1 /* GNU Emacs routines to deal with syntax tables; also word and list parsing.
2 Copyright (C) 1985, 1987, 1993, 1994, 1995, 1997, 1998, 1999, 2001,
3 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010
4 Free Software Foundation, Inc.
6 This file is part of GNU Emacs.
8 GNU Emacs is free software: you can redistribute it and/or modify
9 it under the terms of the GNU General Public License as published by
10 the Free Software Foundation, either version 3 of the License, or
11 (at your option) any later version.
13 GNU Emacs is distributed in the hope that it will be useful,
14 but WITHOUT ANY WARRANTY; without even the implied warranty of
15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 GNU General Public License for more details.
18 You should have received a copy of the GNU General Public License
19 along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>. */
28 #include "character.h"
32 /* Make syntax table lookup grant data in gl_state. */
33 #define SYNTAX_ENTRY_VIA_PROPERTY
36 #include "intervals.h"
38 /* We use these constants in place for comment-style and
39 string-ender-char to distinguish comments/strings started by
40 comment_fence and string_fence codes. */
42 #define ST_COMMENT_STYLE (256 + 1)
43 #define ST_STRING_STYLE (256 + 2)
46 Lisp_Object Qsyntax_table_p
, Qsyntax_table
, Qscan_error
;
48 int words_include_escapes
;
49 int parse_sexp_lookup_properties
;
51 /* Nonzero means `scan-sexps' treat all multibyte characters as symbol. */
52 int multibyte_syntax_as_symbol
;
54 /* Used as a temporary in SYNTAX_ENTRY and other macros in syntax.h,
55 if not compiled with GCC. No need to mark it, since it is used
56 only very temporarily. */
57 Lisp_Object syntax_temp
;
59 /* Non-zero means an open parenthesis in column 0 is always considered
60 to be the start of a defun. Zero means an open parenthesis in
61 column 0 has no special meaning. */
63 int open_paren_in_column_0_is_defun_start
;
65 /* This is the internal form of the parse state used in parse-partial-sexp. */
67 struct lisp_parse_state
69 int depth
; /* Depth at end of parsing. */
70 int instring
; /* -1 if not within string, else desired terminator. */
71 int incomment
; /* -1 if in unnestable comment else comment nesting */
72 int comstyle
; /* comment style a=0, or b=1, or ST_COMMENT_STYLE. */
73 int quoted
; /* Nonzero if just after an escape char at end of parsing */
74 int mindepth
; /* Minimum depth seen while scanning. */
75 /* Char number of most recent start-of-expression at current level */
76 EMACS_INT thislevelstart
;
77 /* Char number of start of containing expression */
78 EMACS_INT prevlevelstart
;
79 EMACS_INT location
; /* Char number at which parsing stopped. */
80 EMACS_INT comstr_start
; /* Position of last comment/string starter. */
81 Lisp_Object levelstarts
; /* Char numbers of starts-of-expression
82 of levels (starting from outermost). */
85 /* These variables are a cache for finding the start of a defun.
86 find_start_pos is the place for which the defun start was found.
87 find_start_value is the defun start position found for it.
88 find_start_value_byte is the corresponding byte position.
89 find_start_buffer is the buffer it was found in.
90 find_start_begv is the BEGV value when it was found.
91 find_start_modiff is the value of MODIFF when it was found. */
93 static EMACS_INT find_start_pos
;
94 static EMACS_INT find_start_value
;
95 static EMACS_INT find_start_value_byte
;
96 static struct buffer
*find_start_buffer
;
97 static EMACS_INT find_start_begv
;
98 static int find_start_modiff
;
101 static Lisp_Object
skip_chars (int, Lisp_Object
, Lisp_Object
, int);
102 static Lisp_Object
skip_syntaxes (int, Lisp_Object
, Lisp_Object
);
103 static Lisp_Object
scan_lists (EMACS_INT
, EMACS_INT
, EMACS_INT
, int);
104 static void scan_sexps_forward (struct lisp_parse_state
*,
105 EMACS_INT
, EMACS_INT
, EMACS_INT
, int,
106 int, Lisp_Object
, int);
107 static int in_classes (int, Lisp_Object
);
110 struct gl_state_s gl_state
; /* Global state of syntax parser. */
112 INTERVAL
interval_of (int, Lisp_Object
);
113 #define INTERVALS_AT_ONCE 10 /* 1 + max-number of intervals
114 to scan to property-change. */
116 /* Update gl_state to an appropriate interval which contains CHARPOS. The
117 sign of COUNT give the relative position of CHARPOS wrt the previously
118 valid interval. If INIT, only [be]_property fields of gl_state are
119 valid at start, the rest is filled basing on OBJECT.
121 `gl_state.*_i' are the intervals, and CHARPOS is further in the search
122 direction than the intervals - or in an interval. We update the
123 current syntax-table basing on the property of this interval, and
124 update the interval to start further than CHARPOS - or be
125 NULL_INTERVAL. We also update lim_property to be the next value of
126 charpos to call this subroutine again - or be before/after the
127 start/end of OBJECT. */
130 update_syntax_table (int charpos
, int count
, int init
, Lisp_Object object
)
132 Lisp_Object tmp_table
;
133 int cnt
= 0, invalidate
= 1;
138 gl_state
.old_prop
= Qnil
;
139 gl_state
.start
= gl_state
.b_property
;
140 gl_state
.stop
= gl_state
.e_property
;
141 i
= interval_of (charpos
, object
);
142 gl_state
.backward_i
= gl_state
.forward_i
= i
;
144 if (NULL_INTERVAL_P (i
))
146 /* interval_of updates only ->position of the return value, so
147 update the parents manually to speed up update_interval. */
148 while (!NULL_PARENT (i
))
150 if (AM_RIGHT_CHILD (i
))
151 INTERVAL_PARENT (i
)->position
= i
->position
152 - LEFT_TOTAL_LENGTH (i
) + TOTAL_LENGTH (i
) /* right end */
153 - TOTAL_LENGTH (INTERVAL_PARENT (i
))
154 + LEFT_TOTAL_LENGTH (INTERVAL_PARENT (i
));
156 INTERVAL_PARENT (i
)->position
= i
->position
- LEFT_TOTAL_LENGTH (i
)
158 i
= INTERVAL_PARENT (i
);
160 i
= gl_state
.forward_i
;
161 gl_state
.b_property
= i
->position
- gl_state
.offset
;
162 gl_state
.e_property
= INTERVAL_LAST_POS (i
) - gl_state
.offset
;
165 i
= count
> 0 ? gl_state
.forward_i
: gl_state
.backward_i
;
167 /* We are guaranteed to be called with CHARPOS either in i,
169 if (NULL_INTERVAL_P (i
))
170 error ("Error in syntax_table logic for to-the-end intervals");
171 else if (charpos
< i
->position
) /* Move left. */
174 error ("Error in syntax_table logic for intervals <-");
175 /* Update the interval. */
176 i
= update_interval (i
, charpos
);
177 if (INTERVAL_LAST_POS (i
) != gl_state
.b_property
)
180 gl_state
.forward_i
= i
;
181 gl_state
.e_property
= INTERVAL_LAST_POS (i
) - gl_state
.offset
;
184 else if (charpos
>= INTERVAL_LAST_POS (i
)) /* Move right. */
187 error ("Error in syntax_table logic for intervals ->");
188 /* Update the interval. */
189 i
= update_interval (i
, charpos
);
190 if (i
->position
!= gl_state
.e_property
)
193 gl_state
.backward_i
= i
;
194 gl_state
.b_property
= i
->position
- gl_state
.offset
;
199 tmp_table
= textget (i
->plist
, Qsyntax_table
);
202 invalidate
= !EQ (tmp_table
, gl_state
.old_prop
); /* Need to invalidate? */
204 if (invalidate
) /* Did not get to adjacent interval. */
205 { /* with the same table => */
206 /* invalidate the old range. */
209 gl_state
.backward_i
= i
;
210 gl_state
.b_property
= i
->position
- gl_state
.offset
;
214 gl_state
.forward_i
= i
;
215 gl_state
.e_property
= INTERVAL_LAST_POS (i
) - gl_state
.offset
;
219 if (!EQ (tmp_table
, gl_state
.old_prop
))
221 gl_state
.current_syntax_table
= tmp_table
;
222 gl_state
.old_prop
= tmp_table
;
223 if (EQ (Fsyntax_table_p (tmp_table
), Qt
))
225 gl_state
.use_global
= 0;
227 else if (CONSP (tmp_table
))
229 gl_state
.use_global
= 1;
230 gl_state
.global_code
= tmp_table
;
234 gl_state
.use_global
= 0;
235 gl_state
.current_syntax_table
= current_buffer
->syntax_table
;
239 while (!NULL_INTERVAL_P (i
))
241 if (cnt
&& !EQ (tmp_table
, textget (i
->plist
, Qsyntax_table
)))
245 gl_state
.e_property
= i
->position
- gl_state
.offset
;
246 gl_state
.forward_i
= i
;
251 = i
->position
+ LENGTH (i
) - gl_state
.offset
;
252 gl_state
.backward_i
= i
;
256 else if (cnt
== INTERVALS_AT_ONCE
)
261 = i
->position
+ LENGTH (i
) - gl_state
.offset
262 /* e_property at EOB is not set to ZV but to ZV+1, so that
263 we can do INC(from);UPDATE_SYNTAX_TABLE_FORWARD without
264 having to check eob between the two. */
265 + (NULL_INTERVAL_P (next_interval (i
)) ? 1 : 0);
266 gl_state
.forward_i
= i
;
270 gl_state
.b_property
= i
->position
- gl_state
.offset
;
271 gl_state
.backward_i
= i
;
276 i
= count
> 0 ? next_interval (i
) : previous_interval (i
);
278 eassert (NULL_INTERVAL_P (i
)); /* This property goes to the end. */
280 gl_state
.e_property
= gl_state
.stop
;
282 gl_state
.b_property
= gl_state
.start
;
285 /* Returns TRUE if char at CHARPOS is quoted.
286 Global syntax-table data should be set up already to be good at CHARPOS
287 or after. On return global syntax data is good for lookup at CHARPOS. */
290 char_quoted (EMACS_INT charpos
, EMACS_INT bytepos
)
292 register enum syntaxcode code
;
293 register EMACS_INT beg
= BEGV
;
294 register int quoted
= 0;
295 EMACS_INT orig
= charpos
;
297 while (charpos
> beg
)
300 DEC_BOTH (charpos
, bytepos
);
302 UPDATE_SYNTAX_TABLE_BACKWARD (charpos
);
303 c
= FETCH_CHAR_AS_MULTIBYTE (bytepos
);
305 if (! (code
== Scharquote
|| code
== Sescape
))
311 UPDATE_SYNTAX_TABLE (orig
);
315 /* Return the bytepos one character after BYTEPOS.
316 We assume that BYTEPOS is not at the end of the buffer. */
319 inc_bytepos (EMACS_INT bytepos
)
321 if (NILP (current_buffer
->enable_multibyte_characters
))
328 /* Return the bytepos one character before BYTEPOS.
329 We assume that BYTEPOS is not at the start of the buffer. */
332 dec_bytepos (EMACS_INT bytepos
)
334 if (NILP (current_buffer
->enable_multibyte_characters
))
341 /* Return a defun-start position before POS and not too far before.
342 It should be the last one before POS, or nearly the last.
344 When open_paren_in_column_0_is_defun_start is nonzero,
345 only the beginning of the buffer is treated as a defun-start.
347 We record the information about where the scan started
348 and what its result was, so that another call in the same area
349 can return the same value very quickly.
351 There is no promise at which position the global syntax data is
352 valid on return from the subroutine, so the caller should explicitly
353 update the global data. */
356 find_defun_start (EMACS_INT pos
, EMACS_INT pos_byte
)
358 EMACS_INT opoint
= PT
, opoint_byte
= PT_BYTE
;
360 if (!open_paren_in_column_0_is_defun_start
)
362 find_start_value_byte
= BEGV_BYTE
;
366 /* Use previous finding, if it's valid and applies to this inquiry. */
367 if (current_buffer
== find_start_buffer
368 /* Reuse the defun-start even if POS is a little farther on.
369 POS might be in the next defun, but that's ok.
370 Our value may not be the best possible, but will still be usable. */
371 && pos
<= find_start_pos
+ 1000
372 && pos
>= find_start_value
373 && BEGV
== find_start_begv
374 && MODIFF
== find_start_modiff
)
375 return find_start_value
;
377 /* Back up to start of line. */
378 scan_newline (pos
, pos_byte
, BEGV
, BEGV_BYTE
, -1, 1);
380 /* We optimize syntax-table lookup for rare updates. Thus we accept
381 only those `^\s(' which are good in global _and_ text-property
383 SETUP_BUFFER_SYNTAX_TABLE ();
388 /* Open-paren at start of line means we may have found our
390 c
= FETCH_CHAR_AS_MULTIBYTE (PT_BYTE
);
391 if (SYNTAX (c
) == Sopen
)
393 SETUP_SYNTAX_TABLE (PT
+ 1, -1); /* Try again... */
394 c
= FETCH_CHAR_AS_MULTIBYTE (PT_BYTE
);
395 if (SYNTAX (c
) == Sopen
)
397 /* Now fallback to the default value. */
398 SETUP_BUFFER_SYNTAX_TABLE ();
400 /* Move to beg of previous line. */
401 scan_newline (PT
, PT_BYTE
, BEGV
, BEGV_BYTE
, -2, 1);
404 /* Record what we found, for the next try. */
405 find_start_value
= PT
;
406 find_start_value_byte
= PT_BYTE
;
407 find_start_buffer
= current_buffer
;
408 find_start_modiff
= MODIFF
;
409 find_start_begv
= BEGV
;
410 find_start_pos
= pos
;
412 TEMP_SET_PT_BOTH (opoint
, opoint_byte
);
414 return find_start_value
;
417 /* Return the SYNTAX_COMEND_FIRST of the character before POS, POS_BYTE. */
420 prev_char_comend_first (int pos
, int pos_byte
)
424 DEC_BOTH (pos
, pos_byte
);
425 UPDATE_SYNTAX_TABLE_BACKWARD (pos
);
426 c
= FETCH_CHAR (pos_byte
);
427 val
= SYNTAX_COMEND_FIRST (c
);
428 UPDATE_SYNTAX_TABLE_FORWARD (pos
+ 1);
432 /* Return the SYNTAX_COMSTART_FIRST of the character before POS, POS_BYTE. */
435 * prev_char_comstart_first (pos, pos_byte)
440 * DEC_BOTH (pos, pos_byte);
441 * UPDATE_SYNTAX_TABLE_BACKWARD (pos);
442 * c = FETCH_CHAR (pos_byte);
443 * val = SYNTAX_COMSTART_FIRST (c);
444 * UPDATE_SYNTAX_TABLE_FORWARD (pos + 1);
448 /* Checks whether charpos FROM is at the end of a comment.
449 FROM_BYTE is the bytepos corresponding to FROM.
450 Do not move back before STOP.
452 Return a positive value if we find a comment ending at FROM/FROM_BYTE;
455 If successful, store the charpos of the comment's beginning
456 into *CHARPOS_PTR, and the bytepos into *BYTEPOS_PTR.
458 Global syntax data remains valid for backward search starting at
459 the returned value (or at FROM, if the search was not successful). */
462 back_comment (EMACS_INT from
, EMACS_INT from_byte
, EMACS_INT stop
, int comnested
, int comstyle
, EMACS_INT
*charpos_ptr
, EMACS_INT
*bytepos_ptr
)
464 /* Look back, counting the parity of string-quotes,
465 and recording the comment-starters seen.
466 When we reach a safe place, assume that's not in a string;
467 then step the main scan to the earliest comment-starter seen
468 an even number of string quotes away from the safe place.
470 OFROM[I] is position of the earliest comment-starter seen
471 which is I+2X quotes from the comment-end.
472 PARITY is current parity of quotes from the comment end. */
473 int string_style
= -1; /* Presumed outside of any string. */
474 int string_lossage
= 0;
475 /* Not a real lossage: indicates that we have passed a matching comment
476 starter plus a non-matching comment-ender, meaning that any matching
477 comment-starter we might see later could be a false positive (hidden
478 inside another comment).
479 Test case: { a (* b } c (* d *) */
480 int comment_lossage
= 0;
481 EMACS_INT comment_end
= from
;
482 EMACS_INT comment_end_byte
= from_byte
;
483 EMACS_INT comstart_pos
= 0;
484 EMACS_INT comstart_byte
;
485 /* Place where the containing defun starts,
486 or 0 if we didn't come across it yet. */
487 EMACS_INT defun_start
= 0;
488 EMACS_INT defun_start_byte
= 0;
489 register enum syntaxcode code
;
490 int nesting
= 1; /* current comment nesting */
494 /* FIXME: A }} comment-ender style leads to incorrect behavior
495 in the case of {{ c }}} because we ignore the last two chars which are
496 assumed to be comment-enders although they aren't. */
498 /* At beginning of range to scan, we're outside of strings;
499 that determines quote parity to the comment-end. */
502 int temp_byte
, prev_syntax
;
503 int com2start
, com2end
;
505 /* Move back and examine a character. */
506 DEC_BOTH (from
, from_byte
);
507 UPDATE_SYNTAX_TABLE_BACKWARD (from
);
509 prev_syntax
= syntax
;
510 c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
511 syntax
= SYNTAX_WITH_FLAGS (c
);
514 /* Check for 2-char comment markers. */
515 com2start
= (SYNTAX_FLAGS_COMSTART_FIRST (syntax
)
516 && SYNTAX_FLAGS_COMSTART_SECOND (prev_syntax
)
517 && comstyle
== SYNTAX_FLAGS_COMMENT_STYLE (prev_syntax
)
518 && (SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax
)
519 || SYNTAX_FLAGS_COMMENT_NESTED (syntax
)) == comnested
);
520 com2end
= (SYNTAX_FLAGS_COMEND_FIRST (syntax
)
521 && SYNTAX_FLAGS_COMEND_SECOND (prev_syntax
));
523 /* Nasty cases with overlapping 2-char comment markers:
524 - snmp-mode: -- c -- foo -- c --
532 /* If a 2-char comment sequence partly overlaps with another,
533 we don't try to be clever. */
534 if (from
> stop
&& (com2end
|| com2start
))
536 int next
= from
, next_byte
= from_byte
, next_c
, next_syntax
;
537 DEC_BOTH (next
, next_byte
);
538 UPDATE_SYNTAX_TABLE_BACKWARD (next
);
539 next_c
= FETCH_CHAR_AS_MULTIBYTE (next_byte
);
540 next_syntax
= SYNTAX_WITH_FLAGS (next_c
);
541 if (((com2start
|| comnested
)
542 && SYNTAX_FLAGS_COMEND_SECOND (syntax
)
543 && SYNTAX_FLAGS_COMEND_FIRST (next_syntax
))
544 || ((com2end
|| comnested
)
545 && SYNTAX_FLAGS_COMSTART_SECOND (syntax
)
546 && comstyle
== SYNTAX_FLAGS_COMMENT_STYLE (syntax
)
547 && SYNTAX_FLAGS_COMSTART_FIRST (next_syntax
)))
549 /* UPDATE_SYNTAX_TABLE_FORWARD (next + 1); */
552 if (com2start
&& comstart_pos
== 0)
553 /* We're looking at a comment starter. But it might be a comment
554 ender as well (see snmp-mode). The first time we see one, we
555 need to consider it as a comment starter,
556 and the subsequent times as a comment ender. */
559 /* Turn a 2-char comment sequences into the appropriate syntax. */
564 /* Ignore comment starters of a different style. */
565 else if (code
== Scomment
566 && (comstyle
!= SYNTAX_FLAGS_COMMENT_STYLE (syntax
)
567 || SYNTAX_FLAGS_COMMENT_NESTED (syntax
) != comnested
))
570 /* Ignore escaped characters, except comment-enders. */
571 if (code
!= Sendcomment
&& char_quoted (from
, from_byte
))
578 c
= (code
== Sstring_fence
? ST_STRING_STYLE
: ST_COMMENT_STYLE
);
580 /* Track parity of quotes. */
581 if (string_style
== -1)
582 /* Entering a string. */
584 else if (string_style
== c
)
585 /* Leaving the string. */
588 /* If we have two kinds of string delimiters.
589 There's no way to grok this scanning backwards. */
594 /* We've already checked that it is the relevant comstyle. */
595 if (string_style
!= -1 || comment_lossage
|| string_lossage
)
596 /* There are odd string quotes involved, so let's be careful.
597 Test case in Pascal: " { " a { " } */
602 /* Record best comment-starter so far. */
604 comstart_byte
= from_byte
;
606 else if (--nesting
<= 0)
607 /* nested comments have to be balanced, so we don't need to
608 keep looking for earlier ones. We use here the same (slightly
609 incorrect) reasoning as below: since it is followed by uniform
610 paired string quotes, this comment-start has to be outside of
611 strings, else the comment-end itself would be inside a string. */
616 if (SYNTAX_FLAGS_COMMENT_STYLE (syntax
) == comstyle
617 && ((com2end
&& SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax
))
618 || SYNTAX_FLAGS_COMMENT_NESTED (syntax
)) == comnested
)
619 /* This is the same style of comment ender as ours. */
624 /* Anything before that can't count because it would match
625 this comment-ender rather than ours. */
626 from
= stop
; /* Break out of the loop. */
628 else if (comstart_pos
!= 0 || c
!= '\n')
629 /* We're mixing comment styles here, so we'd better be careful.
630 The (comstart_pos != 0 || c != '\n') check is not quite correct
631 (we should just always set comment_lossage), but removing it
632 would imply that any multiline comment in C would go through
633 lossage, which seems overkill.
634 The failure should only happen in the rare cases such as
640 /* Assume a defun-start point is outside of strings. */
641 if (open_paren_in_column_0_is_defun_start
643 || (temp_byte
= dec_bytepos (from_byte
),
644 FETCH_CHAR (temp_byte
) == '\n')))
647 defun_start_byte
= from_byte
;
648 from
= stop
; /* Break out of the loop. */
657 if (comstart_pos
== 0)
660 from_byte
= comment_end_byte
;
661 UPDATE_SYNTAX_TABLE_FORWARD (comment_end
- 1);
663 /* If comstart_pos is set and we get here (ie. didn't jump to `lossage'
664 or `done'), then we've found the beginning of the non-nested comment. */
665 else if (1) /* !comnested */
668 from_byte
= comstart_byte
;
669 UPDATE_SYNTAX_TABLE_FORWARD (from
- 1);
673 struct lisp_parse_state state
;
675 /* We had two kinds of string delimiters mixed up
676 together. Decode this going forwards.
677 Scan fwd from a known safe place (beginning-of-defun)
678 to the one in question; this records where we
679 last passed a comment starter. */
680 /* If we did not already find the defun start, find it now. */
681 if (defun_start
== 0)
683 defun_start
= find_defun_start (comment_end
, comment_end_byte
);
684 defun_start_byte
= find_start_value_byte
;
688 scan_sexps_forward (&state
,
689 defun_start
, defun_start_byte
,
690 comment_end
, -10000, 0, Qnil
, 0);
691 defun_start
= comment_end
;
692 if (state
.incomment
== (comnested
? 1 : -1)
693 && state
.comstyle
== comstyle
)
694 from
= state
.comstr_start
;
699 /* If comment_end is inside some other comment, maybe ours
700 is nested, so we need to try again from within the
701 surrounding comment. Example: { a (* " *) */
703 /* FIXME: We should advance by one or two chars. */
704 defun_start
= state
.comstr_start
+ 2;
705 defun_start_byte
= CHAR_TO_BYTE (defun_start
);
708 } while (defun_start
< comment_end
);
710 from_byte
= CHAR_TO_BYTE (from
);
711 UPDATE_SYNTAX_TABLE_FORWARD (from
- 1);
716 *bytepos_ptr
= from_byte
;
718 return (from
== comment_end
) ? -1 : from
;
721 DEFUN ("syntax-table-p", Fsyntax_table_p
, Ssyntax_table_p
, 1, 1, 0,
722 doc
: /* Return t if OBJECT is a syntax table.
723 Currently, any char-table counts as a syntax table. */)
727 if (CHAR_TABLE_P (object
)
728 && EQ (XCHAR_TABLE (object
)->purpose
, Qsyntax_table
))
734 check_syntax_table (Lisp_Object obj
)
736 CHECK_TYPE (CHAR_TABLE_P (obj
) && EQ (XCHAR_TABLE (obj
)->purpose
, Qsyntax_table
),
737 Qsyntax_table_p
, obj
);
740 DEFUN ("syntax-table", Fsyntax_table
, Ssyntax_table
, 0, 0, 0,
741 doc
: /* Return the current syntax table.
742 This is the one specified by the current buffer. */)
745 return current_buffer
->syntax_table
;
748 DEFUN ("standard-syntax-table", Fstandard_syntax_table
,
749 Sstandard_syntax_table
, 0, 0, 0,
750 doc
: /* Return the standard syntax table.
751 This is the one used for new buffers. */)
754 return Vstandard_syntax_table
;
757 DEFUN ("copy-syntax-table", Fcopy_syntax_table
, Scopy_syntax_table
, 0, 1, 0,
758 doc
: /* Construct a new syntax table and return it.
759 It is a copy of the TABLE, which defaults to the standard syntax table. */)
766 check_syntax_table (table
);
768 table
= Vstandard_syntax_table
;
770 copy
= Fcopy_sequence (table
);
772 /* Only the standard syntax table should have a default element.
773 Other syntax tables should inherit from parents instead. */
774 XCHAR_TABLE (copy
)->defalt
= Qnil
;
776 /* Copied syntax tables should all have parents.
777 If we copied one with no parent, such as the standard syntax table,
778 use the standard syntax table as the copy's parent. */
779 if (NILP (XCHAR_TABLE (copy
)->parent
))
780 Fset_char_table_parent (copy
, Vstandard_syntax_table
);
784 DEFUN ("set-syntax-table", Fset_syntax_table
, Sset_syntax_table
, 1, 1, 0,
785 doc
: /* Select a new syntax table for the current buffer.
786 One argument, a syntax table. */)
791 check_syntax_table (table
);
792 current_buffer
->syntax_table
= table
;
793 /* Indicate that this buffer now has a specified syntax table. */
794 idx
= PER_BUFFER_VAR_IDX (syntax_table
);
795 SET_PER_BUFFER_VALUE_P (current_buffer
, idx
, 1);
799 /* Convert a letter which signifies a syntax code
800 into the code it signifies.
801 This is used by modify-syntax-entry, and other things. */
803 unsigned char syntax_spec_code
[0400] =
804 { 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
805 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
806 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
807 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
808 (char) Swhitespace
, (char) Scomment_fence
, (char) Sstring
, 0377,
809 (char) Smath
, 0377, 0377, (char) Squote
,
810 (char) Sopen
, (char) Sclose
, 0377, 0377,
811 0377, (char) Swhitespace
, (char) Spunct
, (char) Scharquote
,
812 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
813 0377, 0377, 0377, 0377,
814 (char) Scomment
, 0377, (char) Sendcomment
, 0377,
815 (char) Sinherit
, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* @, A ... */
816 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
817 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword
,
818 0377, 0377, 0377, 0377, (char) Sescape
, 0377, 0377, (char) Ssymbol
,
819 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* `, a, ... */
820 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
821 0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword
,
822 0377, 0377, 0377, 0377, (char) Sstring_fence
, 0377, 0377, 0377
825 /* Indexed by syntax code, give the letter that describes it. */
827 char syntax_code_spec
[16] =
829 ' ', '.', 'w', '_', '(', ')', '\'', '\"', '$', '\\', '/', '<', '>', '@',
833 /* Indexed by syntax code, give the object (cons of syntax code and
834 nil) to be stored in syntax table. Since these objects can be
835 shared among syntax tables, we generate them in advance. By
836 sharing objects, the function `describe-syntax' can give a more
838 static Lisp_Object Vsyntax_code_object
;
841 DEFUN ("char-syntax", Fchar_syntax
, Schar_syntax
, 1, 1, 0,
842 doc
: /* Return the syntax code of CHARACTER, described by a character.
843 For example, if CHARACTER is a word constituent, the
844 character `w' (119) is returned.
845 The characters that correspond to various syntax codes
846 are listed in the documentation of `modify-syntax-entry'. */)
848 Lisp_Object character
;
851 CHECK_CHARACTER (character
);
852 char_int
= XINT (character
);
853 SETUP_BUFFER_SYNTAX_TABLE ();
854 return make_number (syntax_code_spec
[(int) SYNTAX (char_int
)]);
857 DEFUN ("matching-paren", Fmatching_paren
, Smatching_paren
, 1, 1, 0,
858 doc
: /* Return the matching parenthesis of CHARACTER, or nil if none. */)
860 Lisp_Object character
;
863 CHECK_NUMBER (character
);
864 char_int
= XINT (character
);
865 SETUP_BUFFER_SYNTAX_TABLE ();
866 code
= SYNTAX (char_int
);
867 if (code
== Sopen
|| code
== Sclose
)
868 return SYNTAX_MATCH (char_int
);
872 DEFUN ("string-to-syntax", Fstring_to_syntax
, Sstring_to_syntax
, 1, 1, 0,
873 doc
: /* Convert a syntax specification STRING into syntax cell form.
874 STRING should be a string as it is allowed as argument of
875 `modify-syntax-entry'. Value is the equivalent cons cell
876 \(CODE . MATCHING-CHAR) that can be used as value of a `syntax-table'
881 register const unsigned char *p
;
882 register enum syntaxcode code
;
886 CHECK_STRING (string
);
889 code
= (enum syntaxcode
) syntax_spec_code
[*p
++];
890 if (((int) code
& 0377) == 0377)
891 error ("Invalid syntax description letter: %c", p
[-1]);
893 if (code
== Sinherit
)
899 int character
= STRING_CHAR_AND_LENGTH (p
, len
);
900 XSETINT (match
, character
);
901 if (XFASTINT (match
) == ' ')
941 if (val
< XVECTOR (Vsyntax_code_object
)->size
&& NILP (match
))
942 return XVECTOR (Vsyntax_code_object
)->contents
[val
];
944 /* Since we can't use a shared object, let's make a new one. */
945 return Fcons (make_number (val
), match
);
948 /* I really don't know why this is interactive
949 help-form should at least be made useful whilst reading the second arg. */
950 DEFUN ("modify-syntax-entry", Fmodify_syntax_entry
, Smodify_syntax_entry
, 2, 3,
951 "cSet syntax for character: \nsSet syntax for %s to: ",
952 doc
: /* Set syntax for character CHAR according to string NEWENTRY.
953 The syntax is changed only for table SYNTAX-TABLE, which defaults to
954 the current buffer's syntax table.
955 CHAR may be a cons (MIN . MAX), in which case, syntaxes of all characters
956 in the range MIN to MAX are changed.
957 The first character of NEWENTRY should be one of the following:
958 Space or - whitespace syntax. w word constituent.
959 _ symbol constituent. . punctuation.
960 ( open-parenthesis. ) close-parenthesis.
961 " string quote. \\ escape.
962 $ paired delimiter. ' expression quote or prefix operator.
963 < comment starter. > comment ender.
964 / character-quote. @ inherit from `standard-syntax-table'.
965 | generic string fence. ! generic comment fence.
967 Only single-character comment start and end sequences are represented thus.
968 Two-character sequences are represented as described below.
969 The second character of NEWENTRY is the matching parenthesis,
970 used only if the first character is `(' or `)'.
971 Any additional characters are flags.
972 Defined flags are the characters 1, 2, 3, 4, b, p, and n.
973 1 means CHAR is the start of a two-char comment start sequence.
974 2 means CHAR is the second character of such a sequence.
975 3 means CHAR is the start of a two-char comment end sequence.
976 4 means CHAR is the second character of such a sequence.
978 There can be up to two orthogonal comment sequences. This is to support
979 language modes such as C++. By default, all comment sequences are of style
980 a, but you can set the comment sequence style to b (on the second character
981 of a comment-start, or the first character of a comment-end sequence) using
983 b means CHAR is part of comment sequence b.
984 n means CHAR is part of a nestable comment sequence.
986 p means CHAR is a prefix character for `backward-prefix-chars';
987 such characters are treated as whitespace when they occur
989 usage: (modify-syntax-entry CHAR NEWENTRY &optional SYNTAX-TABLE) */)
990 (c
, newentry
, syntax_table
)
991 Lisp_Object c
, newentry
, syntax_table
;
995 CHECK_CHARACTER_CAR (c
);
996 CHECK_CHARACTER_CDR (c
);
1001 if (NILP (syntax_table
))
1002 syntax_table
= current_buffer
->syntax_table
;
1004 check_syntax_table (syntax_table
);
1006 newentry
= Fstring_to_syntax (newentry
);
1008 SET_RAW_SYNTAX_ENTRY_RANGE (syntax_table
, c
, newentry
);
1010 SET_RAW_SYNTAX_ENTRY (syntax_table
, XINT (c
), newentry
);
1012 /* We clear the regexp cache, since character classes can now have
1013 different values from those in the compiled regexps.*/
1014 clear_regexp_cache ();
1019 /* Dump syntax table to buffer in human-readable format */
1021 DEFUN ("internal-describe-syntax-value", Finternal_describe_syntax_value
,
1022 Sinternal_describe_syntax_value
, 1, 1, 0,
1023 doc
: /* Insert a description of the internal syntax description SYNTAX at point. */)
1027 register enum syntaxcode code
;
1028 char desc
, start1
, start2
, end1
, end2
, prefix
, comstyle
, comnested
;
1030 Lisp_Object first
, match_lisp
, value
= syntax
;
1034 insert_string ("default");
1038 if (CHAR_TABLE_P (value
))
1040 insert_string ("deeper char-table ...");
1046 insert_string ("invalid");
1050 first
= XCAR (value
);
1051 match_lisp
= XCDR (value
);
1053 if (!INTEGERP (first
) || !(NILP (match_lisp
) || INTEGERP (match_lisp
)))
1055 insert_string ("invalid");
1059 code
= (enum syntaxcode
) (XINT (first
) & 0377);
1060 start1
= (XINT (first
) >> 16) & 1;
1061 start2
= (XINT (first
) >> 17) & 1;
1062 end1
= (XINT (first
) >> 18) & 1;
1063 end2
= (XINT (first
) >> 19) & 1;
1064 prefix
= (XINT (first
) >> 20) & 1;
1065 comstyle
= (XINT (first
) >> 21) & 1;
1066 comnested
= (XINT (first
) >> 22) & 1;
1068 if ((int) code
< 0 || (int) code
>= (int) Smax
)
1070 insert_string ("invalid");
1073 desc
= syntax_code_spec
[(int) code
];
1075 str
[0] = desc
, str
[1] = 0;
1078 if (NILP (match_lisp
))
1081 insert_char (XINT (match_lisp
));
1100 insert_string ("\twhich means: ");
1102 switch (SWITCH_ENUM_CAST (code
))
1105 insert_string ("whitespace"); break;
1107 insert_string ("punctuation"); break;
1109 insert_string ("word"); break;
1111 insert_string ("symbol"); break;
1113 insert_string ("open"); break;
1115 insert_string ("close"); break;
1117 insert_string ("prefix"); break;
1119 insert_string ("string"); break;
1121 insert_string ("math"); break;
1123 insert_string ("escape"); break;
1125 insert_string ("charquote"); break;
1127 insert_string ("comment"); break;
1129 insert_string ("endcomment"); break;
1131 insert_string ("inherit"); break;
1132 case Scomment_fence
:
1133 insert_string ("comment fence"); break;
1135 insert_string ("string fence"); break;
1137 insert_string ("invalid");
1141 if (!NILP (match_lisp
))
1143 insert_string (", matches ");
1144 insert_char (XINT (match_lisp
));
1148 insert_string (",\n\t is the first character of a comment-start sequence");
1150 insert_string (",\n\t is the second character of a comment-start sequence");
1153 insert_string (",\n\t is the first character of a comment-end sequence");
1155 insert_string (",\n\t is the second character of a comment-end sequence");
1157 insert_string (" (comment style b)");
1159 insert_string (" (nestable)");
1162 insert_string (",\n\t is a prefix character for `backward-prefix-chars'");
1167 int parse_sexp_ignore_comments
;
1169 /* Char-table of functions that find the next or previous word
1171 Lisp_Object Vfind_word_boundary_function_table
;
1173 /* Return the position across COUNT words from FROM.
1174 If that many words cannot be found before the end of the buffer, return 0.
1175 COUNT negative means scan backward and stop at word beginning. */
1178 scan_words (register int from
, register int count
)
1180 register int beg
= BEGV
;
1181 register int end
= ZV
;
1182 register int from_byte
= CHAR_TO_BYTE (from
);
1183 register enum syntaxcode code
;
1185 Lisp_Object func
, script
, pos
;
1190 SETUP_SYNTAX_TABLE (from
, count
);
1201 UPDATE_SYNTAX_TABLE_FORWARD (from
);
1202 ch0
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
1203 code
= SYNTAX (ch0
);
1204 INC_BOTH (from
, from_byte
);
1205 if (words_include_escapes
1206 && (code
== Sescape
|| code
== Scharquote
))
1211 /* Now CH0 is a character which begins a word and FROM is the
1212 position of the next character. */
1213 func
= CHAR_TABLE_REF (Vfind_word_boundary_function_table
, ch0
);
1214 if (! NILP (Ffboundp (func
)))
1216 pos
= call2 (func
, make_number (from
- 1), make_number (end
));
1217 if (INTEGERP (pos
) && XINT (pos
) > from
)
1220 from_byte
= CHAR_TO_BYTE (from
);
1225 script
= CHAR_TABLE_REF (Vchar_script_table
, ch0
);
1228 if (from
== end
) break;
1229 UPDATE_SYNTAX_TABLE_FORWARD (from
);
1230 ch1
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
1231 code
= SYNTAX (ch1
);
1233 && (! words_include_escapes
1234 || (code
!= Sescape
&& code
!= Scharquote
)))
1235 || word_boundary_p (ch0
, ch1
))
1237 INC_BOTH (from
, from_byte
);
1252 DEC_BOTH (from
, from_byte
);
1253 UPDATE_SYNTAX_TABLE_BACKWARD (from
);
1254 ch1
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
1255 code
= SYNTAX (ch1
);
1256 if (words_include_escapes
1257 && (code
== Sescape
|| code
== Scharquote
))
1262 /* Now CH1 is a character which ends a word and FROM is the
1264 func
= CHAR_TABLE_REF (Vfind_word_boundary_function_table
, ch1
);
1265 if (! NILP (Ffboundp (func
)))
1267 pos
= call2 (func
, make_number (from
), make_number (beg
));
1268 if (INTEGERP (pos
) && XINT (pos
) < from
)
1271 from_byte
= CHAR_TO_BYTE (from
);
1276 script
= CHAR_TABLE_REF (Vchar_script_table
, ch1
);
1281 DEC_BOTH (from
, from_byte
);
1282 UPDATE_SYNTAX_TABLE_BACKWARD (from
);
1283 ch0
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
1284 code
= SYNTAX (ch0
);
1286 && (! words_include_escapes
1287 || (code
!= Sescape
&& code
!= Scharquote
)))
1288 || word_boundary_p (ch0
, ch1
))
1290 INC_BOTH (from
, from_byte
);
1304 DEFUN ("forward-word", Fforward_word
, Sforward_word
, 0, 1, "^p",
1305 doc
: /* Move point forward ARG words (backward if ARG is negative).
1307 If an edge of the buffer or a field boundary is reached, point is left there
1308 and the function returns nil. Field boundaries are not noticed if
1309 `inhibit-field-text-motion' is non-nil. */)
1317 XSETFASTINT (arg
, 1);
1321 val
= orig_val
= scan_words (PT
, XINT (arg
));
1323 val
= XINT (arg
) > 0 ? ZV
: BEGV
;
1325 /* Avoid jumping out of an input field. */
1326 tmp
= Fconstrain_to_field (make_number (val
), make_number (PT
),
1328 val
= XFASTINT (tmp
);
1331 return val
== orig_val
? Qt
: Qnil
;
1334 Lisp_Object
skip_chars (int, Lisp_Object
, Lisp_Object
, int);
1336 DEFUN ("skip-chars-forward", Fskip_chars_forward
, Sskip_chars_forward
, 1, 2, 0,
1337 doc
: /* Move point forward, stopping before a char not in STRING, or at pos LIM.
1338 STRING is like the inside of a `[...]' in a regular expression
1339 except that `]' is never special and `\\' quotes `^', `-' or `\\'
1340 (but not at the end of a range; quoting is never needed there).
1341 Thus, with arg "a-zA-Z", this skips letters stopping before first nonletter.
1342 With arg "^a-zA-Z", skips nonletters stopping before first letter.
1343 Char classes, e.g. `[:alpha:]', are supported.
1345 Returns the distance traveled, either zero or positive. */)
1347 Lisp_Object string
, lim
;
1349 return skip_chars (1, string
, lim
, 1);
1352 DEFUN ("skip-chars-backward", Fskip_chars_backward
, Sskip_chars_backward
, 1, 2, 0,
1353 doc
: /* Move point backward, stopping after a char not in STRING, or at pos LIM.
1354 See `skip-chars-forward' for details.
1355 Returns the distance traveled, either zero or negative. */)
1357 Lisp_Object string
, lim
;
1359 return skip_chars (0, string
, lim
, 1);
1362 DEFUN ("skip-syntax-forward", Fskip_syntax_forward
, Sskip_syntax_forward
, 1, 2, 0,
1363 doc
: /* Move point forward across chars in specified syntax classes.
1364 SYNTAX is a string of syntax code characters.
1365 Stop before a char whose syntax is not in SYNTAX, or at position LIM.
1366 If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX.
1367 This function returns the distance traveled, either zero or positive. */)
1369 Lisp_Object syntax
, lim
;
1371 return skip_syntaxes (1, syntax
, lim
);
1374 DEFUN ("skip-syntax-backward", Fskip_syntax_backward
, Sskip_syntax_backward
, 1, 2, 0,
1375 doc
: /* Move point backward across chars in specified syntax classes.
1376 SYNTAX is a string of syntax code characters.
1377 Stop on reaching a char whose syntax is not in SYNTAX, or at position LIM.
1378 If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX.
1379 This function returns the distance traveled, either zero or negative. */)
1381 Lisp_Object syntax
, lim
;
1383 return skip_syntaxes (0, syntax
, lim
);
1387 skip_chars (int forwardp
, Lisp_Object string
, Lisp_Object lim
, int handle_iso_classes
)
1389 register unsigned int c
;
1390 unsigned char fastmap
[0400];
1391 /* Store the ranges of non-ASCII characters. */
1393 int n_char_ranges
= 0;
1395 register int i
, i_byte
;
1396 /* Set to 1 if the current buffer is multibyte and the region
1397 contains non-ASCII chars. */
1399 /* Set to 1 if STRING is multibyte and it contains non-ASCII
1401 int string_multibyte
;
1403 const unsigned char *str
;
1405 Lisp_Object iso_classes
;
1407 CHECK_STRING (string
);
1411 XSETINT (lim
, forwardp
? ZV
: BEGV
);
1413 CHECK_NUMBER_COERCE_MARKER (lim
);
1415 /* In any case, don't allow scan outside bounds of buffer. */
1416 if (XINT (lim
) > ZV
)
1417 XSETFASTINT (lim
, ZV
);
1418 if (XINT (lim
) < BEGV
)
1419 XSETFASTINT (lim
, BEGV
);
1421 multibyte
= (!NILP (current_buffer
->enable_multibyte_characters
)
1422 && (XINT (lim
) - PT
!= CHAR_TO_BYTE (XINT (lim
)) - PT_BYTE
));
1423 string_multibyte
= SBYTES (string
) > SCHARS (string
);
1425 bzero (fastmap
, sizeof fastmap
);
1427 str
= SDATA (string
);
1428 size_byte
= SBYTES (string
);
1431 if (i_byte
< size_byte
1432 && SREF (string
, 0) == '^')
1434 negate
= 1; i_byte
++;
1437 /* Find the characters specified and set their elements of fastmap.
1438 Handle backslashes and ranges specially.
1440 If STRING contains non-ASCII characters, setup char_ranges for
1441 them and use fastmap only for their leading codes. */
1443 if (! string_multibyte
)
1445 int string_has_eight_bit
= 0;
1447 /* At first setup fastmap. */
1448 while (i_byte
< size_byte
)
1452 if (handle_iso_classes
&& c
== '['
1453 && i_byte
< size_byte
1454 && str
[i_byte
] == ':')
1456 const unsigned char *class_beg
= str
+ i_byte
+ 1;
1457 const unsigned char *class_end
= class_beg
;
1458 const unsigned char *class_limit
= str
+ size_byte
- 2;
1459 /* Leave room for the null. */
1460 unsigned char class_name
[CHAR_CLASS_MAX_LENGTH
+ 1];
1463 if (class_limit
- class_beg
> CHAR_CLASS_MAX_LENGTH
)
1464 class_limit
= class_beg
+ CHAR_CLASS_MAX_LENGTH
;
1466 while (class_end
< class_limit
1467 && *class_end
>= 'a' && *class_end
<= 'z')
1470 if (class_end
== class_beg
1471 || *class_end
!= ':' || class_end
[1] != ']')
1472 goto not_a_class_name
;
1474 bcopy (class_beg
, class_name
, class_end
- class_beg
);
1475 class_name
[class_end
- class_beg
] = 0;
1477 cc
= re_wctype (class_name
);
1479 error ("Invalid ISO C character class");
1481 iso_classes
= Fcons (make_number (cc
), iso_classes
);
1483 i_byte
= class_end
+ 2 - str
;
1490 if (i_byte
== size_byte
)
1495 /* Treat `-' as range character only if another character
1497 if (i_byte
+ 1 < size_byte
1498 && str
[i_byte
] == '-')
1502 /* Skip over the dash. */
1505 /* Get the end of the range. */
1508 && i_byte
< size_byte
)
1515 if (! ASCII_CHAR_P (c2
))
1516 string_has_eight_bit
= 1;
1522 if (! ASCII_CHAR_P (c
))
1523 string_has_eight_bit
= 1;
1527 /* If the current range is multibyte and STRING contains
1528 eight-bit chars, arrange fastmap and setup char_ranges for
1529 the corresponding multibyte chars. */
1530 if (multibyte
&& string_has_eight_bit
)
1532 unsigned char fastmap2
[0400];
1533 int range_start_byte
, range_start_char
;
1535 bcopy (fastmap2
+ 0200, fastmap
+ 0200, 0200);
1536 bzero (fastmap
+ 0200, 0200);
1537 /* We are sure that this loop stops. */
1538 for (i
= 0200; ! fastmap2
[i
]; i
++);
1539 c
= BYTE8_TO_CHAR (i
);
1540 fastmap
[CHAR_LEADING_CODE (c
)] = 1;
1541 range_start_byte
= i
;
1542 range_start_char
= c
;
1543 char_ranges
= (int *) alloca (sizeof (int) * 128 * 2);
1544 for (i
= 129; i
< 0400; i
++)
1546 c
= BYTE8_TO_CHAR (i
);
1547 fastmap
[CHAR_LEADING_CODE (c
)] = 1;
1548 if (i
- range_start_byte
!= c
- range_start_char
)
1550 char_ranges
[n_char_ranges
++] = range_start_char
;
1551 char_ranges
[n_char_ranges
++] = ((i
- 1 - range_start_byte
)
1552 + range_start_char
);
1553 range_start_byte
= i
;
1554 range_start_char
= c
;
1557 char_ranges
[n_char_ranges
++] = range_start_char
;
1558 char_ranges
[n_char_ranges
++] = ((i
- 1 - range_start_byte
)
1559 + range_start_char
);
1562 else /* STRING is multibyte */
1564 char_ranges
= (int *) alloca (sizeof (int) * SCHARS (string
) * 2);
1566 while (i_byte
< size_byte
)
1568 unsigned char leading_code
;
1570 leading_code
= str
[i_byte
];
1571 c
= STRING_CHAR_AND_LENGTH (str
+ i_byte
, len
);
1574 if (handle_iso_classes
&& c
== '['
1575 && i_byte
< size_byte
1576 && STRING_CHAR (str
+ i_byte
) == ':')
1578 const unsigned char *class_beg
= str
+ i_byte
+ 1;
1579 const unsigned char *class_end
= class_beg
;
1580 const unsigned char *class_limit
= str
+ size_byte
- 2;
1581 /* Leave room for the null. */
1582 unsigned char class_name
[CHAR_CLASS_MAX_LENGTH
+ 1];
1585 if (class_limit
- class_beg
> CHAR_CLASS_MAX_LENGTH
)
1586 class_limit
= class_beg
+ CHAR_CLASS_MAX_LENGTH
;
1588 while (class_end
< class_limit
1589 && *class_end
>= 'a' && *class_end
<= 'z')
1592 if (class_end
== class_beg
1593 || *class_end
!= ':' || class_end
[1] != ']')
1594 goto not_a_class_name_multibyte
;
1596 bcopy (class_beg
, class_name
, class_end
- class_beg
);
1597 class_name
[class_end
- class_beg
] = 0;
1599 cc
= re_wctype (class_name
);
1601 error ("Invalid ISO C character class");
1603 iso_classes
= Fcons (make_number (cc
), iso_classes
);
1605 i_byte
= class_end
+ 2 - str
;
1609 not_a_class_name_multibyte
:
1612 if (i_byte
== size_byte
)
1615 leading_code
= str
[i_byte
];
1616 c
= STRING_CHAR_AND_LENGTH (str
+ i_byte
, len
);
1619 /* Treat `-' as range character only if another character
1621 if (i_byte
+ 1 < size_byte
1622 && str
[i_byte
] == '-')
1625 unsigned char leading_code2
;
1627 /* Skip over the dash. */
1630 /* Get the end of the range. */
1631 leading_code2
= str
[i_byte
];
1632 c2
= STRING_CHAR_AND_LENGTH (str
+ i_byte
, len
);
1636 && i_byte
< size_byte
)
1638 leading_code2
= str
[i_byte
];
1639 c2
=STRING_CHAR_AND_LENGTH (str
+ i_byte
, len
);
1645 if (ASCII_CHAR_P (c
))
1647 while (c
<= c2
&& c
< 0x80)
1649 leading_code
= CHAR_LEADING_CODE (c
);
1651 if (! ASCII_CHAR_P (c
))
1653 while (leading_code
<= leading_code2
)
1654 fastmap
[leading_code
++] = 1;
1657 char_ranges
[n_char_ranges
++] = c
;
1658 char_ranges
[n_char_ranges
++] = c2
;
1664 if (ASCII_CHAR_P (c
))
1668 fastmap
[leading_code
] = 1;
1669 char_ranges
[n_char_ranges
++] = c
;
1670 char_ranges
[n_char_ranges
++] = c
;
1675 /* If the current range is unibyte and STRING contains non-ASCII
1676 chars, arrange fastmap for the corresponding unibyte
1679 if (! multibyte
&& n_char_ranges
> 0)
1681 bzero (fastmap
+ 0200, 0200);
1682 for (i
= 0; i
< n_char_ranges
; i
+= 2)
1684 int c1
= char_ranges
[i
];
1685 int c2
= char_ranges
[i
+ 1];
1687 for (; c1
<= c2
; c1
++)
1689 int b
= CHAR_TO_BYTE_SAFE (c1
);
1697 /* If ^ was the first character, complement the fastmap. */
1701 for (i
= 0; i
< sizeof fastmap
; i
++)
1705 for (i
= 0; i
< 0200; i
++)
1707 /* All non-ASCII chars possibly match. */
1708 for (; i
< sizeof fastmap
; i
++)
1714 int start_point
= PT
;
1716 int pos_byte
= PT_BYTE
;
1717 unsigned char *p
= PT_ADDR
, *endp
, *stop
;
1721 endp
= (XINT (lim
) == GPT
) ? GPT_ADDR
: CHAR_POS_ADDR (XINT (lim
));
1722 stop
= (pos
< GPT
&& GPT
< XINT (lim
)) ? GPT_ADDR
: endp
;
1726 endp
= CHAR_POS_ADDR (XINT (lim
));
1727 stop
= (pos
>= GPT
&& GPT
> XINT (lim
)) ? GAP_END_ADDR
: endp
;
1731 /* This code may look up syntax tables using macros that rely on the
1732 gl_state object. To make sure this object is not out of date,
1733 let's initialize it manually.
1734 We ignore syntax-table text-properties for now, since that's
1735 what we've done in the past. */
1736 SETUP_BUFFER_SYNTAX_TABLE ();
1751 c
= STRING_CHAR_AND_LENGTH (p
, nbytes
);
1752 if (! NILP (iso_classes
) && in_classes (c
, iso_classes
))
1762 if (! ASCII_CHAR_P (c
))
1764 /* As we are looking at a multibyte character, we
1765 must look up the character in the table
1766 CHAR_RANGES. If there's no data in the table,
1767 that character is not what we want to skip. */
1769 /* The following code do the right thing even if
1770 n_char_ranges is zero (i.e. no data in
1772 for (i
= 0; i
< n_char_ranges
; i
+= 2)
1773 if (c
>= char_ranges
[i
] && c
<= char_ranges
[i
+ 1])
1775 if (!(negate
^ (i
< n_char_ranges
)))
1779 p
+= nbytes
, pos
++, pos_byte
+= nbytes
;
1792 if (!NILP (iso_classes
) && in_classes (*p
, iso_classes
))
1797 goto fwd_unibyte_ok
;
1803 p
++, pos
++, pos_byte
++;
1811 unsigned char *prev_p
;
1821 while (--p
>= stop
&& ! CHAR_HEAD_P (*p
));
1822 c
= STRING_CHAR (p
);
1824 if (! NILP (iso_classes
) && in_classes (c
, iso_classes
))
1834 if (! ASCII_CHAR_P (c
))
1836 /* See the comment in the previous similar code. */
1837 for (i
= 0; i
< n_char_ranges
; i
+= 2)
1838 if (c
>= char_ranges
[i
] && c
<= char_ranges
[i
+ 1])
1840 if (!(negate
^ (i
< n_char_ranges
)))
1844 pos
--, pos_byte
-= prev_p
- p
;
1857 if (! NILP (iso_classes
) && in_classes (p
[-1], iso_classes
))
1862 goto back_unibyte_ok
;
1865 if (!fastmap
[p
[-1]])
1868 p
--, pos
--, pos_byte
--;
1872 SET_PT_BOTH (pos
, pos_byte
);
1875 return make_number (PT
- start_point
);
1881 skip_syntaxes (int forwardp
, Lisp_Object string
, Lisp_Object lim
)
1883 register unsigned int c
;
1884 unsigned char fastmap
[0400];
1886 register int i
, i_byte
;
1891 CHECK_STRING (string
);
1894 XSETINT (lim
, forwardp
? ZV
: BEGV
);
1896 CHECK_NUMBER_COERCE_MARKER (lim
);
1898 /* In any case, don't allow scan outside bounds of buffer. */
1899 if (XINT (lim
) > ZV
)
1900 XSETFASTINT (lim
, ZV
);
1901 if (XINT (lim
) < BEGV
)
1902 XSETFASTINT (lim
, BEGV
);
1904 if (forwardp
? (PT
>= XFASTINT (lim
)) : (PT
<= XFASTINT (lim
)))
1905 return make_number (0);
1907 multibyte
= (!NILP (current_buffer
->enable_multibyte_characters
)
1908 && (XINT (lim
) - PT
!= CHAR_TO_BYTE (XINT (lim
)) - PT_BYTE
));
1910 bzero (fastmap
, sizeof fastmap
);
1912 if (SBYTES (string
) > SCHARS (string
))
1913 /* As this is very rare case (syntax spec is ASCII only), don't
1914 consider efficiency. */
1915 string
= string_make_unibyte (string
);
1917 str
= SDATA (string
);
1918 size_byte
= SBYTES (string
);
1921 if (i_byte
< size_byte
1922 && SREF (string
, 0) == '^')
1924 negate
= 1; i_byte
++;
1927 /* Find the syntaxes specified and set their elements of fastmap. */
1929 while (i_byte
< size_byte
)
1932 fastmap
[syntax_spec_code
[c
]] = 1;
1935 /* If ^ was the first character, complement the fastmap. */
1937 for (i
= 0; i
< sizeof fastmap
; i
++)
1941 int start_point
= PT
;
1943 int pos_byte
= PT_BYTE
;
1944 unsigned char *p
= PT_ADDR
, *endp
, *stop
;
1948 endp
= (XINT (lim
) == GPT
) ? GPT_ADDR
: CHAR_POS_ADDR (XINT (lim
));
1949 stop
= (pos
< GPT
&& GPT
< XINT (lim
)) ? GPT_ADDR
: endp
;
1953 endp
= CHAR_POS_ADDR (XINT (lim
));
1954 stop
= (pos
>= GPT
&& GPT
> XINT (lim
)) ? GAP_END_ADDR
: endp
;
1958 SETUP_SYNTAX_TABLE (pos
, forwardp
? 1 : -1);
1974 c
= STRING_CHAR_AND_LENGTH (p
, nbytes
);
1975 if (! fastmap
[(int) SYNTAX (c
)])
1977 p
+= nbytes
, pos
++, pos_byte
+= nbytes
;
1978 UPDATE_SYNTAX_TABLE_FORWARD (pos
);
1992 if (! fastmap
[(int) SYNTAX (*p
)])
1994 p
++, pos
++, pos_byte
++;
1995 UPDATE_SYNTAX_TABLE_FORWARD (pos
);
2005 unsigned char *prev_p
;
2014 UPDATE_SYNTAX_TABLE_BACKWARD (pos
- 1);
2016 while (--p
>= stop
&& ! CHAR_HEAD_P (*p
));
2017 c
= STRING_CHAR (p
);
2018 if (! fastmap
[(int) SYNTAX (c
)])
2020 pos
--, pos_byte
-= prev_p
- p
;
2034 UPDATE_SYNTAX_TABLE_BACKWARD (pos
- 1);
2035 if (! fastmap
[(int) SYNTAX (p
[-1])])
2037 p
--, pos
--, pos_byte
--;
2042 SET_PT_BOTH (pos
, pos_byte
);
2045 return make_number (PT
- start_point
);
2049 /* Return 1 if character C belongs to one of the ISO classes
2050 in the list ISO_CLASSES. Each class is represented by an
2051 integer which is its type according to re_wctype. */
2054 in_classes (int c
, Lisp_Object iso_classes
)
2058 while (CONSP (iso_classes
))
2061 elt
= XCAR (iso_classes
);
2062 iso_classes
= XCDR (iso_classes
);
2064 if (re_iswctype (c
, XFASTINT (elt
)))
2071 /* Jump over a comment, assuming we are at the beginning of one.
2072 FROM is the current position.
2073 FROM_BYTE is the bytepos corresponding to FROM.
2074 Do not move past STOP (a charpos).
2075 The comment over which we have to jump is of style STYLE
2076 (either SYNTAX_COMMENT_STYLE(foo) or ST_COMMENT_STYLE).
2077 NESTING should be positive to indicate the nesting at the beginning
2078 for nested comments and should be zero or negative else.
2079 ST_COMMENT_STYLE cannot be nested.
2080 PREV_SYNTAX is the SYNTAX_WITH_FLAGS of the previous character
2081 (or 0 If the search cannot start in the middle of a two-character).
2083 If successful, return 1 and store the charpos of the comment's end
2084 into *CHARPOS_PTR and the corresponding bytepos into *BYTEPOS_PTR.
2085 Else, return 0 and store the charpos STOP into *CHARPOS_PTR, the
2086 corresponding bytepos into *BYTEPOS_PTR and the current nesting
2087 (as defined for state.incomment) in *INCOMMENT_PTR.
2089 The comment end is the last character of the comment rather than the
2090 character just after the comment.
2092 Global syntax data is assumed to initially be valid for FROM and
2093 remains valid for forward search starting at the returned position. */
2096 forw_comment (EMACS_INT from
, EMACS_INT from_byte
, EMACS_INT stop
,
2097 int nesting
, int style
, int prev_syntax
,
2098 EMACS_INT
*charpos_ptr
, EMACS_INT
*bytepos_ptr
,
2102 register enum syntaxcode code
;
2103 register int syntax
;
2105 if (nesting
<= 0) nesting
= -1;
2107 /* Enter the loop in the middle so that we find
2108 a 2-char comment ender if we start in the middle of it. */
2109 syntax
= prev_syntax
;
2110 if (syntax
!= 0) goto forw_incomment
;
2116 *incomment_ptr
= nesting
;
2117 *charpos_ptr
= from
;
2118 *bytepos_ptr
= from_byte
;
2121 c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
2122 syntax
= SYNTAX_WITH_FLAGS (c
);
2123 code
= syntax
& 0xff;
2124 if (code
== Sendcomment
2125 && SYNTAX_FLAGS_COMMENT_STYLE (syntax
) == style
2126 && (SYNTAX_FLAGS_COMMENT_NESTED (syntax
) ?
2127 (nesting
> 0 && --nesting
== 0) : nesting
< 0))
2128 /* we have encountered a comment end of the same style
2129 as the comment sequence which began this comment
2132 if (code
== Scomment_fence
2133 && style
== ST_COMMENT_STYLE
)
2134 /* we have encountered a comment end of the same style
2135 as the comment sequence which began this comment
2140 && SYNTAX_FLAGS_COMMENT_NESTED (syntax
)
2141 && SYNTAX_FLAGS_COMMENT_STYLE (syntax
) == style
)
2142 /* we have encountered a nested comment of the same style
2143 as the comment sequence which began this comment section */
2145 INC_BOTH (from
, from_byte
);
2146 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2149 if (from
< stop
&& SYNTAX_FLAGS_COMEND_FIRST (syntax
)
2150 && SYNTAX_FLAGS_COMMENT_STYLE (syntax
) == style
2151 && (c1
= FETCH_CHAR_AS_MULTIBYTE (from_byte
),
2152 SYNTAX_COMEND_SECOND (c1
))
2153 && ((SYNTAX_FLAGS_COMMENT_NESTED (syntax
) ||
2154 SYNTAX_COMMENT_NESTED (c1
)) ? nesting
> 0 : nesting
< 0))
2157 /* we have encountered a comment end of the same style
2158 as the comment sequence which began this comment
2163 INC_BOTH (from
, from_byte
);
2164 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2169 && SYNTAX_FLAGS_COMSTART_FIRST (syntax
)
2170 && (c1
= FETCH_CHAR_AS_MULTIBYTE (from_byte
),
2171 SYNTAX_COMMENT_STYLE (c1
) == style
2172 && SYNTAX_COMSTART_SECOND (c1
))
2173 && (SYNTAX_FLAGS_COMMENT_NESTED (syntax
) ||
2174 SYNTAX_COMMENT_NESTED (c1
)))
2175 /* we have encountered a nested comment of the same style
2176 as the comment sequence which began this comment
2179 INC_BOTH (from
, from_byte
);
2180 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2184 *charpos_ptr
= from
;
2185 *bytepos_ptr
= from_byte
;
2189 DEFUN ("forward-comment", Fforward_comment
, Sforward_comment
, 1, 1, 0,
2191 Move forward across up to COUNT comments. If COUNT is negative, move backward.
2192 Stop scanning if we find something other than a comment or whitespace.
2193 Set point to where scanning stops.
2194 If COUNT comments are found as expected, with nothing except whitespace
2195 between them, return t; otherwise return nil. */)
2199 register EMACS_INT from
;
2200 EMACS_INT from_byte
;
2201 register EMACS_INT stop
;
2203 register enum syntaxcode code
;
2204 int comstyle
= 0; /* style of comment encountered */
2205 int comnested
= 0; /* whether the comment is nestable or not */
2208 EMACS_INT out_charpos
, out_bytepos
;
2211 CHECK_NUMBER (count
);
2212 count1
= XINT (count
);
2213 stop
= count1
> 0 ? ZV
: BEGV
;
2219 from_byte
= PT_BYTE
;
2221 SETUP_SYNTAX_TABLE (from
, count1
);
2230 SET_PT_BOTH (from
, from_byte
);
2234 c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
2236 comstart_first
= SYNTAX_COMSTART_FIRST (c
);
2237 comnested
= SYNTAX_COMMENT_NESTED (c
);
2238 comstyle
= SYNTAX_COMMENT_STYLE (c
);
2239 INC_BOTH (from
, from_byte
);
2240 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2241 if (from
< stop
&& comstart_first
2242 && (c1
= FETCH_CHAR_AS_MULTIBYTE (from_byte
),
2243 SYNTAX_COMSTART_SECOND (c1
)))
2245 /* We have encountered a comment start sequence and we
2246 are ignoring all text inside comments. We must record
2247 the comment style this sequence begins so that later,
2248 only a comment end of the same style actually ends
2249 the comment section. */
2251 comstyle
= SYNTAX_COMMENT_STYLE (c1
);
2252 comnested
= comnested
|| SYNTAX_COMMENT_NESTED (c1
);
2253 INC_BOTH (from
, from_byte
);
2254 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2257 while (code
== Swhitespace
|| (code
== Sendcomment
&& c
== '\n'));
2259 if (code
== Scomment_fence
)
2260 comstyle
= ST_COMMENT_STYLE
;
2261 else if (code
!= Scomment
)
2264 DEC_BOTH (from
, from_byte
);
2265 SET_PT_BOTH (from
, from_byte
);
2268 /* We're at the start of a comment. */
2269 found
= forw_comment (from
, from_byte
, stop
, comnested
, comstyle
, 0,
2270 &out_charpos
, &out_bytepos
, &dummy
);
2271 from
= out_charpos
; from_byte
= out_bytepos
;
2275 SET_PT_BOTH (from
, from_byte
);
2278 INC_BOTH (from
, from_byte
);
2279 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2280 /* We have skipped one comment. */
2292 SET_PT_BOTH (BEGV
, BEGV_BYTE
);
2297 DEC_BOTH (from
, from_byte
);
2298 /* char_quoted does UPDATE_SYNTAX_TABLE_BACKWARD (from). */
2299 quoted
= char_quoted (from
, from_byte
);
2300 c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
2303 comnested
= SYNTAX_COMMENT_NESTED (c
);
2304 if (code
== Sendcomment
)
2305 comstyle
= SYNTAX_COMMENT_STYLE (c
);
2306 if (from
> stop
&& SYNTAX_COMEND_SECOND (c
)
2307 && prev_char_comend_first (from
, from_byte
)
2308 && !char_quoted (from
- 1, dec_bytepos (from_byte
)))
2310 /* We must record the comment style encountered so that
2311 later, we can match only the proper comment begin
2312 sequence of the same style. */
2313 DEC_BOTH (from
, from_byte
);
2315 /* Calling char_quoted, above, set up global syntax position
2316 at the new value of FROM. */
2317 c1
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
2318 comstyle
= SYNTAX_COMMENT_STYLE (c1
);
2319 comnested
= comnested
|| SYNTAX_COMMENT_NESTED (c1
);
2322 if (code
== Scomment_fence
)
2324 /* Skip until first preceding unquoted comment_fence. */
2325 int found
= 0, ini
= from
, ini_byte
= from_byte
;
2329 DEC_BOTH (from
, from_byte
);
2330 UPDATE_SYNTAX_TABLE_BACKWARD (from
);
2331 c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
2332 if (SYNTAX (c
) == Scomment_fence
2333 && !char_quoted (from
, from_byte
))
2338 else if (from
== stop
)
2343 from
= ini
; /* Set point to ini + 1. */
2344 from_byte
= ini_byte
;
2348 /* We have skipped one comment. */
2351 else if (code
== Sendcomment
)
2353 found
= back_comment (from
, from_byte
, stop
, comnested
, comstyle
,
2354 &out_charpos
, &out_bytepos
);
2358 /* This end-of-line is not an end-of-comment.
2359 Treat it like a whitespace.
2360 CC-mode (and maybe others) relies on this behavior. */
2364 /* Failure: we should go back to the end of this
2365 not-quite-endcomment. */
2366 if (SYNTAX(c
) != code
)
2367 /* It was a two-char Sendcomment. */
2368 INC_BOTH (from
, from_byte
);
2374 /* We have skipped one comment. */
2375 from
= out_charpos
, from_byte
= out_bytepos
;
2379 else if (code
!= Swhitespace
|| quoted
)
2383 INC_BOTH (from
, from_byte
);
2384 SET_PT_BOTH (from
, from_byte
);
2392 SET_PT_BOTH (from
, from_byte
);
2397 /* Return syntax code of character C if C is an ASCII character
2398 or `multibyte_symbol_p' is zero. Otherwise, return Ssymbol. */
2400 #define SYNTAX_WITH_MULTIBYTE_CHECK(c) \
2401 ((ASCII_CHAR_P (c) || !multibyte_symbol_p) \
2402 ? SYNTAX (c) : Ssymbol)
2405 scan_lists (register EMACS_INT from
, EMACS_INT count
, EMACS_INT depth
, int sexpflag
)
2408 register EMACS_INT stop
= count
> 0 ? ZV
: BEGV
;
2413 register enum syntaxcode code
, temp_code
;
2414 int min_depth
= depth
; /* Err out if depth gets less than this. */
2415 int comstyle
= 0; /* style of comment encountered */
2416 int comnested
= 0; /* whether the comment is nestable or not */
2418 EMACS_INT last_good
= from
;
2420 EMACS_INT from_byte
;
2421 EMACS_INT out_bytepos
, out_charpos
;
2423 int multibyte_symbol_p
= sexpflag
&& multibyte_syntax_as_symbol
;
2425 if (depth
> 0) min_depth
= 0;
2427 if (from
> ZV
) from
= ZV
;
2428 if (from
< BEGV
) from
= BEGV
;
2430 from_byte
= CHAR_TO_BYTE (from
);
2435 SETUP_SYNTAX_TABLE (from
, count
);
2440 int comstart_first
, prefix
;
2441 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2442 c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
2443 code
= SYNTAX_WITH_MULTIBYTE_CHECK (c
);
2444 comstart_first
= SYNTAX_COMSTART_FIRST (c
);
2445 comnested
= SYNTAX_COMMENT_NESTED (c
);
2446 comstyle
= SYNTAX_COMMENT_STYLE (c
);
2447 prefix
= SYNTAX_PREFIX (c
);
2448 if (depth
== min_depth
)
2450 INC_BOTH (from
, from_byte
);
2451 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2452 if (from
< stop
&& comstart_first
2453 && (c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
),
2454 SYNTAX_COMSTART_SECOND (c
))
2455 && parse_sexp_ignore_comments
)
2457 /* we have encountered a comment start sequence and we
2458 are ignoring all text inside comments. We must record
2459 the comment style this sequence begins so that later,
2460 only a comment end of the same style actually ends
2461 the comment section */
2463 c1
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
2464 comstyle
= SYNTAX_COMMENT_STYLE (c1
);
2465 comnested
= comnested
|| SYNTAX_COMMENT_NESTED (c1
);
2466 INC_BOTH (from
, from_byte
);
2467 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2473 switch (SWITCH_ENUM_CAST (code
))
2479 INC_BOTH (from
, from_byte
);
2480 /* treat following character as a word constituent */
2483 if (depth
|| !sexpflag
) break;
2484 /* This word counts as a sexp; return at end of it. */
2487 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2489 /* Some compilers can't handle this inside the switch. */
2490 c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
2491 temp
= SYNTAX_WITH_MULTIBYTE_CHECK (c
);
2496 INC_BOTH (from
, from_byte
);
2507 INC_BOTH (from
, from_byte
);
2511 case Scomment_fence
:
2512 comstyle
= ST_COMMENT_STYLE
;
2515 if (!parse_sexp_ignore_comments
) break;
2516 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2517 found
= forw_comment (from
, from_byte
, stop
,
2518 comnested
, comstyle
, 0,
2519 &out_charpos
, &out_bytepos
, &dummy
);
2520 from
= out_charpos
, from_byte
= out_bytepos
;
2527 INC_BOTH (from
, from_byte
);
2528 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2534 if (from
!= stop
&& c
== FETCH_CHAR_AS_MULTIBYTE (from_byte
))
2536 INC_BOTH (from
, from_byte
);
2546 if (!++depth
) goto done
;
2551 if (!--depth
) goto done
;
2552 if (depth
< min_depth
)
2553 xsignal3 (Qscan_error
,
2554 build_string ("Containing expression ends prematurely"),
2555 make_number (last_good
), make_number (from
));
2560 temp_pos
= dec_bytepos (from_byte
);
2561 stringterm
= FETCH_CHAR_AS_MULTIBYTE (temp_pos
);
2566 UPDATE_SYNTAX_TABLE_FORWARD (from
);
2567 c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
2570 && SYNTAX_WITH_MULTIBYTE_CHECK (c
) == Sstring
)
2571 : SYNTAX_WITH_MULTIBYTE_CHECK (c
) == Sstring_fence
)
2574 /* Some compilers can't handle this inside the switch. */
2575 temp
= SYNTAX_WITH_MULTIBYTE_CHECK (c
);
2580 INC_BOTH (from
, from_byte
);
2582 INC_BOTH (from
, from_byte
);
2584 INC_BOTH (from
, from_byte
);
2585 if (!depth
&& sexpflag
) goto done
;
2588 /* Ignore whitespace, punctuation, quote, endcomment. */
2593 /* Reached end of buffer. Error if within object, return nil if between */
2600 /* End of object reached */
2609 DEC_BOTH (from
, from_byte
);
2610 UPDATE_SYNTAX_TABLE_BACKWARD (from
);
2611 c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
2612 code
= SYNTAX_WITH_MULTIBYTE_CHECK (c
);
2613 if (depth
== min_depth
)
2616 comnested
= SYNTAX_COMMENT_NESTED (c
);
2617 if (code
== Sendcomment
)
2618 comstyle
= SYNTAX_COMMENT_STYLE (c
);
2619 if (from
> stop
&& SYNTAX_COMEND_SECOND (c
)
2620 && prev_char_comend_first (from
, from_byte
)
2621 && parse_sexp_ignore_comments
)
2623 /* We must record the comment style encountered so that
2624 later, we can match only the proper comment begin
2625 sequence of the same style. */
2626 DEC_BOTH (from
, from_byte
);
2627 UPDATE_SYNTAX_TABLE_BACKWARD (from
);
2629 c1
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
2630 comstyle
= SYNTAX_COMMENT_STYLE (c1
);
2631 comnested
= comnested
|| SYNTAX_COMMENT_NESTED (c1
);
2634 /* Quoting turns anything except a comment-ender
2635 into a word character. Note that this cannot be true
2636 if we decremented FROM in the if-statement above. */
2637 if (code
!= Sendcomment
&& char_quoted (from
, from_byte
))
2639 DEC_BOTH (from
, from_byte
);
2642 else if (SYNTAX_PREFIX (c
))
2645 switch (SWITCH_ENUM_CAST (code
))
2651 if (depth
|| !sexpflag
) break;
2652 /* This word counts as a sexp; count object finished
2653 after passing it. */
2656 temp_pos
= from_byte
;
2657 if (! NILP (current_buffer
->enable_multibyte_characters
))
2661 UPDATE_SYNTAX_TABLE_BACKWARD (from
- 1);
2662 c1
= FETCH_CHAR_AS_MULTIBYTE (temp_pos
);
2663 temp_code
= SYNTAX_WITH_MULTIBYTE_CHECK (c1
);
2664 /* Don't allow comment-end to be quoted. */
2665 if (temp_code
== Sendcomment
)
2667 quoted
= char_quoted (from
- 1, temp_pos
);
2670 DEC_BOTH (from
, from_byte
);
2671 temp_pos
= dec_bytepos (temp_pos
);
2672 UPDATE_SYNTAX_TABLE_BACKWARD (from
- 1);
2674 c1
= FETCH_CHAR_AS_MULTIBYTE (temp_pos
);
2675 temp_code
= SYNTAX_WITH_MULTIBYTE_CHECK (c1
);
2676 if (! (quoted
|| temp_code
== Sword
2677 || temp_code
== Ssymbol
2678 || temp_code
== Squote
))
2680 DEC_BOTH (from
, from_byte
);
2687 temp_pos
= dec_bytepos (from_byte
);
2688 UPDATE_SYNTAX_TABLE_BACKWARD (from
- 1);
2689 if (from
!= stop
&& c
== FETCH_CHAR_AS_MULTIBYTE (temp_pos
))
2690 DEC_BOTH (from
, from_byte
);
2699 if (!++depth
) goto done2
;
2704 if (!--depth
) goto done2
;
2705 if (depth
< min_depth
)
2706 xsignal3 (Qscan_error
,
2707 build_string ("Containing expression ends prematurely"),
2708 make_number (last_good
), make_number (from
));
2712 if (!parse_sexp_ignore_comments
)
2714 found
= back_comment (from
, from_byte
, stop
, comnested
, comstyle
,
2715 &out_charpos
, &out_bytepos
);
2716 /* FIXME: if found == -1, then it really wasn't a comment-end.
2717 For single-char Sendcomment, we can't do much about it apart
2718 from skipping the char.
2719 For 2-char endcomments, we could try again, taking both
2720 chars as separate entities, but it's a lot of trouble
2721 for very little gain, so we don't bother either. -sm */
2723 from
= out_charpos
, from_byte
= out_bytepos
;
2726 case Scomment_fence
:
2732 DEC_BOTH (from
, from_byte
);
2733 UPDATE_SYNTAX_TABLE_BACKWARD (from
);
2734 if (!char_quoted (from
, from_byte
)
2735 && (c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
),
2736 SYNTAX_WITH_MULTIBYTE_CHECK (c
) == code
))
2739 if (code
== Sstring_fence
&& !depth
&& sexpflag
) goto done2
;
2743 stringterm
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
2748 DEC_BOTH (from
, from_byte
);
2749 UPDATE_SYNTAX_TABLE_BACKWARD (from
);
2750 if (!char_quoted (from
, from_byte
)
2752 == (c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
)))
2753 && SYNTAX_WITH_MULTIBYTE_CHECK (c
) == Sstring
)
2756 if (!depth
&& sexpflag
) goto done2
;
2759 /* Ignore whitespace, punctuation, quote, endcomment. */
2764 /* Reached start of buffer. Error if within object, return nil if between */
2777 XSETFASTINT (val
, from
);
2781 xsignal3 (Qscan_error
,
2782 build_string ("Unbalanced parentheses"),
2783 make_number (last_good
), make_number (from
));
2786 DEFUN ("scan-lists", Fscan_lists
, Sscan_lists
, 3, 3, 0,
2787 doc
: /* Scan from character number FROM by COUNT lists.
2788 Returns the character number of the position thus found.
2790 If DEPTH is nonzero, paren depth begins counting from that value,
2791 only places where the depth in parentheses becomes zero
2792 are candidates for stopping; COUNT such places are counted.
2793 Thus, a positive value for DEPTH means go out levels.
2795 Comments are ignored if `parse-sexp-ignore-comments' is non-nil.
2797 If the beginning or end of (the accessible part of) the buffer is reached
2798 and the depth is wrong, an error is signaled.
2799 If the depth is right but the count is not used up, nil is returned. */)
2800 (from
, count
, depth
)
2801 Lisp_Object from
, count
, depth
;
2803 CHECK_NUMBER (from
);
2804 CHECK_NUMBER (count
);
2805 CHECK_NUMBER (depth
);
2807 return scan_lists (XINT (from
), XINT (count
), XINT (depth
), 0);
2810 DEFUN ("scan-sexps", Fscan_sexps
, Sscan_sexps
, 2, 2, 0,
2811 doc
: /* Scan from character number FROM by COUNT balanced expressions.
2812 If COUNT is negative, scan backwards.
2813 Returns the character number of the position thus found.
2815 Comments are ignored if `parse-sexp-ignore-comments' is non-nil.
2817 If the beginning or end of (the accessible part of) the buffer is reached
2818 in the middle of a parenthetical grouping, an error is signaled.
2819 If the beginning or end is reached between groupings
2820 but before count is used up, nil is returned. */)
2822 Lisp_Object from
, count
;
2824 CHECK_NUMBER (from
);
2825 CHECK_NUMBER (count
);
2827 return scan_lists (XINT (from
), XINT (count
), 0, 1);
2830 DEFUN ("backward-prefix-chars", Fbackward_prefix_chars
, Sbackward_prefix_chars
,
2832 doc
: /* Move point backward over any number of chars with prefix syntax.
2833 This includes chars with "quote" or "prefix" syntax (' or p). */)
2838 int opoint_byte
= PT_BYTE
;
2840 int pos_byte
= PT_BYTE
;
2845 SET_PT_BOTH (opoint
, opoint_byte
);
2850 SETUP_SYNTAX_TABLE (pos
, -1);
2852 DEC_BOTH (pos
, pos_byte
);
2854 while (!char_quoted (pos
, pos_byte
)
2855 /* Previous statement updates syntax table. */
2856 && ((c
= FETCH_CHAR_AS_MULTIBYTE (pos_byte
), SYNTAX (c
) == Squote
)
2857 || SYNTAX_PREFIX (c
)))
2860 opoint_byte
= pos_byte
;
2863 DEC_BOTH (pos
, pos_byte
);
2866 SET_PT_BOTH (opoint
, opoint_byte
);
2871 /* Parse forward from FROM / FROM_BYTE to END,
2872 assuming that FROM has state OLDSTATE (nil means FROM is start of function),
2873 and return a description of the state of the parse at END.
2874 If STOPBEFORE is nonzero, stop at the start of an atom.
2875 If COMMENTSTOP is 1, stop at the start of a comment.
2876 If COMMENTSTOP is -1, stop at the start or end of a comment,
2877 after the beginning of a string, or after the end of a string. */
2880 scan_sexps_forward (stateptr
, from
, from_byte
, end
, targetdepth
,
2881 stopbefore
, oldstate
, commentstop
)
2882 struct lisp_parse_state
*stateptr
;
2883 register EMACS_INT from
;
2884 EMACS_INT from_byte
, end
;
2885 int targetdepth
, stopbefore
;
2886 Lisp_Object oldstate
;
2889 struct lisp_parse_state state
;
2891 register enum syntaxcode code
;
2894 struct level
{ int last
, prev
; };
2895 struct level levelstart
[100];
2896 register struct level
*curlevel
= levelstart
;
2897 struct level
*endlevel
= levelstart
+ 100;
2898 register int depth
; /* Paren depth of current scanning location.
2899 level - levelstart equals this except
2900 when the depth becomes negative. */
2901 int mindepth
; /* Lowest DEPTH value seen. */
2902 int start_quoted
= 0; /* Nonzero means starting after a char quote */
2904 EMACS_INT prev_from
; /* Keep one character before FROM. */
2905 EMACS_INT prev_from_byte
;
2906 int prev_from_syntax
;
2907 int boundary_stop
= commentstop
== -1;
2910 EMACS_INT out_bytepos
, out_charpos
;
2914 prev_from_byte
= from_byte
;
2916 DEC_BOTH (prev_from
, prev_from_byte
);
2918 /* Use this macro instead of `from++'. */
2920 do { prev_from = from; \
2921 prev_from_byte = from_byte; \
2922 temp = FETCH_CHAR_AS_MULTIBYTE (prev_from_byte); \
2923 prev_from_syntax = SYNTAX_WITH_FLAGS (temp); \
2924 INC_BOTH (from, from_byte); \
2926 UPDATE_SYNTAX_TABLE_FORWARD (from); \
2932 if (NILP (oldstate
))
2935 state
.instring
= -1;
2936 state
.incomment
= 0;
2937 state
.comstyle
= 0; /* comment style a by default. */
2938 state
.comstr_start
= -1; /* no comment/string seen. */
2942 tem
= Fcar (oldstate
);
2948 oldstate
= Fcdr (oldstate
);
2949 oldstate
= Fcdr (oldstate
);
2950 oldstate
= Fcdr (oldstate
);
2951 tem
= Fcar (oldstate
);
2952 /* Check whether we are inside string_fence-style string: */
2953 state
.instring
= (!NILP (tem
)
2954 ? (INTEGERP (tem
) ? XINT (tem
) : ST_STRING_STYLE
)
2957 oldstate
= Fcdr (oldstate
);
2958 tem
= Fcar (oldstate
);
2959 state
.incomment
= (!NILP (tem
)
2960 ? (INTEGERP (tem
) ? XINT (tem
) : -1)
2963 oldstate
= Fcdr (oldstate
);
2964 tem
= Fcar (oldstate
);
2965 start_quoted
= !NILP (tem
);
2967 /* if the eighth element of the list is nil, we are in comment
2968 style a. If it is non-nil, we are in comment style b */
2969 oldstate
= Fcdr (oldstate
);
2970 oldstate
= Fcdr (oldstate
);
2971 tem
= Fcar (oldstate
);
2972 state
.comstyle
= NILP (tem
) ? 0 : (EQ (tem
, Qsyntax_table
)
2973 ? ST_COMMENT_STYLE
: 1);
2975 oldstate
= Fcdr (oldstate
);
2976 tem
= Fcar (oldstate
);
2977 state
.comstr_start
= NILP (tem
) ? -1 : XINT (tem
) ;
2978 oldstate
= Fcdr (oldstate
);
2979 tem
= Fcar (oldstate
);
2980 while (!NILP (tem
)) /* >= second enclosing sexps. */
2982 /* curlevel++->last ran into compiler bug on Apollo */
2983 curlevel
->last
= XINT (Fcar (tem
));
2984 if (++curlevel
== endlevel
)
2985 curlevel
--; /* error ("Nesting too deep for parser"); */
2986 curlevel
->prev
= -1;
2987 curlevel
->last
= -1;
2994 curlevel
->prev
= -1;
2995 curlevel
->last
= -1;
2997 SETUP_SYNTAX_TABLE (prev_from
, 1);
2998 temp
= FETCH_CHAR (prev_from_byte
);
2999 prev_from_syntax
= SYNTAX_WITH_FLAGS (temp
);
3000 UPDATE_SYNTAX_TABLE_FORWARD (from
);
3002 /* Enter the loop at a place appropriate for initial state. */
3004 if (state
.incomment
)
3005 goto startincomment
;
3006 if (state
.instring
>= 0)
3008 nofence
= state
.instring
!= ST_STRING_STYLE
;
3010 goto startquotedinstring
;
3013 else if (start_quoted
)
3019 code
= prev_from_syntax
& 0xff;
3022 && SYNTAX_FLAGS_COMSTART_FIRST (prev_from_syntax
)
3023 && (c1
= FETCH_CHAR (from_byte
),
3024 SYNTAX_COMSTART_SECOND (c1
)))
3025 /* Duplicate code to avoid a complex if-expression
3026 which causes trouble for the SGI compiler. */
3028 /* Record the comment style we have entered so that only
3029 the comment-end sequence of the same style actually
3030 terminates the comment section. */
3031 state
.comstyle
= SYNTAX_COMMENT_STYLE (c1
);
3032 comnested
= SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax
);
3033 comnested
= comnested
|| SYNTAX_COMMENT_NESTED (c1
);
3034 state
.incomment
= comnested
? 1 : -1;
3035 state
.comstr_start
= prev_from
;
3039 else if (code
== Scomment_fence
)
3041 /* Record the comment style we have entered so that only
3042 the comment-end sequence of the same style actually
3043 terminates the comment section. */
3044 state
.comstyle
= ST_COMMENT_STYLE
;
3045 state
.incomment
= -1;
3046 state
.comstr_start
= prev_from
;
3049 else if (code
== Scomment
)
3051 state
.comstyle
= SYNTAX_FLAGS_COMMENT_STYLE (prev_from_syntax
);
3052 state
.incomment
= (SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax
) ?
3054 state
.comstr_start
= prev_from
;
3057 if (SYNTAX_FLAGS_PREFIX (prev_from_syntax
))
3059 switch (SWITCH_ENUM_CAST (code
))
3063 if (stopbefore
) goto stop
; /* this arg means stop at sexp start */
3064 curlevel
->last
= prev_from
;
3066 if (from
== end
) goto endquoted
;
3069 /* treat following character as a word constituent */
3072 if (stopbefore
) goto stop
; /* this arg means stop at sexp start */
3073 curlevel
->last
= prev_from
;
3077 /* Some compilers can't handle this inside the switch. */
3078 temp
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
3079 temp
= SYNTAX (temp
);
3085 if (from
== end
) goto endquoted
;
3097 curlevel
->prev
= curlevel
->last
;
3100 case Scomment_fence
: /* Can't happen because it's handled above. */
3102 if (commentstop
|| boundary_stop
) goto done
;
3104 /* The (from == BEGV) test was to enter the loop in the middle so
3105 that we find a 2-char comment ender even if we start in the
3106 middle of it. We don't want to do that if we're just at the
3107 beginning of the comment (think of (*) ... (*)). */
3108 found
= forw_comment (from
, from_byte
, end
,
3109 state
.incomment
, state
.comstyle
,
3110 (from
== BEGV
|| from
< state
.comstr_start
+ 3)
3111 ? 0 : prev_from_syntax
,
3112 &out_charpos
, &out_bytepos
, &state
.incomment
);
3113 from
= out_charpos
; from_byte
= out_bytepos
;
3114 /* Beware! prev_from and friends are invalid now.
3115 Luckily, the `done' doesn't use them and the INC_FROM
3116 sets them to a sane value without looking at them. */
3117 if (!found
) goto done
;
3119 state
.incomment
= 0;
3120 state
.comstyle
= 0; /* reset the comment style */
3121 if (boundary_stop
) goto done
;
3125 if (stopbefore
) goto stop
; /* this arg means stop at sexp start */
3127 /* curlevel++->last ran into compiler bug on Apollo */
3128 curlevel
->last
= prev_from
;
3129 if (++curlevel
== endlevel
)
3130 curlevel
--; /* error ("Nesting too deep for parser"); */
3131 curlevel
->prev
= -1;
3132 curlevel
->last
= -1;
3133 if (targetdepth
== depth
) goto done
;
3138 if (depth
< mindepth
)
3140 if (curlevel
!= levelstart
)
3142 curlevel
->prev
= curlevel
->last
;
3143 if (targetdepth
== depth
) goto done
;
3148 state
.comstr_start
= from
- 1;
3149 if (stopbefore
) goto stop
; /* this arg means stop at sexp start */
3150 curlevel
->last
= prev_from
;
3151 state
.instring
= (code
== Sstring
3152 ? (FETCH_CHAR_AS_MULTIBYTE (prev_from_byte
))
3154 if (boundary_stop
) goto done
;
3157 nofence
= state
.instring
!= ST_STRING_STYLE
;
3163 if (from
>= end
) goto done
;
3164 c
= FETCH_CHAR_AS_MULTIBYTE (from_byte
);
3165 /* Some compilers can't handle this inside the switch. */
3168 /* Check TEMP here so that if the char has
3169 a syntax-table property which says it is NOT
3170 a string character, it does not end the string. */
3171 if (nofence
&& c
== state
.instring
&& temp
== Sstring
)
3177 if (!nofence
) goto string_end
;
3182 startquotedinstring
:
3183 if (from
>= end
) goto endquoted
;
3189 state
.instring
= -1;
3190 curlevel
->prev
= curlevel
->last
;
3192 if (boundary_stop
) goto done
;
3196 /* FIXME: We should do something with it. */
3199 /* Ignore whitespace, punctuation, quote, endcomment. */
3205 stop
: /* Here if stopping before start of sexp. */
3206 from
= prev_from
; /* We have just fetched the char that starts it; */
3207 goto done
; /* but return the position before it. */
3212 state
.depth
= depth
;
3213 state
.mindepth
= mindepth
;
3214 state
.thislevelstart
= curlevel
->prev
;
3215 state
.prevlevelstart
3216 = (curlevel
== levelstart
) ? -1 : (curlevel
- 1)->last
;
3217 state
.location
= from
;
3218 state
.levelstarts
= Qnil
;
3219 while (--curlevel
>= levelstart
)
3220 state
.levelstarts
= Fcons (make_number (curlevel
->last
),
3227 DEFUN ("parse-partial-sexp", Fparse_partial_sexp
, Sparse_partial_sexp
, 2, 6, 0,
3228 doc
: /* Parse Lisp syntax starting at FROM until TO; return status of parse at TO.
3229 Parsing stops at TO or when certain criteria are met;
3230 point is set to where parsing stops.
3231 If fifth arg OLDSTATE is omitted or nil,
3232 parsing assumes that FROM is the beginning of a function.
3233 Value is a list of elements describing final state of parsing:
3235 1. character address of start of innermost containing list; nil if none.
3236 2. character address of start of last complete sexp terminated.
3237 3. non-nil if inside a string.
3238 (it is the character that will terminate the string,
3239 or t if the string should be terminated by a generic string delimiter.)
3240 4. nil if outside a comment, t if inside a non-nestable comment,
3241 else an integer (the current comment nesting).
3242 5. t if following a quote character.
3243 6. the minimum paren-depth encountered during this scan.
3244 7. t if in a comment of style b; symbol `syntax-table' if the comment
3245 should be terminated by a generic comment delimiter.
3246 8. character address of start of comment or string; nil if not in one.
3247 9. Intermediate data for continuation of parsing (subject to change).
3248 If third arg TARGETDEPTH is non-nil, parsing stops if the depth
3249 in parentheses becomes equal to TARGETDEPTH.
3250 Fourth arg STOPBEFORE non-nil means stop when come to
3251 any character that starts a sexp.
3252 Fifth arg OLDSTATE is a list like what this function returns.
3253 It is used to initialize the state of the parse. Elements number 1, 2, 6
3255 Sixth arg COMMENTSTOP non-nil means stop at the start of a comment.
3256 If it is symbol `syntax-table', stop after the start of a comment or a
3257 string, or after end of a comment or a string. */)
3258 (from
, to
, targetdepth
, stopbefore
, oldstate
, commentstop
)
3259 Lisp_Object from
, to
, targetdepth
, stopbefore
, oldstate
, commentstop
;
3261 struct lisp_parse_state state
;
3264 if (!NILP (targetdepth
))
3266 CHECK_NUMBER (targetdepth
);
3267 target
= XINT (targetdepth
);
3270 target
= -100000; /* We won't reach this depth */
3272 validate_region (&from
, &to
);
3273 scan_sexps_forward (&state
, XINT (from
), CHAR_TO_BYTE (XINT (from
)),
3275 target
, !NILP (stopbefore
), oldstate
,
3277 ? 0 : (EQ (commentstop
, Qsyntax_table
) ? -1 : 1)));
3279 SET_PT (state
.location
);
3281 return Fcons (make_number (state
.depth
),
3282 Fcons (state
.prevlevelstart
< 0 ? Qnil
: make_number (state
.prevlevelstart
),
3283 Fcons (state
.thislevelstart
< 0 ? Qnil
: make_number (state
.thislevelstart
),
3284 Fcons (state
.instring
>= 0
3285 ? (state
.instring
== ST_STRING_STYLE
3286 ? Qt
: make_number (state
.instring
)) : Qnil
,
3287 Fcons (state
.incomment
< 0 ? Qt
:
3288 (state
.incomment
== 0 ? Qnil
:
3289 make_number (state
.incomment
)),
3290 Fcons (state
.quoted
? Qt
: Qnil
,
3291 Fcons (make_number (state
.mindepth
),
3292 Fcons ((state
.comstyle
3293 ? (state
.comstyle
== ST_COMMENT_STYLE
3294 ? Qsyntax_table
: Qt
) :
3296 Fcons (((state
.incomment
3297 || (state
.instring
>= 0))
3298 ? make_number (state
.comstr_start
)
3300 Fcons (state
.levelstarts
, Qnil
))))))))));
3304 init_syntax_once (void)
3309 /* This has to be done here, before we call Fmake_char_table. */
3310 Qsyntax_table
= intern_c_string ("syntax-table");
3311 staticpro (&Qsyntax_table
);
3313 /* Intern_C_String this now in case it isn't already done.
3314 Setting this variable twice is harmless.
3315 But don't staticpro it here--that is done in alloc.c. */
3316 Qchar_table_extra_slots
= intern_c_string ("char-table-extra-slots");
3318 /* Create objects which can be shared among syntax tables. */
3319 Vsyntax_code_object
= Fmake_vector (make_number (Smax
), Qnil
);
3320 for (i
= 0; i
< XVECTOR (Vsyntax_code_object
)->size
; i
++)
3321 XVECTOR (Vsyntax_code_object
)->contents
[i
]
3322 = Fcons (make_number (i
), Qnil
);
3324 /* Now we are ready to set up this property, so we can
3325 create syntax tables. */
3326 Fput (Qsyntax_table
, Qchar_table_extra_slots
, make_number (0));
3328 temp
= XVECTOR (Vsyntax_code_object
)->contents
[(int) Swhitespace
];
3330 Vstandard_syntax_table
= Fmake_char_table (Qsyntax_table
, temp
);
3332 /* Control characters should not be whitespace. */
3333 temp
= XVECTOR (Vsyntax_code_object
)->contents
[(int) Spunct
];
3334 for (i
= 0; i
<= ' ' - 1; i
++)
3335 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, i
, temp
);
3336 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, 0177, temp
);
3338 /* Except that a few really are whitespace. */
3339 temp
= XVECTOR (Vsyntax_code_object
)->contents
[(int) Swhitespace
];
3340 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, ' ', temp
);
3341 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, '\t', temp
);
3342 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, '\n', temp
);
3343 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, 015, temp
);
3344 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, 014, temp
);
3346 temp
= XVECTOR (Vsyntax_code_object
)->contents
[(int) Sword
];
3347 for (i
= 'a'; i
<= 'z'; i
++)
3348 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, i
, temp
);
3349 for (i
= 'A'; i
<= 'Z'; i
++)
3350 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, i
, temp
);
3351 for (i
= '0'; i
<= '9'; i
++)
3352 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, i
, temp
);
3354 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, '$', temp
);
3355 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, '%', temp
);
3357 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, '(',
3358 Fcons (make_number (Sopen
), make_number (')')));
3359 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, ')',
3360 Fcons (make_number (Sclose
), make_number ('(')));
3361 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, '[',
3362 Fcons (make_number (Sopen
), make_number (']')));
3363 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, ']',
3364 Fcons (make_number (Sclose
), make_number ('[')));
3365 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, '{',
3366 Fcons (make_number (Sopen
), make_number ('}')));
3367 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, '}',
3368 Fcons (make_number (Sclose
), make_number ('{')));
3369 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, '"',
3370 Fcons (make_number ((int) Sstring
), Qnil
));
3371 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, '\\',
3372 Fcons (make_number ((int) Sescape
), Qnil
));
3374 temp
= XVECTOR (Vsyntax_code_object
)->contents
[(int) Ssymbol
];
3375 for (i
= 0; i
< 10; i
++)
3377 c
= "_-+*/&|<>="[i
];
3378 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, c
, temp
);
3381 temp
= XVECTOR (Vsyntax_code_object
)->contents
[(int) Spunct
];
3382 for (i
= 0; i
< 12; i
++)
3384 c
= ".,;:?!#@~^'`"[i
];
3385 SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table
, c
, temp
);
3388 /* All multibyte characters have syntax `word' by default. */
3389 temp
= XVECTOR (Vsyntax_code_object
)->contents
[(int) Sword
];
3390 char_table_set_range (Vstandard_syntax_table
, 0x80, MAX_CHAR
, temp
);
3394 syms_of_syntax (void)
3396 Qsyntax_table_p
= intern_c_string ("syntax-table-p");
3397 staticpro (&Qsyntax_table_p
);
3399 staticpro (&Vsyntax_code_object
);
3401 staticpro (&gl_state
.object
);
3402 staticpro (&gl_state
.global_code
);
3403 staticpro (&gl_state
.current_syntax_table
);
3404 staticpro (&gl_state
.old_prop
);
3406 /* Defined in regex.c */
3407 staticpro (&re_match_object
);
3409 Qscan_error
= intern_c_string ("scan-error");
3410 staticpro (&Qscan_error
);
3411 Fput (Qscan_error
, Qerror_conditions
,
3412 pure_cons (Qscan_error
, pure_cons (Qerror
, Qnil
)));
3413 Fput (Qscan_error
, Qerror_message
,
3414 make_pure_c_string ("Scan error"));
3416 DEFVAR_BOOL ("parse-sexp-ignore-comments", &parse_sexp_ignore_comments
,
3417 doc
: /* Non-nil means `forward-sexp', etc., should treat comments as whitespace. */);
3419 DEFVAR_BOOL ("parse-sexp-lookup-properties", &parse_sexp_lookup_properties
,
3420 doc
: /* Non-nil means `forward-sexp', etc., obey `syntax-table' property.
3421 Otherwise, that text property is simply ignored.
3422 See the info node `(elisp)Syntax Properties' for a description of the
3423 `syntax-table' property. */);
3425 words_include_escapes
= 0;
3426 DEFVAR_BOOL ("words-include-escapes", &words_include_escapes
,
3427 doc
: /* Non-nil means `forward-word', etc., should treat escape chars part of words. */);
3429 DEFVAR_BOOL ("multibyte-syntax-as-symbol", &multibyte_syntax_as_symbol
,
3430 doc
: /* Non-nil means `scan-sexps' treats all multibyte characters as symbol. */);
3431 multibyte_syntax_as_symbol
= 0;
3433 DEFVAR_BOOL ("open-paren-in-column-0-is-defun-start",
3434 &open_paren_in_column_0_is_defun_start
,
3435 doc
: /* *Non-nil means an open paren in column 0 denotes the start of a defun. */);
3436 open_paren_in_column_0_is_defun_start
= 1;
3439 DEFVAR_LISP ("find-word-boundary-function-table",
3440 &Vfind_word_boundary_function_table
,
3442 Char table of functions to search for the word boundary.
3443 Each function is called with two arguments; POS and LIMIT.
3444 POS and LIMIT are character positions in the current buffer.
3446 If POS is less than LIMIT, POS is at the first character of a word,
3447 and the return value of a function is a position after the last
3448 character of that word.
3450 If POS is not less than LIMIT, POS is at the last character of a word,
3451 and the return value of a function is a position at the first
3452 character of that word.
3454 In both cases, LIMIT bounds the search. */);
3455 Vfind_word_boundary_function_table
= Fmake_char_table (Qnil
, Qnil
);
3457 defsubr (&Ssyntax_table_p
);
3458 defsubr (&Ssyntax_table
);
3459 defsubr (&Sstandard_syntax_table
);
3460 defsubr (&Scopy_syntax_table
);
3461 defsubr (&Sset_syntax_table
);
3462 defsubr (&Schar_syntax
);
3463 defsubr (&Smatching_paren
);
3464 defsubr (&Sstring_to_syntax
);
3465 defsubr (&Smodify_syntax_entry
);
3466 defsubr (&Sinternal_describe_syntax_value
);
3468 defsubr (&Sforward_word
);
3470 defsubr (&Sskip_chars_forward
);
3471 defsubr (&Sskip_chars_backward
);
3472 defsubr (&Sskip_syntax_forward
);
3473 defsubr (&Sskip_syntax_backward
);
3475 defsubr (&Sforward_comment
);
3476 defsubr (&Sscan_lists
);
3477 defsubr (&Sscan_sexps
);
3478 defsubr (&Sbackward_prefix_chars
);
3479 defsubr (&Sparse_partial_sexp
);
3482 /* arch-tag: 3e297b9f-088e-4b64-8f4c-fb0b3443e412
3483 (do not change this comment) */