syntax.c 99.6 KB
Newer Older
Richard M. Stallman's avatar
Richard M. Stallman committed
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,
Glenn Morris's avatar
Glenn Morris committed
3
                 2002, 2003, 2004, 2005, 2006, 2007, 2008
4
                 Free Software Foundation, Inc.
Richard M. Stallman's avatar
Richard M. Stallman committed
5 6 7

This file is part of GNU Emacs.

8
GNU Emacs is free software: you can redistribute it and/or modify
Richard M. Stallman's avatar
Richard M. Stallman committed
9
it under the terms of the GNU General Public License as published by
10 11
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Richard M. Stallman's avatar
Richard M. Stallman committed
12 13 14 15 16 17 18

GNU Emacs is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

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/>.  */
Richard M. Stallman's avatar
Richard M. Stallman committed
20 21


22
#include <config.h>
Richard M. Stallman's avatar
Richard M. Stallman committed
23 24 25 26
#include <ctype.h>
#include "lisp.h"
#include "commands.h"
#include "buffer.h"
27
#include "character.h"
Stefan Monnier's avatar
Stefan Monnier committed
28
#include "keymap.h"
Richard M. Stallman's avatar
Richard M. Stallman committed
29
#include "regex.h"
30 31 32 33

/* Make syntax table lookup grant data in gl_state.  */
#define SYNTAX_ENTRY_VIA_PROPERTY

Richard M. Stallman's avatar
Richard M. Stallman committed
34
#include "syntax.h"
35 36 37 38 39 40 41 42
#include "intervals.h"

/* We use these constants in place for comment-style and
   string-ender-char to distinguish  comments/strings started by
   comment_fence and string_fence codes.  */

#define ST_COMMENT_STYLE (256 + 1)
#define ST_STRING_STYLE (256 + 2)
Karl Heuer's avatar
Karl Heuer committed
43
#include "category.h"
Richard M. Stallman's avatar
Richard M. Stallman committed
44

45
Lisp_Object Qsyntax_table_p, Qsyntax_table, Qscan_error;
Richard M. Stallman's avatar
Richard M. Stallman committed
46 47

int words_include_escapes;
48
int parse_sexp_lookup_properties;
Richard M. Stallman's avatar
Richard M. Stallman committed
49

50 51 52
/* Nonzero means `scan-sexps' treat all multibyte characters as symbol.  */
int multibyte_syntax_as_symbol;

53 54 55 56 57
/* Used as a temporary in SYNTAX_ENTRY and other macros in syntax.h,
   if not compiled with GCC.  No need to mark it, since it is used
   only very temporarily.  */
Lisp_Object syntax_temp;

58 59 60 61 62 63
/* Non-zero means an open parenthesis in column 0 is always considered
   to be the start of a defun.  Zero means an open parenthesis in
   column 0 has no special meaning.  */

int open_paren_in_column_0_is_defun_start;

64 65 66 67
/* This is the internal form of the parse state used in parse-partial-sexp.  */

struct lisp_parse_state
  {
68 69
    int depth;		/* Depth at end of parsing.  */
    int instring;	/* -1 if not within string, else desired terminator.  */
70
    int incomment;	/* -1 if in unnestable comment else comment nesting */
71
    int comstyle;	/* comment style a=0, or b=1, or ST_COMMENT_STYLE.  */
72 73 74
    int quoted;		/* Nonzero if just after an escape char at end of parsing */
    int thislevelstart;	/* Char number of most recent start-of-expression at current level */
    int prevlevelstart; /* Char number of start of containing expression */
75
    int location;	/* Char number at which parsing stopped.  */
76
    int mindepth;	/* Minimum depth seen while scanning.  */
77
    int comstr_start;	/* Position just after last comment/string starter.  */
78 79
    Lisp_Object levelstarts;	/* Char numbers of starts-of-expression
				   of levels (starting from outermost).  */
80 81
  };

82 83 84
/* These variables are a cache for finding the start of a defun.
   find_start_pos is the place for which the defun start was found.
   find_start_value is the defun start position found for it.
85
   find_start_value_byte is the corresponding byte position.
86 87 88 89
   find_start_buffer is the buffer it was found in.
   find_start_begv is the BEGV value when it was found.
   find_start_modiff is the value of MODIFF when it was found.  */

90 91 92
static EMACS_INT find_start_pos;
static EMACS_INT find_start_value;
static EMACS_INT find_start_value_byte;
93
static struct buffer *find_start_buffer;
94
static EMACS_INT find_start_begv;
95
static int find_start_modiff;
96 97


98
static int find_defun_start P_ ((EMACS_INT, EMACS_INT));
99 100
static int back_comment P_ ((EMACS_INT, EMACS_INT, EMACS_INT, int, int,
			     EMACS_INT *, EMACS_INT *));
101
static int char_quoted P_ ((int, int));
102
static Lisp_Object skip_chars P_ ((int, Lisp_Object, Lisp_Object, int));
103
static Lisp_Object skip_syntaxes P_ ((int, Lisp_Object, Lisp_Object));
104
static Lisp_Object scan_lists P_ ((EMACS_INT, EMACS_INT, EMACS_INT, int));
105
static void scan_sexps_forward P_ ((struct lisp_parse_state *,
106
				    EMACS_INT, EMACS_INT, EMACS_INT, int,
107
				    int, Lisp_Object, int));
Richard M. Stallman's avatar
Richard M. Stallman committed
108
static int in_classes P_ ((int, Lisp_Object));
109 110 111 112 113 114 115 116


struct gl_state_s gl_state;		/* Global state of syntax parser.  */

INTERVAL interval_of ();
#define INTERVALS_AT_ONCE 10		/* 1 + max-number of intervals
					   to scan to property-change.  */

117 118
/* Update gl_state to an appropriate interval which contains CHARPOS.  The
   sign of COUNT give the relative position of CHARPOS wrt the previously
119
   valid interval.  If INIT, only [be]_property fields of gl_state are
120
   valid at start, the rest is filled basing on OBJECT.
121

122
   `gl_state.*_i' are the intervals, and CHARPOS is further in the search
123 124
   direction than the intervals - or in an interval.  We update the
   current syntax-table basing on the property of this interval, and
125
   update the interval to start further than CHARPOS - or be
126
   NULL_INTERVAL.  We also update lim_property to be the next value of
127
   charpos to call this subroutine again - or be before/after the
128 129 130
   start/end of OBJECT.  */

void
131 132
update_syntax_table (charpos, count, init, object)
     int charpos, count, init;
133 134 135
     Lisp_Object object;
{
  Lisp_Object tmp_table;
136
  int cnt = 0, invalidate = 1;
137
  INTERVAL i;
138 139 140

  if (init)
    {
141
      gl_state.old_prop = Qnil;
142 143
      gl_state.start = gl_state.b_property;
      gl_state.stop = gl_state.e_property;
144 145
      i = interval_of (charpos, object);
      gl_state.backward_i = gl_state.forward_i = i;
146 147 148
      invalidate = 0;
      if (NULL_INTERVAL_P (i))
	return;
149
      /* interval_of updates only ->position of the return value, so
150
	 update the parents manually to speed up update_interval.  */
151
      while (!NULL_PARENT (i))
152 153
	{
	  if (AM_RIGHT_CHILD (i))
154
	    INTERVAL_PARENT (i)->position = i->position
155
	      - LEFT_TOTAL_LENGTH (i) + TOTAL_LENGTH (i) /* right end */
156 157
	      - TOTAL_LENGTH (INTERVAL_PARENT (i))
	      + LEFT_TOTAL_LENGTH (INTERVAL_PARENT (i));
158
	  else
159
	    INTERVAL_PARENT (i)->position = i->position - LEFT_TOTAL_LENGTH (i)
160
	      + TOTAL_LENGTH (i);
161
	  i = INTERVAL_PARENT (i);
162 163
	}
      i = gl_state.forward_i;
164
      gl_state.b_property = i->position - gl_state.offset;
165
      gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset;
166 167
      goto update;
    }
168
  i = count > 0 ? gl_state.forward_i : gl_state.backward_i;
169

170
  /* We are guaranteed to be called with CHARPOS either in i,
171
     or further off.  */
172 173
  if (NULL_INTERVAL_P (i))
    error ("Error in syntax_table logic for to-the-end intervals");
174
  else if (charpos < i->position)		/* Move left.  */
175 176
    {
      if (count > 0)
177
	error ("Error in syntax_table logic for intervals <-");
178
      /* Update the interval.  */
179
      i = update_interval (i, charpos);
180
      if (INTERVAL_LAST_POS (i) != gl_state.b_property)
181 182 183
	{
	  invalidate = 0;
	  gl_state.forward_i = i;
184
	  gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset;
185
	}
186
    }
187
  else if (charpos >= INTERVAL_LAST_POS (i)) /* Move right.  */
188 189
    {
      if (count < 0)
190
	error ("Error in syntax_table logic for intervals ->");
191
      /* Update the interval.  */
192
      i = update_interval (i, charpos);
193
      if (i->position != gl_state.e_property)
194 195 196
	{
	  invalidate = 0;
	  gl_state.backward_i = i;
197
	  gl_state.b_property = i->position - gl_state.offset;
198 199 200 201 202 203 204 205
	}
    }

  update:
  tmp_table = textget (i->plist, Qsyntax_table);

  if (invalidate)
    invalidate = !EQ (tmp_table, gl_state.old_prop); /* Need to invalidate? */
Juanma Barranquero's avatar
Juanma Barranquero committed
206

207 208 209
  if (invalidate)		/* Did not get to adjacent interval.  */
    {				/* with the same table => */
				/* invalidate the old range.  */
210 211 212
      if (count > 0)
	{
	  gl_state.backward_i = i;
213 214 215
	  gl_state.b_property = i->position - gl_state.offset;
	}
      else
216
	{
217
	  gl_state.forward_i = i;
218
	  gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset;
219 220
	}
    }
221

222
  if (!EQ (tmp_table, gl_state.old_prop))
223
    {
224 225 226 227 228
      gl_state.current_syntax_table = tmp_table;
      gl_state.old_prop = tmp_table;
      if (EQ (Fsyntax_table_p (tmp_table), Qt))
	{
	  gl_state.use_global = 0;
Juanma Barranquero's avatar
Juanma Barranquero committed
229
	}
230 231 232 233 234
      else if (CONSP (tmp_table))
	{
	  gl_state.use_global = 1;
	  gl_state.global_code = tmp_table;
	}
Juanma Barranquero's avatar
Juanma Barranquero committed
235
      else
236 237 238 239
	{
	  gl_state.use_global = 0;
	  gl_state.current_syntax_table = current_buffer->syntax_table;
	}
240 241 242 243 244 245 246
    }

  while (!NULL_INTERVAL_P (i))
    {
      if (cnt && !EQ (tmp_table, textget (i->plist, Qsyntax_table)))
	{
	  if (count > 0)
247 248 249 250 251 252
	    {
	      gl_state.e_property = i->position - gl_state.offset;
	      gl_state.forward_i = i;
	    }
	  else
	    {
253 254
	      gl_state.b_property
		= i->position + LENGTH (i) - gl_state.offset;
255 256 257
	      gl_state.backward_i = i;
	    }
	  return;
258
	}
Juanma Barranquero's avatar
Juanma Barranquero committed
259
      else if (cnt == INTERVALS_AT_ONCE)
260 261
	{
	  if (count > 0)
262
	    {
263 264 265 266 267 268
	      gl_state.e_property
		= i->position + LENGTH (i) - gl_state.offset
		/* e_property at EOB is not set to ZV but to ZV+1, so that
		   we can do INC(from);UPDATE_SYNTAX_TABLE_FORWARD without
		   having to check eob between the two.  */
		+ (NULL_INTERVAL_P (next_interval (i)) ? 1 : 0);
269 270 271 272 273 274 275 276
	      gl_state.forward_i = i;
	    }
	  else
	    {
	      gl_state.b_property = i->position - gl_state.offset;
	      gl_state.backward_i = i;
	    }
	  return;
277 278 279 280
	}
      cnt++;
      i = count > 0 ? next_interval (i) : previous_interval (i);
    }
281 282 283 284 285
  eassert (NULL_INTERVAL_P (i)); /* This property goes to the end.  */
  if (count > 0)
    gl_state.e_property = gl_state.stop;
  else
    gl_state.b_property = gl_state.start;
286 287
}

288 289 290
/* Returns TRUE if char at CHARPOS is quoted.
   Global syntax-table data should be set up already to be good at CHARPOS
   or after.  On return global syntax data is good for lookup at CHARPOS. */
291 292

static int
293 294
char_quoted (charpos, bytepos)
     register int charpos, bytepos;
295 296 297 298
{
  register enum syntaxcode code;
  register int beg = BEGV;
  register int quoted = 0;
299 300
  int orig = charpos;

301
  while (charpos > beg)
302
    {
303
      int c;
304
      DEC_BOTH (charpos, bytepos);
305

306
      UPDATE_SYNTAX_TABLE_BACKWARD (charpos);
307
      c = FETCH_CHAR_AS_MULTIBYTE (bytepos);
308
      code = SYNTAX (c);
309 310 311 312
      if (! (code == Scharquote || code == Sescape))
	break;

      quoted = !quoted;
313
    }
314 315

  UPDATE_SYNTAX_TABLE (orig);
316 317
  return quoted;
}
318 319 320 321 322 323 324 325

/* Return the bytepos one character after BYTEPOS.
   We assume that BYTEPOS is not at the end of the buffer.  */

INLINE int
inc_bytepos (bytepos)
     int bytepos;
{
326 327 328
  if (NILP (current_buffer->enable_multibyte_characters))
    return bytepos + 1;

329 330 331 332 333 334 335 336 337 338 339
  INC_POS (bytepos);
  return bytepos;
}

/* Return the bytepos one character before BYTEPOS.
   We assume that BYTEPOS is not at the start of the buffer.  */

INLINE int
dec_bytepos (bytepos)
     int bytepos;
{
340 341 342
  if (NILP (current_buffer->enable_multibyte_characters))
    return bytepos - 1;

343 344 345
  DEC_POS (bytepos);
  return bytepos;
}
346

Richard M. Stallman's avatar
Richard M. Stallman committed
347 348 349 350
/* Return a defun-start position before before POS and not too far before.
   It should be the last one before POS, or nearly the last.

   When open_paren_in_column_0_is_defun_start is nonzero,
351
   only the beginning of the buffer is treated as a defun-start.
Richard M. Stallman's avatar
Richard M. Stallman committed
352 353 354 355

   We record the information about where the scan started
   and what its result was, so that another call in the same area
   can return the same value very quickly.
356 357 358 359

   There is no promise at which position the global syntax data is
   valid on return from the subroutine, so the caller should explicitly
   update the global data.  */
360 361

static int
362
find_defun_start (pos, pos_byte)
363
     EMACS_INT pos, pos_byte;
364
{
365
  EMACS_INT opoint = PT, opoint_byte = PT_BYTE;
366

367 368 369 370 371 372
  if (!open_paren_in_column_0_is_defun_start)
    {
      find_start_value_byte = BEGV_BYTE;
      return BEGV;
    }

373 374 375 376 377 378 379 380 381 382 383 384
  /* Use previous finding, if it's valid and applies to this inquiry.  */
  if (current_buffer == find_start_buffer
      /* Reuse the defun-start even if POS is a little farther on.
	 POS might be in the next defun, but that's ok.
	 Our value may not be the best possible, but will still be usable.  */
      && pos <= find_start_pos + 1000
      && pos >= find_start_value
      && BEGV == find_start_begv
      && MODIFF == find_start_modiff)
    return find_start_value;

  /* Back up to start of line.  */
385
  scan_newline (pos, pos_byte, BEGV, BEGV_BYTE, -1, 1);
386

387 388 389 390 391
  /* We optimize syntax-table lookup for rare updates.  Thus we accept
     only those `^\s(' which are good in global _and_ text-property
     syntax-tables.  */
  gl_state.current_syntax_table = current_buffer->syntax_table;
  gl_state.use_global = 0;
392
  while (PT > BEGV)
393
    {
394 395
      int c;

396 397
      /* Open-paren at start of line means we may have found our
	 defun-start.  */
398
      c = FETCH_CHAR_AS_MULTIBYTE (PT_BYTE);
399
      if (SYNTAX (c) == Sopen)
400
	{
401
	  SETUP_SYNTAX_TABLE (PT + 1, -1);	/* Try again... */
402
	  c = FETCH_CHAR_AS_MULTIBYTE (PT_BYTE);
403
	  if (SYNTAX (c) == Sopen)
404 405 406 407
	    break;
	  /* Now fallback to the default value.  */
	  gl_state.current_syntax_table = current_buffer->syntax_table;
	  gl_state.use_global = 0;
408
	}
409 410
      /* Move to beg of previous line.  */
      scan_newline (PT, PT_BYTE, BEGV, BEGV_BYTE, -2, 1);
411 412 413
    }

  /* Record what we found, for the next try.  */
414 415
  find_start_value = PT;
  find_start_value_byte = PT_BYTE;
416 417 418 419 420
  find_start_buffer = current_buffer;
  find_start_modiff = MODIFF;
  find_start_begv = BEGV;
  find_start_pos = pos;

421 422
  TEMP_SET_PT_BOTH (opoint, opoint_byte);

423 424 425
  return find_start_value;
}

426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443
/* Return the SYNTAX_COMEND_FIRST of the character before POS, POS_BYTE.  */

static int
prev_char_comend_first (pos, pos_byte)
     int pos, pos_byte;
{
  int c, val;

  DEC_BOTH (pos, pos_byte);
  UPDATE_SYNTAX_TABLE_BACKWARD (pos);
  c = FETCH_CHAR (pos_byte);
  val = SYNTAX_COMEND_FIRST (c);
  UPDATE_SYNTAX_TABLE_FORWARD (pos + 1);
  return val;
}

/* Return the SYNTAX_COMSTART_FIRST of the character before POS, POS_BYTE.  */

444 445 446 447 448
/* static int
 * prev_char_comstart_first (pos, pos_byte)
 *      int pos, pos_byte;
 * {
 *   int c, val;
Juanma Barranquero's avatar
Juanma Barranquero committed
449
 *
450 451 452 453 454 455 456
 *   DEC_BOTH (pos, pos_byte);
 *   UPDATE_SYNTAX_TABLE_BACKWARD (pos);
 *   c = FETCH_CHAR (pos_byte);
 *   val = SYNTAX_COMSTART_FIRST (c);
 *   UPDATE_SYNTAX_TABLE_FORWARD (pos + 1);
 *   return val;
 * } */
457

458 459 460 461 462 463 464 465 466
/* Checks whether charpos FROM is at the end of a comment.
   FROM_BYTE is the bytepos corresponding to FROM.
   Do not move back before STOP.

   Return a positive value if we find a comment ending at FROM/FROM_BYTE;
   return -1 otherwise.

   If successful, store the charpos of the comment's beginning
   into *CHARPOS_PTR, and the bytepos into *BYTEPOS_PTR.
467 468 469

   Global syntax data remains valid for backward search starting at
   the returned value (or at FROM, if the search was not successful).  */
470 471

static int
472
back_comment (from, from_byte, stop, comnested, comstyle, charpos_ptr, bytepos_ptr)
473
     EMACS_INT from, from_byte, stop;
474
     int comnested, comstyle;
475
     EMACS_INT *charpos_ptr, *bytepos_ptr;
476 477 478 479 480 481 482 483 484 485
{
  /* Look back, counting the parity of string-quotes,
     and recording the comment-starters seen.
     When we reach a safe place, assume that's not in a string;
     then step the main scan to the earliest comment-starter seen
     an even number of string quotes away from the safe place.

     OFROM[I] is position of the earliest comment-starter seen
     which is I+2X quotes from the comment-end.
     PARITY is current parity of quotes from the comment end.  */
486
  int string_style = -1;	/* Presumed outside of any string. */
487
  int string_lossage = 0;
488
  /* Not a real lossage: indicates that we have passed a matching comment
Juanma Barranquero's avatar
Juanma Barranquero committed
489
     starter plus a non-matching comment-ender, meaning that any matching
490 491 492 493
     comment-starter we might see later could be a false positive (hidden
     inside another comment).
     Test case:  { a (* b } c (* d *) */
  int comment_lossage = 0;
494 495 496 497
  EMACS_INT comment_end = from;
  EMACS_INT comment_end_byte = from_byte;
  EMACS_INT comstart_pos = 0;
  EMACS_INT comstart_byte;
498 499
  /* Place where the containing defun starts,
     or 0 if we didn't come across it yet.  */
500 501
  EMACS_INT defun_start = 0;
  EMACS_INT defun_start_byte = 0;
502
  register enum syntaxcode code;
503
  int nesting = 1;		/* current comment nesting */
504
  int c;
505 506 507 508 509
  int syntax = 0;

  /* FIXME: A }} comment-ender style leads to incorrect behavior
     in the case of {{ c }}} because we ignore the last two chars which are
     assumed to be comment-enders although they aren't.  */
510 511 512 513 514

  /* At beginning of range to scan, we're outside of strings;
     that determines quote parity to the comment-end.  */
  while (from != stop)
    {
515 516
      int temp_byte, prev_syntax;
      int com2start, com2end;
517

518
      /* Move back and examine a character.  */
519
      DEC_BOTH (from, from_byte);
520 521
      UPDATE_SYNTAX_TABLE_BACKWARD (from);

522
      prev_syntax = syntax;
523
      c = FETCH_CHAR_AS_MULTIBYTE (from_byte);
524
      syntax = SYNTAX_WITH_FLAGS (c);
525 526
      code = SYNTAX (c);

527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547
      /* Check for 2-char comment markers.  */
      com2start = (SYNTAX_FLAGS_COMSTART_FIRST (syntax)
		   && SYNTAX_FLAGS_COMSTART_SECOND (prev_syntax)
		   && comstyle == SYNTAX_FLAGS_COMMENT_STYLE (prev_syntax)
		   && (SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax)
		       || SYNTAX_FLAGS_COMMENT_NESTED (syntax)) == comnested);
      com2end = (SYNTAX_FLAGS_COMEND_FIRST (syntax)
		 && SYNTAX_FLAGS_COMEND_SECOND (prev_syntax));

      /* Nasty cases with overlapping 2-char comment markers:
	 - snmp-mode: -- c -- foo -- c --
	              --- c --
		      ------ c --
	 - c-mode:    *||*
		      |* *|* *|
		      |*| |* |*|
		      ///   */

      /* If a 2-char comment sequence partly overlaps with another,
	 we don't try to be clever.  */
      if (from > stop && (com2end || com2start))
548
	{
549 550 551
	  int next = from, next_byte = from_byte, next_c, next_syntax;
	  DEC_BOTH (next, next_byte);
	  UPDATE_SYNTAX_TABLE_BACKWARD (next);
552
	  next_c = FETCH_CHAR_AS_MULTIBYTE (next_byte);
553 554 555 556 557 558 559 560 561 562
	  next_syntax = SYNTAX_WITH_FLAGS (next_c);
	  if (((com2start || comnested)
	       && SYNTAX_FLAGS_COMEND_SECOND (syntax)
	       && SYNTAX_FLAGS_COMEND_FIRST (next_syntax))
	      || ((com2end || comnested)
		  && SYNTAX_FLAGS_COMSTART_SECOND (syntax)
		  && comstyle == SYNTAX_FLAGS_COMMENT_STYLE (syntax)
		  && SYNTAX_FLAGS_COMSTART_FIRST (next_syntax)))
	    goto lossage;
	  /* UPDATE_SYNTAX_TABLE_FORWARD (next + 1); */
563
	}
564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580

      if (com2start && comstart_pos == 0)
	/* We're looking at a comment starter.  But it might be a comment
	   ender as well (see snmp-mode).  The first time we see one, we
	   need to consider it as a comment starter,
	   and the subsequent times as a comment ender.  */
	com2end = 0;

      /* Turn a 2-char comment sequences into the appropriate syntax.  */
      if (com2end)
	code = Sendcomment;
      else if (com2start)
	code = Scomment;
      /* Ignore comment starters of a different style.  */
      else if (code == Scomment
	       && (comstyle != SYNTAX_FLAGS_COMMENT_STYLE (syntax)
		   || SYNTAX_FLAGS_COMMENT_NESTED (syntax) != comnested))
581
	continue;
582

583 584
      /* Ignore escaped characters, except comment-enders.  */
      if (code != Sendcomment && char_quoted (from, from_byte))
585 586
	continue;

587
      switch (code)
588
	{
589 590 591 592 593 594 595 596 597 598 599 600 601 602
	case Sstring_fence:
	case Scomment_fence:
	  c = (code == Sstring_fence ? ST_STRING_STYLE : ST_COMMENT_STYLE);
	case Sstring:
	  /* Track parity of quotes.  */
	  if (string_style == -1)
	    /* Entering a string.  */
	    string_style = c;
	  else if (string_style == c)
	    /* Leaving the string.  */
	    string_style = -1;
	  else
	    /* If we have two kinds of string delimiters.
	       There's no way to grok this scanning backwards.  */
603
	    string_lossage = 1;
604
	  break;
Juanma Barranquero's avatar
Juanma Barranquero committed
605

606 607
	case Scomment:
	  /* We've already checked that it is the relevant comstyle.  */
608
	  if (string_style != -1 || comment_lossage || string_lossage)
609 610 611 612
	    /* There are odd string quotes involved, so let's be careful.
	       Test case in Pascal: " { " a { " } */
	    goto lossage;

613 614 615 616 617 618 619
	  if (!comnested)
	    {
	      /* Record best comment-starter so far.  */
	      comstart_pos = from;
	      comstart_byte = from_byte;
	    }
	  else if (--nesting <= 0)
620 621 622 623 624 625
	    /* nested comments have to be balanced, so we don't need to
	       keep looking for earlier ones.  We use here the same (slightly
	       incorrect) reasoning as below:  since it is followed by uniform
	       paired string quotes, this comment-start has to be outside of
	       strings, else the comment-end itself would be inside a string. */
	    goto done;
626 627
	  break;

628
	case Sendcomment:
629
	  if (SYNTAX_FLAGS_COMMENT_STYLE (syntax) == comstyle
630
	      && ((com2end && SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax))
631
		  || SYNTAX_FLAGS_COMMENT_NESTED (syntax)) == comnested)
632 633 634 635 636 637 638 639 640
	    /* This is the same style of comment ender as ours. */
	    {
	      if (comnested)
		nesting++;
	      else
		/* Anything before that can't count because it would match
		   this comment-ender rather than ours.  */
		from = stop;	/* Break out of the loop.  */
	    }
641 642 643 644 645 646 647 648 649
	  else if (comstart_pos != 0 || c != '\n')
	    /* We're mixing comment styles here, so we'd better be careful.
	       The (comstart_pos != 0 || c != '\n') check is not quite correct
	       (we should just always set comment_lossage), but removing it
	       would imply that any multiline comment in C would go through
	       lossage, which seems overkill.
	       The failure should only happen in the rare cases such as
	         { (* } *)   */
	    comment_lossage = 1;
650
	  break;
651

652 653 654 655 656 657 658 659 660 661 662
	case Sopen:
	  /* Assume a defun-start point is outside of strings.  */
	  if (open_paren_in_column_0_is_defun_start
	      && (from == stop
		  || (temp_byte = dec_bytepos (from_byte),
		      FETCH_CHAR (temp_byte) == '\n')))
	    {
	      defun_start = from;
	      defun_start_byte = from_byte;
	      from = stop;	/* Break out of the loop.  */
	    }
663
	  break;
664 665

	default:
666
	  break;
667
	}
668 669 670 671 672
    }

  if (comstart_pos == 0)
    {
      from = comment_end;
673
      from_byte = comment_end_byte;
674 675
      UPDATE_SYNTAX_TABLE_FORWARD (comment_end - 1);
    }
676 677 678
  /* If comstart_pos is set and we get here (ie. didn't jump to `lossage'
     or `done'), then we've found the beginning of the non-nested comment.  */
  else if (1)	/* !comnested */
679 680
    {
      from = comstart_pos;
681
      from_byte = comstart_byte;
682
      UPDATE_SYNTAX_TABLE_FORWARD (from - 1);
683 684 685
    }
  else
    {
686 687
      struct lisp_parse_state state;
    lossage:
688 689
      /* We had two kinds of string delimiters mixed up
	 together.  Decode this going forwards.
690
	 Scan fwd from a known safe place (beginning-of-defun)
691 692
	 to the one in question; this records where we
	 last passed a comment starter.  */
693 694 695 696 697 698
      /* If we did not already find the defun start, find it now.  */
      if (defun_start == 0)
	{
	  defun_start = find_defun_start (comment_end, comment_end_byte);
	  defun_start_byte = find_start_value_byte;
	}
699
      do
700
	{
701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722
	  scan_sexps_forward (&state,
			      defun_start, defun_start_byte,
			      comment_end, -10000, 0, Qnil, 0);
	  defun_start = comment_end;
	  if (state.incomment == (comnested ? 1 : -1)
	      && state.comstyle == comstyle)
	    from = state.comstr_start;
	  else
	    {
	      from = comment_end;
	      if (state.incomment)
		/* If comment_end is inside some other comment, maybe ours
		   is nested, so we need to try again from within the
		   surrounding comment.  Example: { a (* " *)  */
		{
		  /* FIXME: We should advance by one or two chars. */
		  defun_start = state.comstr_start + 2;
		  defun_start_byte = CHAR_TO_BYTE (defun_start);
		}
	    }
	} while (defun_start < comment_end);

723
      from_byte = CHAR_TO_BYTE (from);
724 725
      UPDATE_SYNTAX_TABLE_FORWARD (from - 1);
    }
Juanma Barranquero's avatar
Juanma Barranquero committed
726

727
 done:
728 729 730
  *charpos_ptr = from;
  *bytepos_ptr = from_byte;

731
  return (from == comment_end) ? -1 : from;
732 733
}

Richard M. Stallman's avatar
Richard M. Stallman committed
734
DEFUN ("syntax-table-p", Fsyntax_table_p, Ssyntax_table_p, 1, 1, 0,
735 736 737
       doc: /* Return t if OBJECT is a syntax table.
Currently, any char-table counts as a syntax table.  */)
     (object)
738
     Lisp_Object object;
Richard M. Stallman's avatar
Richard M. Stallman committed
739
{
740
  if (CHAR_TABLE_P (object)
741
      && EQ (XCHAR_TABLE (object)->purpose, Qsyntax_table))
Richard M. Stallman's avatar
Richard M. Stallman committed
742 743 744 745
    return Qt;
  return Qnil;
}

746
static void
Richard M. Stallman's avatar
Richard M. Stallman committed
747 748 749
check_syntax_table (obj)
     Lisp_Object obj;
{
750 751
  CHECK_TYPE (CHAR_TABLE_P (obj) && EQ (XCHAR_TABLE (obj)->purpose, Qsyntax_table),
	      Qsyntax_table_p, obj);
Juanma Barranquero's avatar
Juanma Barranquero committed
752
}
Richard M. Stallman's avatar
Richard M. Stallman committed
753 754

DEFUN ("syntax-table", Fsyntax_table, Ssyntax_table, 0, 0, 0,
755 756 757
       doc: /* Return the current syntax table.
This is the one specified by the current buffer.  */)
     ()
Richard M. Stallman's avatar
Richard M. Stallman committed
758 759 760 761 762 763
{
  return current_buffer->syntax_table;
}

DEFUN ("standard-syntax-table", Fstandard_syntax_table,
   Sstandard_syntax_table, 0, 0, 0,
764 765 766
       doc: /* Return the standard syntax table.
This is the one used for new buffers.  */)
     ()
Richard M. Stallman's avatar
Richard M. Stallman committed
767 768 769 770 771
{
  return Vstandard_syntax_table;
}

DEFUN ("copy-syntax-table", Fcopy_syntax_table, Scopy_syntax_table, 0, 1, 0,
772 773 774
       doc: /* Construct a new syntax table and return it.
It is a copy of the TABLE, which defaults to the standard syntax table.  */)
     (table)
Richard M. Stallman's avatar
Richard M. Stallman committed
775 776
     Lisp_Object table;
{
777 778
  Lisp_Object copy;

Jim Blandy's avatar
Jim Blandy committed
779
  if (!NILP (table))
780 781 782 783 784
    check_syntax_table (table);
  else
    table = Vstandard_syntax_table;

  copy = Fcopy_sequence (table);
785 786 787 788 789 790 791 792 793 794

  /* Only the standard syntax table should have a default element.
     Other syntax tables should inherit from parents instead.  */
  XCHAR_TABLE (copy)->defalt = Qnil;

  /* Copied syntax tables should all have parents.
     If we copied one with no parent, such as the standard syntax table,
     use the standard syntax table as the copy's parent.  */
  if (NILP (XCHAR_TABLE (copy)->parent))
    Fset_char_table_parent (copy, Vstandard_syntax_table);
795
  return copy;
Richard M. Stallman's avatar
Richard M. Stallman committed
796 797 798
}

DEFUN ("set-syntax-table", Fset_syntax_table, Sset_syntax_table, 1, 1, 0,
799 800 801
       doc: /* Select a new syntax table for the current buffer.
One argument, a syntax table.  */)
     (table)
Richard M. Stallman's avatar
Richard M. Stallman committed
802 803
     Lisp_Object table;
{
804
  int idx;
805
  check_syntax_table (table);
Richard M. Stallman's avatar
Richard M. Stallman committed
806 807
  current_buffer->syntax_table = table;
  /* Indicate that this buffer now has a specified syntax table.  */
Gerd Moellmann's avatar
Gerd Moellmann committed
808 809
  idx = PER_BUFFER_VAR_IDX (syntax_table);
  SET_PER_BUFFER_VALUE_P (current_buffer, idx, 1);
Richard M. Stallman's avatar
Richard M. Stallman committed
810 811 812 813 814
  return table;
}

/* Convert a letter which signifies a syntax code
 into the code it signifies.
815
 This is used by modify-syntax-entry, and other things.  */
Richard M. Stallman's avatar
Richard M. Stallman committed
816 817 818 819 820 821

unsigned char syntax_spec_code[0400] =
  { 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
    0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
    0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
    0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
822
    (char) Swhitespace, (char) Scomment_fence, (char) Sstring, 0377,
Richard M. Stallman's avatar
Richard M. Stallman committed
823 824 825 826 827 828
        (char) Smath, 0377, 0377, (char) Squote,
    (char) Sopen, (char) Sclose, 0377, 0377,
	0377, (char) Swhitespace, (char) Spunct, (char) Scharquote,
    0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
    0377, 0377, 0377, 0377,
	(char) Scomment, 0377, (char) Sendcomment, 0377,
829
    (char) Sinherit, 0377, 0377, 0377, 0377, 0377, 0377, 0377,   /* @, A ... */
Richard M. Stallman's avatar
Richard M. Stallman committed
830 831 832 833 834 835
    0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
    0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword,
    0377, 0377, 0377, 0377, (char) Sescape, 0377, 0377, (char) Ssymbol,
    0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,   /* `, a, ... */
    0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
    0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword,
836
    0377, 0377, 0377, 0377, (char) Sstring_fence, 0377, 0377, 0377
Richard M. Stallman's avatar
Richard M. Stallman committed
837 838
  };

839
/* Indexed by syntax code, give the letter that describes it.  */
Richard M. Stallman's avatar
Richard M. Stallman committed
840

841
char syntax_code_spec[16] =
Richard M. Stallman's avatar
Richard M. Stallman committed
842
  {
843 844
    ' ', '.', 'w', '_', '(', ')', '\'', '\"', '$', '\\', '/', '<', '>', '@',
    '!', '|'
Richard M. Stallman's avatar
Richard M. Stallman committed
845
  };
Karl Heuer's avatar
Karl Heuer committed
846 847 848 849 850 851 852 853

/* Indexed by syntax code, give the object (cons of syntax code and
   nil) to be stored in syntax table.  Since these objects can be
   shared among syntax tables, we generate them in advance.  By
   sharing objects, the function `describe-syntax' can give a more
   compact listing.  */
static Lisp_Object Vsyntax_code_object;

Richard M. Stallman's avatar
Richard M. Stallman committed
854 855

DEFUN ("char-syntax", Fchar_syntax, Schar_syntax, 1, 1, 0,
856 857 858 859 860 861
       doc: /* Return the syntax code of CHARACTER, described by a character.
For example, if CHARACTER is a word constituent,
the character `w' is returned.
The characters that correspond to various syntax codes
are listed in the documentation of `modify-syntax-entry'.  */)
     (character)
862
     Lisp_Object character;
Richard M. Stallman's avatar
Richard M. Stallman committed
863
{
864
  int char_int;
865 866 867
  gl_state.current_syntax_table = current_buffer->syntax_table;

  gl_state.use_global = 0;
868
  CHECK_NUMBER (character);
869
  char_int = XINT (character);
870
  return make_number (syntax_code_spec[(int) SYNTAX (char_int)]);
871 872 873
}

DEFUN ("matching-paren", Fmatching_paren, Smatching_paren, 1, 1, 0,
874 875
       doc: /* Return the matching parenthesis of CHARACTER, or nil if none.  */)
     (character)
876
     Lisp_Object character;
877
{
878
  int char_int, code;
879 880
  gl_state.current_syntax_table = current_buffer->syntax_table;
  gl_state.use_global = 0;
881
  CHECK_NUMBER (character);
882
  char_int = XINT (character);
883
  code = SYNTAX (char_int);
884
  if (code == Sopen || code == Sclose)
885
    return SYNTAX_MATCH (char_int);
886
  return Qnil;
Richard M. Stallman's avatar
Richard M. Stallman committed
887 888
}

889
DEFUN ("string-to-syntax", Fstring_to_syntax, Sstring_to_syntax, 1, 1, 0,
890 891 892
       doc: /* Convert a syntax specification STRING into syntax cell form.
STRING should be a string as it is allowed as argument of
`modify-syntax-entry'.  Value is the equivalent cons cell
893
\(CODE . MATCHING-CHAR) that can be used as value of a `syntax-table'
894 895
text property.  */)
     (string)
896
     Lisp_Object string;
Richard M. Stallman's avatar
Richard M. Stallman committed
897
{
898
  register const unsigned char *p;
Richard M. Stallman's avatar
Richard M. Stallman committed
899
  register enum syntaxcode code;
900
  int val;
901
  Lisp_Object match;
Richard M. Stallman's avatar
Richard M. Stallman committed
902

903
  CHECK_STRING (string);
Richard M. Stallman's avatar
Richard M. Stallman committed
904

905
  p = SDATA (string);
Richard M. Stallman's avatar
Richard M. Stallman committed
906 907
  code = (enum syntaxcode) syntax_spec_code[*p++];
  if (((int) code & 0377) == 0377)
908
    error ("Invalid syntax description letter: %c", p[-1]);
Richard M. Stallman's avatar
Richard M. Stallman committed
909

910
  if (code == Sinherit)
911
    return Qnil;
912 913

  if (*p)
914
    {
Karl Heuer's avatar
Karl Heuer committed
915
      int len;
916
      int character = (STRING_CHAR_AND_LENGTH
917
		       (p, SBYTES (string) - 1, len));
Karl Heuer's avatar
Karl Heuer committed
918
      XSETINT (match, character);
919 920
      if (XFASTINT (match) == ' ')
	match = Qnil;
Karl Heuer's avatar
Karl Heuer committed
921
      p += len;
922 923
    }
  else
924
    match = Qnil;
Richard M. Stallman's avatar
Richard M. Stallman committed
925

926
  val = (int) code;
Richard M. Stallman's avatar
Richard M. Stallman committed
927 928 929 930
  while (*p)
    switch (*p++)
      {
      case '1':
931
	val |= 1 << 16;
Richard M. Stallman's avatar
Richard M. Stallman committed
932 933 934
	break;

      case '2':
935
	val |= 1 << 17;
Richard M. Stallman's avatar
Richard M. Stallman committed
936 937 938
	break;

      case '3':
939
	val |= 1 << 18;
Richard M. Stallman's avatar
Richard M. Stallman committed
940 941 942
	break;

      case '4':
943
	val |= 1 << 19;
Richard M. Stallman's avatar
Richard M. Stallman committed
944 945 946
	break;

      case 'p':
947
	val |= 1 << 20;
Richard M. Stallman's avatar
Richard M. Stallman committed
948
	break;
949 950

      case 'b':
951
	val |= 1 << 21;
952
	break;
953 954 955 956

      case 'n':
	val |= 1 << 22;
	break;
Richard M. Stallman's avatar
Richard M. Stallman committed
957
      }
Juanma Barranquero's avatar
Juanma Barranquero committed
958

Karl Heuer's avatar
Karl Heuer committed
959
  if (val < XVECTOR (Vsyntax_code_object)->size && NILP (match))
960
    return XVECTOR (Vsyntax_code_object)->contents[val];
Karl Heuer's avatar
Karl Heuer committed
961 962
  else
    /* Since we can't use a shared object, let's make a new one.  */
963 964 965
    return Fcons (make_number (val), match);
}

966
/* I really don't know why this is interactive
Stefan Monnier's avatar
Stefan Monnier committed
967
   help-form should at least be made useful whilst reading the second arg.  */
Juanma Barranquero's avatar
Juanma Barranquero committed
968
DEFUN ("modify-syntax-entry", Fmodify_syntax_entry, Smodify_syntax_entry, 2, 3,
969
  "cSet syntax for character: \nsSet syntax for %s to: ",
970
       doc: /* Set syntax for character CHAR according to string NEWENTRY.
971
The syntax is changed only for table SYNTAX-TABLE, which defaults to
972
 the current buffer's syntax table.
973 974
CHAR may be a cons (MIN . MAX), in which case, syntaxes of all characters
in the range MIN and MAX are changed.
975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990
The first character of NEWENTRY should be one of the following:
  Space or -  whitespace syntax.    w   word constituent.
  _           symbol constituent.   .   punctuation.
  (           open-parenthesis.     )   close-parenthesis.
  "           string quote.         \\   escape.
  $           paired delimiter.     '   expression quote or prefix operator.
  <           comment starter.      >   comment ender.
  /           character-quote.      @   inherit from `standard-syntax-table'.
  |           generic string fence. !   generic comment fence.

Only single-character comment start and end sequences are represented thus.
Two-character sequences are represented as described below.
The second character of NEWENTRY is the matching parenthesis,
 used only if the first character is `(' or `)'.
Any additional characters are flags.
Defined flags are the characters 1, 2, 3, 4, b, p, and n.
991 992 993 994
 1 means CHAR is the start of a two-char comment start sequence.
 2 means CHAR is the second character of such a sequence.
 3 means CHAR is the start of a two-char comment end sequence.
 4 means CHAR is the second character of such a sequence.
995 996 997 998 999 1000

There can be up to two orthogonal comment sequences.  This is to support
language modes such as C++.  By default, all comment sequences are of style
a, but you can set the comment sequence style to b (on the second character
of a comment-start, or the first character of a comment-end sequence) using
this flag:
1001 1002
 b means CHAR is part of comment sequence b.
 n means CHAR is part of a nestable comment sequence.
1003

1004
 p means CHAR is a prefix character for `backward-prefix-chars';
1005
   such characters are treated as whitespace when they occur
1006 1007
   between expressions.
usage: (modify-syntax-entry CHAR NEWENTRY &optional SYNTAX-TABLE) */)
1008
     (c, newentry, syntax_table)
1009 1010
     Lisp_Object c, newentry, syntax_table;
{
1011 1012
  if (CONSP (c))
    {
Kenichi Handa's avatar
Kenichi Handa committed
1013 1014
      CHECK_CHARACTER_CAR (c);
      CHECK_CHARACTER_CDR (c);
1015 1016 1017
    }
  else
    CHECK_CHARACTER (c);
1018 1019 1020 1021 1022

  if (NILP (syntax_table))
    syntax_table = current_buffer->syntax_table;
  else
    check_syntax_table (syntax_table);
Richard M. Stallman's avatar
Richard M. Stallman committed
1023

1024 1025 1026 1027 1028
  newentry = Fstring_to_syntax (newentry);
  if (CONSP (c))
    SET_RAW_SYNTAX_ENTRY_RANGE (syntax_table, c, newentry);
  else
    SET_RAW_SYNTAX_ENTRY (syntax_table, XINT (c), newentry);
1029 1030 1031 1032 1033

  /* We clear the regexp cache, since character classes can now have
     different values from those in the compiled regexps.*/
  clear_regexp_cache ();

Richard M. Stallman's avatar
Richard M. Stallman committed
1034 1035 1036 1037 1038
  return Qnil;
}

/* Dump syntax table to buffer in human-readable format */

1039 1040 1041 1042 1043
DEFUN ("internal-describe-syntax-value", Finternal_describe_syntax_value,
       Sinternal_describe_syntax_value, 1, 1, 0,
       doc: /* Insert a description of the internal syntax description SYNTAX at point.  */)
     (syntax)
     Lisp_Object syntax;
Richard M. Stallman's avatar
Richard M. Stallman committed
1044 1045
{
  register enum syntaxcode code;
1046
  char desc, start1, start2, end1, end2, prefix, comstyle, comnested;
Richard M. Stallman's avatar
Richard M. Stallman committed
1047
  char str[2];
1048
  Lisp_Object first, match_lisp, value = syntax;
Richard M. Stallman's avatar
Richard M. Stallman committed
1049

1050 1051
  if (NILP (value))
    {
1052 1053
      insert_string ("default");
      return syntax;
1054 1055
    }

1056 1057
  if (CHAR_TABLE_P (value))
    {
1058 1059
      insert_string ("deeper char-table ...");
      return syntax;
1060 1061
    }

1062 1063
  if (!CONSP (value))
    {
1064