Commit | Line | Data |
---|---|---|
8489eb67 | 1 | /* GNU Emacs routines to deal with syntax tables; also word and list parsing. |
0220c518 | 2 | Copyright (C) 1985, 1987, 1993, 1994, 1995 Free Software Foundation, Inc. |
8489eb67 RS |
3 | |
4 | This file is part of GNU Emacs. | |
5 | ||
6 | GNU Emacs is free software; you can redistribute it and/or modify | |
7 | it under the terms of the GNU General Public License as published by | |
a4275ad1 | 8 | the Free Software Foundation; either version 2, or (at your option) |
8489eb67 RS |
9 | any later version. |
10 | ||
11 | GNU Emacs is distributed in the hope that it will be useful, | |
12 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
13 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
14 | GNU General Public License for more details. | |
15 | ||
16 | You should have received a copy of the GNU General Public License | |
17 | along with GNU Emacs; see the file COPYING. If not, write to | |
3b7ad313 EN |
18 | the Free Software Foundation, Inc., 59 Temple Place - Suite 330, |
19 | Boston, MA 02111-1307, USA. */ | |
8489eb67 RS |
20 | |
21 | ||
18160b98 | 22 | #include <config.h> |
8489eb67 RS |
23 | #include <ctype.h> |
24 | #include "lisp.h" | |
25 | #include "commands.h" | |
26 | #include "buffer.h" | |
27 | #include "syntax.h" | |
28 | ||
5ebaddf5 | 29 | Lisp_Object Qsyntax_table_p, Qsyntax_table; |
8489eb67 | 30 | |
340f92b5 RS |
31 | static void scan_sexps_forward (); |
32 | static int char_quoted (); | |
33 | ||
8489eb67 RS |
34 | int words_include_escapes; |
35 | ||
8ea151b2 RS |
36 | /* Used as a temporary in SYNTAX_ENTRY and other macros in syntax.h, |
37 | if not compiled with GCC. No need to mark it, since it is used | |
38 | only very temporarily. */ | |
39 | Lisp_Object syntax_temp; | |
40 | ||
e5d4f4dc RS |
41 | /* This is the internal form of the parse state used in parse-partial-sexp. */ |
42 | ||
43 | struct lisp_parse_state | |
44 | { | |
45 | int depth; /* Depth at end of parsing */ | |
46 | int instring; /* -1 if not within string, else desired terminator. */ | |
47 | int incomment; /* Nonzero if within a comment at end of parsing */ | |
48 | int comstyle; /* comment style a=0, or b=1 */ | |
49 | int quoted; /* Nonzero if just after an escape char at end of parsing */ | |
50 | int thislevelstart; /* Char number of most recent start-of-expression at current level */ | |
51 | int prevlevelstart; /* Char number of start of containing expression */ | |
52 | int location; /* Char number at which parsing stopped. */ | |
53 | int mindepth; /* Minimum depth seen while scanning. */ | |
54 | int comstart; /* Position just after last comment starter. */ | |
55 | }; | |
56 | \f | |
37bef230 RS |
57 | /* These variables are a cache for finding the start of a defun. |
58 | find_start_pos is the place for which the defun start was found. | |
59 | find_start_value is the defun start position found for it. | |
60 | find_start_buffer is the buffer it was found in. | |
61 | find_start_begv is the BEGV value when it was found. | |
62 | find_start_modiff is the value of MODIFF when it was found. */ | |
63 | ||
64 | static int find_start_pos; | |
65 | static int find_start_value; | |
66 | static struct buffer *find_start_buffer; | |
67 | static int find_start_begv; | |
68 | static int find_start_modiff; | |
69 | ||
70 | /* Find a defun-start that is the last one before POS (or nearly the last). | |
71 | We record what we find, so that another call in the same area | |
72 | can return the same value right away. */ | |
73 | ||
74 | static int | |
75 | find_defun_start (pos) | |
76 | int pos; | |
77 | { | |
78 | int tem; | |
79 | int shortage; | |
80 | ||
81 | /* Use previous finding, if it's valid and applies to this inquiry. */ | |
82 | if (current_buffer == find_start_buffer | |
83 | /* Reuse the defun-start even if POS is a little farther on. | |
84 | POS might be in the next defun, but that's ok. | |
85 | Our value may not be the best possible, but will still be usable. */ | |
86 | && pos <= find_start_pos + 1000 | |
87 | && pos >= find_start_value | |
88 | && BEGV == find_start_begv | |
89 | && MODIFF == find_start_modiff) | |
90 | return find_start_value; | |
91 | ||
92 | /* Back up to start of line. */ | |
be5357e2 | 93 | tem = scan_buffer ('\n', pos, BEGV, -1, &shortage, 1); |
37bef230 RS |
94 | |
95 | while (tem > BEGV) | |
96 | { | |
97 | /* Open-paren at start of line means we found our defun-start. */ | |
98 | if (SYNTAX (FETCH_CHAR (tem)) == Sopen) | |
99 | break; | |
100 | /* Move to beg of previous line. */ | |
be5357e2 | 101 | tem = scan_buffer ('\n', tem, BEGV, -2, &shortage, 1); |
37bef230 RS |
102 | } |
103 | ||
104 | /* Record what we found, for the next try. */ | |
105 | find_start_value = tem; | |
106 | find_start_buffer = current_buffer; | |
107 | find_start_modiff = MODIFF; | |
108 | find_start_begv = BEGV; | |
109 | find_start_pos = pos; | |
110 | ||
111 | return find_start_value; | |
112 | } | |
113 | \f | |
8489eb67 | 114 | DEFUN ("syntax-table-p", Fsyntax_table_p, Ssyntax_table_p, 1, 1, 0, |
2203e1e8 | 115 | "Return t if OBJECT is a syntax table.\n\ |
8ea151b2 | 116 | Currently, any char-table counts as a syntax table.") |
2203e1e8 EN |
117 | (object) |
118 | Lisp_Object object; | |
8489eb67 | 119 | { |
2203e1e8 EN |
120 | if (CHAR_TABLE_P (object) |
121 | && XCHAR_TABLE (object)->purpose == Qsyntax_table) | |
8489eb67 RS |
122 | return Qt; |
123 | return Qnil; | |
124 | } | |
125 | ||
8ea151b2 | 126 | static void |
8489eb67 RS |
127 | check_syntax_table (obj) |
128 | Lisp_Object obj; | |
129 | { | |
d1be9f0f RS |
130 | if (!(CHAR_TABLE_P (obj) |
131 | && XCHAR_TABLE (obj)->purpose == Qsyntax_table)) | |
132 | wrong_type_argument (Qsyntax_table_p, obj); | |
8489eb67 RS |
133 | } |
134 | ||
8489eb67 RS |
135 | DEFUN ("syntax-table", Fsyntax_table, Ssyntax_table, 0, 0, 0, |
136 | "Return the current syntax table.\n\ | |
137 | This is the one specified by the current buffer.") | |
138 | () | |
139 | { | |
140 | return current_buffer->syntax_table; | |
141 | } | |
142 | ||
143 | DEFUN ("standard-syntax-table", Fstandard_syntax_table, | |
144 | Sstandard_syntax_table, 0, 0, 0, | |
145 | "Return the standard syntax table.\n\ | |
146 | This is the one used for new buffers.") | |
147 | () | |
148 | { | |
149 | return Vstandard_syntax_table; | |
150 | } | |
151 | ||
152 | DEFUN ("copy-syntax-table", Fcopy_syntax_table, Scopy_syntax_table, 0, 1, 0, | |
153 | "Construct a new syntax table and return it.\n\ | |
154 | It is a copy of the TABLE, which defaults to the standard syntax table.") | |
155 | (table) | |
156 | Lisp_Object table; | |
157 | { | |
8ea151b2 RS |
158 | Lisp_Object copy; |
159 | ||
265a9e55 | 160 | if (!NILP (table)) |
8ea151b2 RS |
161 | check_syntax_table (table); |
162 | else | |
163 | table = Vstandard_syntax_table; | |
164 | ||
165 | copy = Fcopy_sequence (table); | |
166 | Fset_char_table_parent (copy, Vstandard_syntax_table); | |
167 | return copy; | |
8489eb67 RS |
168 | } |
169 | ||
170 | DEFUN ("set-syntax-table", Fset_syntax_table, Sset_syntax_table, 1, 1, 0, | |
171 | "Select a new syntax table for the current buffer.\n\ | |
172 | One argument, a syntax table.") | |
173 | (table) | |
174 | Lisp_Object table; | |
175 | { | |
8ea151b2 | 176 | check_syntax_table (table); |
8489eb67 RS |
177 | current_buffer->syntax_table = table; |
178 | /* Indicate that this buffer now has a specified syntax table. */ | |
f110a664 RS |
179 | current_buffer->local_var_flags |
180 | |= XFASTINT (buffer_local_flags.syntax_table); | |
8489eb67 RS |
181 | return table; |
182 | } | |
183 | \f | |
184 | /* Convert a letter which signifies a syntax code | |
185 | into the code it signifies. | |
186 | This is used by modify-syntax-entry, and other things. */ | |
187 | ||
188 | unsigned char syntax_spec_code[0400] = | |
189 | { 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
190 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
191 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
192 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
193 | (char) Swhitespace, 0377, (char) Sstring, 0377, | |
194 | (char) Smath, 0377, 0377, (char) Squote, | |
195 | (char) Sopen, (char) Sclose, 0377, 0377, | |
196 | 0377, (char) Swhitespace, (char) Spunct, (char) Scharquote, | |
197 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
198 | 0377, 0377, 0377, 0377, | |
199 | (char) Scomment, 0377, (char) Sendcomment, 0377, | |
6cb71bf6 | 200 | (char) Sinherit, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* @, A ... */ |
8489eb67 RS |
201 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, |
202 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword, | |
203 | 0377, 0377, 0377, 0377, (char) Sescape, 0377, 0377, (char) Ssymbol, | |
204 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* `, a, ... */ | |
205 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, | |
206 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword, | |
207 | 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377 | |
208 | }; | |
209 | ||
210 | /* Indexed by syntax code, give the letter that describes it. */ | |
211 | ||
6cb71bf6 | 212 | char syntax_code_spec[14] = |
8489eb67 | 213 | { |
6cb71bf6 | 214 | ' ', '.', 'w', '_', '(', ')', '\'', '\"', '$', '\\', '/', '<', '>', '@' |
8489eb67 RS |
215 | }; |
216 | \f | |
8ea151b2 RS |
217 | /* Look up the value for CHARACTER in syntax table TABLE's parent |
218 | and its parents. SYNTAX_ENTRY calls this, when TABLE itself has nil | |
219 | for CHARACTER. It's actually used only when not compiled with GCC. */ | |
220 | ||
221 | Lisp_Object | |
222 | syntax_parent_lookup (table, character) | |
223 | Lisp_Object table; | |
224 | int character; | |
225 | { | |
226 | Lisp_Object value; | |
227 | ||
228 | while (1) | |
229 | { | |
230 | table = XCHAR_TABLE (table)->parent; | |
231 | if (NILP (table)) | |
232 | return Qnil; | |
233 | ||
234 | value = XCHAR_TABLE (table)->contents[character]; | |
235 | if (!NILP (value)) | |
236 | return value; | |
237 | } | |
238 | } | |
239 | ||
8489eb67 | 240 | DEFUN ("char-syntax", Fchar_syntax, Schar_syntax, 1, 1, 0, |
2203e1e8 EN |
241 | "Return the syntax code of CHARACTER, described by a character.\n\ |
242 | For example, if CHARACTER is a word constituent,\n\ | |
243 | the character `w' is returned.\n\ | |
8489eb67 RS |
244 | The characters that correspond to various syntax codes\n\ |
245 | are listed in the documentation of `modify-syntax-entry'.") | |
2203e1e8 EN |
246 | (character) |
247 | Lisp_Object character; | |
8489eb67 | 248 | { |
8ea151b2 | 249 | int char_int; |
2203e1e8 EN |
250 | CHECK_NUMBER (character, 0); |
251 | char_int = XINT (character); | |
8ea151b2 | 252 | return make_number (syntax_code_spec[(int) SYNTAX (char_int)]); |
beefa22e RS |
253 | } |
254 | ||
255 | DEFUN ("matching-paren", Fmatching_paren, Smatching_paren, 1, 1, 0, | |
2203e1e8 EN |
256 | "Return the matching parenthesis of CHARACTER, or nil if none.") |
257 | (character) | |
258 | Lisp_Object character; | |
beefa22e | 259 | { |
8ea151b2 | 260 | int char_int, code; |
2203e1e8 EN |
261 | CHECK_NUMBER (character, 0); |
262 | char_int = XINT (character); | |
8ea151b2 | 263 | code = SYNTAX (char_int); |
a8bd7cd8 | 264 | if (code == Sopen || code == Sclose) |
8ea151b2 | 265 | return make_number (SYNTAX_MATCH (char_int)); |
beefa22e | 266 | return Qnil; |
8489eb67 RS |
267 | } |
268 | ||
269 | /* This comment supplies the doc string for modify-syntax-entry, | |
270 | for make-docfile to see. We cannot put this in the real DEFUN | |
271 | due to limits in the Unix cpp. | |
272 | ||
2fcea188 | 273 | DEFUN ("modify-syntax-entry", foo, bar, 2, 3, 0, |
8489eb67 RS |
274 | "Set syntax for character CHAR according to string S.\n\ |
275 | The syntax is changed only for table TABLE, which defaults to\n\ | |
276 | the current buffer's syntax table.\n\ | |
277 | The first character of S should be one of the following:\n\ | |
32676c08 JB |
278 | Space or - whitespace syntax. w word constituent.\n\ |
279 | _ symbol constituent. . punctuation.\n\ | |
280 | ( open-parenthesis. ) close-parenthesis.\n\ | |
281 | \" string quote. \\ escape.\n\ | |
282 | $ paired delimiter. ' expression quote or prefix operator.\n\ | |
abae13bd | 283 | < comment starter. > comment ender.\n\ |
6cb71bf6 | 284 | / character-quote. @ inherit from `standard-syntax-table'.\n\ |
abae13bd | 285 | \n\ |
8489eb67 RS |
286 | Only single-character comment start and end sequences are represented thus.\n\ |
287 | Two-character sequences are represented as described below.\n\ | |
288 | The second character of S is the matching parenthesis,\n\ | |
289 | used only if the first character is `(' or `)'.\n\ | |
290 | Any additional characters are flags.\n\ | |
e5d4f4dc | 291 | Defined flags are the characters 1, 2, 3, 4, b, and p.\n\ |
2203e1e8 EN |
292 | 1 means CHAR is the start of a two-char comment start sequence.\n\ |
293 | 2 means CHAR is the second character of such a sequence.\n\ | |
294 | 3 means CHAR is the start of a two-char comment end sequence.\n\ | |
295 | 4 means CHAR is the second character of such a sequence.\n\ | |
e5d4f4dc RS |
296 | \n\ |
297 | There can be up to two orthogonal comment sequences. This is to support\n\ | |
298 | language modes such as C++. By default, all comment sequences are of style\n\ | |
abae13bd RS |
299 | a, but you can set the comment sequence style to b (on the second character\n\ |
300 | of a comment-start, or the first character of a comment-end sequence) using\n\ | |
e5d4f4dc | 301 | this flag:\n\ |
2203e1e8 | 302 | b means CHAR is part of comment sequence b.\n\ |
e5d4f4dc | 303 | \n\ |
2203e1e8 | 304 | p means CHAR is a prefix character for `backward-prefix-chars';\n\ |
e5d4f4dc | 305 | such characters are treated as whitespace when they occur\n\ |
8489eb67 | 306 | between expressions.") |
2fcea188 | 307 | (char, s, table) |
8489eb67 RS |
308 | */ |
309 | ||
310 | DEFUN ("modify-syntax-entry", Fmodify_syntax_entry, Smodify_syntax_entry, 2, 3, | |
311 | /* I really don't know why this is interactive | |
312 | help-form should at least be made useful whilst reading the second arg | |
313 | */ | |
314 | "cSet syntax for character: \nsSet syntax for %s to: ", | |
315 | 0 /* See immediately above */) | |
316 | (c, newentry, syntax_table) | |
317 | Lisp_Object c, newentry, syntax_table; | |
318 | { | |
8ea151b2 | 319 | register unsigned char *p; |
8489eb67 | 320 | register enum syntaxcode code; |
247e20a8 | 321 | int val; |
8ea151b2 | 322 | Lisp_Object match; |
8489eb67 RS |
323 | |
324 | CHECK_NUMBER (c, 0); | |
325 | CHECK_STRING (newentry, 1); | |
8ea151b2 | 326 | |
265a9e55 | 327 | if (NILP (syntax_table)) |
8489eb67 RS |
328 | syntax_table = current_buffer->syntax_table; |
329 | else | |
8ea151b2 | 330 | check_syntax_table (syntax_table); |
8489eb67 RS |
331 | |
332 | p = XSTRING (newentry)->data; | |
333 | code = (enum syntaxcode) syntax_spec_code[*p++]; | |
334 | if (((int) code & 0377) == 0377) | |
335 | error ("invalid syntax description letter: %c", c); | |
336 | ||
8ea151b2 RS |
337 | if (code == Sinherit) |
338 | { | |
339 | SET_RAW_SYNTAX_ENTRY (syntax_table, c, Qnil); | |
340 | return Qnil; | |
341 | } | |
342 | ||
343 | if (*p) | |
d1be9f0f RS |
344 | { |
345 | XSETINT (match, *p++); | |
346 | if (XFASTINT (match) == ' ') | |
347 | match = Qnil; | |
348 | } | |
349 | else | |
8ea151b2 | 350 | match = Qnil; |
8489eb67 | 351 | |
8ea151b2 | 352 | val = (int) code; |
8489eb67 RS |
353 | while (*p) |
354 | switch (*p++) | |
355 | { | |
356 | case '1': | |
247e20a8 | 357 | val |= 1 << 16; |
8489eb67 RS |
358 | break; |
359 | ||
360 | case '2': | |
247e20a8 | 361 | val |= 1 << 17; |
8489eb67 RS |
362 | break; |
363 | ||
364 | case '3': | |
247e20a8 | 365 | val |= 1 << 18; |
8489eb67 RS |
366 | break; |
367 | ||
368 | case '4': | |
247e20a8 | 369 | val |= 1 << 19; |
8489eb67 RS |
370 | break; |
371 | ||
372 | case 'p': | |
247e20a8 | 373 | val |= 1 << 20; |
8489eb67 | 374 | break; |
e5d4f4dc RS |
375 | |
376 | case 'b': | |
247e20a8 | 377 | val |= 1 << 21; |
e5d4f4dc | 378 | break; |
8489eb67 RS |
379 | } |
380 | ||
8ea151b2 RS |
381 | SET_RAW_SYNTAX_ENTRY (syntax_table, c, |
382 | Fcons (make_number (val), match)); | |
8489eb67 RS |
383 | |
384 | return Qnil; | |
385 | } | |
386 | \f | |
387 | /* Dump syntax table to buffer in human-readable format */ | |
388 | ||
340f92b5 | 389 | static void |
8489eb67 RS |
390 | describe_syntax (value) |
391 | Lisp_Object value; | |
392 | { | |
393 | register enum syntaxcode code; | |
e5d4f4dc | 394 | char desc, match, start1, start2, end1, end2, prefix, comstyle; |
8489eb67 | 395 | char str[2]; |
8ea151b2 | 396 | Lisp_Object first, match_lisp; |
8489eb67 RS |
397 | |
398 | Findent_to (make_number (16), make_number (1)); | |
399 | ||
8ea151b2 RS |
400 | if (NILP (value)) |
401 | { | |
402 | insert_string ("inherit"); | |
403 | return; | |
404 | } | |
405 | ||
406 | if (!CONSP (value)) | |
407 | { | |
408 | insert_string ("invalid"); | |
409 | return; | |
410 | } | |
411 | ||
412 | first = XCONS (value)->car; | |
413 | match_lisp = XCONS (value)->cdr; | |
414 | ||
415 | if (!INTEGERP (first) || !(NILP (match_lisp) || INTEGERP (match_lisp))) | |
8489eb67 RS |
416 | { |
417 | insert_string ("invalid"); | |
418 | return; | |
419 | } | |
420 | ||
8ea151b2 RS |
421 | code = (enum syntaxcode) (first & 0377); |
422 | start1 = (XINT (first) >> 16) & 1; | |
423 | start2 = (XINT (first) >> 17) & 1; | |
424 | end1 = (XINT (first) >> 18) & 1; | |
425 | end2 = (XINT (first) >> 19) & 1; | |
426 | prefix = (XINT (first) >> 20) & 1; | |
427 | comstyle = (XINT (first) >> 21) & 1; | |
8489eb67 RS |
428 | |
429 | if ((int) code < 0 || (int) code >= (int) Smax) | |
430 | { | |
431 | insert_string ("invalid"); | |
432 | return; | |
433 | } | |
434 | desc = syntax_code_spec[(int) code]; | |
435 | ||
436 | str[0] = desc, str[1] = 0; | |
437 | insert (str, 1); | |
438 | ||
8ea151b2 | 439 | str[0] = !NILP (match_lisp) ? XINT (match_lisp) : ' '; |
8489eb67 RS |
440 | insert (str, 1); |
441 | ||
8489eb67 RS |
442 | if (start1) |
443 | insert ("1", 1); | |
444 | if (start2) | |
445 | insert ("2", 1); | |
446 | ||
447 | if (end1) | |
448 | insert ("3", 1); | |
449 | if (end2) | |
450 | insert ("4", 1); | |
451 | ||
452 | if (prefix) | |
453 | insert ("p", 1); | |
e5d4f4dc RS |
454 | if (comstyle) |
455 | insert ("b", 1); | |
8489eb67 RS |
456 | |
457 | insert_string ("\twhich means: "); | |
458 | ||
0220c518 | 459 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
460 | { |
461 | case Swhitespace: | |
462 | insert_string ("whitespace"); break; | |
463 | case Spunct: | |
464 | insert_string ("punctuation"); break; | |
465 | case Sword: | |
466 | insert_string ("word"); break; | |
467 | case Ssymbol: | |
468 | insert_string ("symbol"); break; | |
469 | case Sopen: | |
470 | insert_string ("open"); break; | |
471 | case Sclose: | |
472 | insert_string ("close"); break; | |
473 | case Squote: | |
474 | insert_string ("quote"); break; | |
475 | case Sstring: | |
476 | insert_string ("string"); break; | |
477 | case Smath: | |
478 | insert_string ("math"); break; | |
479 | case Sescape: | |
480 | insert_string ("escape"); break; | |
481 | case Scharquote: | |
482 | insert_string ("charquote"); break; | |
483 | case Scomment: | |
484 | insert_string ("comment"); break; | |
485 | case Sendcomment: | |
486 | insert_string ("endcomment"); break; | |
487 | default: | |
488 | insert_string ("invalid"); | |
489 | return; | |
490 | } | |
491 | ||
8ea151b2 | 492 | if (!NILP (match_lisp)) |
8489eb67 RS |
493 | { |
494 | insert_string (", matches "); | |
8ea151b2 | 495 | insert_char (XINT (match_lisp)); |
8489eb67 RS |
496 | } |
497 | ||
498 | if (start1) | |
499 | insert_string (",\n\t is the first character of a comment-start sequence"); | |
500 | if (start2) | |
501 | insert_string (",\n\t is the second character of a comment-start sequence"); | |
502 | ||
503 | if (end1) | |
504 | insert_string (",\n\t is the first character of a comment-end sequence"); | |
505 | if (end2) | |
506 | insert_string (",\n\t is the second character of a comment-end sequence"); | |
e5d4f4dc RS |
507 | if (comstyle) |
508 | insert_string (" (comment style b)"); | |
509 | ||
8489eb67 RS |
510 | if (prefix) |
511 | insert_string (",\n\t is a prefix character for `backward-prefix-chars'"); | |
512 | ||
513 | insert_string ("\n"); | |
514 | } | |
515 | ||
340f92b5 | 516 | static Lisp_Object |
8489eb67 RS |
517 | describe_syntax_1 (vector) |
518 | Lisp_Object vector; | |
519 | { | |
520 | struct buffer *old = current_buffer; | |
521 | set_buffer_internal (XBUFFER (Vstandard_output)); | |
5588734e | 522 | describe_vector (vector, Qnil, describe_syntax, 0, Qnil, Qnil); |
36cd82fe | 523 | call0 (intern ("help-mode")); |
8489eb67 RS |
524 | set_buffer_internal (old); |
525 | return Qnil; | |
526 | } | |
527 | ||
528 | DEFUN ("describe-syntax", Fdescribe_syntax, Sdescribe_syntax, 0, 0, "", | |
529 | "Describe the syntax specifications in the syntax table.\n\ | |
530 | The descriptions are inserted in a buffer, which is then displayed.") | |
531 | () | |
532 | { | |
533 | internal_with_output_to_temp_buffer | |
534 | ("*Help*", describe_syntax_1, current_buffer->syntax_table); | |
535 | ||
536 | return Qnil; | |
537 | } | |
538 | \f | |
539 | /* Return the position across COUNT words from FROM. | |
540 | If that many words cannot be found before the end of the buffer, return 0. | |
541 | COUNT negative means scan backward and stop at word beginning. */ | |
542 | ||
543 | scan_words (from, count) | |
544 | register int from, count; | |
545 | { | |
546 | register int beg = BEGV; | |
547 | register int end = ZV; | |
548 | register int code; | |
8ea151b2 | 549 | int charcode; |
8489eb67 RS |
550 | |
551 | immediate_quit = 1; | |
552 | QUIT; | |
553 | ||
554 | while (count > 0) | |
555 | { | |
556 | while (1) | |
557 | { | |
558 | if (from == end) | |
559 | { | |
560 | immediate_quit = 0; | |
561 | return 0; | |
562 | } | |
8ea151b2 RS |
563 | charcode = FETCH_CHAR (from); |
564 | code = SYNTAX (charcode); | |
8489eb67 RS |
565 | if (words_include_escapes |
566 | && (code == Sescape || code == Scharquote)) | |
567 | break; | |
568 | if (code == Sword) | |
569 | break; | |
570 | from++; | |
571 | } | |
572 | while (1) | |
573 | { | |
574 | if (from == end) break; | |
8ea151b2 RS |
575 | charcode = FETCH_CHAR (from); |
576 | code = SYNTAX (charcode); | |
8489eb67 RS |
577 | if (!(words_include_escapes |
578 | && (code == Sescape || code == Scharquote))) | |
579 | if (code != Sword) | |
580 | break; | |
581 | from++; | |
582 | } | |
583 | count--; | |
584 | } | |
585 | while (count < 0) | |
586 | { | |
587 | while (1) | |
588 | { | |
589 | if (from == beg) | |
590 | { | |
591 | immediate_quit = 0; | |
592 | return 0; | |
593 | } | |
8ea151b2 RS |
594 | charcode = FETCH_CHAR (from - 1); |
595 | code = SYNTAX (charcode); | |
8489eb67 RS |
596 | if (words_include_escapes |
597 | && (code == Sescape || code == Scharquote)) | |
598 | break; | |
599 | if (code == Sword) | |
600 | break; | |
601 | from--; | |
602 | } | |
603 | while (1) | |
604 | { | |
605 | if (from == beg) break; | |
8ea151b2 RS |
606 | charcode = FETCH_CHAR (from - 1); |
607 | code = SYNTAX (charcode); | |
8489eb67 RS |
608 | if (!(words_include_escapes |
609 | && (code == Sescape || code == Scharquote))) | |
610 | if (code != Sword) | |
611 | break; | |
612 | from--; | |
613 | } | |
614 | count++; | |
615 | } | |
616 | ||
617 | immediate_quit = 0; | |
618 | ||
619 | return from; | |
620 | } | |
621 | ||
622 | DEFUN ("forward-word", Fforward_word, Sforward_word, 1, 1, "p", | |
623 | "Move point forward ARG words (backward if ARG is negative).\n\ | |
624 | Normally returns t.\n\ | |
625 | If an edge of the buffer is reached, point is left there\n\ | |
626 | and nil is returned.") | |
627 | (count) | |
628 | Lisp_Object count; | |
629 | { | |
630 | int val; | |
631 | CHECK_NUMBER (count, 0); | |
632 | ||
633 | if (!(val = scan_words (point, XINT (count)))) | |
634 | { | |
635 | SET_PT (XINT (count) > 0 ? ZV : BEGV); | |
636 | return Qnil; | |
637 | } | |
638 | SET_PT (val); | |
639 | return Qt; | |
640 | } | |
641 | \f | |
b3cfe0c8 RS |
642 | DEFUN ("forward-comment", Fforward_comment, Sforward_comment, 1, 1, 0, |
643 | "Move forward across up to N comments. If N is negative, move backward.\n\ | |
b3cfe0c8 | 644 | Stop scanning if we find something other than a comment or whitespace.\n\ |
c81a3712 | 645 | Set point to where scanning stops.\n\ |
b3cfe0c8 RS |
646 | If N comments are found as expected, with nothing except whitespace\n\ |
647 | between them, return t; otherwise return nil.") | |
648 | (count) | |
840f481c | 649 | Lisp_Object count; |
b3cfe0c8 RS |
650 | { |
651 | register int from; | |
652 | register int stop; | |
8ea151b2 | 653 | register int c, c1; |
b3cfe0c8 RS |
654 | register enum syntaxcode code; |
655 | int comstyle = 0; /* style of comment encountered */ | |
be720845 | 656 | int found; |
840f481c RS |
657 | int count1; |
658 | ||
659 | CHECK_NUMBER (count, 0); | |
660 | count1 = XINT (count); | |
b3cfe0c8 RS |
661 | |
662 | immediate_quit = 1; | |
663 | QUIT; | |
664 | ||
665 | from = PT; | |
666 | ||
840f481c | 667 | while (count1 > 0) |
b3cfe0c8 RS |
668 | { |
669 | stop = ZV; | |
04882296 | 670 | do |
b3cfe0c8 | 671 | { |
04882296 KH |
672 | if (from == stop) |
673 | { | |
674 | SET_PT (from); | |
b7e6e612 | 675 | immediate_quit = 0; |
04882296 KH |
676 | return Qnil; |
677 | } | |
b3cfe0c8 RS |
678 | c = FETCH_CHAR (from); |
679 | code = SYNTAX (c); | |
680 | from++; | |
681 | comstyle = 0; | |
682 | if (from < stop && SYNTAX_COMSTART_FIRST (c) | |
8ea151b2 RS |
683 | && (c1 = FETCH_CHAR (from), |
684 | SYNTAX_COMSTART_SECOND (c1))) | |
b3cfe0c8 | 685 | { |
7fc8191e RS |
686 | /* We have encountered a comment start sequence and we |
687 | are ignoring all text inside comments. We must record | |
b3cfe0c8 RS |
688 | the comment style this sequence begins so that later, |
689 | only a comment end of the same style actually ends | |
7fc8191e | 690 | the comment section. */ |
b3cfe0c8 | 691 | code = Scomment; |
8ea151b2 | 692 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
b3cfe0c8 RS |
693 | from++; |
694 | } | |
04882296 KH |
695 | } |
696 | while (code == Swhitespace || code == Sendcomment); | |
697 | if (code != Scomment) | |
698 | { | |
699 | immediate_quit = 0; | |
700 | SET_PT (from - 1); | |
701 | return Qnil; | |
702 | } | |
703 | /* We're at the start of a comment. */ | |
704 | while (1) | |
705 | { | |
706 | if (from == stop) | |
b3cfe0c8 RS |
707 | { |
708 | immediate_quit = 0; | |
04882296 | 709 | SET_PT (from); |
b3cfe0c8 RS |
710 | return Qnil; |
711 | } | |
04882296 | 712 | c = FETCH_CHAR (from); |
2eb1d6a2 | 713 | from++; |
04882296 KH |
714 | if (SYNTAX (c) == Sendcomment |
715 | && SYNTAX_COMMENT_STYLE (c) == comstyle) | |
716 | /* we have encountered a comment end of the same style | |
717 | as the comment sequence which began this comment | |
718 | section */ | |
719 | break; | |
04882296 | 720 | if (from < stop && SYNTAX_COMEND_FIRST (c) |
8ea151b2 RS |
721 | && (c1 = FETCH_CHAR (from), |
722 | SYNTAX_COMEND_SECOND (c1)) | |
04882296 KH |
723 | && SYNTAX_COMMENT_STYLE (c) == comstyle) |
724 | /* we have encountered a comment end of the same style | |
725 | as the comment sequence which began this comment | |
726 | section */ | |
727 | { from++; break; } | |
b3cfe0c8 | 728 | } |
04882296 | 729 | /* We have skipped one comment. */ |
840f481c | 730 | count1--; |
b3cfe0c8 RS |
731 | } |
732 | ||
840f481c | 733 | while (count1 < 0) |
b3cfe0c8 RS |
734 | { |
735 | stop = BEGV; | |
736 | while (from > stop) | |
737 | { | |
738 | int quoted; | |
739 | ||
740 | from--; | |
741 | quoted = char_quoted (from); | |
742 | if (quoted) | |
743 | from--; | |
744 | c = FETCH_CHAR (from); | |
745 | code = SYNTAX (c); | |
746 | comstyle = 0; | |
7fc8191e RS |
747 | if (code == Sendcomment) |
748 | comstyle = SYNTAX_COMMENT_STYLE (c); | |
b3cfe0c8 | 749 | if (from > stop && SYNTAX_COMEND_SECOND (c) |
8ea151b2 RS |
750 | && (c1 = FETCH_CHAR (from - 1), |
751 | SYNTAX_COMEND_FIRST (c1)) | |
b3cfe0c8 RS |
752 | && !char_quoted (from - 1)) |
753 | { | |
7fc8191e | 754 | /* We must record the comment style encountered so that |
b3cfe0c8 | 755 | later, we can match only the proper comment begin |
7fc8191e | 756 | sequence of the same style. */ |
b3cfe0c8 | 757 | code = Sendcomment; |
8ea151b2 | 758 | comstyle = SYNTAX_COMMENT_STYLE (c1); |
b3cfe0c8 RS |
759 | from--; |
760 | } | |
761 | ||
762 | if (code == Sendcomment && !quoted) | |
763 | { | |
8f9dc2ed | 764 | #if 0 |
b3cfe0c8 RS |
765 | if (code != SYNTAX (c)) |
766 | /* For a two-char comment ender, we can assume | |
767 | it does end a comment. So scan back in a simple way. */ | |
768 | { | |
769 | if (from != stop) from--; | |
770 | while (1) | |
771 | { | |
8ea151b2 RS |
772 | if ((c = FETCH_CHAR (from), |
773 | SYNTAX (c) == Scomment) | |
b3cfe0c8 RS |
774 | && SYNTAX_COMMENT_STYLE (c) == comstyle) |
775 | break; | |
776 | if (from == stop) | |
777 | { | |
778 | immediate_quit = 0; | |
c81a3712 | 779 | SET_PT (from); |
b3cfe0c8 RS |
780 | return Qnil; |
781 | } | |
782 | from--; | |
783 | if (SYNTAX_COMSTART_SECOND (c) | |
8ea151b2 RS |
784 | && (c1 = FETCH_CHAR (from), |
785 | SYNTAX_COMSTART_FIRST (c1)) | |
b3cfe0c8 RS |
786 | && SYNTAX_COMMENT_STYLE (c) == comstyle |
787 | && !char_quoted (from)) | |
788 | break; | |
789 | } | |
790 | break; | |
791 | } | |
8f9dc2ed | 792 | #endif /* 0 */ |
b3cfe0c8 RS |
793 | |
794 | /* Look back, counting the parity of string-quotes, | |
795 | and recording the comment-starters seen. | |
796 | When we reach a safe place, assume that's not in a string; | |
797 | then step the main scan to the earliest comment-starter seen | |
798 | an even number of string quotes away from the safe place. | |
799 | ||
800 | OFROM[I] is position of the earliest comment-starter seen | |
801 | which is I+2X quotes from the comment-end. | |
802 | PARITY is current parity of quotes from the comment end. */ | |
803 | { | |
804 | int parity = 0; | |
805 | char my_stringend = 0; | |
806 | int string_lossage = 0; | |
807 | int comment_end = from; | |
808 | int comstart_pos = 0; | |
809 | int comstart_parity = 0; | |
4841063f | 810 | int scanstart = from - 1; |
b3cfe0c8 RS |
811 | |
812 | /* At beginning of range to scan, we're outside of strings; | |
813 | that determines quote parity to the comment-end. */ | |
814 | while (from != stop) | |
815 | { | |
816 | /* Move back and examine a character. */ | |
817 | from--; | |
818 | ||
819 | c = FETCH_CHAR (from); | |
820 | code = SYNTAX (c); | |
821 | ||
822 | /* If this char is the second of a 2-char comment sequence, | |
823 | back up and give the pair the appropriate syntax. */ | |
824 | if (from > stop && SYNTAX_COMEND_SECOND (c) | |
825 | && SYNTAX_COMEND_FIRST (FETCH_CHAR (from - 1))) | |
826 | { | |
827 | code = Sendcomment; | |
828 | from--; | |
4841063f | 829 | c = FETCH_CHAR (from); |
b3cfe0c8 RS |
830 | } |
831 | ||
4841063f RS |
832 | /* If this char starts a 2-char comment start sequence, |
833 | treat it like a 1-char comment starter. */ | |
834 | if (from < scanstart && SYNTAX_COMSTART_FIRST (c) | |
835 | && SYNTAX_COMSTART_SECOND (FETCH_CHAR (from + 1)) | |
836 | && comstyle == SYNTAX_COMMENT_STYLE (FETCH_CHAR (from + 1))) | |
837 | code = Scomment; | |
b3cfe0c8 RS |
838 | |
839 | /* Ignore escaped characters. */ | |
840 | if (char_quoted (from)) | |
841 | continue; | |
842 | ||
843 | /* Track parity of quotes. */ | |
844 | if (code == Sstring) | |
845 | { | |
846 | parity ^= 1; | |
847 | if (my_stringend == 0) | |
848 | my_stringend = c; | |
849 | /* If we have two kinds of string delimiters. | |
850 | There's no way to grok this scanning backwards. */ | |
851 | else if (my_stringend != c) | |
852 | string_lossage = 1; | |
853 | } | |
854 | ||
855 | /* Record comment-starters according to that | |
856 | quote-parity to the comment-end. */ | |
857 | if (code == Scomment) | |
858 | { | |
859 | comstart_parity = parity; | |
860 | comstart_pos = from; | |
861 | } | |
862 | ||
863 | /* If we find another earlier comment-ender, | |
eb8c3be9 | 864 | any comment-starts earlier than that don't count |
b3cfe0c8 RS |
865 | (because they go with the earlier comment-ender). */ |
866 | if (code == Sendcomment | |
867 | && SYNTAX_COMMENT_STYLE (FETCH_CHAR (from)) == comstyle) | |
868 | break; | |
869 | ||
870 | /* Assume a defun-start point is outside of strings. */ | |
871 | if (code == Sopen | |
872 | && (from == stop || FETCH_CHAR (from - 1) == '\n')) | |
873 | break; | |
874 | } | |
875 | ||
876 | if (comstart_pos == 0) | |
877 | from = comment_end; | |
878 | /* If the earliest comment starter | |
879 | is followed by uniform paired string quotes or none, | |
880 | we know it can't be inside a string | |
881 | since if it were then the comment ender would be inside one. | |
882 | So it does start a comment. Skip back to it. */ | |
883 | else if (comstart_parity == 0 && !string_lossage) | |
884 | from = comstart_pos; | |
885 | else | |
886 | { | |
887 | /* We had two kinds of string delimiters mixed up | |
888 | together. Decode this going forwards. | |
889 | Scan fwd from the previous comment ender | |
890 | to the one in question; this records where we | |
891 | last passed a comment starter. */ | |
892 | struct lisp_parse_state state; | |
893 | scan_sexps_forward (&state, find_defun_start (comment_end), | |
340f92b5 | 894 | comment_end - 1, -10000, 0, Qnil, 0); |
b3cfe0c8 RS |
895 | if (state.incomment) |
896 | from = state.comstart; | |
897 | else | |
898 | /* We can't grok this as a comment; scan it normally. */ | |
899 | from = comment_end; | |
900 | } | |
901 | } | |
7fc8191e RS |
902 | /* We have skipped one comment. */ |
903 | break; | |
b3cfe0c8 | 904 | } |
c81a3712 | 905 | else if ((code != Swhitespace && code != Scomment) || quoted) |
b3cfe0c8 RS |
906 | { |
907 | immediate_quit = 0; | |
c81a3712 | 908 | SET_PT (from + 1); |
b3cfe0c8 RS |
909 | return Qnil; |
910 | } | |
911 | } | |
912 | ||
840f481c | 913 | count1++; |
b3cfe0c8 RS |
914 | } |
915 | ||
916 | SET_PT (from); | |
917 | immediate_quit = 0; | |
918 | return Qt; | |
919 | } | |
920 | \f | |
8489eb67 RS |
921 | int parse_sexp_ignore_comments; |
922 | ||
923 | Lisp_Object | |
924 | scan_lists (from, count, depth, sexpflag) | |
925 | register int from; | |
926 | int count, depth, sexpflag; | |
927 | { | |
928 | Lisp_Object val; | |
929 | register int stop; | |
930 | register int c; | |
ee384542 | 931 | unsigned char stringterm; |
8489eb67 RS |
932 | int quoted; |
933 | int mathexit = 0; | |
934 | register enum syntaxcode code; | |
935 | int min_depth = depth; /* Err out if depth gets less than this. */ | |
e5d4f4dc | 936 | int comstyle = 0; /* style of comment encountered */ |
8489eb67 RS |
937 | |
938 | if (depth > 0) min_depth = 0; | |
939 | ||
940 | immediate_quit = 1; | |
941 | QUIT; | |
942 | ||
943 | while (count > 0) | |
944 | { | |
945 | stop = ZV; | |
946 | while (from < stop) | |
947 | { | |
948 | c = FETCH_CHAR (from); | |
e5d4f4dc | 949 | code = SYNTAX (c); |
8489eb67 RS |
950 | from++; |
951 | if (from < stop && SYNTAX_COMSTART_FIRST (c) | |
952 | && SYNTAX_COMSTART_SECOND (FETCH_CHAR (from)) | |
953 | && parse_sexp_ignore_comments) | |
e5d4f4dc RS |
954 | { |
955 | /* we have encountered a comment start sequence and we | |
956 | are ignoring all text inside comments. we must record | |
957 | the comment style this sequence begins so that later, | |
958 | only a comment end of the same style actually ends | |
959 | the comment section */ | |
960 | code = Scomment; | |
961 | comstyle = SYNTAX_COMMENT_STYLE (FETCH_CHAR (from)); | |
962 | from++; | |
963 | } | |
964 | ||
8489eb67 RS |
965 | if (SYNTAX_PREFIX (c)) |
966 | continue; | |
967 | ||
0220c518 | 968 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
969 | { |
970 | case Sescape: | |
971 | case Scharquote: | |
972 | if (from == stop) goto lose; | |
973 | from++; | |
974 | /* treat following character as a word constituent */ | |
975 | case Sword: | |
976 | case Ssymbol: | |
977 | if (depth || !sexpflag) break; | |
978 | /* This word counts as a sexp; return at end of it. */ | |
979 | while (from < stop) | |
980 | { | |
0220c518 | 981 | switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from)))) |
8489eb67 RS |
982 | { |
983 | case Scharquote: | |
984 | case Sescape: | |
985 | from++; | |
986 | if (from == stop) goto lose; | |
987 | break; | |
988 | case Sword: | |
989 | case Ssymbol: | |
990 | case Squote: | |
991 | break; | |
992 | default: | |
993 | goto done; | |
994 | } | |
995 | from++; | |
996 | } | |
997 | goto done; | |
998 | ||
999 | case Scomment: | |
1000 | if (!parse_sexp_ignore_comments) break; | |
1001 | while (1) | |
1002 | { | |
c01f7ec0 RS |
1003 | if (from == stop) |
1004 | { | |
1005 | if (depth == 0) | |
1006 | goto done; | |
1007 | goto lose; | |
1008 | } | |
e5d4f4dc RS |
1009 | c = FETCH_CHAR (from); |
1010 | if (SYNTAX (c) == Sendcomment | |
1011 | && SYNTAX_COMMENT_STYLE (c) == comstyle) | |
1012 | /* we have encountered a comment end of the same style | |
1013 | as the comment sequence which began this comment | |
1014 | section */ | |
8489eb67 RS |
1015 | break; |
1016 | from++; | |
1017 | if (from < stop && SYNTAX_COMEND_FIRST (c) | |
e5d4f4dc RS |
1018 | && SYNTAX_COMEND_SECOND (FETCH_CHAR (from)) |
1019 | && SYNTAX_COMMENT_STYLE (c) == comstyle) | |
1020 | /* we have encountered a comment end of the same style | |
1021 | as the comment sequence which began this comment | |
1022 | section */ | |
8489eb67 RS |
1023 | { from++; break; } |
1024 | } | |
1025 | break; | |
1026 | ||
1027 | case Smath: | |
1028 | if (!sexpflag) | |
1029 | break; | |
1030 | if (from != stop && c == FETCH_CHAR (from)) | |
1031 | from++; | |
1032 | if (mathexit) | |
1033 | { | |
1034 | mathexit = 0; | |
1035 | goto close1; | |
1036 | } | |
1037 | mathexit = 1; | |
1038 | ||
1039 | case Sopen: | |
1040 | if (!++depth) goto done; | |
1041 | break; | |
1042 | ||
1043 | case Sclose: | |
1044 | close1: | |
1045 | if (!--depth) goto done; | |
1046 | if (depth < min_depth) | |
1047 | error ("Containing expression ends prematurely"); | |
1048 | break; | |
1049 | ||
1050 | case Sstring: | |
1051 | stringterm = FETCH_CHAR (from - 1); | |
1052 | while (1) | |
1053 | { | |
1054 | if (from >= stop) goto lose; | |
1055 | if (FETCH_CHAR (from) == stringterm) break; | |
0220c518 | 1056 | switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from)))) |
8489eb67 RS |
1057 | { |
1058 | case Scharquote: | |
1059 | case Sescape: | |
1060 | from++; | |
1061 | } | |
1062 | from++; | |
1063 | } | |
1064 | from++; | |
1065 | if (!depth && sexpflag) goto done; | |
1066 | break; | |
1067 | } | |
1068 | } | |
1069 | ||
1070 | /* Reached end of buffer. Error if within object, return nil if between */ | |
1071 | if (depth) goto lose; | |
1072 | ||
1073 | immediate_quit = 0; | |
1074 | return Qnil; | |
1075 | ||
1076 | /* End of object reached */ | |
1077 | done: | |
1078 | count--; | |
1079 | } | |
1080 | ||
1081 | while (count < 0) | |
1082 | { | |
1083 | stop = BEGV; | |
1084 | while (from > stop) | |
1085 | { | |
1086 | from--; | |
1087 | if (quoted = char_quoted (from)) | |
1088 | from--; | |
1089 | c = FETCH_CHAR (from); | |
1090 | code = SYNTAX (c); | |
7fc8191e RS |
1091 | comstyle = 0; |
1092 | if (code == Sendcomment) | |
1093 | comstyle = SYNTAX_COMMENT_STYLE (c); | |
8489eb67 RS |
1094 | if (from > stop && SYNTAX_COMEND_SECOND (c) |
1095 | && SYNTAX_COMEND_FIRST (FETCH_CHAR (from - 1)) | |
1096 | && !char_quoted (from - 1) | |
1097 | && parse_sexp_ignore_comments) | |
e5d4f4dc RS |
1098 | { |
1099 | /* we must record the comment style encountered so that | |
1100 | later, we can match only the proper comment begin | |
1101 | sequence of the same style */ | |
1102 | code = Sendcomment; | |
1103 | comstyle = SYNTAX_COMMENT_STYLE (FETCH_CHAR (from - 1)); | |
1104 | from--; | |
1105 | } | |
1106 | ||
8489eb67 RS |
1107 | if (SYNTAX_PREFIX (c)) |
1108 | continue; | |
1109 | ||
0220c518 | 1110 | switch (SWITCH_ENUM_CAST (quoted ? Sword : code)) |
8489eb67 RS |
1111 | { |
1112 | case Sword: | |
1113 | case Ssymbol: | |
1114 | if (depth || !sexpflag) break; | |
1115 | /* This word counts as a sexp; count object finished after passing it. */ | |
1116 | while (from > stop) | |
1117 | { | |
1118 | quoted = char_quoted (from - 1); | |
1119 | if (quoted) | |
1120 | from--; | |
e5d4f4dc RS |
1121 | if (! (quoted || SYNTAX (FETCH_CHAR (from - 1)) == Sword |
1122 | || SYNTAX (FETCH_CHAR (from - 1)) == Ssymbol | |
1123 | || SYNTAX (FETCH_CHAR (from - 1)) == Squote)) | |
8489eb67 RS |
1124 | goto done2; |
1125 | from--; | |
1126 | } | |
1127 | goto done2; | |
1128 | ||
1129 | case Smath: | |
1130 | if (!sexpflag) | |
1131 | break; | |
1132 | if (from != stop && c == FETCH_CHAR (from - 1)) | |
1133 | from--; | |
1134 | if (mathexit) | |
1135 | { | |
1136 | mathexit = 0; | |
1137 | goto open2; | |
1138 | } | |
1139 | mathexit = 1; | |
1140 | ||
1141 | case Sclose: | |
1142 | if (!++depth) goto done2; | |
1143 | break; | |
1144 | ||
1145 | case Sopen: | |
1146 | open2: | |
1147 | if (!--depth) goto done2; | |
1148 | if (depth < min_depth) | |
1149 | error ("Containing expression ends prematurely"); | |
1150 | break; | |
1151 | ||
1152 | case Sendcomment: | |
1153 | if (!parse_sexp_ignore_comments) | |
1154 | break; | |
8f9dc2ed | 1155 | #if 0 |
37bef230 RS |
1156 | if (code != SYNTAX (c)) |
1157 | /* For a two-char comment ender, we can assume | |
1158 | it does end a comment. So scan back in a simple way. */ | |
1159 | { | |
1160 | if (from != stop) from--; | |
1161 | while (1) | |
1162 | { | |
1163 | if (SYNTAX (c = FETCH_CHAR (from)) == Scomment | |
1164 | && SYNTAX_COMMENT_STYLE (c) == comstyle) | |
1165 | break; | |
c01f7ec0 RS |
1166 | if (from == stop) |
1167 | { | |
1168 | if (depth == 0) | |
1169 | goto done2; | |
1170 | goto lose; | |
1171 | } | |
37bef230 RS |
1172 | from--; |
1173 | if (SYNTAX_COMSTART_SECOND (c) | |
1174 | && SYNTAX_COMSTART_FIRST (FETCH_CHAR (from)) | |
1175 | && SYNTAX_COMMENT_STYLE (c) == comstyle | |
1176 | && !char_quoted (from)) | |
1177 | break; | |
1178 | } | |
1179 | break; | |
1180 | } | |
8f9dc2ed | 1181 | #endif /* 0 */ |
37bef230 | 1182 | |
8489eb67 RS |
1183 | /* Look back, counting the parity of string-quotes, |
1184 | and recording the comment-starters seen. | |
1185 | When we reach a safe place, assume that's not in a string; | |
1186 | then step the main scan to the earliest comment-starter seen | |
1187 | an even number of string quotes away from the safe place. | |
1188 | ||
1189 | OFROM[I] is position of the earliest comment-starter seen | |
1190 | which is I+2X quotes from the comment-end. | |
1191 | PARITY is current parity of quotes from the comment end. */ | |
1192 | { | |
8489eb67 | 1193 | int parity = 0; |
e5d4f4dc RS |
1194 | char my_stringend = 0; |
1195 | int string_lossage = 0; | |
1196 | int comment_end = from; | |
37bef230 RS |
1197 | int comstart_pos = 0; |
1198 | int comstart_parity = 0; | |
4841063f | 1199 | int scanstart = from - 1; |
8489eb67 RS |
1200 | |
1201 | /* At beginning of range to scan, we're outside of strings; | |
1202 | that determines quote parity to the comment-end. */ | |
1203 | while (from != stop) | |
1204 | { | |
1205 | /* Move back and examine a character. */ | |
1206 | from--; | |
1207 | ||
1208 | c = FETCH_CHAR (from); | |
1209 | code = SYNTAX (c); | |
1210 | ||
1211 | /* If this char is the second of a 2-char comment sequence, | |
1212 | back up and give the pair the appropriate syntax. */ | |
1213 | if (from > stop && SYNTAX_COMEND_SECOND (c) | |
1214 | && SYNTAX_COMEND_FIRST (FETCH_CHAR (from - 1))) | |
e5d4f4dc RS |
1215 | { |
1216 | code = Sendcomment; | |
1217 | from--; | |
4841063f | 1218 | c = FETCH_CHAR (from); |
e5d4f4dc RS |
1219 | } |
1220 | ||
4841063f RS |
1221 | /* If this char starts a 2-char comment start sequence, |
1222 | treat it like a 1-char comment starter. */ | |
1223 | if (from < scanstart && SYNTAX_COMSTART_FIRST (c) | |
1224 | && SYNTAX_COMSTART_SECOND (FETCH_CHAR (from + 1)) | |
1225 | && comstyle == SYNTAX_COMMENT_STYLE (FETCH_CHAR (from + 1))) | |
1226 | code = Scomment; | |
8489eb67 RS |
1227 | |
1228 | /* Ignore escaped characters. */ | |
1229 | if (char_quoted (from)) | |
1230 | continue; | |
1231 | ||
37bef230 | 1232 | /* Track parity of quotes. */ |
8489eb67 | 1233 | if (code == Sstring) |
e5d4f4dc RS |
1234 | { |
1235 | parity ^= 1; | |
1236 | if (my_stringend == 0) | |
1237 | my_stringend = c; | |
37bef230 | 1238 | /* If we have two kinds of string delimiters. |
e5d4f4dc RS |
1239 | There's no way to grok this scanning backwards. */ |
1240 | else if (my_stringend != c) | |
1241 | string_lossage = 1; | |
1242 | } | |
8489eb67 RS |
1243 | |
1244 | /* Record comment-starters according to that | |
1245 | quote-parity to the comment-end. */ | |
1246 | if (code == Scomment) | |
37bef230 RS |
1247 | { |
1248 | comstart_parity = parity; | |
1249 | comstart_pos = from; | |
1250 | } | |
8489eb67 | 1251 | |
37bef230 | 1252 | /* If we find another earlier comment-ender, |
eb8c3be9 | 1253 | any comment-starts earlier than that don't count |
37bef230 | 1254 | (because they go with the earlier comment-ender). */ |
e5d4f4dc RS |
1255 | if (code == Sendcomment |
1256 | && SYNTAX_COMMENT_STYLE (FETCH_CHAR (from)) == comstyle) | |
8489eb67 | 1257 | break; |
37bef230 RS |
1258 | |
1259 | /* Assume a defun-start point is outside of strings. */ | |
1260 | if (code == Sopen | |
1261 | && (from == stop || FETCH_CHAR (from - 1) == '\n')) | |
1262 | break; | |
8489eb67 | 1263 | } |
37bef230 RS |
1264 | |
1265 | if (comstart_pos == 0) | |
1266 | from = comment_end; | |
1267 | /* If the earliest comment starter | |
1268 | is followed by uniform paired string quotes or none, | |
1269 | we know it can't be inside a string | |
1270 | since if it were then the comment ender would be inside one. | |
1271 | So it does start a comment. Skip back to it. */ | |
1272 | else if (comstart_parity == 0 && !string_lossage) | |
1273 | from = comstart_pos; | |
1274 | else | |
e5d4f4dc RS |
1275 | { |
1276 | /* We had two kinds of string delimiters mixed up | |
1277 | together. Decode this going forwards. | |
1278 | Scan fwd from the previous comment ender | |
1279 | to the one in question; this records where we | |
1280 | last passed a comment starter. */ | |
1281 | struct lisp_parse_state state; | |
37bef230 | 1282 | scan_sexps_forward (&state, find_defun_start (comment_end), |
340f92b5 | 1283 | comment_end - 1, -10000, 0, Qnil, 0); |
e5d4f4dc RS |
1284 | if (state.incomment) |
1285 | from = state.comstart; | |
1286 | else | |
1287 | /* We can't grok this as a comment; scan it normally. */ | |
1288 | from = comment_end; | |
1289 | } | |
8489eb67 RS |
1290 | } |
1291 | break; | |
1292 | ||
1293 | case Sstring: | |
1294 | stringterm = FETCH_CHAR (from); | |
1295 | while (1) | |
1296 | { | |
1297 | if (from == stop) goto lose; | |
1298 | if (!char_quoted (from - 1) | |
1299 | && stringterm == FETCH_CHAR (from - 1)) | |
1300 | break; | |
1301 | from--; | |
1302 | } | |
1303 | from--; | |
1304 | if (!depth && sexpflag) goto done2; | |
1305 | break; | |
1306 | } | |
1307 | } | |
1308 | ||
1309 | /* Reached start of buffer. Error if within object, return nil if between */ | |
1310 | if (depth) goto lose; | |
1311 | ||
1312 | immediate_quit = 0; | |
1313 | return Qnil; | |
1314 | ||
1315 | done2: | |
1316 | count++; | |
1317 | } | |
1318 | ||
1319 | ||
1320 | immediate_quit = 0; | |
1e142fb7 | 1321 | XSETFASTINT (val, from); |
8489eb67 RS |
1322 | return val; |
1323 | ||
1324 | lose: | |
1325 | error ("Unbalanced parentheses"); | |
1326 | /* NOTREACHED */ | |
1327 | } | |
1328 | ||
340f92b5 | 1329 | static int |
8489eb67 RS |
1330 | char_quoted (pos) |
1331 | register int pos; | |
1332 | { | |
1333 | register enum syntaxcode code; | |
1334 | register int beg = BEGV; | |
1335 | register int quoted = 0; | |
1336 | ||
1337 | while (pos > beg | |
1338 | && ((code = SYNTAX (FETCH_CHAR (pos - 1))) == Scharquote | |
1339 | || code == Sescape)) | |
1340 | pos--, quoted = !quoted; | |
1341 | return quoted; | |
1342 | } | |
1343 | ||
1344 | DEFUN ("scan-lists", Fscan_lists, Sscan_lists, 3, 3, 0, | |
1345 | "Scan from character number FROM by COUNT lists.\n\ | |
1346 | Returns the character number of the position thus found.\n\ | |
1347 | \n\ | |
1348 | If DEPTH is nonzero, paren depth begins counting from that value,\n\ | |
1349 | only places where the depth in parentheses becomes zero\n\ | |
1350 | are candidates for stopping; COUNT such places are counted.\n\ | |
1351 | Thus, a positive value for DEPTH means go out levels.\n\ | |
1352 | \n\ | |
1353 | Comments are ignored if `parse-sexp-ignore-comments' is non-nil.\n\ | |
1354 | \n\ | |
1355 | If the beginning or end of (the accessible part of) the buffer is reached\n\ | |
1356 | and the depth is wrong, an error is signaled.\n\ | |
1357 | If the depth is right but the count is not used up, nil is returned.") | |
1358 | (from, count, depth) | |
1359 | Lisp_Object from, count, depth; | |
1360 | { | |
1361 | CHECK_NUMBER (from, 0); | |
1362 | CHECK_NUMBER (count, 1); | |
1363 | CHECK_NUMBER (depth, 2); | |
1364 | ||
1365 | return scan_lists (XINT (from), XINT (count), XINT (depth), 0); | |
1366 | } | |
1367 | ||
1368 | DEFUN ("scan-sexps", Fscan_sexps, Sscan_sexps, 2, 2, 0, | |
1369 | "Scan from character number FROM by COUNT balanced expressions.\n\ | |
1370 | If COUNT is negative, scan backwards.\n\ | |
1371 | Returns the character number of the position thus found.\n\ | |
1372 | \n\ | |
1373 | Comments are ignored if `parse-sexp-ignore-comments' is non-nil.\n\ | |
1374 | \n\ | |
1375 | If the beginning or end of (the accessible part of) the buffer is reached\n\ | |
1376 | in the middle of a parenthetical grouping, an error is signaled.\n\ | |
1377 | If the beginning or end is reached between groupings\n\ | |
1378 | but before count is used up, nil is returned.") | |
1379 | (from, count) | |
1380 | Lisp_Object from, count; | |
1381 | { | |
1382 | CHECK_NUMBER (from, 0); | |
1383 | CHECK_NUMBER (count, 1); | |
1384 | ||
1385 | return scan_lists (XINT (from), XINT (count), 0, 1); | |
1386 | } | |
1387 | ||
1388 | DEFUN ("backward-prefix-chars", Fbackward_prefix_chars, Sbackward_prefix_chars, | |
1389 | 0, 0, 0, | |
1390 | "Move point backward over any number of chars with prefix syntax.\n\ | |
1391 | This includes chars with \"quote\" or \"prefix\" syntax (' or p).") | |
1392 | () | |
1393 | { | |
1394 | int beg = BEGV; | |
1395 | int pos = point; | |
1396 | ||
1397 | while (pos > beg && !char_quoted (pos - 1) | |
1398 | && (SYNTAX (FETCH_CHAR (pos - 1)) == Squote | |
1399 | || SYNTAX_PREFIX (FETCH_CHAR (pos - 1)))) | |
1400 | pos--; | |
1401 | ||
1402 | SET_PT (pos); | |
1403 | ||
1404 | return Qnil; | |
1405 | } | |
1406 | \f | |
8489eb67 | 1407 | /* Parse forward from FROM to END, |
e5d4f4dc RS |
1408 | assuming that FROM has state OLDSTATE (nil means FROM is start of function), |
1409 | and return a description of the state of the parse at END. | |
c81a3712 RS |
1410 | If STOPBEFORE is nonzero, stop at the start of an atom. |
1411 | If COMMENTSTOP is nonzero, stop at the start of a comment. */ | |
8489eb67 | 1412 | |
340f92b5 | 1413 | static void |
c81a3712 RS |
1414 | scan_sexps_forward (stateptr, from, end, targetdepth, |
1415 | stopbefore, oldstate, commentstop) | |
e5d4f4dc | 1416 | struct lisp_parse_state *stateptr; |
8489eb67 RS |
1417 | register int from; |
1418 | int end, targetdepth, stopbefore; | |
1419 | Lisp_Object oldstate; | |
c81a3712 | 1420 | int commentstop; |
8489eb67 RS |
1421 | { |
1422 | struct lisp_parse_state state; | |
1423 | ||
1424 | register enum syntaxcode code; | |
1425 | struct level { int last, prev; }; | |
1426 | struct level levelstart[100]; | |
1427 | register struct level *curlevel = levelstart; | |
1428 | struct level *endlevel = levelstart + 100; | |
1429 | char prev; | |
1430 | register int depth; /* Paren depth of current scanning location. | |
1431 | level - levelstart equals this except | |
1432 | when the depth becomes negative. */ | |
1433 | int mindepth; /* Lowest DEPTH value seen. */ | |
1434 | int start_quoted = 0; /* Nonzero means starting after a char quote */ | |
1435 | Lisp_Object tem; | |
1436 | ||
1437 | immediate_quit = 1; | |
1438 | QUIT; | |
1439 | ||
265a9e55 | 1440 | if (NILP (oldstate)) |
8489eb67 RS |
1441 | { |
1442 | depth = 0; | |
1443 | state.instring = -1; | |
1444 | state.incomment = 0; | |
e5d4f4dc | 1445 | state.comstyle = 0; /* comment style a by default */ |
8489eb67 RS |
1446 | } |
1447 | else | |
1448 | { | |
1449 | tem = Fcar (oldstate); | |
265a9e55 | 1450 | if (!NILP (tem)) |
8489eb67 RS |
1451 | depth = XINT (tem); |
1452 | else | |
1453 | depth = 0; | |
1454 | ||
1455 | oldstate = Fcdr (oldstate); | |
1456 | oldstate = Fcdr (oldstate); | |
1457 | oldstate = Fcdr (oldstate); | |
1458 | tem = Fcar (oldstate); | |
265a9e55 | 1459 | state.instring = !NILP (tem) ? XINT (tem) : -1; |
8489eb67 RS |
1460 | |
1461 | oldstate = Fcdr (oldstate); | |
1462 | tem = Fcar (oldstate); | |
265a9e55 | 1463 | state.incomment = !NILP (tem); |
8489eb67 RS |
1464 | |
1465 | oldstate = Fcdr (oldstate); | |
1466 | tem = Fcar (oldstate); | |
265a9e55 | 1467 | start_quoted = !NILP (tem); |
e5d4f4dc RS |
1468 | |
1469 | /* if the eight element of the list is nil, we are in comment | |
1470 | style a. if it is non-nil, we are in comment style b */ | |
1471 | oldstate = Fcdr (oldstate); | |
1472 | oldstate = Fcdr (oldstate); | |
e5d4f4dc RS |
1473 | tem = Fcar (oldstate); |
1474 | state.comstyle = !NILP (tem); | |
8489eb67 RS |
1475 | } |
1476 | state.quoted = 0; | |
1477 | mindepth = depth; | |
1478 | ||
1479 | curlevel->prev = -1; | |
1480 | curlevel->last = -1; | |
1481 | ||
1482 | /* Enter the loop at a place appropriate for initial state. */ | |
1483 | ||
1484 | if (state.incomment) goto startincomment; | |
1485 | if (state.instring >= 0) | |
1486 | { | |
1487 | if (start_quoted) goto startquotedinstring; | |
1488 | goto startinstring; | |
1489 | } | |
1490 | if (start_quoted) goto startquoted; | |
1491 | ||
1492 | while (from < end) | |
1493 | { | |
e5d4f4dc | 1494 | code = SYNTAX (FETCH_CHAR (from)); |
8489eb67 | 1495 | from++; |
8f9dc2ed RS |
1496 | if (code == Scomment) |
1497 | state.comstart = from-1; | |
1498 | ||
1499 | else if (from < end && SYNTAX_COMSTART_FIRST (FETCH_CHAR (from - 1)) | |
1500 | && SYNTAX_COMSTART_SECOND (FETCH_CHAR (from))) | |
e5d4f4dc RS |
1501 | { |
1502 | /* Record the comment style we have entered so that only | |
1503 | the comment-end sequence of the same style actually | |
1504 | terminates the comment section. */ | |
1505 | code = Scomment; | |
1506 | state.comstyle = SYNTAX_COMMENT_STYLE (FETCH_CHAR (from)); | |
8f9dc2ed | 1507 | state.comstart = from-1; |
e5d4f4dc RS |
1508 | from++; |
1509 | } | |
1510 | ||
8489eb67 RS |
1511 | if (SYNTAX_PREFIX (FETCH_CHAR (from - 1))) |
1512 | continue; | |
0220c518 | 1513 | switch (SWITCH_ENUM_CAST (code)) |
8489eb67 RS |
1514 | { |
1515 | case Sescape: | |
1516 | case Scharquote: | |
1517 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
1518 | curlevel->last = from - 1; | |
1519 | startquoted: | |
1520 | if (from == end) goto endquoted; | |
1521 | from++; | |
1522 | goto symstarted; | |
1523 | /* treat following character as a word constituent */ | |
1524 | case Sword: | |
1525 | case Ssymbol: | |
1526 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
1527 | curlevel->last = from - 1; | |
1528 | symstarted: | |
1529 | while (from < end) | |
1530 | { | |
0220c518 | 1531 | switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from)))) |
8489eb67 RS |
1532 | { |
1533 | case Scharquote: | |
1534 | case Sescape: | |
1535 | from++; | |
1536 | if (from == end) goto endquoted; | |
1537 | break; | |
1538 | case Sword: | |
1539 | case Ssymbol: | |
1540 | case Squote: | |
1541 | break; | |
1542 | default: | |
1543 | goto symdone; | |
1544 | } | |
1545 | from++; | |
1546 | } | |
1547 | symdone: | |
1548 | curlevel->prev = curlevel->last; | |
1549 | break; | |
1550 | ||
5a28e48c RS |
1551 | startincomment: |
1552 | if (commentstop) | |
1553 | goto done; | |
1554 | if (from != BEGV) | |
1555 | { | |
1556 | /* Enter the loop in the middle so that we find | |
1557 | a 2-char comment ender if we start in the middle of it. */ | |
1558 | prev = FETCH_CHAR (from - 1); | |
1559 | goto startincomment_1; | |
1560 | } | |
1561 | /* At beginning of buffer, enter the loop the ordinary way. */ | |
1562 | ||
8489eb67 RS |
1563 | case Scomment: |
1564 | state.incomment = 1; | |
c81a3712 RS |
1565 | if (commentstop) |
1566 | goto done; | |
8489eb67 RS |
1567 | while (1) |
1568 | { | |
1569 | if (from == end) goto done; | |
e5d4f4dc RS |
1570 | prev = FETCH_CHAR (from); |
1571 | if (SYNTAX (prev) == Sendcomment | |
1572 | && SYNTAX_COMMENT_STYLE (prev) == state.comstyle) | |
1573 | /* Only terminate the comment section if the endcomment | |
1574 | of the same style as the start sequence has been | |
1575 | encountered. */ | |
8489eb67 RS |
1576 | break; |
1577 | from++; | |
5a28e48c | 1578 | startincomment_1: |
8489eb67 | 1579 | if (from < end && SYNTAX_COMEND_FIRST (prev) |
e5d4f4dc RS |
1580 | && SYNTAX_COMEND_SECOND (FETCH_CHAR (from)) |
1581 | && SYNTAX_COMMENT_STYLE (prev) == state.comstyle) | |
1582 | /* Only terminate the comment section if the end-comment | |
1583 | sequence of the same style as the start sequence has | |
1584 | been encountered. */ | |
8489eb67 RS |
1585 | { from++; break; } |
1586 | } | |
1587 | state.incomment = 0; | |
e5d4f4dc | 1588 | state.comstyle = 0; /* reset the comment style */ |
8489eb67 RS |
1589 | break; |
1590 | ||
1591 | case Sopen: | |
1592 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
1593 | depth++; | |
1594 | /* curlevel++->last ran into compiler bug on Apollo */ | |
1595 | curlevel->last = from - 1; | |
1596 | if (++curlevel == endlevel) | |
1597 | error ("Nesting too deep for parser"); | |
1598 | curlevel->prev = -1; | |
1599 | curlevel->last = -1; | |
30844415 | 1600 | if (targetdepth == depth) goto done; |
8489eb67 RS |
1601 | break; |
1602 | ||
1603 | case Sclose: | |
1604 | depth--; | |
1605 | if (depth < mindepth) | |
1606 | mindepth = depth; | |
1607 | if (curlevel != levelstart) | |
1608 | curlevel--; | |
1609 | curlevel->prev = curlevel->last; | |
30844415 | 1610 | if (targetdepth == depth) goto done; |
8489eb67 RS |
1611 | break; |
1612 | ||
1613 | case Sstring: | |
1614 | if (stopbefore) goto stop; /* this arg means stop at sexp start */ | |
1615 | curlevel->last = from - 1; | |
1616 | state.instring = FETCH_CHAR (from - 1); | |
1617 | startinstring: | |
1618 | while (1) | |
1619 | { | |
1620 | if (from >= end) goto done; | |
1621 | if (FETCH_CHAR (from) == state.instring) break; | |
0220c518 | 1622 | switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from)))) |
8489eb67 RS |
1623 | { |
1624 | case Scharquote: | |
1625 | case Sescape: | |
1626 | from++; | |
1627 | startquotedinstring: | |
1628 | if (from >= end) goto endquoted; | |
1629 | } | |
1630 | from++; | |
1631 | } | |
1632 | state.instring = -1; | |
1633 | curlevel->prev = curlevel->last; | |
1634 | from++; | |
1635 | break; | |
1636 | ||
1637 | case Smath: | |
1638 | break; | |
1639 | } | |
1640 | } | |
1641 | goto done; | |
1642 | ||
1643 | stop: /* Here if stopping before start of sexp. */ | |
1644 | from--; /* We have just fetched the char that starts it; */ | |
1645 | goto done; /* but return the position before it. */ | |
1646 | ||
1647 | endquoted: | |
1648 | state.quoted = 1; | |
1649 | done: | |
1650 | state.depth = depth; | |
1651 | state.mindepth = mindepth; | |
1652 | state.thislevelstart = curlevel->prev; | |
1653 | state.prevlevelstart | |
1654 | = (curlevel == levelstart) ? -1 : (curlevel - 1)->last; | |
1655 | state.location = from; | |
1656 | immediate_quit = 0; | |
1657 | ||
e5d4f4dc | 1658 | *stateptr = state; |
8489eb67 RS |
1659 | } |
1660 | ||
1661 | /* This comment supplies the doc string for parse-partial-sexp, | |
1662 | for make-docfile to see. We cannot put this in the real DEFUN | |
1663 | due to limits in the Unix cpp. | |
1664 | ||
c81a3712 | 1665 | DEFUN ("parse-partial-sexp", Ffoo, Sfoo, 2, 6, 0, |
8489eb67 RS |
1666 | "Parse Lisp syntax starting at FROM until TO; return status of parse at TO.\n\ |
1667 | Parsing stops at TO or when certain criteria are met;\n\ | |
1668 | point is set to where parsing stops.\n\ | |
1669 | If fifth arg STATE is omitted or nil,\n\ | |
1670 | parsing assumes that FROM is the beginning of a function.\n\ | |
e5d4f4dc | 1671 | Value is a list of eight elements describing final state of parsing:\n\ |
af50f9e5 RS |
1672 | 0. depth in parens.\n\ |
1673 | 1. character address of start of innermost containing list; nil if none.\n\ | |
1674 | 2. character address of start of last complete sexp terminated.\n\ | |
1675 | 3. non-nil if inside a string.\n\ | |
8489eb67 | 1676 | (it is the character that will terminate the string.)\n\ |
af50f9e5 RS |
1677 | 4. t if inside a comment.\n\ |
1678 | 5. t if following a quote character.\n\ | |
1679 | 6. the minimum paren-depth encountered during this scan.\n\ | |
1680 | 7. t if in a comment of style `b'.\n\ | |
8489eb67 RS |
1681 | If third arg TARGETDEPTH is non-nil, parsing stops if the depth\n\ |
1682 | in parentheses becomes equal to TARGETDEPTH.\n\ | |
1683 | Fourth arg STOPBEFORE non-nil means stop when come to\n\ | |
1684 | any character that starts a sexp.\n\ | |
ec11639d | 1685 | Fifth arg STATE is an eight-list like what this function returns.\n\ |
a4275ad1 | 1686 | It is used to initialize the state of the parse. Its second and third |
c81a3712 RS |
1687 | elements are ignored. |
1688 | Sixth args COMMENTSTOP non-nil means stop at the start of a comment.") | |
1689 | (from, to, targetdepth, stopbefore, state, commentstop) | |
8489eb67 RS |
1690 | */ |
1691 | ||
c81a3712 | 1692 | DEFUN ("parse-partial-sexp", Fparse_partial_sexp, Sparse_partial_sexp, 2, 6, 0, |
8489eb67 | 1693 | 0 /* See immediately above */) |
c81a3712 RS |
1694 | (from, to, targetdepth, stopbefore, oldstate, commentstop) |
1695 | Lisp_Object from, to, targetdepth, stopbefore, oldstate, commentstop; | |
8489eb67 RS |
1696 | { |
1697 | struct lisp_parse_state state; | |
1698 | int target; | |
1699 | ||
265a9e55 | 1700 | if (!NILP (targetdepth)) |
8489eb67 RS |
1701 | { |
1702 | CHECK_NUMBER (targetdepth, 3); | |
1703 | target = XINT (targetdepth); | |
1704 | } | |
1705 | else | |
1706 | target = -100000; /* We won't reach this depth */ | |
1707 | ||
1708 | validate_region (&from, &to); | |
e5d4f4dc | 1709 | scan_sexps_forward (&state, XINT (from), XINT (to), |
c81a3712 RS |
1710 | target, !NILP (stopbefore), oldstate, |
1711 | !NILP (commentstop)); | |
8489eb67 RS |
1712 | |
1713 | SET_PT (state.location); | |
1714 | ||
1715 | return Fcons (make_number (state.depth), | |
1716 | Fcons (state.prevlevelstart < 0 ? Qnil : make_number (state.prevlevelstart), | |
1717 | Fcons (state.thislevelstart < 0 ? Qnil : make_number (state.thislevelstart), | |
1718 | Fcons (state.instring >= 0 ? make_number (state.instring) : Qnil, | |
1719 | Fcons (state.incomment ? Qt : Qnil, | |
1720 | Fcons (state.quoted ? Qt : Qnil, | |
e5d4f4dc RS |
1721 | Fcons (make_number (state.mindepth), |
1722 | Fcons (state.comstyle ? Qt : Qnil, | |
1723 | Qnil)))))))); | |
8489eb67 RS |
1724 | } |
1725 | \f | |
1726 | init_syntax_once () | |
1727 | { | |
1728 | register int i; | |
8ea151b2 | 1729 | Lisp_Object temp; |
8489eb67 | 1730 | |
5ebaddf5 RS |
1731 | /* This has to be done here, before we call Fmake_char_table. */ |
1732 | Qsyntax_table = intern ("syntax-table"); | |
1733 | staticpro (&Qsyntax_table); | |
1734 | ||
1735 | /* Intern this now in case it isn't already done. | |
1736 | Setting this variable twice is harmless. | |
1737 | But don't staticpro it here--that is done in alloc.c. */ | |
1738 | Qchar_table_extra_slots = intern ("char-table-extra-slots"); | |
1739 | ||
1740 | /* Now we are ready to set up this property, so we can | |
1741 | create syntax tables. */ | |
1742 | Fput (Qsyntax_table, Qchar_table_extra_slots, make_number (0)); | |
1743 | ||
8ea151b2 | 1744 | temp = Fcons (make_number ((int) Swhitespace), Qnil); |
8489eb67 | 1745 | |
5ebaddf5 | 1746 | Vstandard_syntax_table = Fmake_char_table (Qsyntax_table, temp); |
8489eb67 | 1747 | |
8ea151b2 | 1748 | temp = Fcons (make_number ((int) Sword), Qnil); |
8489eb67 | 1749 | for (i = 'a'; i <= 'z'; i++) |
8ea151b2 | 1750 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
8489eb67 | 1751 | for (i = 'A'; i <= 'Z'; i++) |
8ea151b2 | 1752 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
8489eb67 | 1753 | for (i = '0'; i <= '9'; i++) |
8ea151b2 RS |
1754 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, i, temp); |
1755 | ||
1756 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '$', temp); | |
1757 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '%', temp); | |
1758 | ||
1759 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '(', | |
1760 | Fcons (make_number (Sopen), make_number (')'))); | |
1761 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ')', | |
1762 | Fcons (make_number (Sclose), make_number ('('))); | |
1763 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '[', | |
1764 | Fcons (make_number (Sopen), make_number (']'))); | |
1765 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ']', | |
1766 | Fcons (make_number (Sclose), make_number ('['))); | |
1767 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '{', | |
1768 | Fcons (make_number (Sopen), make_number ('}'))); | |
1769 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '}', | |
1770 | Fcons (make_number (Sclose), make_number ('{'))); | |
1771 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '"', | |
1772 | Fcons (make_number ((int) Sstring), Qnil)); | |
1773 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, '\\', | |
1774 | Fcons (make_number ((int) Sescape), Qnil)); | |
1775 | ||
1776 | temp = Fcons (make_number ((int) Ssymbol), Qnil); | |
8489eb67 | 1777 | for (i = 0; i < 10; i++) |
8ea151b2 | 1778 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, "_-+*/&|<>="[i], temp); |
8489eb67 | 1779 | |
8ea151b2 | 1780 | temp = Fcons (make_number ((int) Spunct), Qnil); |
8489eb67 | 1781 | for (i = 0; i < 12; i++) |
8ea151b2 | 1782 | SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, ".,;:?!#@~^'`"[i], temp); |
8489eb67 RS |
1783 | } |
1784 | ||
1785 | syms_of_syntax () | |
1786 | { | |
1787 | Qsyntax_table_p = intern ("syntax-table-p"); | |
1788 | staticpro (&Qsyntax_table_p); | |
1789 | ||
1790 | DEFVAR_BOOL ("parse-sexp-ignore-comments", &parse_sexp_ignore_comments, | |
1791 | "Non-nil means `forward-sexp', etc., should treat comments as whitespace."); | |
1792 | ||
1793 | words_include_escapes = 0; | |
1794 | DEFVAR_BOOL ("words-include-escapes", &words_include_escapes, | |
1795 | "Non-nil means `forward-word', etc., should treat escape chars part of words."); | |
1796 | ||
1797 | defsubr (&Ssyntax_table_p); | |
1798 | defsubr (&Ssyntax_table); | |
1799 | defsubr (&Sstandard_syntax_table); | |
1800 | defsubr (&Scopy_syntax_table); | |
1801 | defsubr (&Sset_syntax_table); | |
1802 | defsubr (&Schar_syntax); | |
beefa22e | 1803 | defsubr (&Smatching_paren); |
8489eb67 RS |
1804 | defsubr (&Smodify_syntax_entry); |
1805 | defsubr (&Sdescribe_syntax); | |
1806 | ||
1807 | defsubr (&Sforward_word); | |
1808 | ||
b3cfe0c8 | 1809 | defsubr (&Sforward_comment); |
8489eb67 RS |
1810 | defsubr (&Sscan_lists); |
1811 | defsubr (&Sscan_sexps); | |
1812 | defsubr (&Sbackward_prefix_chars); | |
1813 | defsubr (&Sparse_partial_sexp); | |
1814 | } |