb82f93 PJ |
794 | doc: /* Return the standard syntax table. |
795 | This is the one used for new buffers. */) | |
5842a27b | 796 | (void) |
8489eb67 RS |
797 | { |
798 | return Vstandard_syntax_table; | |
799 | } | |
800 | ||
801 | DEFUN ("copy-syntax-table", Fcopy_syntax_table, Scopy_syntax_table, 0, 1, 0, | |
fdb82f93 PJ |
802 | doc: /* Construct a new syntax table and return it. |
803 | It is a copy of the TABLE, which defaults to the standard syntax table. */) | |
5842a27b | 804 | (Lisp_Object table) |
8489eb67 | 805 | { |
8ea151b2 RS |
806 | Lisp_Object copy; |
807 | ||
265a9e55 | 808 | if (!NILP (table)) |
8ea151b2 RS |
809 | check_syntax_table (table); |
810 | else | |
811 | table = Vstandard_syntax_table; | |
812 | ||
813 | copy = Fcopy_sequence (table); | |
0f867324 RS |
814 | |
815 | /* Only the standard syntax table should have a default element. | |
816 | Other syntax tables should inherit from parents instead. */ | |
817 | XCHAR_TABLE (copy)->defalt = Qnil; | |
818 | ||
819 | /* Copied syntax tables should all have parents. | |
820 | If we copied one with no parent, such as the standard syntax table, | |
821 | use the standard syntax table as the copy's parent. */ | |
822 | if (NILP (XCHAR_TABLE (copy)->parent)) | |
823 | Fset_char_table_parent (copy, Vstandard_syntax_table); | |
8ea151b2 | 824 | return copy; |
8489eb67 RS |
825 | } |
826 | ||
827 | DEFUN ("set-syntax-table", Fset_syntax_table, Sset_syntax_table, 1, 1, 0, | |
fdb82f93 PJ |
828 | doc: /* Select a new syntax table for the current buffer. |
829 | One argument, a syntax table. */) | |
5842a27b | 830 | (Lisp_Object table) |
8489eb67 | 831 | { |
4e4dfa78 | 832 | int idx; |
8ea151b2 | 833 | check_syntax_table (table); |
4b4deea2 | 834 | BVAR (current_buffer, syntax_table) = table; |
8489eb67 | 835 | /* Indicate that this buffer now has a specified syntax table. */ |
f6cd0527 GM |
836 | idx = PER_BUFFER_VAR_IDX (syntax_table); |
837 | SET_PER_BUFFER_VALUE_P (current_buffer, idx, 1); | |
8489eb67 RS |
838 | return table; |
839 | } | |
840 | \f | |
841 | /* Convert a letter which signifies a syntax code | |
842 | into the code it signifies. | |
195d1361 | 843 | This is used by modify-syntax-entry, and other things. */ |
8489eb67 RS |
844 | |
845 | unsigned char syntax_spec_code[0400] = | |
846 | { 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
847 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
848 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
849 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
195d1361 | 850 | (char) Swhitespace, (char) Scomment_fence, (char) Sstring, 0377, |
8489eb67 RS |
851 | (char) Smath, 0377, 0377, (char) Squote, |
852 | (char) Sopen, (char) Sclose, 0377, 0377, | |
853 | 0377, (char) Swhitespace, (char) Spunct, (char) Scharquote, | |
854 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
855 | 0377, 0377, 0377, 0377, | |
856 | (char) Scomment, 0377, (char) Sendcomment, 0377, | |
6cb71bf6 | 857 | (char) Sinherit, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* @, A ... */ |
8489eb67 RS |
858 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, |
859 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword, | |
860 | 0377, 0377, 0377, 0377, (char) Sescape, 0377, 0377, (char) Ssymbol, | |
861 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* `, a, ... */ | |
862 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
863 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword, | |
195d1361 | 864 | 0377, 0377, 0377, 0377, (char) Sstring_fence, 0377, 0377, 0377 |
8489eb67 RS |
865 | }; |
866 | ||
195d1361 | 867 | /* Indexed by syntax code, give the letter that describes it. */ |
8489eb67 | 868 | |
195d1361 | 869 | char syntax_code_spec[16] = |
8489eb67 | 870 | { |
195d1361 RS |
871 | ' ', '.', 'w', '_', '(', ')', '\'', '\"', '$', '\\', '/', '<', '>', '@', |
872 | '!', '|' | |
8489eb67 | 873 | }; |
93da5fff KH |
874 | |
875 | /* Indexed by syntax code, give the object (cons of syntax code and | |
876 | nil) to be stored in syntax table. Since these objects can be | |
877 | shared among syntax tables, we generate them in advance. By | |
878 | sharing objects, the function `describe-syntax' can give a more | |
879 | compact listing. */ | |
880 | static Lisp_Object Vsyntax_code_object; | |
881 | ||
8489eb67 RS |
882 | \f |
883 | DEFUN ("char-syntax", Fchar_syntax, Schar_syntax, 1, 1, 0, | |
fdb82f93 | 884 | doc: /* Return the syntax code of CHARACTER, described by a character. |
e2b6daf4 CY |
885 | For example, if CHARACTER is a word constituent, the |
886 | character `w' (119) is returned. | |
fdb82f93 PJ |
887 | The characters that correspond to various syntax codes |
888 | are listed in the documentation of `modify-syntax-entry'. */) | |
5842a27b | 889 | (Lisp_Object character) |
8489eb67 | 890 | { |
8ea151b2 | 891 | int char_int; |
774b9a60 | 892 | CHECK_CHARACTER (character); |
2203e1e8 | 893 | char_int = XINT (character); |
d48cd3f4 | 894 | SETUP_BUFFER_SYNTAX_TABLE (); |
8ea151b2 | 895 | return make_number (syntax_code_spec[(int) SYNTAX (char_int)]); |
beefa22e RS |
896 | } |
897 | ||
898 | DEFUN ("matching-paren", Fmatching_paren, Smatching_paren, 1, 1, 0, | |
fdb82f93 | 899 | doc: /* Return the matching parenthesis of CHARACTER, or nil if none. */) |
5842a27b | 900 | (Lisp_Object character) |
beefa22e | 901 | { |
8ea151b2 | 902 | int char_int, code; |
b7826503 | 903 | CHECK_NUMBER (character); |
2203e1e8 | 904 | char_int = XINT (character); |
d48cd3f4 | 905 | SETUP_BUFFER_SYNTAX_TABLE (); |
8ea151b2 | 906 | code = SYNTAX (char_int); |
a8bd7cd8 | 907 | if (code == Sopen || code == Sclose) |
2e34157c | 908 | return SYNTAX_MATCH (char_int); |
beefa22e | 909 | return Qnil; |
8489eb67 RS |
910 | } |
911 | ||
c65adb44 | 912 | DEFUN ("string-to-syntax", Fstring_to_syntax, Sstring_to_syntax, 1, 1, 0, |
fdb82f93 PJ |
913 | doc: /* Convert a syntax specification STRING into syntax cell form. |
914 | STRING should be a string as it is allowed as argument of | |
915 | `modify-syntax-entry'. Value is the equivalent cons cell | |
bce3dc36 | 916 | \(CODE . MATCHING-CHAR) that can be used as value of a `syntax-table' |
fdb82f93 | 917 | text property. */) |
5842a27b | 918 | (Lisp_Object string) |
8489eb67 | 919 | { |
2e567bd3 | 920 | register const unsigned char *p; |
8489eb67 | 921 | register enum syntaxcode code; |
247e20a8 | 922 | int val; |
8ea151b2 | 923 | Lisp_Object match; |
8489eb67 | 924 | |
b7826503 | 925 | CHECK_STRING (string); |
8489eb67 | 926 | |
d5db4077 | 927 | p = SDATA (string); |
8489eb67 RS |
928 | code = (enum syntaxcode) syntax_spec_code[*p++]; |
929 | if (((int) code & 0377) == 0377) | |
b764a653 | 930 | error ("Invalid syntax description letter: %c", p[-1]); |
8489eb67 | 931 | |
8ea151b2 | 932 | if (code == Sinherit) |
c65adb44 | 933 | return Qnil; |
8ea151b2 RS |
934 | |
935 | if (*p) | |
d1be9f0f | 936 | { |
93da5fff | 937 | int len; |
62a6e103 | 938 | int character = STRING_CHAR_AND_LENGTH (p, len); |
93da5fff | 939 | XSETINT (match, character); |
d1be9f0f RS |
940 | if (XFASTINT (match) == ' ') |
941 | match = Qnil; | |
93da5fff | 942 | p += len; |
d1be9f0f RS |
943 | } |
944 | else | |
8ea151b2 | 945 | match = Qnil; |
8489eb67 | 946 | |
8ea151b2 | 947 | val = (int) code; |
8489eb67 RS |
948 | while (*p) |
949 | switch (*p++) | |
950 | { | |
951 | case '1': | |
247e20a8 | 952 | val |= 1 << 16; |
8489eb67 RS |
953 | break; |
954 | ||
955 | case '2': | |
247e20a8 | 956 | val |= 1 << 17; |
8489eb67 RS |
957 | break; |
958 | ||
959 | case '3': | |
247e20a8 | 960 | val |= 1 << 18; |
8489eb67 RS |
961 | break; |
962 | ||
963 | case '4': | |
247e20a8 | 964 | val |= 1 << 19; |
8489eb67 RS |
965 | break; |
966 | ||
967 | case 'p': | |
247e20a8 | 968 | val |= 1 << 20; |
8489eb67 | 969 | break; |
e5d4f4dc RS |
970 | |
971 | case 'b': | |
247e20a8 | 972 | val |= 1 << 21; |
e5d4f4dc | 973 | break; |
95ff8dfc RS |
974 | |
975 | case 'n': | |
976 | val |= 1 << 22; | |
977 | break; | |
c5683ceb SM |
978 | |
979 | case 'c': | |
980 | val |= 1 << 23; | |
981 | break; | |
8489eb67 | 982 | } |
7d0393cf | 983 | |
77b37c05 | 984 | if (val < ASIZE (Vsyntax_code_object) && NILP (match)) |
c65adb44 | 985 | return XVECTOR (Vsyntax_code_object)->contents[val]; |
93da5fff KH |
986 | else |
987 | /* Since we can't use a shared object, let's make a new one. */ | |
c65adb44 SM |
988 | return Fcons (make_number (val), match); |
989 | } | |
990 | ||
fdb82f93 | 991 | /* I really don't know why this is interactive |
2c9e1900 | 992 | help-form should at least be made useful whilst reading the second arg. */ |
7d0393cf | 993 | DEFUN ("modify-syntax-entry", Fmodify_syntax_entry, Smodify_syntax_entry, 2, 3, |
c65adb44 | 994 | "cSet syntax for character: \nsSet syntax for %s to: ", |
a50a10a0 | 995 | doc: /* Set syntax for character CHAR according to string NEWENTRY. |
d7ee9fab | 996 | The syntax is changed only for table SYNTAX-TABLE, which defaults to |
fdb82f93 | 997 | the current buffer's syntax table. |
5c7b02ab | 998 | CHAR may be a cons (MIN . MAX), in which case, syntaxes of all characters |
fedc6ab5 | 999 | in the range MIN to MAX are changed. |
fdb82f93 PJ |
1000 | The first character of NEWENTRY should be one of the following: |
1001 | Space or - whitespace syntax. w word constituent. | |
1002 | _ symbol constituent. . punctuation. | |
1003 | ( open-parenthesis. ) close-parenthesis. | |
1004 | " string quote. \\ escape. | |
1005 | $ paired delimiter. ' expression quote or prefix operator. | |
1006 | < comment starter. > comment ender. | |
1007 | / character-quote. @ inherit from `standard-syntax-table'. | |
1008 | | generic string fence. ! generic comment fence. | |
1009 | ||
1010 | Only single-character comment start and end sequences are represented thus. | |
1011 | Two-character sequences are represented as described below. | |
1012 | The second character of NEWENTRY is the matching parenthesis, | |
1013 | used only if the first character is `(' or `)'. | |
1014 | Any additional characters are flags. | |
1015 | Defined flags are the characters 1, 2, 3, 4, b, p, and n. | |
a50a10a0 PJ |
1016 | 1 means CHAR is the start of a two-char comment start sequence. |
1017 | 2 means CHAR is the second character of such a sequence. | |
1018 | 3 means CHAR is the start of a two-char comment end sequence. | |
1019 | 4 means CHAR is the second character of such a sequence. | |
fdb82f93 | 1020 | |
c5683ceb | 1021 | There can be several orthogonal comment sequences. This is to support |
fdb82f93 PJ |
1022 | language modes such as C++. By default, all comment sequences are of style |
1023 | a, but you can set the comment sequence style to b (on the second character | |
c5683ceb SM |
1024 | of a comment-start, and the first character of a comment-end sequence) and/or |
1025 | c (on any of its chars) using this flag: | |
a50a10a0 | 1026 | b means CHAR is part of comment sequence b. |
c5683ceb | 1027 | c means CHAR is part of comment sequence c. |
a50a10a0 | 1028 | n means CHAR is part of a nestable comment sequence. |
fdb82f93 | 1029 | |
a50a10a0 | 1030 | p means CHAR is a prefix character for `backward-prefix-chars'; |
fdb82f93 | 1031 | such characters are treated as whitespace when they occur |
a50a10a0 | 1032 | between expressions. |
fedc6ab5 | 1033 | usage: (modify-syntax-entry CHAR NEWENTRY &optional SYNTAX-TABLE) */) |
5842a27b | 1034 | (Lisp_Object c, Lisp_Object newentry, Lisp_Object syntax_table) |
c65adb44 | 1035 | { |
5c7b02ab KH |
1036 | if (CONSP (c)) |
1037 | { | |
8f924df7 KH |
1038 | CHECK_CHARACTER_CAR (c); |
1039 | CHECK_CHARACTER_CDR (c); | |
5c7b02ab KH |
1040 | } |
1041 | else | |
1042 | CHECK_CHARACTER (c); | |
c65adb44 SM |
1043 | |
1044 | if (NILP (syntax_table)) | |
4b4deea2 | 1045 | syntax_table = BVAR (current_buffer, syntax_table); |
c65adb44 SM |
1046 | else |
1047 | check_syntax_table (syntax_table); | |
8489eb67 | 1048 | |
5c7b02ab KH |
1049 | newentry = Fstring_to_syntax (newentry); |
1050 | if (CONSP (c)) | |
1051 | SET_RAW_SYNTAX_ENTRY_RANGE (syntax_table, c, newentry); | |
1052 | else | |
1053 | SET_RAW_SYNTAX_ENTRY (syntax_table, XINT (c), newentry); | |
e5b94d44 CY |
1054 | |
1055 | /* We clear the regexp cache, since character classes can now have | |
1056 | different values from those in the compiled regexps.*/ | |
1057 | clear_regexp_cache (); | |
1058 | ||
8489eb67 RS |
1059 | return Qnil; |
1060 | } | |
1061 | \f | |
1062 | /* Dump syntax table to buffer in human-readable format */ | |
1063 | ||
62abe9cb SM |
1064 | DEFUN ("internal-describe-syntax-value", Finternal_describe_syntax_value, |
1065 | Sinternal_describe_syntax_value, 1, 1, 0, | |
1066 | doc: /* Insert a description of the internal syntax description SYNTAX at point. */) | |
5842a27b | 1067 | (Lisp_Object syntax) |
8489eb67 RS |
1068 | { |
1069 | register enum syntaxcode code; | |
c5683ceb SM |
1070 | int syntax_code; |
1071 | char desc, start1, start2, end1, end2, prefix, | |
1072 | comstyleb, comstylec, comnested; | |
8489eb67 | 1073 | char str[2]; |
62abe9cb | 1074 | Lisp_Object first, match_lisp, value = syntax; |
8489eb67 | 1075 | |
8ea151b2 RS |
1076 | if (NILP (value)) |
1077 | { | |
62abe9cb SM |
1078 | insert_string ("default"); |
1079 | return syntax; | |
8ea151b2 RS |
1080 | } |
1081 | ||
908b7fea KH |
1082 | if (CHAR_TABLE_P (value)) |
1083 | { | |
62abe9cb SM |
1084 | insert_string ("deeper char-table ..."); |
1085 | return syntax; | |
908b7fea KH |
1086 | } |
1087 | ||
8ea151b2 RS |
1088 | if (!CONSP (value)) |
1089 | { | |
62abe9cb SM |
1090 | insert_string ("invalid"); |
1091 | return syntax; | |
8ea151b2 RS |
1092 | } |
1093 | ||
c1d497be KR |
1094 | first = XCAR (value); |
1095 | match_lisp = XCDR (value); | |
8ea151b2 RS |
1096 | |
1097 | if (!INTEGERP (first) || !(NILP (match_lisp) || INTEGERP (match_lisp))) | |
8489eb67 | 1098 | { |
62abe9cb SM |
1099 | insert_string ("invalid"); |
1100 | return syntax; | |
8489eb67 RS |
1101 | } |
1102 | ||
c5683ceb SM |
1103 | syntax_code = XINT (first); |
1104 | code = (enum syntaxcode) (syntax_code & 0377); | |
1105 | start1 = SYNTAX_FLAGS_COMSTART_FIRST (syntax_code); | |
1106 | start2 = SYNTAX_FLAGS_COMSTART_SECOND (syntax_code);; | |
1107 | end1 = SYNTAX_FLAGS_COMEND_FIRST (syntax_code); | |
1108 | end2 = SYNTAX_FLAGS_COMEND_SECOND (syntax_code); | |
1109 | prefix = SYNTAX_FLAGS_PREFIX (syntax_code); | |
1110 | comstyleb = SYNTAX_FLAGS_COMMENT_STYLEB (syntax_code); | |
1111 | comstylec = SYNTAX_FLAGS_COMMENT_STYLEC (syntax_code); | |
1112 | comnested = SYNTAX_FLAGS_COMMENT_NESTED (syntax_code); | |
8489eb67 RS |
1113 | |
1114 | if ((int) code < 0 || (int) code >= (int) Smax) | |
1115 | { | |
1116 | insert_string ("invalid"); | |
62abe9cb | 1117 | return syntax; |
8489eb67 RS |
1118 | } |
1119 | desc = syntax_code_spec[(int) code]; | |
1120 | ||
1121 | str[0] = desc, str[1] = 0; | |
1122 | insert (str, 1); | |
1123 | ||
93da5fff KH |
1124 | if (NILP (match_lisp)) |
1125 | insert (" ", 1); | |
1126 | else | |
1127 | insert_char (XINT (match_lisp)); | |
8489eb67 | 1128 | |
8489eb67 RS |
1129 | if (start1) |
1130 | insert ("1", 1); | |
1131 | if (start2) | |
1132 | insert ("2", 1); | |
1133 | ||
1134 | if (end1) | |
1135 | insert ("3", 1); | |
1136 | if (end2) | |
1137 | insert ("4", 1); | |
1138 | ||
1139 | if (prefix) | |
1140 | insert ("p", 1); | |
c5683ceb | 1141 | if (comstyleb) |
e5d4f4dc | 1142 | insert ("b", 1); |
c5683ceb SM |
1143 | if (comstylec) |
1144 | insert ("c", 1); | |
e6365855 SM |
1145 | if (comnested) |
1146 | insert ("n", 1); | |
8489eb67 RS |
1147 | |
1148 | insert_string ("\twhich means: "); | |
1149 | ||
0220c518 | 1150 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
1151 | { |
1152 | case Swhitespace: | |
1153 | insert_string ("whitespace"); break; | |
1154 | case Spunct: | |
1155 | insert_string ("punctuation"); break; | |
1156 | case Sword: | |
1157 | insert_string ("word"); break; | |
1158 | case Ssymbol: | |
1159 | insert_string ("symbol"); break; | |
1160 | case Sopen: | |
1161 | insert_string ("open"); break; | |
1162 | case Sclose: | |
1163 | insert_string ("close"); break; | |
1164 | case Squote: | |
d671382e | 1165 | insert_string ("prefix"); break; |
8489eb67 RS |
1166 | case Sstring: |
1167 | insert_string ("string"); break; | |
1168 | case Smath: | |
1169 | insert_string ("math"); break; | |
1170 | case Sescape: | |
1171 | insert_string ("escape"); break; | |
1172 | case Scharquote: | |
1173 | insert_string ("charquote"); break; | |
1174 | case Scomment: | |
1175 | insert_string ("comment"); break; | |
1176 | case Sendcomment: | |
1177 | insert_string ("endcomment"); break; | |
d671382e SM |
1178 | case Sinherit: |
1179 | insert_string ("inherit"); break; | |
1180 | case Scomment_fence: | |
1181 | insert_string ("comment fence"); break; | |
1182 | case Sstring_fence: | |
1183 | insert_string ("string fence"); break; | |
8489eb67 RS |
1184 | default: |
1185 | insert_string ("invalid"); | |
62abe9cb | 1186 | return syntax; |
8489eb67 RS |
1187 | } |
1188 | ||
8ea151b2 | 1189 | if (!NILP (match_lisp)) |
8489eb67 RS |
1190 | { |
1191 | insert_string (", matches "); | |
8ea151b2 | 1192 | insert_char (XINT (match_lisp)); |
8489eb67 RS |
1193 | } |
1194 | ||
1195 | if (start1) | |
1196 | insert_string (",\n\t is the first character of a comment-start sequence"); | |
1197 | if (start2) | |
1198 | insert_string (",\n\t is the second character of a comment-start sequence"); | |
1199 | ||
1200 | if (end1) | |
1201 | insert_string (",\n\t is the first character of a comment-end sequence"); | |
1202 | if (end2) | |
1203 | insert_string (",\n\t is the second character of a comment-end sequence"); | |
c5683ceb | 1204 | if (comstyleb) |
e5d4f4dc | 1205 | insert_string (" (comment style b)"); |
c5683ceb SM |
1206 | if (comstylec) |
1207 | insert_string (" (comment style c)"); | |
e6365855 SM |
1208 | if (comnested) |
1209 | insert_string (" (nestable)"); | |
e5d4f4dc | 1210 | |
8489eb67 RS |
1211 | if (prefix) |
1212 | insert_string (",\n\t is a prefix character for `backward-prefix-chars'"); | |
1213 | ||
62abe9cb SM |
1214 | return syntax; |
1215 | } | |
8489eb67 RS |
1216 | \f |
1217 | /* Return the position across COUNT words from FROM. | |
1218 | If that many words cannot be found before the end of the buffer, return 0. | |
1219 | COUNT negative means scan backward and stop at word beginning. */ | |
1220 | ||
4f3a2f8d EZ |
1221 | EMACS_INT |
1222 | scan_words (register EMACS_INT from, register EMACS_INT count) | |
8489eb67 | 1223 | { |
4f3a2f8d EZ |
1224 | register EMACS_INT beg = BEGV; |
1225 | register EMACS_INT end = ZV; | |
1226 | register EMACS_INT from_byte = CHAR_TO_BYTE (from); | |
93da5fff KH |
1227 | register enum syntaxcode code; |
1228 | int ch0, ch1; | |
fca8fe46 | 1229 | Lisp_Object func, pos; |
8489eb67 RS |
1230 | |
1231 | immediate_quit = 1; | |
1232 | QUIT; | |
1233 | ||
195d1361 RS |
1234 | SETUP_SYNTAX_TABLE (from, count); |
1235 | ||
8489eb67 RS |
1236 | while (count > 0) |
1237 | { | |
1238 | while (1) | |
1239 | { | |
1240 | if (from == end) | |
1241 | { | |
1242 | immediate_quit = 0; | |
1243 | return 0; | |
1244 | } | |
195d1361 | 1245 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 1246 | ch0 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
93da5fff | 1247 | code = SYNTAX (ch0); |
6a140a74 | 1248 | INC_BOTH (from, from_byte); |
8489eb67 RS |
1249 | if (words_include_escapes |
1250 | && (code == Sescape || code == Scharquote)) | |
1251 | break; | |
1252 | if (code == Sword) | |
1253 | break; | |
8489eb67 | 1254 | } |
93da5fff KH |
1255 | /* Now CH0 is a character which begins a word and FROM is the |
1256 | position of the next character. */ | |
8f924df7 | 1257 | func = CHAR_TABLE_REF (Vfind_word_boundary_function_table, ch0); |
869bb237 | 1258 | if (! NILP (Ffboundp (func))) |
8489eb67 | 1259 | { |
869bb237 | 1260 | pos = call2 (func, make_number (from - 1), make_number (end)); |
8f924df7 KH |
1261 | if (INTEGERP (pos) && XINT (pos) > from) |
1262 | { | |
1263 | from = XINT (pos); | |
1264 | from_byte = CHAR_TO_BYTE (from); | |
1265 | } | |
8489eb67 | 1266 | } |
869bb237 | 1267 | else |
8cb8232a | 1268 | { |
8cb8232a KH |
1269 | while (1) |
1270 | { | |
1271 | if (from == end) break; | |
1272 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
b7dbcc19 | 1273 | ch1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8cb8232a KH |
1274 | code = SYNTAX (ch1); |
1275 | if ((code != Sword | |
1276 | && (! words_include_escapes | |
1277 | || (code != Sescape && code != Scharquote))) | |
2b70a11c | 1278 | || word_boundary_p (ch0, ch1)) |
869bb237 | 1279 | break; |
8cb8232a KH |
1280 | INC_BOTH (from, from_byte); |
1281 | ch0 = ch1; | |
1282 | } | |
8489eb67 RS |
1283 | } |
1284 | count--; | |
1285 | } | |
1286 | while (count < 0) | |
1287 | { | |
1288 | while (1) | |
1289 | { | |
1290 | if (from == beg) | |
1291 | { | |
1292 | immediate_quit = 0; | |
1293 | return 0; | |
1294 | } | |
6a140a74 | 1295 | DEC_BOTH (from, from_byte); |
195d1361 | 1296 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 1297 | ch1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
93da5fff | 1298 | code = SYNTAX (ch1); |
8489eb67 RS |
1299 | if (words_include_escapes |
1300 | && (code == Sescape || code == Scharquote)) | |
1301 | break; | |
1302 | if (code == Sword) | |
1303 | break; | |
8489eb67 | 1304 | } |
93da5fff KH |
1305 | /* Now CH1 is a character which ends a word and FROM is the |
1306 | position of it. */ | |
8f924df7 | 1307 | func = CHAR_TABLE_REF (Vfind_word_boundary_function_table, ch1); |
869bb237 | 1308 | if (! NILP (Ffboundp (func))) |
8f924df7 | 1309 | { |
869bb237 | 1310 | pos = call2 (func, make_number (from), make_number (beg)); |
8f924df7 KH |
1311 | if (INTEGERP (pos) && XINT (pos) < from) |
1312 | { | |
1313 | from = XINT (pos); | |
1314 | from_byte = CHAR_TO_BYTE (from); | |
1315 | } | |
8489eb67 | 1316 | } |
869bb237 | 1317 | else |
8489eb67 | 1318 | { |
8cb8232a KH |
1319 | while (1) |
1320 | { | |
8cb8232a KH |
1321 | if (from == beg) |
1322 | break; | |
262be72a | 1323 | DEC_BOTH (from, from_byte); |
8cb8232a | 1324 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
262be72a | 1325 | ch0 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8cb8232a KH |
1326 | code = SYNTAX (ch0); |
1327 | if ((code != Sword | |
1328 | && (! words_include_escapes | |
1329 | || (code != Sescape && code != Scharquote))) | |
2b70a11c | 1330 | || word_boundary_p (ch0, ch1)) |
262be72a MB |
1331 | { |
1332 | INC_BOTH (from, from_byte); | |
1333 | break; | |
1334 | } | |
8cb8232a KH |
1335 | ch1 = ch0; |
1336 | } | |
8489eb67 RS |
1337 | } |
1338 | count++; | |
1339 | } | |
1340 | ||
1341 | immediate_quit = 0; | |
1342 | ||
1343 | return from; | |
1344 | } | |
1345 | ||
a7ca3326 | 1346 | DEFUN ("forward-word", Fforward_word, Sforward_word, 0, 1, "^p", |
fdb82f93 PJ |
1347 | doc: /* Move point forward ARG words (backward if ARG is negative). |
1348 | Normally returns t. | |
1349 | If an edge of the buffer or a field boundary is reached, point is left there | |
1350 | and the function returns nil. Field boundaries are not noticed if | |
1351 | `inhibit-field-text-motion' is non-nil. */) | |
5842a27b | 1352 | (Lisp_Object arg) |
8489eb67 | 1353 | { |
c096ae4d | 1354 | Lisp_Object tmp; |
2c6ea900 | 1355 | int orig_val, val; |
8489eb67 | 1356 | |
839966f3 KH |
1357 | if (NILP (arg)) |
1358 | XSETFASTINT (arg, 1); | |
1359 | else | |
1360 | CHECK_NUMBER (arg); | |
1361 | ||
1362 | val = orig_val = scan_words (PT, XINT (arg)); | |
2c6ea900 | 1363 | if (! orig_val) |
839966f3 | 1364 | val = XINT (arg) > 0 ? ZV : BEGV; |
5878ee6f | 1365 | |
b8855607 | 1366 | /* Avoid jumping out of an input field. */ |
c096ae4d SM |
1367 | tmp = Fconstrain_to_field (make_number (val), make_number (PT), |
1368 | Qt, Qnil, Qnil); | |
1369 | val = XFASTINT (tmp); | |
7d0393cf | 1370 | |
8489eb67 | 1371 | SET_PT (val); |
e6d8341f | 1372 | return val == orig_val ? Qt : Qnil; |
8489eb67 RS |
1373 | } |
1374 | \f | |
a7ca3326 | 1375 | DEFUN ("skip-chars-forward", Fskip_chars_forward, Sskip_chars_forward, 1, 2, 0, |
fdb82f93 PJ |
1376 | doc: /* Move point forward, stopping before a char not in STRING, or at pos LIM. |
1377 | STRING is like the inside of a `[...]' in a regular expression | |
1378 | except that `]' is never special and `\\' quotes `^', `-' or `\\' | |
7087d5e9 | 1379 | (but not at the end of a range; quoting is never needed there). |
fdb82f93 PJ |
1380 | Thus, with arg "a-zA-Z", this skips letters stopping before first nonletter. |
1381 | With arg "^a-zA-Z", skips nonletters stopping before first letter. | |
a1bc88d4 RS |
1382 | Char classes, e.g. `[:alpha:]', are supported. |
1383 | ||
1384 | Returns the distance traveled, either zero or positive. */) | |
5842a27b | 1385 | (Lisp_Object string, Lisp_Object lim) |
195d1361 | 1386 | { |
327719ee | 1387 | return skip_chars (1, string, lim, 1); |
195d1361 RS |
1388 | } |
1389 | ||
a7ca3326 | 1390 | DEFUN ("skip-chars-backward", Fskip_chars_backward, Sskip_chars_backward, 1, 2, 0, |
fdb82f93 PJ |
1391 | doc: /* Move point backward, stopping after a char not in STRING, or at pos LIM. |
1392 | See `skip-chars-forward' for details. | |
1393 | Returns the distance traveled, either zero or negative. */) | |
5842a27b | 1394 | (Lisp_Object string, Lisp_Object lim) |
195d1361 | 1395 | { |
327719ee | 1396 | return skip_chars (0, string, lim, 1); |
195d1361 RS |
1397 | } |
1398 | ||
1399 | DEFUN ("skip-syntax-forward", Fskip_syntax_forward, Sskip_syntax_forward, 1, 2, 0, | |
fdb82f93 PJ |
1400 | doc: /* Move point forward across chars in specified syntax classes. |
1401 | SYNTAX is a string of syntax code characters. | |
1402 | Stop before a char whose syntax is not in SYNTAX, or at position LIM. | |
1403 | If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX. | |
1404 | This function returns the distance traveled, either zero or positive. */) | |
5842a27b | 1405 | (Lisp_Object syntax, Lisp_Object lim) |
195d1361 | 1406 | { |
b7dbcc19 | 1407 | return skip_syntaxes (1, syntax, lim); |
195d1361 RS |
1408 | } |
1409 | ||
1410 | DEFUN ("skip-syntax-backward", Fskip_syntax_backward, Sskip_syntax_backward, 1, 2, 0, | |
fdb82f93 PJ |
1411 | doc: /* Move point backward across chars in specified syntax classes. |
1412 | SYNTAX is a string of syntax code characters. | |
1413 | Stop on reaching a char whose syntax is not in SYNTAX, or at position LIM. | |
1414 | If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX. | |
1415 | This function returns the distance traveled, either zero or negative. */) | |
5842a27b | 1416 | (Lisp_Object syntax, Lisp_Object lim) |
195d1361 | 1417 | { |
b7dbcc19 | 1418 | return skip_syntaxes (0, syntax, lim); |
195d1361 RS |
1419 | } |
1420 | ||
6a140a74 | 1421 | static Lisp_Object |
971de7fb | 1422 | skip_chars (int forwardp, Lisp_Object string, Lisp_Object lim, int handle_iso_classes) |
195d1361 | 1423 | { |
195d1361 RS |
1424 | register unsigned int c; |
1425 | unsigned char fastmap[0400]; | |
b7dbcc19 | 1426 | /* Store the ranges of non-ASCII characters. */ |
4f63c6bb | 1427 | int *char_ranges IF_LINT (= NULL); |
7e68b0ea | 1428 | int n_char_ranges = 0; |
195d1361 | 1429 | int negate = 0; |
4f3a2f8d | 1430 | register EMACS_INT i, i_byte; |
b7dbcc19 KH |
1431 | /* Set to 1 if the current buffer is multibyte and the region |
1432 | contains non-ASCII chars. */ | |
1433 | int multibyte; | |
1434 | /* Set to 1 if STRING is multibyte and it contains non-ASCII | |
1435 | chars. */ | |
1674d9a2 | 1436 | int string_multibyte; |
4f3a2f8d | 1437 | EMACS_INT size_byte; |
2e567bd3 | 1438 | const unsigned char *str; |
82d497fc | 1439 | int len; |
a1bc88d4 | 1440 | Lisp_Object iso_classes; |
195d1361 | 1441 | |
b7826503 | 1442 | CHECK_STRING (string); |
a1bc88d4 | 1443 | iso_classes = Qnil; |
82d497fc | 1444 | |
195d1361 RS |
1445 | if (NILP (lim)) |
1446 | XSETINT (lim, forwardp ? ZV : BEGV); | |
1447 | else | |
b7826503 | 1448 | CHECK_NUMBER_COERCE_MARKER (lim); |
195d1361 RS |
1449 | |
1450 | /* In any case, don't allow scan outside bounds of buffer. */ | |
195d1361 RS |
1451 | if (XINT (lim) > ZV) |
1452 | XSETFASTINT (lim, ZV); | |
1453 | if (XINT (lim) < BEGV) | |
1454 | XSETFASTINT (lim, BEGV); | |
1455 | ||
4b4deea2 | 1456 | multibyte = (!NILP (BVAR (current_buffer, enable_multibyte_characters)) |
92eaa22e | 1457 | && (XINT (lim) - PT != CHAR_TO_BYTE (XINT (lim)) - PT_BYTE)); |
8f924df7 | 1458 | string_multibyte = SBYTES (string) > SCHARS (string); |
b7dbcc19 | 1459 | |
72af86bd | 1460 | memset (fastmap, 0, sizeof fastmap); |
195d1361 | 1461 | |
8f924df7 KH |
1462 | str = SDATA (string); |
1463 | size_byte = SBYTES (string); | |
4101e6fe | 1464 | |
82d497fc | 1465 | i_byte = 0; |
13090112 | 1466 | if (i_byte < size_byte |
d5db4077 | 1467 | && SREF (string, 0) == '^') |
195d1361 | 1468 | { |
82d497fc | 1469 | negate = 1; i_byte++; |
195d1361 RS |
1470 | } |
1471 | ||
1472 | /* Find the characters specified and set their elements of fastmap. | |
b7dbcc19 | 1473 | Handle backslashes and ranges specially. |
195d1361 | 1474 | |
b7dbcc19 KH |
1475 | If STRING contains non-ASCII characters, setup char_ranges for |
1476 | them and use fastmap only for their leading codes. */ | |
4101e6fe | 1477 | |
b7dbcc19 | 1478 | if (! string_multibyte) |
195d1361 | 1479 | { |
b7dbcc19 | 1480 | int string_has_eight_bit = 0; |
4101e6fe | 1481 | |
b7dbcc19 KH |
1482 | /* At first setup fastmap. */ |
1483 | while (i_byte < size_byte) | |
1484 | { | |
1485 | c = str[i_byte++]; | |
1486 | ||
a1bc88d4 RS |
1487 | if (handle_iso_classes && c == '[' |
1488 | && i_byte < size_byte | |
327719ee | 1489 | && str[i_byte] == ':') |
a1bc88d4 RS |
1490 | { |
1491 | const unsigned char *class_beg = str + i_byte + 1; | |
1492 | const unsigned char *class_end = class_beg; | |
b3bda4fd | 1493 | const unsigned char *class_limit = str + size_byte - 2; |
db9cd97a | 1494 | /* Leave room for the null. */ |
a1bc88d4 RS |
1495 | unsigned char class_name[CHAR_CLASS_MAX_LENGTH + 1]; |
1496 | re_wctype_t cc; | |
1497 | ||
1498 | if (class_limit - class_beg > CHAR_CLASS_MAX_LENGTH) | |
1499 | class_limit = class_beg + CHAR_CLASS_MAX_LENGTH; | |
1500 | ||
b3bda4fd KS |
1501 | while (class_end < class_limit |
1502 | && *class_end >= 'a' && *class_end <= 'z') | |
a1bc88d4 RS |
1503 | class_end++; |
1504 | ||
b3bda4fd KS |
1505 | if (class_end == class_beg |
1506 | || *class_end != ':' || class_end[1] != ']') | |
1507 | goto not_a_class_name; | |
a1bc88d4 | 1508 | |
72af86bd | 1509 | memcpy (class_name, class_beg, class_end - class_beg); |
a1bc88d4 RS |
1510 | class_name[class_end - class_beg] = 0; |
1511 | ||
1512 | cc = re_wctype (class_name); | |
1513 | if (cc == 0) | |
1514 | error ("Invalid ISO C character class"); | |
1515 | ||
1516 | iso_classes = Fcons (make_number (cc), iso_classes); | |
1517 | ||
1518 | i_byte = class_end + 2 - str; | |
1519 | continue; | |
1520 | } | |
1521 | ||
b3bda4fd | 1522 | not_a_class_name: |
b7dbcc19 KH |
1523 | if (c == '\\') |
1524 | { | |
13090112 | 1525 | if (i_byte == size_byte) |
4101e6fe RS |
1526 | break; |
1527 | ||
5c7b02ab | 1528 | c = str[i_byte++]; |
195d1361 | 1529 | } |
839966f3 KH |
1530 | /* Treat `-' as range character only if another character |
1531 | follows. */ | |
1532 | if (i_byte + 1 < size_byte | |
82d497fc | 1533 | && str[i_byte] == '-') |
195d1361 | 1534 | { |
9690d026 | 1535 | unsigned int c2; |
4101e6fe RS |
1536 | |
1537 | /* Skip over the dash. */ | |
82d497fc | 1538 | i_byte++; |
4101e6fe | 1539 | |
4101e6fe | 1540 | /* Get the end of the range. */ |
5c7b02ab | 1541 | c2 = str[i_byte++]; |
b7dbcc19 KH |
1542 | if (c2 == '\\' |
1543 | && i_byte < size_byte) | |
1544 | c2 = str[i_byte++]; | |
1545 | ||
8f924df7 KH |
1546 | if (c <= c2) |
1547 | { | |
1e69125e PE |
1548 | unsigned lim2 = c2 + 1; |
1549 | while (c < lim2) | |
8f924df7 KH |
1550 | fastmap[c++] = 1; |
1551 | if (! ASCII_CHAR_P (c2)) | |
1552 | string_has_eight_bit = 1; | |
1553 | } | |
195d1361 RS |
1554 | } |
1555 | else | |
b7dbcc19 KH |
1556 | { |
1557 | fastmap[c] = 1; | |
1558 | if (! ASCII_CHAR_P (c)) | |
1559 | string_has_eight_bit = 1; | |
1560 | } | |
1561 | } | |
1562 | ||
1563 | /* If the current range is multibyte and STRING contains | |
1564 | eight-bit chars, arrange fastmap and setup char_ranges for | |
1565 | the corresponding multibyte chars. */ | |
1566 | if (multibyte && string_has_eight_bit) | |
1567 | { | |
1568 | unsigned char fastmap2[0400]; | |
1569 | int range_start_byte, range_start_char; | |
1570 | ||
72af86bd AS |
1571 | memcpy (fastmap + 0200, fastmap2 + 0200, 0200); |
1572 | memset (fastmap + 0200, 0, 0200); | |
b7dbcc19 KH |
1573 | /* We are sure that this loop stops. */ |
1574 | for (i = 0200; ! fastmap2[i]; i++); | |
4c0354d7 | 1575 | c = BYTE8_TO_CHAR (i); |
b7dbcc19 KH |
1576 | fastmap[CHAR_LEADING_CODE (c)] = 1; |
1577 | range_start_byte = i; | |
1578 | range_start_char = c; | |
f003ca54 | 1579 | char_ranges = (int *) alloca (sizeof (int) * 128 * 2); |
b7dbcc19 KH |
1580 | for (i = 129; i < 0400; i++) |
1581 | { | |
4c0354d7 | 1582 | c = BYTE8_TO_CHAR (i); |
b7dbcc19 KH |
1583 | fastmap[CHAR_LEADING_CODE (c)] = 1; |
1584 | if (i - range_start_byte != c - range_start_char) | |
1585 | { | |
1586 | char_ranges[n_char_ranges++] = range_start_char; | |
1587 | char_ranges[n_char_ranges++] = ((i - 1 - range_start_byte) | |
1588 | + range_start_char); | |
1589 | range_start_byte = i; | |
1590 | range_start_char = c; | |
8f924df7 | 1591 | } |
b7dbcc19 KH |
1592 | } |
1593 | char_ranges[n_char_ranges++] = range_start_char; | |
1594 | char_ranges[n_char_ranges++] = ((i - 1 - range_start_byte) | |
1595 | + range_start_char); | |
195d1361 RS |
1596 | } |
1597 | } | |
f003ca54 | 1598 | else /* STRING is multibyte */ |
b7dbcc19 | 1599 | { |
f003ca54 KH |
1600 | char_ranges = (int *) alloca (sizeof (int) * SCHARS (string) * 2); |
1601 | ||
b7dbcc19 | 1602 | while (i_byte < size_byte) |
195d1361 | 1603 | { |
b7dbcc19 | 1604 | unsigned char leading_code; |
5c7b02ab | 1605 | |
b7dbcc19 | 1606 | leading_code = str[i_byte]; |
62a6e103 | 1607 | c = STRING_CHAR_AND_LENGTH (str + i_byte, len); |
b7dbcc19 | 1608 | i_byte += len; |
5c7b02ab | 1609 | |
327719ee MB |
1610 | if (handle_iso_classes && c == '[' |
1611 | && i_byte < size_byte | |
62a6e103 | 1612 | && STRING_CHAR (str + i_byte) == ':') |
327719ee MB |
1613 | { |
1614 | const unsigned char *class_beg = str + i_byte + 1; | |
1615 | const unsigned char *class_end = class_beg; | |
1616 | const unsigned char *class_limit = str + size_byte - 2; | |
1617 | /* Leave room for the null. */ | |
1618 | unsigned char class_name[CHAR_CLASS_MAX_LENGTH + 1]; | |
1619 | re_wctype_t cc; | |
1620 | ||
1621 | if (class_limit - class_beg > CHAR_CLASS_MAX_LENGTH) | |
1622 | class_limit = class_beg + CHAR_CLASS_MAX_LENGTH; | |
1623 | ||
1624 | while (class_end < class_limit | |
1625 | && *class_end >= 'a' && *class_end <= 'z') | |
1626 | class_end++; | |
1627 | ||
1628 | if (class_end == class_beg | |
1629 | || *class_end != ':' || class_end[1] != ']') | |
1630 | goto not_a_class_name_multibyte; | |
1631 | ||
72af86bd | 1632 | memcpy (class_name, class_beg, class_end - class_beg); |
327719ee MB |
1633 | class_name[class_end - class_beg] = 0; |
1634 | ||
1635 | cc = re_wctype (class_name); | |
1636 | if (cc == 0) | |
1637 | error ("Invalid ISO C character class"); | |
1638 | ||
1639 | iso_classes = Fcons (make_number (cc), iso_classes); | |
1640 | ||
1641 | i_byte = class_end + 2 - str; | |
1642 | continue; | |
1643 | } | |
1644 | ||
1645 | not_a_class_name_multibyte: | |
195d1361 RS |
1646 | if (c == '\\') |
1647 | { | |
13090112 | 1648 | if (i_byte == size_byte) |
4101e6fe RS |
1649 | break; |
1650 | ||
b7dbcc19 | 1651 | leading_code = str[i_byte]; |
62a6e103 | 1652 | c = STRING_CHAR_AND_LENGTH (str + i_byte, len); |
82d497fc | 1653 | i_byte += len; |
195d1361 | 1654 | } |
839966f3 KH |
1655 | /* Treat `-' as range character only if another character |
1656 | follows. */ | |
1657 | if (i_byte + 1 < size_byte | |
82d497fc | 1658 | && str[i_byte] == '-') |
195d1361 | 1659 | { |
9690d026 | 1660 | unsigned int c2; |
b7dbcc19 | 1661 | unsigned char leading_code2; |
4101e6fe RS |
1662 | |
1663 | /* Skip over the dash. */ | |
82d497fc | 1664 | i_byte++; |
4101e6fe | 1665 | |
4101e6fe | 1666 | /* Get the end of the range. */ |
b7dbcc19 | 1667 | leading_code2 = str[i_byte]; |
62a6e103 | 1668 | c2 = STRING_CHAR_AND_LENGTH (str + i_byte, len); |
82d497fc | 1669 | i_byte += len; |
7e68b0ea | 1670 | |
b7dbcc19 KH |
1671 | if (c2 == '\\' |
1672 | && i_byte < size_byte) | |
1673 | { | |
1674 | leading_code2 = str[i_byte]; | |
62a6e103 | 1675 | c2 =STRING_CHAR_AND_LENGTH (str + i_byte, len); |
b7dbcc19 KH |
1676 | i_byte += len; |
1677 | } | |
1678 | ||
f003ca54 KH |
1679 | if (c > c2) |
1680 | continue; | |
b7dbcc19 | 1681 | if (ASCII_CHAR_P (c)) |
06274af5 | 1682 | { |
5c7b02ab KH |
1683 | while (c <= c2 && c < 0x80) |
1684 | fastmap[c++] = 1; | |
b7dbcc19 KH |
1685 | leading_code = CHAR_LEADING_CODE (c); |
1686 | } | |
1687 | if (! ASCII_CHAR_P (c)) | |
1688 | { | |
1e69125e PE |
1689 | unsigned lim2 = leading_code2 + 1; |
1690 | while (leading_code < lim2) | |
b7dbcc19 KH |
1691 | fastmap[leading_code++] = 1; |
1692 | if (c <= c2) | |
e39091b6 | 1693 | { |
b7dbcc19 | 1694 | char_ranges[n_char_ranges++] = c; |
e39091b6 | 1695 | char_ranges[n_char_ranges++] = c2; |
06274af5 KH |
1696 | } |
1697 | } | |
195d1361 RS |
1698 | } |
1699 | else | |
7e68b0ea | 1700 | { |
b7dbcc19 | 1701 | if (ASCII_CHAR_P (c)) |
e39091b6 KH |
1702 | fastmap[c] = 1; |
1703 | else | |
7e68b0ea | 1704 | { |
b7dbcc19 | 1705 | fastmap[leading_code] = 1; |
4101e6fe RS |
1706 | char_ranges[n_char_ranges++] = c; |
1707 | char_ranges[n_char_ranges++] = c; | |
7e68b0ea RS |
1708 | } |
1709 | } | |
195d1361 | 1710 | } |
b7dbcc19 KH |
1711 | |
1712 | /* If the current range is unibyte and STRING contains non-ASCII | |
1713 | chars, arrange fastmap for the corresponding unibyte | |
1714 | chars. */ | |
1715 | ||
1716 | if (! multibyte && n_char_ranges > 0) | |
1717 | { | |
72af86bd | 1718 | memset (fastmap + 0200, 0, 0200); |
b7dbcc19 KH |
1719 | for (i = 0; i < n_char_ranges; i += 2) |
1720 | { | |
1721 | int c1 = char_ranges[i]; | |
1e69125e | 1722 | unsigned lim2 = char_ranges[i + 1] + 1; |
b7dbcc19 | 1723 | |
1e69125e | 1724 | for (; c1 < lim2; c1++) |
2afc21f5 SM |
1725 | { |
1726 | int b = CHAR_TO_BYTE_SAFE (c1); | |
1727 | if (b >= 0) | |
1728 | fastmap[b] = 1; | |
1729 | } | |
b7dbcc19 KH |
1730 | } |
1731 | } | |
195d1361 RS |
1732 | } |
1733 | ||
9690d026 | 1734 | /* If ^ was the first character, complement the fastmap. */ |
195d1361 | 1735 | if (negate) |
b7dbcc19 KH |
1736 | { |
1737 | if (! multibyte) | |
1738 | for (i = 0; i < sizeof fastmap; i++) | |
1739 | fastmap[i] ^= 1; | |
1740 | else | |
1741 | { | |
1742 | for (i = 0; i < 0200; i++) | |
1743 | fastmap[i] ^= 1; | |
1744 | /* All non-ASCII chars possibly match. */ | |
1745 | for (; i < sizeof fastmap; i++) | |
1746 | fastmap[i] = 1; | |
1747 | } | |
1748 | } | |
195d1361 RS |
1749 | |
1750 | { | |
4f3a2f8d EZ |
1751 | EMACS_INT start_point = PT; |
1752 | EMACS_INT pos = PT; | |
1753 | EMACS_INT pos_byte = PT_BYTE; | |
9af7511a KH |
1754 | unsigned char *p = PT_ADDR, *endp, *stop; |
1755 | ||
1756 | if (forwardp) | |
1757 | { | |
1758 | endp = (XINT (lim) == GPT) ? GPT_ADDR : CHAR_POS_ADDR (XINT (lim)); | |
d7ee9fab | 1759 | stop = (pos < GPT && GPT < XINT (lim)) ? GPT_ADDR : endp; |
9af7511a KH |
1760 | } |
1761 | else | |
1762 | { | |
1763 | endp = CHAR_POS_ADDR (XINT (lim)); | |
d7ee9fab | 1764 | stop = (pos >= GPT && GPT > XINT (lim)) ? GAP_END_ADDR : endp; |
9af7511a | 1765 | } |
195d1361 RS |
1766 | |
1767 | immediate_quit = 1; | |
618db430 SM |
1768 | /* This code may look up syntax tables using macros that rely on the |
1769 | gl_state object. To make sure this object is not out of date, | |
1770 | let's initialize it manually. | |
1771 | We ignore syntax-table text-properties for now, since that's | |
1772 | what we've done in the past. */ | |
d48cd3f4 | 1773 | SETUP_BUFFER_SYNTAX_TABLE (); |
b7dbcc19 | 1774 | if (forwardp) |
195d1361 | 1775 | { |
b7dbcc19 | 1776 | if (multibyte) |
8f924df7 | 1777 | while (1) |
b7dbcc19 | 1778 | { |
8f924df7 | 1779 | int nbytes; |
9af7511a | 1780 | |
8f924df7 | 1781 | if (p >= stop) |
9af7511a | 1782 | { |
8f924df7 | 1783 | if (p >= endp) |
9af7511a | 1784 | break; |
8f924df7 KH |
1785 | p = GAP_END_ADDR; |
1786 | stop = endp; | |
9af7511a | 1787 | } |
62a6e103 | 1788 | c = STRING_CHAR_AND_LENGTH (p, nbytes); |
327719ee | 1789 | if (! NILP (iso_classes) && in_classes (c, iso_classes)) |
9af7511a | 1790 | { |
327719ee | 1791 | if (negate) |
9af7511a | 1792 | break; |
327719ee MB |
1793 | else |
1794 | goto fwd_ok; | |
9af7511a | 1795 | } |
9af7511a | 1796 | |
8f924df7 | 1797 | if (! fastmap[*p]) |
b7dbcc19 KH |
1798 | break; |
1799 | if (! ASCII_CHAR_P (c)) | |
9af7511a | 1800 | { |
b7dbcc19 KH |
1801 | /* As we are looking at a multibyte character, we |
1802 | must look up the character in the table | |
1803 | CHAR_RANGES. If there's no data in the table, | |
1804 | that character is not what we want to skip. */ | |
1805 | ||
1806 | /* The following code do the right thing even if | |
1807 | n_char_ranges is zero (i.e. no data in | |
1808 | CHAR_RANGES). */ | |
1809 | for (i = 0; i < n_char_ranges; i += 2) | |
1810 | if (c >= char_ranges[i] && c <= char_ranges[i + 1]) | |
9af7511a | 1811 | break; |
b7dbcc19 KH |
1812 | if (!(negate ^ (i < n_char_ranges))) |
1813 | break; | |
9af7511a | 1814 | } |
327719ee | 1815 | fwd_ok: |
8f924df7 | 1816 | p += nbytes, pos++, pos_byte += nbytes; |
b7dbcc19 KH |
1817 | } |
1818 | else | |
8f924df7 KH |
1819 | while (1) |
1820 | { | |
1821 | if (p >= stop) | |
9af7511a | 1822 | { |
8f924df7 | 1823 | if (p >= endp) |
9af7511a | 1824 | break; |
8f924df7 KH |
1825 | p = GAP_END_ADDR; |
1826 | stop = endp; | |
9af7511a | 1827 | } |
327719ee MB |
1828 | |
1829 | if (!NILP (iso_classes) && in_classes (*p, iso_classes)) | |
1830 | { | |
1831 | if (negate) | |
1832 | break; | |
1833 | else | |
1834 | goto fwd_unibyte_ok; | |
1835 | } | |
1836 | ||
8f924df7 KH |
1837 | if (!fastmap[*p]) |
1838 | break; | |
327719ee | 1839 | fwd_unibyte_ok: |
8f924df7 KH |
1840 | p++, pos++, pos_byte++; |
1841 | } | |
195d1361 RS |
1842 | } |
1843 | else | |
1844 | { | |
b7dbcc19 | 1845 | if (multibyte) |
8f924df7 | 1846 | while (1) |
b7dbcc19 | 1847 | { |
8f924df7 | 1848 | unsigned char *prev_p; |
9af7511a | 1849 | |
8f924df7 | 1850 | if (p <= stop) |
9af7511a | 1851 | { |
8f924df7 | 1852 | if (p <= endp) |
9af7511a | 1853 | break; |
8f924df7 KH |
1854 | p = GPT_ADDR; |
1855 | stop = endp; | |
9af7511a | 1856 | } |
8f924df7 KH |
1857 | prev_p = p; |
1858 | while (--p >= stop && ! CHAR_HEAD_P (*p)); | |
62a6e103 | 1859 | c = STRING_CHAR (p); |
a1bc88d4 | 1860 | |
327719ee MB |
1861 | if (! NILP (iso_classes) && in_classes (c, iso_classes)) |
1862 | { | |
1863 | if (negate) | |
1864 | break; | |
9690d026 | 1865 | else |
327719ee | 1866 | goto back_ok; |
7e68b0ea | 1867 | } |
a1bc88d4 | 1868 | |
8f924df7 | 1869 | if (! fastmap[*p]) |
b7dbcc19 KH |
1870 | break; |
1871 | if (! ASCII_CHAR_P (c)) | |
7e68b0ea | 1872 | { |
b7dbcc19 KH |
1873 | /* See the comment in the previous similar code. */ |
1874 | for (i = 0; i < n_char_ranges; i += 2) | |
1875 | if (c >= char_ranges[i] && c <= char_ranges[i + 1]) | |
1876 | break; | |
1877 | if (!(negate ^ (i < n_char_ranges))) | |
1878 | break; | |
7e68b0ea | 1879 | } |
327719ee | 1880 | back_ok: |
8f924df7 | 1881 | pos--, pos_byte -= prev_p - p; |
b7dbcc19 KH |
1882 | } |
1883 | else | |
8f924df7 KH |
1884 | while (1) |
1885 | { | |
1886 | if (p <= stop) | |
9af7511a | 1887 | { |
8f924df7 | 1888 | if (p <= endp) |
9af7511a | 1889 | break; |
8f924df7 KH |
1890 | p = GPT_ADDR; |
1891 | stop = endp; | |
9af7511a | 1892 | } |
e39091b6 | 1893 | |
327719ee MB |
1894 | if (! NILP (iso_classes) && in_classes (p[-1], iso_classes)) |
1895 | { | |
1896 | if (negate) | |
1897 | break; | |
9af7511a | 1898 | else |
327719ee MB |
1899 | goto back_unibyte_ok; |
1900 | } | |
a1bc88d4 | 1901 | |
8f924df7 KH |
1902 | if (!fastmap[p[-1]]) |
1903 | break; | |
327719ee | 1904 | back_unibyte_ok: |
8f924df7 KH |
1905 | p--, pos--, pos_byte--; |
1906 | } | |
195d1361 | 1907 | } |
7e68b0ea | 1908 | |
b7dbcc19 KH |
1909 | SET_PT_BOTH (pos, pos_byte); |
1910 | immediate_quit = 0; | |
1911 | ||
1912 | return make_number (PT - start_point); | |
1913 | } | |
1914 | } | |
7e68b0ea | 1915 | |
b7dbcc19 KH |
1916 | |
1917 | static Lisp_Object | |
971de7fb | 1918 | skip_syntaxes (int forwardp, Lisp_Object string, Lisp_Object lim) |
b7dbcc19 KH |
1919 | { |
1920 | register unsigned int c; | |
1921 | unsigned char fastmap[0400]; | |
1922 | int negate = 0; | |
4f3a2f8d | 1923 | register EMACS_INT i, i_byte; |
b7dbcc19 | 1924 | int multibyte; |
4f3a2f8d | 1925 | EMACS_INT size_byte; |
b7dbcc19 KH |
1926 | unsigned char *str; |
1927 | ||
1928 | CHECK_STRING (string); | |
1929 | ||
1930 | if (NILP (lim)) | |
1931 | XSETINT (lim, forwardp ? ZV : BEGV); | |
1932 | else | |
1933 | CHECK_NUMBER_COERCE_MARKER (lim); | |
1934 | ||
1935 | /* In any case, don't allow scan outside bounds of buffer. */ | |
1936 | if (XINT (lim) > ZV) | |
1937 | XSETFASTINT (lim, ZV); | |
1938 | if (XINT (lim) < BEGV) | |
1939 | XSETFASTINT (lim, BEGV); | |
1940 | ||
ade8ee9e | 1941 | if (forwardp ? (PT >= XFASTINT (lim)) : (PT <= XFASTINT (lim))) |
49feb1cd | 1942 | return make_number (0); |
8c6e735b | 1943 | |
4b4deea2 | 1944 | multibyte = (!NILP (BVAR (current_buffer, enable_multibyte_characters)) |
92eaa22e | 1945 | && (XINT (lim) - PT != CHAR_TO_BYTE (XINT (lim)) - PT_BYTE)); |
b7dbcc19 | 1946 | |
72af86bd | 1947 | memset (fastmap, 0, sizeof fastmap); |
b7dbcc19 | 1948 | |
8f924df7 KH |
1949 | if (SBYTES (string) > SCHARS (string)) |
1950 | /* As this is very rare case (syntax spec is ASCII only), don't | |
1951 | consider efficiency. */ | |
bc796a59 KH |
1952 | string = string_make_unibyte (string); |
1953 | ||
8f924df7 KH |
1954 | str = SDATA (string); |
1955 | size_byte = SBYTES (string); | |
bc796a59 | 1956 | |
b7dbcc19 KH |
1957 | i_byte = 0; |
1958 | if (i_byte < size_byte | |
8f924df7 | 1959 | && SREF (string, 0) == '^') |
b7dbcc19 KH |
1960 | { |
1961 | negate = 1; i_byte++; | |
1962 | } | |
1963 | ||
b7dbcc19 KH |
1964 | /* Find the syntaxes specified and set their elements of fastmap. */ |
1965 | ||
1966 | while (i_byte < size_byte) | |
1967 | { | |
1968 | c = str[i_byte++]; | |
8f924df7 | 1969 | fastmap[syntax_spec_code[c]] = 1; |
b7dbcc19 | 1970 | } |
195d1361 | 1971 | |
9690d026 | 1972 | /* If ^ was the first character, complement the fastmap. */ |
195d1361 RS |
1973 | if (negate) |
1974 | for (i = 0; i < sizeof fastmap; i++) | |
9690d026 | 1975 | fastmap[i] ^= 1; |
195d1361 RS |
1976 | |
1977 | { | |
4f3a2f8d EZ |
1978 | EMACS_INT start_point = PT; |
1979 | EMACS_INT pos = PT; | |
1980 | EMACS_INT pos_byte = PT_BYTE; | |
8f924df7 KH |
1981 | unsigned char *p = PT_ADDR, *endp, *stop; |
1982 | ||
1983 | if (forwardp) | |
1984 | { | |
1985 | endp = (XINT (lim) == GPT) ? GPT_ADDR : CHAR_POS_ADDR (XINT (lim)); | |
1986 | stop = (pos < GPT && GPT < XINT (lim)) ? GPT_ADDR : endp; | |
1987 | } | |
1988 | else | |
1989 | { | |
1990 | endp = CHAR_POS_ADDR (XINT (lim)); | |
1991 | stop = (pos >= GPT && GPT > XINT (lim)) ? GAP_END_ADDR : endp; | |
1992 | } | |
195d1361 RS |
1993 | |
1994 | immediate_quit = 1; | |
b7dbcc19 KH |
1995 | SETUP_SYNTAX_TABLE (pos, forwardp ? 1 : -1); |
1996 | if (forwardp) | |
195d1361 | 1997 | { |
b7dbcc19 | 1998 | if (multibyte) |
195d1361 | 1999 | { |
8f924df7 | 2000 | while (1) |
8c6e735b | 2001 | { |
8f924df7 KH |
2002 | int nbytes; |
2003 | ||
2004 | if (p >= stop) | |
2005 | { | |
2006 | if (p >= endp) | |
2007 | break; | |
2008 | p = GAP_END_ADDR; | |
2009 | stop = endp; | |
2010 | } | |
62a6e103 | 2011 | c = STRING_CHAR_AND_LENGTH (p, nbytes); |
8f924df7 | 2012 | if (! fastmap[(int) SYNTAX (c)]) |
8c6e735b | 2013 | break; |
8f924df7 | 2014 | p += nbytes, pos++, pos_byte += nbytes; |
8c6e735b KH |
2015 | UPDATE_SYNTAX_TABLE_FORWARD (pos); |
2016 | } | |
195d1361 RS |
2017 | } |
2018 | else | |
2019 | { | |
8c6e735b | 2020 | while (1) |
7e68b0ea | 2021 | { |
8f924df7 KH |
2022 | if (p >= stop) |
2023 | { | |
2024 | if (p >= endp) | |
2025 | break; | |
2026 | p = GAP_END_ADDR; | |
2027 | stop = endp; | |
2028 | } | |
2029 | if (! fastmap[(int) SYNTAX (*p)]) | |
8c6e735b | 2030 | break; |
8f924df7 | 2031 | p++, pos++, pos_byte++; |
b7dbcc19 | 2032 | UPDATE_SYNTAX_TABLE_FORWARD (pos); |
195d1361 RS |
2033 | } |
2034 | } | |
2035 | } | |
2036 | else | |
2037 | { | |
b7dbcc19 | 2038 | if (multibyte) |
195d1361 | 2039 | { |
8c6e735b | 2040 | while (1) |
b7dbcc19 | 2041 | { |
8f924df7 KH |
2042 | unsigned char *prev_p; |
2043 | ||
2044 | if (p <= stop) | |
b7dbcc19 | 2045 | { |
8f924df7 KH |
2046 | if (p <= endp) |
2047 | break; | |
2048 | p = GPT_ADDR; | |
2049 | stop = endp; | |
b7dbcc19 | 2050 | } |
262be72a | 2051 | UPDATE_SYNTAX_TABLE_BACKWARD (pos - 1); |
8f924df7 KH |
2052 | prev_p = p; |
2053 | while (--p >= stop && ! CHAR_HEAD_P (*p)); | |
62a6e103 | 2054 | c = STRING_CHAR (p); |
8f924df7 KH |
2055 | if (! fastmap[(int) SYNTAX (c)]) |
2056 | break; | |
2057 | pos--, pos_byte -= prev_p - p; | |
b7dbcc19 | 2058 | } |
195d1361 RS |
2059 | } |
2060 | else | |
2061 | { | |
8c6e735b KH |
2062 | while (1) |
2063 | { | |
8f924df7 KH |
2064 | if (p <= stop) |
2065 | { | |
2066 | if (p <= endp) | |
2067 | break; | |
2068 | p = GPT_ADDR; | |
2069 | stop = endp; | |
2070 | } | |
102f6132 | 2071 | UPDATE_SYNTAX_TABLE_BACKWARD (pos - 1); |
8f924df7 | 2072 | if (! fastmap[(int) SYNTAX (p[-1])]) |
8c6e735b | 2073 | break; |
8f924df7 | 2074 | p--, pos--, pos_byte--; |
8c6e735b | 2075 | } |
195d1361 RS |
2076 | } |
2077 | } | |
6a140a74 RS |
2078 | |
2079 | SET_PT_BOTH (pos, pos_byte); | |
195d1361 RS |
2080 | immediate_quit = 0; |
2081 | ||
2082 | return make_number (PT - start_point); | |
2083 | } | |
2084 | } | |
a1bc88d4 RS |
2085 | |
2086 | /* Return 1 if character C belongs to one of the ISO classes | |
2087 | in the list ISO_CLASSES. Each class is represented by an | |
2088 | integer which is its type according to re_wctype. */ | |
2089 | ||
2090 | static int | |
971de7fb | 2091 | in_classes (int c, Lisp_Object iso_classes) |
a1bc88d4 RS |
2092 | { |
2093 | int fits_class = 0; | |
2094 | ||
618db430 | 2095 | while (CONSP (iso_classes)) |
a1bc88d4 RS |
2096 | { |
2097 | Lisp_Object elt; | |
2098 | elt = XCAR (iso_classes); | |
2099 | iso_classes = XCDR (iso_classes); | |
2100 | ||
2101 | if (re_iswctype (c, XFASTINT (elt))) | |
2102 | fits_class = 1; | |
2103 | } | |
2104 | ||
2105 | return fits_class; | |
2106 | } | |
195d1361 | 2107 | \f |
95ff8dfc RS |
2108 | /* Jump over a comment, assuming we are at the beginning of one. |
2109 | FROM is the current position. | |
2110 | FROM_BYTE is the bytepos corresponding to FROM. | |
2111 | Do not move past STOP (a charpos). | |
2112 | The comment over which we have to jump is of style STYLE | |
c5683ceb | 2113 | (either SYNTAX_FLAGS_COMMENT_STYLE(foo) or ST_COMMENT_STYLE). |
95ff8dfc RS |
2114 | NESTING should be positive to indicate the nesting at the beginning |
2115 | for nested comments and should be zero or negative else. | |
2116 | ST_COMMENT_STYLE cannot be nested. | |
2117 | PREV_SYNTAX is the SYNTAX_WITH_FLAGS of the previous character | |
2118 | (or 0 If the search cannot start in the middle of a two-character). | |
2119 | ||
2120 | If successful, return 1 and store the charpos of the comment's end | |
2121 | into *CHARPOS_PTR and the corresponding bytepos into *BYTEPOS_PTR. | |
2122 | Else, return 0 and store the charpos STOP into *CHARPOS_PTR, the | |
2123 | corresponding bytepos into *BYTEPOS_PTR and the current nesting | |
2124 | (as defined for state.incomment) in *INCOMMENT_PTR. | |
2125 | ||
2126 | The comment end is the last character of the comment rather than the | |
2127 | character just after the comment. | |
2128 | ||
2129 | Global syntax data is assumed to initially be valid for FROM and | |
2130 | remains valid for forward search starting at the returned position. */ | |
2131 | ||
2132 | static int | |
dd4c5104 DN |
2133 | forw_comment (EMACS_INT from, EMACS_INT from_byte, EMACS_INT stop, |
2134 | int nesting, int style, int prev_syntax, | |
2135 | EMACS_INT *charpos_ptr, EMACS_INT *bytepos_ptr, | |
2136 | int *incomment_ptr) | |
95ff8dfc RS |
2137 | { |
2138 | register int c, c1; | |
2139 | register enum syntaxcode code; | |
c5683ceb | 2140 | register int syntax, other_syntax; |
95ff8dfc RS |
2141 | |
2142 | if (nesting <= 0) nesting = -1; | |
2143 | ||
2144 | /* Enter the loop in the middle so that we find | |
2145 | a 2-char comment ender if we start in the middle of it. */ | |
2146 | syntax = prev_syntax; | |
2147 | if (syntax != 0) goto forw_incomment; | |
2148 | ||
2149 | while (1) | |
2150 | { | |
2151 | if (from == stop) | |
2152 | { | |
2153 | *incomment_ptr = nesting; | |
2154 | *charpos_ptr = from; | |
2155 | *bytepos_ptr = from_byte; | |
2156 | return 0; | |
2157 | } | |
8f924df7 | 2158 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
95ff8dfc RS |
2159 | syntax = SYNTAX_WITH_FLAGS (c); |
2160 | code = syntax & 0xff; | |
2161 | if (code == Sendcomment | |
c5683ceb | 2162 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax, 0) == style |
abf8a9ff SM |
2163 | && (SYNTAX_FLAGS_COMMENT_NESTED (syntax) ? |
2164 | (nesting > 0 && --nesting == 0) : nesting < 0)) | |
95ff8dfc RS |
2165 | /* we have encountered a comment end of the same style |
2166 | as the comment sequence which began this comment | |
2167 | section */ | |
2168 | break; | |
2169 | if (code == Scomment_fence | |
2170 | && style == ST_COMMENT_STYLE) | |
2171 | /* we have encountered a comment end of the same style | |
2172 | as the comment sequence which began this comment | |
2173 | section. */ | |
2174 | break; | |
2175 | if (nesting > 0 | |
2176 | && code == Scomment | |
abf8a9ff | 2177 | && SYNTAX_FLAGS_COMMENT_NESTED (syntax) |
c5683ceb | 2178 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax, 0) == style) |
95ff8dfc RS |
2179 | /* we have encountered a nested comment of the same style |
2180 | as the comment sequence which began this comment section */ | |
2181 | nesting++; | |
2182 | INC_BOTH (from, from_byte); | |
2183 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
7d0393cf | 2184 | |
95ff8dfc RS |
2185 | forw_incomment: |
2186 | if (from < stop && SYNTAX_FLAGS_COMEND_FIRST (syntax) | |
8f924df7 | 2187 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
c5683ceb SM |
2188 | other_syntax = SYNTAX_WITH_FLAGS (c1), |
2189 | SYNTAX_FLAGS_COMEND_SECOND (other_syntax)) | |
2190 | && SYNTAX_FLAGS_COMMENT_STYLE (syntax, other_syntax) == style | |
abf8a9ff | 2191 | && ((SYNTAX_FLAGS_COMMENT_NESTED (syntax) || |
c5683ceb SM |
2192 | SYNTAX_FLAGS_COMMENT_NESTED (other_syntax)) |
2193 | ? nesting > 0 : nesting < 0)) | |
4ffe723b GM |
2194 | { |
2195 | if (--nesting <= 0) | |
2196 | /* we have encountered a comment end of the same style | |
2197 | as the comment sequence which began this comment | |
2198 | section */ | |
2199 | break; | |
2200 | else | |
2201 | { | |
2202 | INC_BOTH (from, from_byte); | |
2203 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
2204 | } | |
2205 | } | |
95ff8dfc RS |
2206 | if (nesting > 0 |
2207 | && from < stop | |
2208 | && SYNTAX_FLAGS_COMSTART_FIRST (syntax) | |
8f924df7 | 2209 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
c5683ceb SM |
2210 | other_syntax = SYNTAX_WITH_FLAGS (c1), |
2211 | SYNTAX_FLAGS_COMMENT_STYLE (other_syntax, syntax) == style | |
2212 | && SYNTAX_FLAGS_COMSTART_SECOND (other_syntax)) | |
abf8a9ff | 2213 | && (SYNTAX_FLAGS_COMMENT_NESTED (syntax) || |
c5683ceb | 2214 | SYNTAX_FLAGS_COMMENT_NESTED (other_syntax))) |
95ff8dfc RS |
2215 | /* we have encountered a nested comment of the same style |
2216 | as the comment sequence which began this comment | |
2217 | section */ | |
2218 | { | |
2219 | INC_BOTH (from, from_byte); | |
2220 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
2221 | nesting++; | |
2222 | } | |
2223 | } | |
2224 | *charpos_ptr = from; | |
2225 | *bytepos_ptr = from_byte; | |
2226 | return 1; | |
2227 | } | |
2228 | ||
b3cfe0c8 | 2229 | DEFUN ("forward-comment", Fforward_comment, Sforward_comment, 1, 1, 0, |
177c0ea7 | 2230 | doc: /* |
ee648c11 | 2231 | Move forward across up to COUNT comments. If COUNT is negative, move backward. |
fdb82f93 PJ |
2232 | Stop scanning if we find something other than a comment or whitespace. |
2233 | Set point to where scanning stops. | |
ee648c11 | 2234 | If COUNT comments are found as expected, with nothing except whitespace |
fdb82f93 | 2235 | between them, return t; otherwise return nil. */) |
5842a27b | 2236 | (Lisp_Object count) |
b3cfe0c8 | 2237 | { |
2312c580 SM |
2238 | register EMACS_INT from; |
2239 | EMACS_INT from_byte; | |
2240 | register EMACS_INT stop; | |
8ea151b2 | 2241 | register int c, c1; |
b3cfe0c8 RS |
2242 | register enum syntaxcode code; |
2243 | int comstyle = 0; /* style of comment encountered */ | |
95ff8dfc | 2244 | int comnested = 0; /* whether the comment is nestable or not */ |
be720845 | 2245 | int found; |
2312c580 SM |
2246 | EMACS_INT count1; |
2247 | EMACS_INT out_charpos, out_bytepos; | |
95ff8dfc | 2248 | int dummy; |
840f481c | 2249 | |
b7826503 | 2250 | CHECK_NUMBER (count); |
840f481c | 2251 | count1 = XINT (count); |
195d1361 | 2252 | stop = count1 > 0 ? ZV : BEGV; |
b3cfe0c8 RS |
2253 | |
2254 | immediate_quit = 1; | |
2255 | QUIT; | |
2256 | ||
2257 | from = PT; | |
6a140a74 | 2258 | from_byte = PT_BYTE; |
b3cfe0c8 | 2259 | |
195d1361 | 2260 | SETUP_SYNTAX_TABLE (from, count1); |
840f481c | 2261 | while (count1 > 0) |
b3cfe0c8 | 2262 | { |
04882296 | 2263 | do |
b3cfe0c8 | 2264 | { |
c5683ceb | 2265 | int comstart_first, syntax, other_syntax; |
f902a008 | 2266 | |
04882296 KH |
2267 | if (from == stop) |
2268 | { | |
ef316cf0 | 2269 | SET_PT_BOTH (from, from_byte); |
b7e6e612 | 2270 | immediate_quit = 0; |
04882296 KH |
2271 | return Qnil; |
2272 | } | |
b7dbcc19 | 2273 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
c5683ceb | 2274 | syntax = SYNTAX_WITH_FLAGS (c); |
b3cfe0c8 | 2275 | code = SYNTAX (c); |
c5683ceb SM |
2276 | comstart_first = SYNTAX_FLAGS_COMSTART_FIRST (syntax); |
2277 | comnested = SYNTAX_FLAGS_COMMENT_NESTED (syntax); | |
2278 | comstyle = SYNTAX_FLAGS_COMMENT_STYLE (syntax, 0); | |
6a140a74 | 2279 | INC_BOTH (from, from_byte); |
f902a008 | 2280 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
f902a008 | 2281 | if (from < stop && comstart_first |
b7dbcc19 | 2282 | && (c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
c5683ceb SM |
2283 | other_syntax = SYNTAX_WITH_FLAGS (c1), |
2284 | SYNTAX_FLAGS_COMSTART_SECOND (other_syntax))) | |
b3cfe0c8 | 2285 | { |
7d0393cf | 2286 | /* We have encountered a comment start sequence and we |
7fc8191e | 2287 | are ignoring all text inside comments. We must record |
b3cfe0c8 RS |
2288 | the comment style this sequence begins so that later, |
2289 | only a comment end of the same style actually ends | |
7fc8191e | 2290 | the comment section. */ |
b3cfe0c8 | 2291 | code = Scomment; |
c5683ceb SM |
2292 | comstyle = SYNTAX_FLAGS_COMMENT_STYLE (other_syntax, syntax); |
2293 | comnested | |
2294 | = comnested || SYNTAX_FLAGS_COMMENT_NESTED (other_syntax); | |
6a140a74 | 2295 | INC_BOTH (from, from_byte); |
95ff8dfc | 2296 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b3cfe0c8 | 2297 | } |
04882296 | 2298 | } |
abf8a9ff | 2299 | while (code == Swhitespace || (code == Sendcomment && c == '\n')); |
6a140a74 | 2300 | |
da71ab91 KH |
2301 | if (code == Scomment_fence) |
2302 | comstyle = ST_COMMENT_STYLE; | |
2303 | else if (code != Scomment) | |
04882296 KH |
2304 | { |
2305 | immediate_quit = 0; | |
6a140a74 | 2306 | DEC_BOTH (from, from_byte); |
ef316cf0 | 2307 | SET_PT_BOTH (from, from_byte); |
04882296 KH |
2308 | return Qnil; |
2309 | } | |
2310 | /* We're at the start of a comment. */ | |
95ff8dfc RS |
2311 | found = forw_comment (from, from_byte, stop, comnested, comstyle, 0, |
2312 | &out_charpos, &out_bytepos, &dummy); | |
2313 | from = out_charpos; from_byte = out_bytepos; | |
2314 | if (!found) | |
04882296 | 2315 | { |
95ff8dfc RS |
2316 | immediate_quit = 0; |
2317 | SET_PT_BOTH (from, from_byte); | |
2318 | return Qnil; | |
b3cfe0c8 | 2319 | } |
95ff8dfc RS |
2320 | INC_BOTH (from, from_byte); |
2321 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
04882296 | 2322 | /* We have skipped one comment. */ |
840f481c | 2323 | count1--; |
b3cfe0c8 RS |
2324 | } |
2325 | ||
840f481c | 2326 | while (count1 < 0) |
b3cfe0c8 | 2327 | { |
b9145dbb | 2328 | while (1) |
b3cfe0c8 | 2329 | { |
c5683ceb | 2330 | int quoted, syntax; |
f902a008 | 2331 | |
b9145dbb RS |
2332 | if (from <= stop) |
2333 | { | |
6a140a74 | 2334 | SET_PT_BOTH (BEGV, BEGV_BYTE); |
b9145dbb RS |
2335 | immediate_quit = 0; |
2336 | return Qnil; | |
2337 | } | |
b3cfe0c8 | 2338 | |
6a140a74 | 2339 | DEC_BOTH (from, from_byte); |
f902a008 | 2340 | /* char_quoted does UPDATE_SYNTAX_TABLE_BACKWARD (from). */ |
6a140a74 | 2341 | quoted = char_quoted (from, from_byte); |
b7dbcc19 | 2342 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
c5683ceb | 2343 | syntax = SYNTAX_WITH_FLAGS (c); |
b3cfe0c8 RS |
2344 | code = SYNTAX (c); |
2345 | comstyle = 0; | |
c5683ceb | 2346 | comnested = SYNTAX_FLAGS_COMMENT_NESTED (syntax); |
7fc8191e | 2347 | if (code == Sendcomment) |
c5683ceb SM |
2348 | comstyle = SYNTAX_FLAGS_COMMENT_STYLE (syntax, 0); |
2349 | if (from > stop && SYNTAX_FLAGS_COMEND_SECOND (syntax) | |
f902a008 | 2350 | && prev_char_comend_first (from, from_byte) |
89c6809a | 2351 | && !char_quoted (from - 1, dec_bytepos (from_byte))) |
b3cfe0c8 | 2352 | { |
c5683ceb | 2353 | int other_syntax; |
7fc8191e | 2354 | /* We must record the comment style encountered so that |
b3cfe0c8 | 2355 | later, we can match only the proper comment begin |
7fc8191e | 2356 | sequence of the same style. */ |
6a140a74 | 2357 | DEC_BOTH (from, from_byte); |
f902a008 RS |
2358 | code = Sendcomment; |
2359 | /* Calling char_quoted, above, set up global syntax position | |
2360 | at the new value of FROM. */ | |
b7dbcc19 | 2361 | c1 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
c5683ceb SM |
2362 | other_syntax = SYNTAX_WITH_FLAGS (c1); |
2363 | comstyle = SYNTAX_FLAGS_COMMENT_STYLE (other_syntax, syntax); | |
2364 | comnested | |
2365 | = comnested || SYNTAX_FLAGS_COMMENT_NESTED (other_syntax); | |
b3cfe0c8 RS |
2366 | } |
2367 | ||
195d1361 RS |
2368 | if (code == Scomment_fence) |
2369 | { | |
2370 | /* Skip until first preceding unquoted comment_fence. */ | |
01f44d5a | 2371 | int fence_found = 0; |
4f3a2f8d | 2372 | EMACS_INT ini = from, ini_byte = from_byte; |
7d0393cf | 2373 | |
6a140a74 | 2374 | while (1) |
195d1361 | 2375 | { |
6a140a74 | 2376 | DEC_BOTH (from, from_byte); |
195d1361 | 2377 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 2378 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
6a140a74 | 2379 | if (SYNTAX (c) == Scomment_fence |
7d0393cf | 2380 | && !char_quoted (from, from_byte)) |
195d1361 | 2381 | { |
01f44d5a | 2382 | fence_found = 1; |
195d1361 RS |
2383 | break; |
2384 | } | |
fcdd4585 SM |
2385 | else if (from == stop) |
2386 | break; | |
195d1361 | 2387 | } |
01f44d5a | 2388 | if (fence_found == 0) |
195d1361 RS |
2389 | { |
2390 | from = ini; /* Set point to ini + 1. */ | |
6a140a74 | 2391 | from_byte = ini_byte; |
195d1361 RS |
2392 | goto leave; |
2393 | } | |
39c41ad4 SM |
2394 | else |
2395 | /* We have skipped one comment. */ | |
2396 | break; | |
195d1361 RS |
2397 | } |
2398 | else if (code == Sendcomment) | |
b3cfe0c8 | 2399 | { |
95ff8dfc | 2400 | found = back_comment (from, from_byte, stop, comnested, comstyle, |
6a140a74 | 2401 | &out_charpos, &out_bytepos); |
1aa963c8 SM |
2402 | if (found == -1) |
2403 | { | |
abf8a9ff SM |
2404 | if (c == '\n') |
2405 | /* This end-of-line is not an end-of-comment. | |
2406 | Treat it like a whitespace. | |
2407 | CC-mode (and maybe others) relies on this behavior. */ | |
2408 | ; | |
2409 | else | |
2410 | { | |
2411 | /* Failure: we should go back to the end of this | |
2412 | not-quite-endcomment. */ | |
c5683ceb | 2413 | if (SYNTAX (c) != code) |
abf8a9ff SM |
2414 | /* It was a two-char Sendcomment. */ |
2415 | INC_BOTH (from, from_byte); | |
2416 | goto leave; | |
2417 | } | |
1aa963c8 | 2418 | } |
2b927d02 | 2419 | else |
abf8a9ff SM |
2420 | { |
2421 | /* We have skipped one comment. */ | |
2422 | from = out_charpos, from_byte = out_bytepos; | |
2423 | break; | |
2424 | } | |
b3cfe0c8 | 2425 | } |
bb0de084 | 2426 | else if (code != Swhitespace || quoted) |
b3cfe0c8 | 2427 | { |
195d1361 | 2428 | leave: |
b3cfe0c8 | 2429 | immediate_quit = 0; |
6a140a74 | 2430 | INC_BOTH (from, from_byte); |
ef316cf0 | 2431 | SET_PT_BOTH (from, from_byte); |
b3cfe0c8 RS |
2432 | return Qnil; |
2433 | } | |
2434 | } | |
2435 | ||
840f481c | 2436 | count1++; |
b3cfe0c8 RS |
2437 | } |
2438 | ||
ef316cf0 | 2439 | SET_PT_BOTH (from, from_byte); |
b3cfe0c8 RS |
2440 | immediate_quit = 0; |
2441 | return Qt; | |
2442 | } | |
2443 | \f | |
b7dbcc19 | 2444 | /* Return syntax code of character C if C is an ASCII character |
db9f48c3 | 2445 | or `multibyte_symbol_p' is zero. Otherwise, return Ssymbol. */ |
bd25db08 | 2446 | |
b7dbcc19 KH |
2447 | #define SYNTAX_WITH_MULTIBYTE_CHECK(c) \ |
2448 | ((ASCII_CHAR_P (c) || !multibyte_symbol_p) \ | |
bd25db08 KH |
2449 | ? SYNTAX (c) : Ssymbol) |
2450 | ||
6a140a74 | 2451 | static Lisp_Object |
971de7fb | 2452 | scan_lists (register EMACS_INT from, EMACS_INT count, EMACS_INT depth, int sexpflag) |
8489eb67 RS |
2453 | { |
2454 | Lisp_Object val; | |
932e6895 | 2455 | register EMACS_INT stop = count > 0 ? ZV : BEGV; |
93da5fff KH |
2456 | register int c, c1; |
2457 | int stringterm; | |
8489eb67 RS |
2458 | int quoted; |
2459 | int mathexit = 0; | |
93da5fff | 2460 | register enum syntaxcode code, temp_code; |
195d1361 | 2461 | int min_depth = depth; /* Err out if depth gets less than this. */ |
e5d4f4dc | 2462 | int comstyle = 0; /* style of comment encountered */ |
95ff8dfc | 2463 | int comnested = 0; /* whether the comment is nestable or not */ |
932e6895 SM |
2464 | EMACS_INT temp_pos; |
2465 | EMACS_INT last_good = from; | |
195d1361 | 2466 | int found; |
932e6895 SM |
2467 | EMACS_INT from_byte; |
2468 | EMACS_INT out_bytepos, out_charpos; | |
95ff8dfc | 2469 | int temp, dummy; |
bd25db08 | 2470 | int multibyte_symbol_p = sexpflag && multibyte_syntax_as_symbol; |
8489eb67 RS |
2471 | |
2472 | if (depth > 0) min_depth = 0; | |
2473 | ||
c3907a7e RS |
2474 | if (from > ZV) from = ZV; |
2475 | if (from < BEGV) from = BEGV; | |
2476 | ||
2477 | from_byte = CHAR_TO_BYTE (from); | |
2478 | ||
8489eb67 RS |
2479 | immediate_quit = 1; |
2480 | QUIT; | |
2481 | ||
195d1361 | 2482 | SETUP_SYNTAX_TABLE (from, count); |
8489eb67 RS |
2483 | while (count > 0) |
2484 | { | |
8489eb67 RS |
2485 | while (from < stop) |
2486 | { | |
c5683ceb | 2487 | int comstart_first, prefix, syntax, other_syntax; |
195d1361 | 2488 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 2489 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
c5683ceb | 2490 | syntax = SYNTAX_WITH_FLAGS (c); |
bd25db08 | 2491 | code = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
c5683ceb SM |
2492 | comstart_first = SYNTAX_FLAGS_COMSTART_FIRST (syntax); |
2493 | comnested = SYNTAX_FLAGS_COMMENT_NESTED (syntax); | |
2494 | comstyle = SYNTAX_FLAGS_COMMENT_STYLE (syntax, 0); | |
2495 | prefix = SYNTAX_FLAGS_PREFIX (syntax); | |
7bf5e9e4 RS |
2496 | if (depth == min_depth) |
2497 | last_good = from; | |
6a140a74 | 2498 | INC_BOTH (from, from_byte); |
195d1361 | 2499 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
f902a008 | 2500 | if (from < stop && comstart_first |
327719ee | 2501 | && (c = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
c5683ceb SM |
2502 | other_syntax = SYNTAX_WITH_FLAGS (c), |
2503 | SYNTAX_FLAGS_COMSTART_SECOND (other_syntax)) | |
8489eb67 | 2504 | && parse_sexp_ignore_comments) |
e5d4f4dc | 2505 | { |
7d0393cf | 2506 | /* we have encountered a comment start sequence and we |
195d1361 | 2507 | are ignoring all text inside comments. We must record |
e5d4f4dc RS |
2508 | the comment style this sequence begins so that later, |
2509 | only a comment end of the same style actually ends | |
2510 | the comment section */ | |
2511 | code = Scomment; | |
c5683ceb SM |
2512 | comstyle = SYNTAX_FLAGS_COMMENT_STYLE (other_syntax, syntax); |
2513 | comnested | |
2514 | = comnested || SYNTAX_FLAGS_COMMENT_NESTED (other_syntax); | |
6a140a74 | 2515 | INC_BOTH (from, from_byte); |
f902a008 | 2516 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
e5d4f4dc | 2517 | } |
7d0393cf | 2518 | |
f902a008 | 2519 | if (prefix) |
8489eb67 RS |
2520 | continue; |
2521 | ||
0220c518 | 2522 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
2523 | { |
2524 | case Sescape: | |
2525 | case Scharquote: | |
8b8bf8e6 SM |
2526 | if (from == stop) |
2527 | goto lose; | |
6a140a74 | 2528 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2529 | /* treat following character as a word constituent */ |
2530 | case Sword: | |
2531 | case Ssymbol: | |
2532 | if (depth || !sexpflag) break; | |
195d1361 | 2533 | /* This word counts as a sexp; return at end of it. */ |
8489eb67 RS |
2534 | while (from < stop) |
2535 | { | |
195d1361 | 2536 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
7c7ca3d2 RS |
2537 | |
2538 | /* Some compilers can't handle this inside the switch. */ | |
b7dbcc19 | 2539 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
bd25db08 | 2540 | temp = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7c7ca3d2 | 2541 | switch (temp) |
8489eb67 RS |
2542 | { |
2543 | case Scharquote: | |
2544 | case Sescape: | |
6a140a74 | 2545 | INC_BOTH (from, from_byte); |
8b8bf8e6 SM |
2546 | if (from == stop) |
2547 | goto lose; | |
8489eb67 RS |
2548 | break; |
2549 | case Sword: | |
2550 | case Ssymbol: | |
2551 | case Squote: | |
2552 | break; | |
2553 | default: | |
2554 | goto done; | |
2555 | } | |
6a140a74 | 2556 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2557 | } |
2558 | goto done; | |
2559 | ||
195d1361 | 2560 | case Scomment_fence: |
95ff8dfc RS |
2561 | comstyle = ST_COMMENT_STYLE; |
2562 | /* FALLTHROUGH */ | |
2563 | case Scomment: | |
8489eb67 | 2564 | if (!parse_sexp_ignore_comments) break; |
95ff8dfc RS |
2565 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
2566 | found = forw_comment (from, from_byte, stop, | |
2567 | comnested, comstyle, 0, | |
2568 | &out_charpos, &out_bytepos, &dummy); | |
2569 | from = out_charpos, from_byte = out_bytepos; | |
2570 | if (!found) | |
8489eb67 | 2571 | { |
95ff8dfc RS |
2572 | if (depth == 0) |
2573 | goto done; | |
2574 | goto lose; | |
8489eb67 | 2575 | } |
95ff8dfc RS |
2576 | INC_BOTH (from, from_byte); |
2577 | UPDATE_SYNTAX_TABLE_FORWARD (from); | |
8489eb67 RS |
2578 | break; |
2579 | ||
2580 | case Smath: | |
2581 | if (!sexpflag) | |
2582 | break; | |
b7dbcc19 | 2583 | if (from != stop && c == FETCH_CHAR_AS_MULTIBYTE (from_byte)) |
6a140a74 RS |
2584 | { |
2585 | INC_BOTH (from, from_byte); | |
2586 | } | |
8489eb67 RS |
2587 | if (mathexit) |
2588 | { | |
2589 | mathexit = 0; | |
2590 | goto close1; | |
2591 | } | |
2592 | mathexit = 1; | |
2593 | ||
2594 | case Sopen: | |
2595 | if (!++depth) goto done; | |
2596 | break; | |
2597 | ||
2598 | case Sclose: | |
2599 | close1: | |
2600 | if (!--depth) goto done; | |
2601 | if (depth < min_depth) | |
9471945b KS |
2602 | xsignal3 (Qscan_error, |
2603 | build_string ("Containing expression ends prematurely"), | |
2604 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2605 | break; |
2606 | ||
2607 | case Sstring: | |
195d1361 | 2608 | case Sstring_fence: |
ef316cf0 | 2609 | temp_pos = dec_bytepos (from_byte); |
b7dbcc19 | 2610 | stringterm = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
8489eb67 RS |
2611 | while (1) |
2612 | { | |
8b8bf8e6 SM |
2613 | if (from >= stop) |
2614 | goto lose; | |
195d1361 | 2615 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
b7dbcc19 | 2616 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
42ebfa31 SM |
2617 | if (code == Sstring |
2618 | ? (c == stringterm | |
2619 | && SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring) | |
bd25db08 | 2620 | : SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring_fence) |
195d1361 | 2621 | break; |
7c7ca3d2 RS |
2622 | |
2623 | /* Some compilers can't handle this inside the switch. */ | |
bd25db08 | 2624 | temp = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7c7ca3d2 | 2625 | switch (temp) |
8489eb67 RS |
2626 | { |
2627 | case Scharquote: | |
2628 | case Sescape: | |
6a140a74 | 2629 | INC_BOTH (from, from_byte); |
8489eb67 | 2630 | } |
6a140a74 | 2631 | INC_BOTH (from, from_byte); |
8489eb67 | 2632 | } |
6a140a74 | 2633 | INC_BOTH (from, from_byte); |
8489eb67 RS |
2634 | if (!depth && sexpflag) goto done; |
2635 | break; | |
240c43e8 SM |
2636 | default: |
2637 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
2638 | break; | |
8489eb67 RS |
2639 | } |
2640 | } | |
2641 | ||
2642 | /* Reached end of buffer. Error if within object, return nil if between */ | |
8b8bf8e6 SM |
2643 | if (depth) |
2644 | goto lose; | |
8489eb67 RS |
2645 | |
2646 | immediate_quit = 0; | |
2647 | return Qnil; | |
2648 | ||
2649 | /* End of object reached */ | |
2650 | done: | |
2651 | count--; | |
2652 | } | |
2653 | ||
2654 | while (count < 0) | |
2655 | { | |
8489eb67 RS |
2656 | while (from > stop) |
2657 | { | |
c5683ceb | 2658 | int syntax; |
6a140a74 | 2659 | DEC_BOTH (from, from_byte); |
195d1361 | 2660 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
b7dbcc19 | 2661 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
c5683ceb | 2662 | syntax= SYNTAX_WITH_FLAGS (c); |
bd25db08 | 2663 | code = SYNTAX_WITH_MULTIBYTE_CHECK (c); |
7bf5e9e4 RS |
2664 | if (depth == min_depth) |
2665 | last_good = from; | |
7fc8191e | 2666 | comstyle = 0; |
c5683ceb | 2667 | comnested = SYNTAX_FLAGS_COMMENT_NESTED (syntax); |
7fc8191e | 2668 | if (code == Sendcomment) |
c5683ceb SM |
2669 | comstyle = SYNTAX_FLAGS_COMMENT_STYLE (syntax, 0); |
2670 | if (from > stop && SYNTAX_FLAGS_COMEND_SECOND (syntax) | |
1674d9a2 | 2671 | && prev_char_comend_first (from, from_byte) |
8489eb67 | 2672 | && parse_sexp_ignore_comments) |
e5d4f4dc | 2673 | { |
f902a008 | 2674 | /* We must record the comment style encountered so that |
e5d4f4dc | 2675 | later, we can match only the proper comment begin |
f902a008 | 2676 | sequence of the same style. */ |
01f44d5a | 2677 | int c2, other_syntax; |
6a140a74 | 2678 | DEC_BOTH (from, from_byte); |
f902a008 RS |
2679 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
2680 | code = Sendcomment; | |
01f44d5a PE |
2681 | c2 = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
2682 | other_syntax = SYNTAX_WITH_FLAGS (c2); | |
c5683ceb SM |
2683 | comstyle = SYNTAX_FLAGS_COMMENT_STYLE (other_syntax, syntax); |
2684 | comnested | |
2685 | = comnested || SYNTAX_FLAGS_COMMENT_NESTED (other_syntax); | |
e5d4f4dc | 2686 | } |
7d0393cf | 2687 | |
9828a477 | 2688 | /* Quoting turns anything except a comment-ender |
e6365855 | 2689 | into a word character. Note that this cannot be true |
f902a008 | 2690 | if we decremented FROM in the if-statement above. */ |
9828a477 | 2691 | if (code != Sendcomment && char_quoted (from, from_byte)) |
240c43e8 SM |
2692 | { |
2693 | DEC_BOTH (from, from_byte); | |
2694 | code = Sword; | |
2695 | } | |
c5683ceb | 2696 | else if (SYNTAX_FLAGS_PREFIX (syntax)) |
8489eb67 RS |
2697 | continue; |
2698 | ||
9828a477 | 2699 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
2700 | { |
2701 | case Sword: | |
2702 | case Ssymbol: | |
9828a477 KH |
2703 | case Sescape: |
2704 | case Scharquote: | |
8489eb67 | 2705 | if (depth || !sexpflag) break; |
195d1361 RS |
2706 | /* This word counts as a sexp; count object finished |
2707 | after passing it. */ | |
8489eb67 RS |
2708 | while (from > stop) |
2709 | { | |
6a140a74 | 2710 | temp_pos = from_byte; |
4b4deea2 | 2711 | if (! NILP (BVAR (current_buffer, enable_multibyte_characters))) |
ef316cf0 RS |
2712 | DEC_POS (temp_pos); |
2713 | else | |
2714 | temp_pos--; | |
6a140a74 | 2715 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
b7dbcc19 | 2716 | c1 = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
bd25db08 | 2717 | temp_code = SYNTAX_WITH_MULTIBYTE_CHECK (c1); |
9828a477 KH |
2718 | /* Don't allow comment-end to be quoted. */ |
2719 | if (temp_code == Sendcomment) | |
2720 | goto done2; | |
6a140a74 | 2721 | quoted = char_quoted (from - 1, temp_pos); |
8489eb67 | 2722 | if (quoted) |
93da5fff | 2723 | { |
6a140a74 | 2724 | DEC_BOTH (from, from_byte); |
ef316cf0 | 2725 | temp_pos = dec_bytepos (temp_pos); |
6a140a74 | 2726 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
93da5fff | 2727 | } |
b7dbcc19 | 2728 | c1 = FETCH_CHAR_AS_MULTIBYTE (temp_pos); |
bd25db08 | 2729 | temp_code = SYNTAX_WITH_MULTIBYTE_CHECK (c1); |
93da5fff KH |
2730 | if (! (quoted || temp_code == Sword |
2731 | || temp_code == Ssymbol | |
2732 | || temp_code == Squote)) | |
8489eb67 | 2733 | goto done2; |
6a140a74 | 2734 | DEC_BOTH (from, from_byte); |
8489eb67 RS |
2735 | } |
2736 | goto done2; | |
2737 | ||
2738 | case Smath: | |
2739 | if (!sexpflag) | |
2740 | break; | |
ef316cf0 | 2741 | temp_pos = dec_bytepos (from_byte); |
6a140a74 | 2742 | UPDATE_SYNTAX_TABLE_BACKWARD (from - 1); |
b7dbcc19 | 2743 | if (from != stop && c == FETCH_CHAR_AS_MULTIBYTE (temp_pos)) |
6a140a74 | 2744 | DEC_BOTH (from, from_byte); |
8489eb67 RS |
2745 | if (mathexit) |
2746 | { | |
2747 | mathexit = 0; | |
2748 | goto open2; | |
2749 | } | |
2750 | mathexit = 1; | |
2751 | ||
2752 | case Sclose: | |
2753 | if (!++depth) goto done2; | |
2754 | break; | |
2755 | ||
2756 | case Sopen: | |
2757 | open2: | |
2758 | if (!--depth) goto done2; | |
2759 | if (depth < min_depth) | |
9471945b KS |
2760 | xsignal3 (Qscan_error, |
2761 | build_string ("Containing expression ends prematurely"), | |
2762 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2763 | break; |
2764 | ||
2765 | case Sendcomment: | |
2766 | if (!parse_sexp_ignore_comments) | |
2767 | break; | |
95ff8dfc | 2768 | found = back_comment (from, from_byte, stop, comnested, comstyle, |
6a140a74 | 2769 | &out_charpos, &out_bytepos); |
1aa963c8 SM |
2770 | /* FIXME: if found == -1, then it really wasn't a comment-end. |
2771 | For single-char Sendcomment, we can't do much about it apart | |
2772 | from skipping the char. | |
2773 | For 2-char endcomments, we could try again, taking both | |
2774 | chars as separate entities, but it's a lot of trouble | |
2775 | for very little gain, so we don't bother either. -sm */ | |
6a140a74 RS |
2776 | if (found != -1) |
2777 | from = out_charpos, from_byte = out_bytepos; | |
8489eb67 RS |
2778 | break; |
2779 | ||
195d1361 RS |
2780 | case Scomment_fence: |
2781 | case Sstring_fence: | |
2782 | while (1) | |
2783 | { | |
8b8bf8e6 SM |
2784 | if (from == stop) |
2785 | goto lose; | |
6b61353c | 2786 | DEC_BOTH (from, from_byte); |
195d1361 | 2787 | UPDATE_SYNTAX_TABLE_BACKWARD (from); |
7d0393cf | 2788 | if (!char_quoted (from, from_byte) |
b7dbcc19 | 2789 | && (c = FETCH_CHAR_AS_MULTIBYTE (from_byte), |
bd25db08 | 2790 | SYNTAX_WITH_MULTIBYTE_CHECK (c) == code)) |
195d1361 RS |
2791 | break; |
2792 | } | |
2793 | if (code == Sstring_fence && !depth && sexpflag) goto done2; | |
2794 | break; | |
7d0393cf | 2795 | |
8489eb67 | 2796 | case Sstring: |
b7dbcc19 | 2797 | stringterm = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
8489eb67 RS |
2798 | while (1) |
2799 | { | |
8b8bf8e6 SM |
2800 | if (from == stop) |
2801 | goto lose; | |
6b61353c KH |
2802 | DEC_BOTH (from, from_byte); |
2803 | UPDATE_SYNTAX_TABLE_BACKWARD (from); | |
2804 | if (!char_quoted (from, from_byte) | |
2805 | && (stringterm | |
2806 | == (c = FETCH_CHAR_AS_MULTIBYTE (from_byte))) | |
42ebfa31 | 2807 | && SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring) |
8489eb67 | 2808 | break; |
8489eb67 | 2809 | } |
8489eb67 RS |
2810 | if (!depth && sexpflag) goto done2; |
2811 | break; | |
240c43e8 SM |
2812 | default: |
2813 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
2814 | break; | |
8489eb67 RS |
2815 | } |
2816 | } | |
2817 | ||
2818 | /* Reached start of buffer. Error if within object, return nil if between */ | |
8b8bf8e6 SM |
2819 | if (depth) |
2820 | goto lose; | |
8489eb67 RS |
2821 | |
2822 | immediate_quit = 0; | |
2823 | return Qnil; | |
2824 | ||
2825 | done2: | |
2826 | count++; | |
2827 | } | |
2828 | ||
2829 | ||
2830 | immediate_quit = 0; | |
1e142fb7 | 2831 | XSETFASTINT (val, from); |
8489eb67 RS |
2832 | return val; |
2833 | ||
2834 | lose: | |
9471945b KS |
2835 | xsignal3 (Qscan_error, |
2836 | build_string ("Unbalanced parentheses"), | |
2837 | make_number (last_good), make_number (from)); | |
8489eb67 RS |
2838 | } |
2839 | ||
8489eb67 | 2840 | DEFUN ("scan-lists", Fscan_lists, Sscan_lists, 3, 3, 0, |
fdb82f93 PJ |
2841 | doc: /* Scan from character number FROM by COUNT lists. |
2842 | Returns the character number of the position thus found. | |
2843 | ||
2844 | If DEPTH is nonzero, paren depth begins counting from that value, | |
2845 | only places where the depth in parentheses becomes zero | |
2846 | are candidates for stopping; COUNT such places are counted. | |
2847 | Thus, a positive value for DEPTH means go out levels. | |
2848 | ||
2849 | Comments are ignored if `parse-sexp-ignore-comments' is non-nil. | |
2850 | ||
2851 | If the beginning or end of (the accessible part of) the buffer is reached | |
2852 | and the depth is wrong, an error is signaled. | |
2853 | If the depth is right but the count is not used up, nil is returned. */) | |
5842a27b | 2854 | (Lisp_Object from, Lisp_Object count, Lisp_Object depth) |
8489eb67 | 2855 | { |
b7826503 PJ |
2856 | CHECK_NUMBER (from); |
2857 | CHECK_NUMBER (count); | |
2858 | CHECK_NUMBER (depth); | |
8489eb67 RS |
2859 | |
2860 | return scan_lists (XINT (from), XINT (count), XINT (depth), 0); | |
2861 | } | |
2862 | ||
2863 | DEFUN ("scan-sexps", Fscan_sexps, Sscan_sexps, 2, 2, 0, | |
fdb82f93 PJ |
2864 | doc: /* Scan from character number FROM by COUNT balanced expressions. |
2865 | If COUNT is negative, scan backwards. | |
2866 | Returns the character number of the position thus found. | |
2867 | ||
2868 | Comments are ignored if `parse-sexp-ignore-comments' is non-nil. | |
2869 | ||
2870 | If the beginning or end of (the accessible part of) the buffer is reached | |
2871 | in the middle of a parenthetical grouping, an error is signaled. | |
2872 | If the beginning or end is reached between groupings | |
2873 | but before count is used up, nil is returned. */) | |
5842a27b | 2874 | (Lisp_Object from, Lisp_Object count) |
8489eb67 | 2875 | { |
b7826503 PJ |
2876 | CHECK_NUMBER (from); |
2877 | CHECK_NUMBER (count); | |
8489eb67 RS |
2878 | |
2879 | return scan_lists (XINT (from), XINT (count), 0, 1); | |
2880 | } | |
2881 | ||
2882 | DEFUN ("backward-prefix-chars", Fbackward_prefix_chars, Sbackward_prefix_chars, | |
fdb82f93 PJ |
2883 | 0, 0, 0, |
2884 | doc: /* Move point backward over any number of chars with prefix syntax. | |
2885 | This includes chars with "quote" or "prefix" syntax (' or p). */) | |
5842a27b | 2886 | (void) |
8489eb67 | 2887 | { |
4f3a2f8d EZ |
2888 | EMACS_INT beg = BEGV; |
2889 | EMACS_INT opoint = PT; | |
2890 | EMACS_INT opoint_byte = PT_BYTE; | |
2891 | EMACS_INT pos = PT; | |
2892 | EMACS_INT pos_byte = PT_BYTE; | |
93da5fff | 2893 | int c; |
93da5fff | 2894 | |
7d0393cf | 2895 | if (pos <= beg) |
195d1361 | 2896 | { |
f902a008 RS |
2897 | SET_PT_BOTH (opoint, opoint_byte); |
2898 | ||
2899 | return Qnil; | |
195d1361 | 2900 | } |
8489eb67 | 2901 | |
f902a008 RS |
2902 | SETUP_SYNTAX_TABLE (pos, -1); |
2903 | ||
6a140a74 RS |
2904 | DEC_BOTH (pos, pos_byte); |
2905 | ||
1fd3172d | 2906 | while (!char_quoted (pos, pos_byte) |
195d1361 | 2907 | /* Previous statement updates syntax table. */ |
b7dbcc19 | 2908 | && ((c = FETCH_CHAR_AS_MULTIBYTE (pos_byte), SYNTAX (c) == Squote) |
93da5fff KH |
2909 | || SYNTAX_PREFIX (c))) |
2910 | { | |
1fd3172d RS |
2911 | opoint = pos; |
2912 | opoint_byte = pos_byte; | |
2913 | ||
2914 | if (pos + 1 > beg) | |
2915 | DEC_BOTH (pos, pos_byte); | |
93da5fff | 2916 | } |
8489eb67 | 2917 | |
6a140a74 | 2918 | SET_PT_BOTH (opoint, opoint_byte); |
8489eb67 RS |
2919 | |
2920 | return Qnil; | |
2921 | } | |
2922 | \f | |
6a140a74 | 2923 | /* Parse forward from FROM / FROM_BYTE to END, |
e5d4f4dc RS |
2924 | assuming that FROM has state OLDSTATE (nil means FROM is start of function), |
2925 | and return a description of the state of the parse at END. | |
c81a3712 | 2926 | If STOPBEFORE is nonzero, stop at the start of an atom. |
644ea4df RS |
2927 | If COMMENTSTOP is 1, stop at the start of a comment. |
2928 | If COMMENTSTOP is -1, stop at the start or end of a comment, | |
2929 | after the beginning of a string, or after the end of a string. */ | |
8489eb67 | 2930 | |
340f92b5 | 2931 | static void |
d5a3eaaf AS |
2932 | scan_sexps_forward (struct lisp_parse_state *stateptr, |
2933 | EMACS_INT from, EMACS_INT from_byte, EMACS_INT end, | |
2934 | int targetdepth, int stopbefore, | |
2935 | Lisp_Object oldstate, int commentstop) | |
8489eb67 RS |
2936 | { |
2937 | struct lisp_parse_state state; | |
2938 | ||
2939 | register enum syntaxcode code; | |
95ff8dfc RS |
2940 | int c1; |
2941 | int comnested; | |
8489eb67 RS |
2942 | struct level { int last, prev; }; |
2943 | struct level levelstart[100]; | |
2944 | register struct level *curlevel = levelstart; | |
2945 | struct level *endlevel = levelstart + 100; | |
8489eb67 RS |
2946 | register int depth; /* Paren depth of current scanning location. |
2947 | level - levelstart equals this except | |
2948 | when the depth becomes negative. */ | |
2949 | int mindepth; /* Lowest DEPTH value seen. */ | |
2950 | int start_quoted = 0; /* Nonzero means starting after a char quote */ | |
2951 | Lisp_Object tem; | |
8e2911c2 AS |
2952 | EMACS_INT prev_from; /* Keep one character before FROM. */ |
2953 | EMACS_INT prev_from_byte; | |
1fd3172d | 2954 | int prev_from_syntax; |
195d1361 RS |
2955 | int boundary_stop = commentstop == -1; |
2956 | int nofence; | |
95ff8dfc | 2957 | int found; |
60c86a83 | 2958 | EMACS_INT out_bytepos, out_charpos; |
7c7ca3d2 | 2959 | int temp; |
93da5fff KH |
2960 | |
2961 | prev_from = from; | |
6a140a74 RS |
2962 | prev_from_byte = from_byte; |
2963 | if (from != BEGV) | |
2964 | DEC_BOTH (prev_from, prev_from_byte); | |
93da5fff KH |
2965 | |
2966 | /* Use this macro instead of `from++'. */ | |
6a140a74 RS |
2967 | #define INC_FROM \ |
2968 | do { prev_from = from; \ | |
2969 | prev_from_byte = from_byte; \ | |
327719ee | 2970 | temp = FETCH_CHAR_AS_MULTIBYTE (prev_from_byte); \ |
ab229fdd | 2971 | prev_from_syntax = SYNTAX_WITH_FLAGS (temp); \ |
ef316cf0 | 2972 | INC_BOTH (from, from_byte); \ |
3ceb4629 SM |
2973 | if (from < end) \ |
2974 | UPDATE_SYNTAX_TABLE_FORWARD (from); \ | |
6a140a74 | 2975 | } while (0) |
8489eb67 RS |
2976 | |
2977 | immediate_quit = 1; | |
2978 | QUIT; | |
2979 | ||
265a9e55 | 2980 | if (NILP (oldstate)) |
8489eb67 RS |
2981 | { |
2982 | depth = 0; | |
2983 | state.instring = -1; | |
2984 | state.incomment = 0; | |
195d1361 RS |
2985 | state.comstyle = 0; /* comment style a by default. */ |
2986 | state.comstr_start = -1; /* no comment/string seen. */ | |
8489eb67 RS |
2987 | } |
2988 | else | |
2989 | { | |
2990 | tem = Fcar (oldstate); | |
265a9e55 | 2991 | if (!NILP (tem)) |
8489eb67 RS |
2992 | depth = XINT (tem); |
2993 | else | |
2994 | depth = 0; | |
2995 | ||
2996 | oldstate = Fcdr (oldstate); | |
2997 | oldstate = Fcdr (oldstate); | |
2998 | oldstate = Fcdr (oldstate); | |
2999 | tem = Fcar (oldstate); | |
195d1361 | 3000 | /* Check whether we are inside string_fence-style string: */ |
7d0393cf JB |
3001 | state.instring = (!NILP (tem) |
3002 | ? (INTEGERP (tem) ? XINT (tem) : ST_STRING_STYLE) | |
e76f1c44 | 3003 | : -1); |
8489eb67 RS |
3004 | |
3005 | oldstate = Fcdr (oldstate); | |
3006 | tem = Fcar (oldstate); | |
e76f1c44 GM |
3007 | state.incomment = (!NILP (tem) |
3008 | ? (INTEGERP (tem) ? XINT (tem) : -1) | |
95ff8dfc | 3009 | : 0); |
8489eb67 RS |
3010 | |
3011 | oldstate = Fcdr (oldstate); | |
3012 | tem = Fcar (oldstate); | |
265a9e55 | 3013 | start_quoted = !NILP (tem); |
e5d4f4dc | 3014 | |
95ff8dfc | 3015 | /* if the eighth element of the list is nil, we are in comment |
195d1361 RS |
3016 | style a. If it is non-nil, we are in comment style b */ |
3017 | oldstate = Fcdr (oldstate); | |
e5d4f4dc | 3018 | oldstate = Fcdr (oldstate); |
195d1361 | 3019 | tem = Fcar (oldstate); |
c5683ceb SM |
3020 | state.comstyle = (NILP (tem) |
3021 | ? 0 | |
3022 | : (EQ (tem, Qsyntax_table) | |
3023 | ? ST_COMMENT_STYLE | |
3024 | : INTEGERP (tem) ? XINT (tem) : 1)); | |
195d1361 | 3025 | |
e5d4f4dc | 3026 | oldstate = Fcdr (oldstate); |
e5d4f4dc | 3027 | tem = Fcar (oldstate); |
195d1361 | 3028 | state.comstr_start = NILP (tem) ? -1 : XINT (tem) ; |
1a102a58 RS |
3029 | oldstate = Fcdr (oldstate); |
3030 | tem = Fcar (oldstate); | |
3031 | while (!NILP (tem)) /* >= second enclosing sexps. */ | |
3032 | { | |
3033 | /* curlevel++->last ran into compiler bug on Apollo */ | |
3034 | curlevel->last = XINT (Fcar (tem)); | |
3035 | if (++curlevel == endlevel) | |
02010917 | 3036 | curlevel--; /* error ("Nesting too deep for parser"); */ |
1a102a58 RS |
3037 | curlevel->prev = -1; |
3038 | curlevel->last = -1; | |
3039 | tem = Fcdr (tem); | |
3040 | } | |
8489eb67 RS |
3041 | } |
3042 | state.quoted = 0; | |
3043 | mindepth = depth; | |
3044 | ||
3045 | curlevel->prev = -1; | |
3046 | curlevel->last = -1; | |
3047 | ||
326b283b | 3048 | SETUP_SYNTAX_TABLE (prev_from, 1); |
ab229fdd AS |
3049 | temp = FETCH_CHAR (prev_from_byte); |
3050 | prev_from_syntax = SYNTAX_WITH_FLAGS (temp); | |
3ceb4629 | 3051 | UPDATE_SYNTAX_TABLE_FORWARD (from); |
326b283b | 3052 | |
195d1361 | 3053 | /* Enter the loop at a place appropriate for initial state. */ |
8489eb67 | 3054 | |
326b283b RS |
3055 | if (state.incomment) |
3056 | goto startincomment; | |
8489eb67 RS |
3057 | if (state.instring >= 0) |
3058 | { | |
195d1361 | 3059 | nofence = state.instring != ST_STRING_STYLE; |
326b283b RS |
3060 | if (start_quoted) |
3061 | goto startquotedinstring; | |
8489eb67 RS |
3062 | goto startinstring; |
3063 | } | |
326b283b RS |
3064 | else if (start_quoted) |
3065 | goto startquoted; | |
1fd3172d | 3066 | |
8489eb67 RS |
3067 | while (from < end) |
3068 | { | |
c5683ceb | 3069 | int syntax; |
93da5fff | 3070 | INC_FROM; |
1fd3172d | 3071 | code = prev_from_syntax & 0xff; |
4c920633 | 3072 | |
c5d4b96f SM |
3073 | if (from < end |
3074 | && SYNTAX_FLAGS_COMSTART_FIRST (prev_from_syntax) | |
3075 | && (c1 = FETCH_CHAR (from_byte), | |
c5683ceb SM |
3076 | syntax = SYNTAX_WITH_FLAGS (c1), |
3077 | SYNTAX_FLAGS_COMSTART_SECOND (syntax))) | |
c5d4b96f SM |
3078 | /* Duplicate code to avoid a complex if-expression |
3079 | which causes trouble for the SGI compiler. */ | |
95ff8dfc | 3080 | { |
c5d4b96f SM |
3081 | /* Record the comment style we have entered so that only |
3082 | the comment-end sequence of the same style actually | |
3083 | terminates the comment section. */ | |
c5683ceb SM |
3084 | state.comstyle |
3085 | = SYNTAX_FLAGS_COMMENT_STYLE (syntax, prev_from_syntax); | |
c5d4b96f | 3086 | comnested = SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax); |
c5683ceb | 3087 | comnested = comnested || SYNTAX_FLAGS_COMMENT_NESTED (syntax); |
c5d4b96f | 3088 | state.incomment = comnested ? 1 : -1; |
95ff8dfc | 3089 | state.comstr_start = prev_from; |
c5d4b96f SM |
3090 | INC_FROM; |
3091 | code = Scomment; | |
95ff8dfc | 3092 | } |
4c920633 | 3093 | else if (code == Scomment_fence) |
e5d4f4dc RS |
3094 | { |
3095 | /* Record the comment style we have entered so that only | |
3096 | the comment-end sequence of the same style actually | |
3097 | terminates the comment section. */ | |
95ff8dfc RS |
3098 | state.comstyle = ST_COMMENT_STYLE; |
3099 | state.incomment = -1; | |
195d1361 | 3100 | state.comstr_start = prev_from; |
e5d4f4dc | 3101 | code = Scomment; |
e5d4f4dc | 3102 | } |
c5d4b96f SM |
3103 | else if (code == Scomment) |
3104 | { | |
c5683ceb | 3105 | state.comstyle = SYNTAX_FLAGS_COMMENT_STYLE (prev_from_syntax, 0); |
c5d4b96f SM |
3106 | state.incomment = (SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax) ? |
3107 | 1 : -1); | |
3108 | state.comstr_start = prev_from; | |
3109 | } | |
e5d4f4dc | 3110 | |
1fd3172d | 3111 | if (SYNTAX_FLAGS_PREFIX (prev_from_syntax)) |
8489eb67 | 3112 | continue; |
0220c518 | 3113 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
3114 | { |
3115 | case Sescape: | |
3116 | case Scharquote: | |
3117 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
93da5fff | 3118 | curlevel->last = prev_from; |
8489eb67 RS |
3119 | startquoted: |
3120 | if (from == end) goto endquoted; | |
93da5fff | 3121 | INC_FROM; |
8489eb67 RS |
3122 | goto symstarted; |
3123 | /* treat following character as a word constituent */ | |
3124 | case Sword: | |
3125 | case Ssymbol: | |
3126 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
93da5fff | 3127 | curlevel->last = prev_from; |
8489eb67 RS |
3128 | symstarted: |
3129 | while (from < end) | |
3130 | { | |
7c7ca3d2 | 3131 | /* Some compilers can't handle this inside the switch. */ |
327719ee | 3132 | temp = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
ab229fdd | 3133 | temp = SYNTAX (temp); |
7c7ca3d2 | 3134 | switch (temp) |
8489eb67 RS |
3135 | { |
3136 | case Scharquote: | |
3137 | case Sescape: | |
93da5fff | 3138 | INC_FROM; |
8489eb67 RS |
3139 | if (from == end) goto endquoted; |
3140 | break; | |
3141 | case Sword: | |
3142 | case Ssymbol: | |
3143 | case Squote: | |
3144 | break; | |
3145 | default: | |
3146 | goto symdone; | |
3147 | } | |
93da5fff | 3148 | INC_FROM; |
8489eb67 RS |
3149 | } |
3150 | symdone: | |
3151 | curlevel->prev = curlevel->last; | |
3152 | break; | |
3153 | ||
240c43e8 | 3154 | case Scomment_fence: /* Can't happen because it's handled above. */ |
8489eb67 | 3155 | case Scomment: |
195d1361 | 3156 | if (commentstop || boundary_stop) goto done; |
d355bd8a SM |
3157 | startincomment: |
3158 | /* The (from == BEGV) test was to enter the loop in the middle so | |
95ff8dfc | 3159 | that we find a 2-char comment ender even if we start in the |
e6365855 | 3160 | middle of it. We don't want to do that if we're just at the |
d355bd8a | 3161 | beginning of the comment (think of (*) ... (*)). */ |
95ff8dfc RS |
3162 | found = forw_comment (from, from_byte, end, |
3163 | state.incomment, state.comstyle, | |
e6365855 SM |
3164 | (from == BEGV || from < state.comstr_start + 3) |
3165 | ? 0 : prev_from_syntax, | |
95ff8dfc RS |
3166 | &out_charpos, &out_bytepos, &state.incomment); |
3167 | from = out_charpos; from_byte = out_bytepos; | |
3168 | /* Beware! prev_from and friends are invalid now. | |
3169 | Luckily, the `done' doesn't use them and the INC_FROM | |
3170 | sets them to a sane value without looking at them. */ | |
3171 | if (!found) goto done; | |
d355bd8a | 3172 | INC_FROM; |
8489eb67 | 3173 | state.incomment = 0; |
e5d4f4dc | 3174 | state.comstyle = 0; /* reset the comment style */ |
195d1361 | 3175 | if (boundary_stop) goto done; |
8489eb67 RS |
3176 | break; |
3177 | ||
3178 | case Sopen: | |
3179 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
3180 | depth++; | |
3181 | /* curlevel++->last ran into compiler bug on Apollo */ | |
93da5fff | 3182 | curlevel->last = prev_from; |
8489eb67 | 3183 | if (++curlevel == endlevel) |
02010917 | 3184 | curlevel--; /* error ("Nesting too deep for parser"); */ |
8489eb67 RS |
3185 | curlevel->prev = -1; |
3186 | curlevel->last = -1; | |
30844415 | 3187 | if (targetdepth == depth) goto done; |
8489eb67 RS |
3188 | break; |
3189 | ||
3190 | case Sclose: | |
3191 | depth--; | |
3192 | if (depth < mindepth) | |
3193 | mindepth = depth; | |
3194 | if (curlevel != levelstart) | |
3195 | curlevel--; | |
3196 | curlevel->prev = curlevel->last; | |
30844415 | 3197 | if (targetdepth == depth) goto done; |
8489eb67 RS |
3198 | break; |
3199 | ||
3200 | case Sstring: | |
195d1361 RS |
3201 | case Sstring_fence: |
3202 | state.comstr_start = from - 1; | |
8489eb67 | 3203 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ |
93da5fff | 3204 | curlevel->last = prev_from; |
7d0393cf | 3205 | state.instring = (code == Sstring |
b7dbcc19 | 3206 | ? (FETCH_CHAR_AS_MULTIBYTE (prev_from_byte)) |
195d1361 RS |
3207 | : ST_STRING_STYLE); |
3208 | if (boundary_stop) goto done; | |
8489eb67 | 3209 | startinstring: |
195d1361 | 3210 | { |
644ea4df | 3211 | nofence = state.instring != ST_STRING_STYLE; |
7d0393cf | 3212 | |
644ea4df RS |
3213 | while (1) |
3214 | { | |
3215 | int c; | |
195d1361 | 3216 | |
644ea4df | 3217 | if (from >= end) goto done; |
b7dbcc19 | 3218 | c = FETCH_CHAR_AS_MULTIBYTE (from_byte); |
7c7ca3d2 RS |
3219 | /* Some compilers can't handle this inside the switch. */ |
3220 | temp = SYNTAX (c); | |
ccf89641 KH |
3221 | |
3222 | /* Check TEMP here so that if the char has | |
3223 | a syntax-table property which says it is NOT | |
3224 | a string character, it does not end the string. */ | |
3225 | if (nofence && c == state.instring && temp == Sstring) | |
3226 | break; | |
3227 | ||
7c7ca3d2 | 3228 | switch (temp) |
644ea4df RS |
3229 | { |
3230 | case Sstring_fence: | |
3231 | if (!nofence) goto string_end; | |
3232 | break; | |
3233 | case Scharquote: | |
3234 | case Sescape: | |
3235 | INC_FROM; | |
3236 | startquotedinstring: | |
3237 | if (from >= end) goto endquoted; | |
195d1361 | 3238 | } |
644ea4df RS |
3239 | INC_FROM; |
3240 | } | |
195d1361 RS |
3241 | } |
3242 | string_end: | |
8489eb67 RS |
3243 | state.instring = -1; |
3244 | curlevel->prev = curlevel->last; | |
93da5fff | 3245 | INC_FROM; |
195d1361 | 3246 | if (boundary_stop) goto done; |
8489eb67 RS |
3247 | break; |
3248 | ||
3249 | case Smath: | |
240c43e8 SM |
3250 | /* FIXME: We should do something with it. */ |
3251 | break; | |
3252 | default: | |
3253 | /* Ignore whitespace, punctuation, quote, endcomment. */ | |
8489eb67 RS |
3254 | break; |
3255 | } | |
3256 | } | |
3257 | goto done; | |
3258 | ||
3259 | stop: /* Here if stopping before start of sexp. */ | |
93da5fff | 3260 | from = prev_from; /* We have just fetched the char that starts it; */ |
8489eb67 RS |
3261 | goto done; /* but return the position before it. */ |
3262 | ||
3263 | endquoted: | |
3264 | state.quoted = 1; | |
3265 | done: | |
3266 | state.depth = depth; | |
3267 | state.mindepth = mindepth; | |
3268 | state.thislevelstart = curlevel->prev; | |
3269 | state.prevlevelstart | |
3270 | = (curlevel == levelstart) ? -1 : (curlevel - 1)->last; | |
3271 | state.location = from; | |
1a102a58 | 3272 | state.levelstarts = Qnil; |
eb4d412d PE |
3273 | while (curlevel > levelstart) |
3274 | state.levelstarts = Fcons (make_number ((--curlevel)->last), | |
3275 | state.levelstarts); | |
8489eb67 RS |
3276 | immediate_quit = 0; |
3277 | ||
e5d4f4dc | 3278 | *stateptr = state; |
8489eb67 RS |
3279 | } |
3280 | ||
c81a3712 | 3281 | DEFUN ("parse-partial-sexp", Fparse_partial_sexp, Sparse_partial_sexp, 2, 6, 0, |
fdb82f93 PJ |
3282 | doc: /* Parse Lisp syntax starting at FROM until TO; return status of parse at TO. |
3283 | Parsing stops at TO or when certain criteria are met; | |
3284 | point is set to where parsing stops. | |
3285 | If fifth arg OLDSTATE is omitted or nil, | |
3286 | parsing assumes that FROM is the beginning of a function. | |
0e6228bc | 3287 | Value is a list of elements describing final state of parsing: |
fdb82f93 PJ |
3288 | 0. depth in parens. |
3289 | 1. character address of start of innermost containing list; nil if none. | |
3290 | 2. character address of start of last complete sexp terminated. | |
3291 | 3. non-nil if inside a string. | |
3292 | (it is the character that will terminate the string, | |
3293 | or t if the string should be terminated by a generic string delimiter.) | |
7d0393cf | 3294 | 4. nil if outside a comment, t if inside a non-nestable comment, |
fdb82f93 PJ |
3295 | else an integer (the current comment nesting). |
3296 | 5. t if following a quote character. | |
3297 | 6. the minimum paren-depth encountered during this scan. | |
c5683ceb | 3298 | 7. style of comment, if any. |
fdb82f93 PJ |
3299 | 8. character address of start of comment or string; nil if not in one. |
3300 | 9. Intermediate data for continuation of parsing (subject to change). | |
3301 | If third arg TARGETDEPTH is non-nil, parsing stops if the depth | |
3302 | in parentheses becomes equal to TARGETDEPTH. | |
3303 | Fourth arg STOPBEFORE non-nil means stop when come to | |
3304 | any character that starts a sexp. | |
0e6228bc | 3305 | Fifth arg OLDSTATE is a list like what this function returns. |
fdb82f93 | 3306 | It is used to initialize the state of the parse. Elements number 1, 2, 6 |
43726c05 | 3307 | and 8 are ignored. |
fdb82f93 PJ |
3308 | Sixth arg COMMENTSTOP non-nil means stop at the start of a comment. |
3309 | If it is symbol `syntax-table', stop after the start of a comment or a | |
3310 | string, or after end of a comment or a string. */) | |
5842a27b | 3311 | (Lisp_Object from, Lisp_Object to, Lisp_Object targetdepth, Lisp_Object stopbefore, Lisp_Object oldstate, Lisp_Object commentstop) |
8489eb67 RS |
3312 | { |
3313 | struct lisp_parse_state state; | |
3314 | int target; | |
3315 | ||
265a9e55 | 3316 | if (!NILP (targetdepth)) |
8489eb67 | 3317 | { |
b7826503 | 3318 | CHECK_NUMBER (targetdepth); |
8489eb67 RS |
3319 | target = XINT (targetdepth); |
3320 | } | |
3321 | else | |
3322 | target = -100000; /* We won't reach this depth */ | |
3323 | ||
3324 | validate_region (&from, &to); | |
6a140a74 RS |
3325 | scan_sexps_forward (&state, XINT (from), CHAR_TO_BYTE (XINT (from)), |
3326 | XINT (to), | |
c81a3712 | 3327 | target, !NILP (stopbefore), oldstate, |
7d0393cf | 3328 | (NILP (commentstop) |
195d1361 | 3329 | ? 0 : (EQ (commentstop, Qsyntax_table) ? -1 : 1))); |
8489eb67 RS |
3330 | |
3331 | SET_PT (state.location); | |
7d0393cf | 3332 | |
8489eb67 | 3333 | return Fcons (make_number (state.depth), |
c5683ceb SM |
3334 | Fcons (state.prevlevelstart < 0 |
3335 | ? Qnil : make_number (state.prevlevelstart), | |
3336 | Fcons (state.thislevelstart < 0 | |
3337 | ? Qnil : make_number (state.thislevelstart), | |
7d0393cf JB |
3338 | Fcons (state.instring >= 0 |
3339 | ? (state.instring == ST_STRING_STYLE | |
195d1361 | 3340 | ? Qt : make_number (state.instring)) : Qnil, |
95ff8dfc RS |
3341 | Fcons (state.incomment < 0 ? Qt : |
3342 | (state.incomment == 0 ? Qnil : | |
3343 | make_number (state.incomment)), | |
8489eb67 | 3344 | Fcons (state.quoted ? Qt : Qnil, |
bff37d2a | 3345 | Fcons (make_number (state.mindepth), |
7d0393cf | 3346 | Fcons ((state.comstyle |
bff37d2a | 3347 | ? (state.comstyle == ST_COMMENT_STYLE |
c5683ceb SM |
3348 | ? Qsyntax_table |
3349 | : make_number (state.comstyle)) | |
3350 | : Qnil), | |
0beaf54f DL |
3351 | Fcons (((state.incomment |
3352 | || (state.instring >= 0)) | |
bff37d2a RS |
3353 | ? make_number (state.comstr_start) |
3354 | : Qnil), | |
1a102a58 | 3355 | Fcons (state.levelstarts, Qnil)))))))))); |
8489eb67 RS |
3356 | } |
3357 | \f | |
dfcf069d | 3358 | void |
971de7fb | 3359 | init_syntax_once (void) |
8489eb67 | 3360 | { |
78f9a1f7 | 3361 | register int i, c; |
8ea151b2 | 3362 | Lisp_Object temp; |
8489eb67 | 3363 | |
5ebaddf5 | 3364 | /* This has to be done here, before we call Fmake_char_table. */ |
d67b4f80 | 3365 | Qsyntax_table = intern_c_string ("syntax-table"); |
5ebaddf5 RS |
3366 | staticpro (&Qsyntax_table); |
3367 | ||
d67b4f80 | 3368 | /* Intern_C_String this now in case it isn't already done. |
5ebaddf5 RS |
3369 | Setting this variable twice is harmless. |
3370 | But don't staticpro it here--that is done in alloc.c. */ | |
d67b4f80 | 3371 | Qchar_table_extra_slots = intern_c_string ("char-table-extra-slots"); |
5ebaddf5 | 3372 | |
93da5fff | 3373 | /* Create objects which can be shared among syntax tables. */ |
42ebfa31 | 3374 | Vsyntax_code_object = Fmake_vector (make_number (Smax), Qnil); |
77b37c05 | 3375 | for (i = 0; i < ASIZE (Vsyntax_code_object); i++) |
93da5fff KH |
3376 | XVECTOR (Vsyntax_code_object)->contents[i] |
3377 | = Fcons (make_number (i), Qnil); | |
3378 | ||
5ebaddf5 RS |
3379 | /* Now we are ready to set up this property, so we can |
3380 | create syntax tables. */ | |
3381 | Fput (Qsyntax_table, Qchar_table_extra_slots, make_number (0)); | |
3382 | ||
93da5fff | 3383 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Swhitespace]; |
8489eb67 | 3384 | |
5ebaddf5 | 3385 | Vstandard_syntax_table = Fmake_char_table (Qsyntax_table, temp); |
8489eb67 | 3386 | |
aa7b08b4 RS |
3387 | /* Control characters should not be whitespace. */ |
3388 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Spunct]; | |
3389 | for (i = 0; i <= ' ' - 1; i++) | |
3390 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); | |
3391 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 0177, temp); | |
3392 | ||
3393 | /* Except that a few really are whitespace. */ | |
3394 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Swhitespace]; | |
3395 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ' ', temp); | |
3396 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\t', temp); | |
3397 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\n', temp); | |
3398 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 015, temp); | |
3399 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, 014, temp); | |
3400 | ||
93da5fff | 3401 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Sword]; |
8489eb67 | 3402 | for (i = 'a'; i <= 'z'; i++) |
8ea151b2 | 3403 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
8489eb67 | 3404 | for (i = 'A'; i <= 'Z'; i++) |
8ea151b2 | 3405 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
8489eb67 | 3406 | for (i = '0'; i <= '9'; i++) |
8ea151b2 RS |
3407 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
3408 | ||
3409 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '$', temp); | |
3410 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '%', temp); | |
3411 | ||
3412 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '(', | |
3413 | Fcons (make_number (Sopen), make_number (')'))); | |
3414 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ')', | |
3415 | Fcons (make_number (Sclose), make_number ('('))); | |
3416 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '[', | |
3417 | Fcons (make_number (Sopen), make_number (']'))); | |
3418 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ']', | |
3419 | Fcons (make_number (Sclose), make_number ('['))); | |
3420 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '{', | |
3421 | Fcons (make_number (Sopen), make_number ('}'))); | |
3422 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '}', | |
3423 | Fcons (make_number (Sclose), make_number ('{'))); | |
3424 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '"', | |
3425 | Fcons (make_number ((int) Sstring), Qnil)); | |
3426 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\\', | |
3427 | Fcons (make_number ((int) Sescape), Qnil)); | |
3428 | ||
93da5fff | 3429 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Ssymbol]; |
8489eb67 | 3430 | for (i = 0; i < 10; i++) |
78f9a1f7 KH |
3431 | { |
3432 | c = "_-+*/&|<>="[i]; | |
3433 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, c, temp); | |
3434 | } | |
8489eb67 | 3435 | |
93da5fff | 3436 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Spunct]; |
8489eb67 | 3437 | for (i = 0; i < 12; i++) |
78f9a1f7 KH |
3438 | { |
3439 | c = ".,;:?!#@~^'`"[i]; | |
3440 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, c, temp); | |
3441 | } | |
bd25db08 KH |
3442 | |
3443 | /* All multibyte characters have syntax `word' by default. */ | |
3444 | temp = XVECTOR (Vsyntax_code_object)->contents[(int) Sword]; | |
5c7b02ab | 3445 | char_table_set_range (Vstandard_syntax_table, 0x80, MAX_CHAR, temp); |
8489eb67 RS |
3446 | } |
3447 | ||
dfcf069d | 3448 | void |
971de7fb | 3449 | syms_of_syntax (void) |
8489eb67 | 3450 | { |
d67b4f80 | 3451 | Qsyntax_table_p = intern_c_string ("syntax-table-p"); |
8489eb67 RS |
3452 | staticpro (&Qsyntax_table_p); |
3453 | ||
93da5fff KH |
3454 | staticpro (&Vsyntax_code_object); |
3455 | ||
a2994c46 KS |
3456 | staticpro (&gl_state.object); |
3457 | staticpro (&gl_state.global_code); | |
3458 | staticpro (&gl_state.current_syntax_table); | |
3459 | staticpro (&gl_state.old_prop); | |
3460 | ||
3461 | /* Defined in regex.c */ | |
3462 | staticpro (&re_match_object); | |
3463 | ||
d67b4f80 | 3464 | Qscan_error = intern_c_string ("scan-error"); |
7bf5e9e4 RS |
3465 | staticpro (&Qscan_error); |
3466 | Fput (Qscan_error, Qerror_conditions, | |
d67b4f80 | 3467 | pure_cons (Qscan_error, pure_cons (Qerror, Qnil))); |
7bf5e9e4 | 3468 | Fput (Qscan_error, Qerror_message, |
d67b4f80 | 3469 | make_pure_c_string ("Scan error")); |
7bf5e9e4 | 3470 | |
29208e82 | 3471 | DEFVAR_BOOL ("parse-sexp-ignore-comments", parse_sexp_ignore_comments, |
fdb82f93 | 3472 | doc: /* Non-nil means `forward-sexp', etc., should treat comments as whitespace. */); |
8489eb67 | 3473 | |
29208e82 | 3474 | DEFVAR_BOOL ("parse-sexp-lookup-properties", parse_sexp_lookup_properties, |
fdb82f93 PJ |
3475 | doc: /* Non-nil means `forward-sexp', etc., obey `syntax-table' property. |
3476 | Otherwise, that text property is simply ignored. | |
3477 | See the info node `(elisp)Syntax Properties' for a description of the | |
3478 | `syntax-table' property. */); | |
195d1361 | 3479 | |
8489eb67 | 3480 | words_include_escapes = 0; |
29208e82 | 3481 | DEFVAR_BOOL ("words-include-escapes", words_include_escapes, |
fdb82f93 | 3482 | doc: /* Non-nil means `forward-word', etc., should treat escape chars part of words. */); |
8489eb67 | 3483 | |
29208e82 | 3484 | DEFVAR_BOOL ("multibyte-syntax-as-symbol", multibyte_syntax_as_symbol, |
fdb82f93 | 3485 | doc: /* Non-nil means `scan-sexps' treats all multibyte characters as symbol. */); |
bd25db08 KH |
3486 | multibyte_syntax_as_symbol = 0; |
3487 | ||
f4ed767f | 3488 | DEFVAR_BOOL ("open-paren-in-column-0-is-defun-start", |
29208e82 | 3489 | open_paren_in_column_0_is_defun_start, |
b222e415 | 3490 | doc: /* *Non-nil means an open paren in column 0 denotes the start of a defun. */); |
f4ed767f GM |
3491 | open_paren_in_column_0_is_defun_start = 1; |
3492 | ||
8f924df7 KH |
3493 | |
3494 | DEFVAR_LISP ("find-word-boundary-function-table", | |
29208e82 | 3495 | Vfind_word_boundary_function_table, |
869bb237 | 3496 | doc: /* |
8f924df7 | 3497 | Char table of functions to search for the word boundary. |
869bb237 KH |
3498 | Each function is called with two arguments; POS and LIMIT. |
3499 | POS and LIMIT are character positions in the current buffer. | |
3500 | ||
3501 | If POS is less than LIMIT, POS is at the first character of a word, | |
3502 | and the return value of a function is a position after the last | |
3503 | character of that word. | |
3504 | ||
3505 | If POS is not less than LIMIT, POS is at the last character of a word, | |
3506 | and the return value of a function is a position at the first | |
3507 | character of that word. | |
3508 | ||
3509 | In both cases, LIMIT bounds the search. */); | |
8f924df7 | 3510 | Vfind_word_boundary_function_table = Fmake_char_table (Qnil, Qnil); |
869bb237 | 3511 | |
8489eb67 RS |
3512 | defsubr (&Ssyntax_table_p); |
3513 | defsubr (&Ssyntax_table); | |
3514 | defsubr (&Sstandard_syntax_table); | |
3515 | defsubr (&Scopy_syntax_table); | |
3516 | defsubr (&Sset_syntax_table); | |
3517 | defsubr (&Schar_syntax); | |
beefa22e | 3518 | defsubr (&Smatching_paren); |
c65adb44 | 3519 | defsubr (&Sstring_to_syntax); |
8489eb67 | 3520 | defsubr (&Smodify_syntax_entry); |
62abe9cb | 3521 | defsubr (&Sinternal_describe_syntax_value); |
8489eb67 RS |
3522 | |
3523 | defsubr (&Sforward_word); | |
3524 | ||
195d1361 RS |
3525 | defsubr (&Sskip_chars_forward); |
3526 | defsubr (&Sskip_chars_backward); | |
3527 | defsubr (&Sskip_syntax_forward); | |
3528 | defsubr (&Sskip_syntax_backward); | |
3529 | ||
b3cfe0c8 | 3530 | defsubr (&Sforward_comment); |
8489eb67 RS |
3531 | defsubr (&Sscan_lists); |
3532 | defsubr (&Sscan_sexps); | |
3533 | defsubr (&Sbackward_prefix_chars); | |
3534 | defsubr (&Sparse_partial_sexp); | |
3535 | } |