composite.c 52.5 KB
Newer Older
Kenichi Handa's avatar
Kenichi Handa committed
1
/* Composite sequence support.
2
   Copyright (C) 2001, 2002, 2003, 2004, 2005,
Glenn Morris's avatar
Glenn Morris committed
3 4
                 2006, 2007, 2008 Free Software Foundation, Inc.
   Copyright (C) 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
Kenichi Handa's avatar
Kenichi Handa committed
5 6
     National Institute of Advanced Industrial Science and Technology (AIST)
     Registration Number H14PRO021
7
   Copyright (C) 2003, 2006
8 9
     National Institute of Advanced Industrial Science and Technology (AIST)
     Registration Number H13PRO009
Kenichi Handa's avatar
Kenichi Handa committed
10 11 12

This file is part of GNU Emacs.

13
GNU Emacs is free software: you can redistribute it and/or modify
Kenichi Handa's avatar
Kenichi Handa committed
14
it under the terms of the GNU General Public License as published by
15 16
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Kenichi Handa's avatar
Kenichi Handa committed
17 18 19 20 21 22 23

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
24
along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.  */
Kenichi Handa's avatar
Kenichi Handa committed
25 26 27 28

#include <config.h>
#include "lisp.h"
#include "buffer.h"
29
#include "character.h"
Kenichi Handa's avatar
Kenichi Handa committed
30
#include "intervals.h"
31 32 33 34
#include "window.h"
#include "frame.h"
#include "dispextern.h"
#include "font.h"
Kenichi Handa's avatar
Kenichi Handa committed
35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101

/* Emacs uses special text property `composition' to support character
   composition.  A sequence of characters that have the same (i.e. eq)
   `composition' property value is treated as a single composite
   sequence (we call it just `composition' here after).  Characters in
   a composition are all composed somehow on the screen.

   The property value has this form when the composition is made:
	((LENGTH . COMPONENTS) . MODIFICATION-FUNC)
   then turns to this form:
	(COMPOSITION-ID . (LENGTH COMPONENTS-VEC . MODIFICATION-FUNC))
   when the composition is registered in composition_hash_table and
   composition_table.  These rather peculiar structures were designed
   to make it easy to distinguish them quickly (we can do that by
   checking only the first element) and to extract LENGTH (from the
   former form) and COMPOSITION-ID (from the latter form).

   We register a composition when it is displayed, or when the width
   is required (for instance, to calculate columns).

   LENGTH -- Length of the composition.  This information is used to
	check the validity of the composition.

   COMPONENTS --  Character, string, vector, list, or nil.

	If it is nil, characters in the text are composed relatively
	according to their metrics in font glyphs.

	If it is a character or a string, the character or characters
	in the string are composed relatively.

	If it is a vector or list of integers, the element is a
	character or an encoded composition rule.  The characters are
	composed according to the rules.  (2N)th elements are
	characters to be composed and (2N+1)th elements are
	composition rules to tell how to compose (2N+2)th element with
	the previously composed 2N glyphs.

   COMPONENTS-VEC -- Vector of integers.  In relative composition, the
	elements are characters to be composed.  In rule-base
	composition, the elements are characters or encoded
	composition rules.

   MODIFICATION-FUNC -- If non nil, it is a function to call when the
	composition gets invalid after a modification in a buffer.  If
	it is nil, a function in `composition-function-table' of the
	first character in the sequence is called.

   COMPOSITION-ID --Identification number of the composition.  It is
	used as an index to composition_table for the composition.

   When Emacs has to display a composition or has to know its
   displaying width, the function get_composition_id is called.  It
   returns COMPOSITION-ID so that the caller can access the
   information about the composition through composition_table.  If a
   COMPOSITION-ID has not yet been assigned to the composition,
   get_composition_id checks the validity of `composition' property,
   and, if valid, assigns a new ID, registers the information in
   composition_hash_table and composition_table, and changes the form
   of the property value.  If the property is invalid, return -1
   without changing the property value.

   We use two tables to keep information about composition;
   composition_hash_table and composition_table.

   The former is a hash table in which keys are COMPONENTS-VECs and
   values are the corresponding COMPOSITION-IDs.  This hash table is
102
   weak, but as each key (COMPONENTS-VEC) is also kept as a value of the
Kenichi Handa's avatar
Kenichi Handa committed
103
   `composition' property, it won't be collected as garbage until all
104
   bits of text that have the same COMPONENTS-VEC are deleted.
Kenichi Handa's avatar
Kenichi Handa committed
105 106

   The latter is a table of pointers to `struct composition' indexed
107
   by COMPOSITION-ID.  This structure keeps the other information (see
Kenichi Handa's avatar
Kenichi Handa committed
108 109 110 111
   composite.h).

   In general, a text property holds information about individual
   characters.  But, a `composition' property holds information about
112
   a sequence of characters (in this sense, it is like the `intangible'
Kenichi Handa's avatar
Kenichi Handa committed
113
   property).  That means that we should not share the property value
114
   in adjacent compositions -- we can't distinguish them if they have the
Kenichi Handa's avatar
Kenichi Handa committed
115 116 117 118 119 120
   same property.  So, after any changes, we call
   `update_compositions' and change a property of one of adjacent
   compositions to a copy of it.  This function also runs a proper
   composition modification function to make a composition that gets
   invalid by the change valid again.

121
   As the value of the `composition' property holds information about a
Kenichi Handa's avatar
Kenichi Handa committed
122
   specific range of text, the value gets invalid if we change the
123
   text in the range.  We treat the `composition' property as always
Kenichi Handa's avatar
Kenichi Handa committed
124 125 126
   rear-nonsticky (currently by setting default-text-properties to
   (rear-nonsticky (composition))) and we never make properties of
   adjacent compositions identical.  Thus, any such changes make the
127
   range just shorter.  So, we can check the validity of the `composition'
Kenichi Handa's avatar
Kenichi Handa committed
128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154
   property by comparing LENGTH information with the actual length of
   the composition.

*/


Lisp_Object Qcomposition;

/* Table of pointers to the structure `composition' indexed by
   COMPOSITION-ID.  This structure is for storing information about
   each composition except for COMPONENTS-VEC.  */
struct composition **composition_table;

/* The current size of `composition_table'.  */
static int composition_table_size;

/* Number of compositions currently made. */
int n_compositions;

/* Hash table for compositions.  The key is COMPONENTS-VEC of
   `composition' property.  The value is the corresponding
   COMPOSITION-ID.  */
Lisp_Object composition_hash_table;

/* Function to call to adjust composition.  */
Lisp_Object Vcompose_chars_after_function;

Kenichi Handa's avatar
Kenichi Handa committed
155 156 157
Lisp_Object Qauto_composed;
Lisp_Object Vauto_composition_function;
Lisp_Object Qauto_composition_function;
158
Lisp_Object Vcomposition_function_table;
159

160 161
EXFUN (Fremove_list_of_text_properties, 4);

Kenichi Handa's avatar
Kenichi Handa committed
162 163
/* Temporary variable used in macros COMPOSITION_XXX.  */
Lisp_Object composition_temp;
164

Kenichi Handa's avatar
Kenichi Handa committed
165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250

/* Return COMPOSITION-ID of a composition at buffer position
   CHARPOS/BYTEPOS and length NCHARS.  The `composition' property of
   the sequence is PROP.  STRING, if non-nil, is a string that
   contains the composition instead of the current buffer.

   If the composition is invalid, return -1.  */

int
get_composition_id (charpos, bytepos, nchars, prop, string)
     int charpos, bytepos, nchars;
     Lisp_Object prop, string;
{
  Lisp_Object id, length, components, key, *key_contents;
  int glyph_len;
  struct Lisp_Hash_Table *hash_table = XHASH_TABLE (composition_hash_table);
  int hash_index;
  unsigned hash_code;
  struct composition *cmp;
  int i, ch;

  /* PROP should be
	Form-A: ((LENGTH . COMPONENTS) . MODIFICATION-FUNC)
     or
	Form-B: (COMPOSITION-ID . (LENGTH COMPONENTS-VEC . MODIFICATION-FUNC))
  */
  if (nchars == 0 || !CONSP (prop))
    goto invalid_composition;

  id = XCAR (prop);
  if (INTEGERP (id))
    {
      /* PROP should be Form-B.  */
      if (XINT (id) < 0 || XINT (id) >= n_compositions)
	goto invalid_composition;
      return XINT (id);
    }

  /* PROP should be Form-A.
     Thus, ID should be (LENGTH . COMPONENTS).  */
  if (!CONSP (id))
    goto invalid_composition;
  length = XCAR (id);
  if (!INTEGERP (length) || XINT (length) != nchars)
    goto invalid_composition;

  components = XCDR (id);

  /* Check if the same composition has already been registered or not
     by consulting composition_hash_table.  The key for this table is
     COMPONENTS (converted to a vector COMPONENTS-VEC) or, if it is
     nil, vector of characters in the composition range.  */
  if (INTEGERP (components))
    key = Fmake_vector (make_number (1), components);
  else if (STRINGP (components) || CONSP (components))
    key = Fvconcat (1, &components);
  else if (VECTORP (components))
    key = components;
  else if (NILP (components))
    {
      key = Fmake_vector (make_number (nchars), Qnil);
      if (STRINGP (string))
	for (i = 0; i < nchars; i++)
	  {
	    FETCH_STRING_CHAR_ADVANCE (ch, string, charpos, bytepos);
	    XVECTOR (key)->contents[i] = make_number (ch);
	  }
      else
	for (i = 0; i < nchars; i++)
	  {
	    FETCH_CHAR_ADVANCE (ch, charpos, bytepos);
	    XVECTOR (key)->contents[i] = make_number (ch);
	  }
    }
  else
    goto invalid_composition;

  hash_index = hash_lookup (hash_table, key, &hash_code);
  if (hash_index >= 0)
    {
      /* We have already registered the same composition.  Change PROP
	 from Form-A above to Form-B while replacing COMPONENTS with
	 COMPONENTS-VEC stored in the hash table.  We can directly
	 modify the cons cell of PROP because it is not shared.  */
      key = HASH_KEY (hash_table, hash_index);
      id = HASH_VALUE (hash_table, hash_index);
251 252
      XSETCAR (prop, id);
      XSETCDR (prop, Fcons (make_number (nchars), Fcons (key, XCDR (prop))));
Kenichi Handa's avatar
Kenichi Handa committed
253 254 255 256
      return XINT (id);
    }

  /* This composition is a new one.  We must register it.  */
257

Kenichi Handa's avatar
Kenichi Handa committed
258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279
  /* Check if we have sufficient memory to store this information.  */
  if (composition_table_size == 0)
    {
      composition_table_size = 256;
      composition_table
	= (struct composition **) xmalloc (sizeof (composition_table[0])
					   * composition_table_size);
    }
  else if (composition_table_size <= n_compositions)
    {
      composition_table_size += 256;
      composition_table
	= (struct composition **) xrealloc (composition_table,
					    sizeof (composition_table[0])
					    * composition_table_size);
    }

  key_contents = XVECTOR (key)->contents;

  /* Check if the contents of COMPONENTS are valid if COMPONENTS is a
     vector or a list.  It should be a sequence of:
	char1 rule1 char2 rule2 char3 ...    ruleN charN+1  */
280

281
  if (VECTORP (components)
282 283 284 285 286 287 288 289 290 291
      && ASIZE (components) >= 2
      && VECTORP (AREF (components, 0)))
    {
      /* COMPONENTS is a glyph-string.  */
      int len = ASIZE (key);

      for (i = 1; i < len; i++)
	if (! VECTORP (AREF (key, i)))
	  goto invalid_composition;
    }
292
  else if (VECTORP (components) || CONSP (components))
Kenichi Handa's avatar
Kenichi Handa committed
293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310
    {
      int len = XVECTOR (key)->size;

      /* The number of elements should be odd.  */
      if ((len % 2) == 0)
	goto invalid_composition;
      /* All elements should be integers (character or encoded
         composition rule).  */
      for (i = 0; i < len; i++)
	{
	  if (!INTEGERP (key_contents[i]))
	    goto invalid_composition;
	}
    }

  /* Change PROP from Form-A above to Form-B.  We can directly modify
     the cons cell of PROP because it is not shared.  */
  XSETFASTINT (id, n_compositions);
311 312
  XSETCAR (prop, id);
  XSETCDR (prop, Fcons (make_number (nchars), Fcons (key, XCDR (prop))));
Kenichi Handa's avatar
Kenichi Handa committed
313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332

  /* Register the composition in composition_hash_table.  */
  hash_index = hash_put (hash_table, key, id, hash_code);

  /* Register the composition in composition_table.  */
  cmp = (struct composition *) xmalloc (sizeof (struct composition));

  cmp->method = (NILP (components)
		 ? COMPOSITION_RELATIVE
		 : ((INTEGERP (components) || STRINGP (components))
		    ? COMPOSITION_WITH_ALTCHARS
		    : COMPOSITION_WITH_RULE_ALTCHARS));
  cmp->hash_index = hash_index;
  glyph_len = (cmp->method == COMPOSITION_WITH_RULE_ALTCHARS
	       ? (XVECTOR (key)->size + 1) / 2
	       : XVECTOR (key)->size);
  cmp->glyph_len = glyph_len;
  cmp->offsets = (short *) xmalloc (sizeof (short) * glyph_len * 2);
  cmp->font = NULL;

333
  if (cmp->method != COMPOSITION_WITH_RULE_ALTCHARS)
Kenichi Handa's avatar
Kenichi Handa committed
334 335 336 337 338 339 340
    {
      /* Relative composition.  */
      cmp->width = 0;
      for (i = 0; i < glyph_len; i++)
	{
	  int this_width;
	  ch = XINT (key_contents[i]);
341
	  this_width = (ch == '\t' ? 1 : CHAR_WIDTH (ch));
Kenichi Handa's avatar
Kenichi Handa committed
342 343 344 345 346 347 348 349 350 351
	  if (cmp->width < this_width)
	    cmp->width = this_width;
	}
    }
  else
    {
      /* Rule-base composition.  */
      float leftmost = 0.0, rightmost;

      ch = XINT (key_contents[0]);
352
      rightmost = ch != '\t' ? CHAR_WIDTH (ch) : 1;
Kenichi Handa's avatar
Kenichi Handa committed
353 354 355

      for (i = 1; i < glyph_len; i += 2)
	{
356
	  int rule, gref, nref, xoff, yoff;
Kenichi Handa's avatar
Kenichi Handa committed
357 358 359 360 361
	  int this_width;
	  float this_left;

	  rule = XINT (key_contents[i]);
	  ch = XINT (key_contents[i + 1]);
362
	  this_width = ch != '\t' ? CHAR_WIDTH (ch) : 1;
Kenichi Handa's avatar
Kenichi Handa committed
363 364 365 366 367 368 369 370 371 372 373 374 375 376 377

	  /* A composition rule is specified by an integer value
	     that encodes global and new reference points (GREF and
	     NREF).  GREF and NREF are specified by numbers as
	     below:
		0---1---2 -- ascent
		|       |
		|       |
		|       |
		9--10--11 -- center
		|       |
	     ---3---4---5--- baseline
		|       |
		6---7---8 -- descent
	  */
378
	  COMPOSITION_DECODE_RULE (rule, gref, nref, xoff, yoff);
Kenichi Handa's avatar
Kenichi Handa committed
379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404
	  this_left = (leftmost
		       + (gref % 3) * (rightmost - leftmost) / 2.0
		       - (nref % 3) * this_width / 2.0);

	  if (this_left < leftmost)
	    leftmost = this_left;
	  if (this_left + this_width > rightmost)
	    rightmost = this_left + this_width;
	}

      cmp->width = rightmost - leftmost;
      if (cmp->width < (rightmost - leftmost))
	/* To get a ceiling integer value.  */
	cmp->width++;
    }

  composition_table[n_compositions] = cmp;

  return n_compositions++;

 invalid_composition:
  /* Would it be better to remove this `composition' property?  */
  return -1;
}


405 406
/* Find a static composition at or nearest to position POS of OBJECT
   (buffer or string).
Kenichi Handa's avatar
Kenichi Handa committed
407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423

   OBJECT defaults to the current buffer.  If there's a composition at
   POS, set *START and *END to the start and end of the sequence,
   *PROP to the `composition' property, and return 1.

   If there's no composition at POS and LIMIT is negative, return 0.

   Otherwise, search for a composition forward (LIMIT > POS) or
   backward (LIMIT < POS).  In this case, LIMIT bounds the search.

   If a composition is found, set *START, *END, and *PROP as above,
   and return 1, else return 0.

   This doesn't check the validity of composition.  */

int
find_composition (pos, limit, start, end, prop, object)
424 425
     int pos, limit;
     EMACS_INT *start, *end;
Kenichi Handa's avatar
Kenichi Handa committed
426 427 428 429 430 431 432 433 434 435 436
     Lisp_Object *prop, object;
{
  Lisp_Object val;

  if (get_property_and_range (pos, Qcomposition, prop, start, end, object))
    return 1;

  if (limit < 0 || limit == pos)
    return 0;

  if (limit > pos)		/* search forward */
437 438 439 440 441 442 443
    {
      val = Fnext_single_property_change (make_number (pos), Qcomposition,
					  object, make_number (limit));
      pos = XINT (val);
      if (pos == limit)
	return 0;
    }
Kenichi Handa's avatar
Kenichi Handa committed
444
  else				/* search backward */
445 446 447 448 449 450 451 452 453 454 455
    {
      if (get_property_and_range (pos - 1, Qcomposition, prop, start, end,
				  object))
	return 1;
      val = Fprevious_single_property_change (make_number (pos), Qcomposition,
					      object, make_number (limit));
      pos = XINT (val);
      if (pos == limit)
	return 0;
      pos--;
    }
Kenichi Handa's avatar
Kenichi Handa committed
456 457 458 459 460 461 462 463 464 465 466 467
  get_property_and_range (pos, Qcomposition, prop, start, end, object);
  return 1;
}

/* Run a proper function to adjust the composition sitting between
   FROM and TO with property PROP.  */

static void
run_composition_function (from, to, prop)
     int from, to;
     Lisp_Object prop;
{
468
  Lisp_Object func;
469
  EMACS_INT start, end;
Kenichi Handa's avatar
Kenichi Handa committed
470 471 472 473 474 475 476 477 478 479 480 481

  func = COMPOSITION_MODIFICATION_FUNC (prop);
  /* If an invalid composition precedes or follows, try to make them
     valid too.  */
  if (from > BEGV
      && find_composition (from - 1, -1, &start, &end, &prop, Qnil)
      && !COMPOSITION_VALID_P (start, end, prop))
    from = start;
  if (to < ZV
      && find_composition (to, -1, &start, &end, &prop, Qnil)
      && !COMPOSITION_VALID_P (start, end, prop))
    to = end;
Kenichi Handa's avatar
Kenichi Handa committed
482
  if (!NILP (Ffboundp (func)))
Kenichi Handa's avatar
Kenichi Handa committed
483 484 485 486 487 488 489
    call2 (func, make_number (from), make_number (to));
}

/* Make invalid compositions adjacent to or inside FROM and TO valid.
   CHECK_MASK is bitwise `or' of mask bits defined by macros
   CHECK_XXX (see the comment in composite.h).

490
   It also resets the text-property `auto-composed' to a proper region
Kenichi Handa's avatar
Kenichi Handa committed
491 492
   so that automatic character composition works correctly later while
   displaying the region.
Kenichi Handa's avatar
Kenichi Handa committed
493

Kenichi Handa's avatar
Kenichi Handa committed
494 495 496 497 498
   This function is called when a buffer text is changed.  If the
   change is deletion, FROM == TO.  Otherwise, FROM < TO.  */

void
update_compositions (from, to, check_mask)
499 500
     EMACS_INT from, to;
     int check_mask;
Kenichi Handa's avatar
Kenichi Handa committed
501
{
502
  Lisp_Object prop;
503
  EMACS_INT start, end;
Kenichi Handa's avatar
Kenichi Handa committed
504 505
  /* The beginning and end of the region to set the property
     `auto-composed' to nil.  */
506
  EMACS_INT min_pos = from, max_pos = to;
Kenichi Handa's avatar
Kenichi Handa committed
507

508 509
  if (inhibit_modification_hooks)
    return;
510

511 512 513 514
  /* If FROM and TO are not in a valid range, do nothing.  */
  if (! (BEGV <= from && from <= to && to <= ZV))
    return;

Kenichi Handa's avatar
Kenichi Handa committed
515 516 517 518 519 520 521 522
  if (check_mask & CHECK_HEAD)
    {
      /* FROM should be at composition boundary.  But, insertion or
	 deletion will make two compositions adjacent and
	 indistinguishable when they have same (eq) property.  To
	 avoid it, in such a case, we change the property of the
	 latter to the copy of it.  */
      if (from > BEGV
523 524
	  && find_composition (from - 1, -1, &start, &end, &prop, Qnil)
	  && COMPOSITION_VALID_P (start, end, prop))
Kenichi Handa's avatar
Kenichi Handa committed
525
	{
Kenichi Handa's avatar
Kenichi Handa committed
526 527 528
	  min_pos = start;
	  if (end > to)
	    max_pos = end;
Kenichi Handa's avatar
Kenichi Handa committed
529 530 531 532 533 534 535
	  if (from < end)
	    Fput_text_property (make_number (from), make_number (end),
				Qcomposition,
				Fcons (XCAR (prop), XCDR (prop)), Qnil);
	  run_composition_function (start, end, prop);
	  from = end;
	}
536
      else if (from < ZV
537
	       && find_composition (from, -1, &start, &from, &prop, Qnil)
538
	       && COMPOSITION_VALID_P (start, from, prop))
Kenichi Handa's avatar
Kenichi Handa committed
539 540 541 542 543
	{
	  if (from > to)
	    max_pos = from;
	  run_composition_function (start, from, prop);
	}
Kenichi Handa's avatar
Kenichi Handa committed
544 545 546 547 548 549 550 551 552
    }

  if (check_mask & CHECK_INSIDE)
    {
      /* In this case, we are sure that (check & CHECK_TAIL) is also
         nonzero.  Thus, here we should check only compositions before
         (to - 1).  */
      while (from < to - 1
	     && find_composition (from, to, &start, &from, &prop, Qnil)
553
	     && COMPOSITION_VALID_P (start, from, prop)
Kenichi Handa's avatar
Kenichi Handa committed
554 555 556 557 558 559 560
	     && from < to - 1)
	run_composition_function (start, from, prop);
    }

  if (check_mask & CHECK_TAIL)
    {
      if (from < to
561 562
	  && find_composition (to - 1, -1, &start, &end, &prop, Qnil)
	  && COMPOSITION_VALID_P (start, end, prop))
Kenichi Handa's avatar
Kenichi Handa committed
563 564 565 566 567 568 569
	{
	  /* TO should be also at composition boundary.  But,
	     insertion or deletion will make two compositions adjacent
	     and indistinguishable when they have same (eq) property.
	     To avoid it, in such a case, we change the property of
	     the former to the copy of it.  */
	  if (to < end)
Kenichi Handa's avatar
Kenichi Handa committed
570 571 572 573 574 575
	    {
	      Fput_text_property (make_number (start), make_number (to),
				  Qcomposition,
				  Fcons (XCAR (prop), XCDR (prop)), Qnil);
	      max_pos = end;
	    }
Kenichi Handa's avatar
Kenichi Handa committed
576 577 578
	  run_composition_function (start, end, prop);
	}
      else if (to < ZV
579 580
	       && find_composition (to, -1, &start, &end, &prop, Qnil)
	       && COMPOSITION_VALID_P (start, end, prop))
Kenichi Handa's avatar
Kenichi Handa committed
581 582 583 584
	{
	  run_composition_function (start, end, prop);
	  max_pos = end;
	}
Kenichi Handa's avatar
Kenichi Handa committed
585
    }
Kenichi Handa's avatar
Kenichi Handa committed
586
  if (min_pos < max_pos)
587 588 589 590 591 592 593 594 595 596 597
    {
      int count = SPECPDL_INDEX ();

      specbind (Qinhibit_read_only, Qt);
      specbind (Qinhibit_modification_hooks, Qt);
      specbind (Qinhibit_point_motion_hooks, Qt);
      Fremove_list_of_text_properties (make_number (min_pos),
				       make_number (max_pos),
				       Fcons (Qauto_composed, Qnil), Qnil);
      unbind_to (count, Qnil);
    }
Kenichi Handa's avatar
Kenichi Handa committed
598 599
}

600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617

/* Modify composition property values in LIST destructively.  LIST is
   a list as returned from text_property_list.  Change values to the
   top-level copies of them so that none of them are `eq'.  */

void
make_composition_value_copy (list)
     Lisp_Object list;
{
  Lisp_Object plist, val;

  for (; CONSP (list); list = XCDR (list))
    {
      plist = XCAR (XCDR (XCDR (XCAR (list))));
      while (CONSP (plist) && CONSP (XCDR (plist)))
	{
	  if (EQ (XCAR (plist), Qcomposition)
	      && (val = XCAR (XCDR (plist)), CONSP (val)))
618
	    XSETCAR (XCDR (plist), Fcons (XCAR (val), XCDR (val)));
619 620 621 622 623 624
	  plist = XCDR (XCDR (plist));
	}
    }
}


Kenichi Handa's avatar
Kenichi Handa committed
625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642
/* Make text in the region between START and END a composition that
   has COMPONENTS and MODIFICATION-FUNC.

   If STRING is non-nil, then operate on characters contained between
   indices START and END in STRING.  */

void
compose_text (start, end, components, modification_func, string)
     int start, end;
     Lisp_Object components, modification_func, string;
{
  Lisp_Object prop;

  prop = Fcons (Fcons (make_number (end - start), components),
		modification_func);
  Fput_text_property  (make_number (start), make_number (end),
		       Qcomposition, prop, string);
}
643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867


static Lisp_Object autocmp_chars P_ ((Lisp_Object, EMACS_INT, EMACS_INT,
				      EMACS_INT, struct window *,
				      struct face *, Lisp_Object));


/* Lisp glyph-string handlers */

/* Hash table for automatic composition.  The key is a header of a
   lgstring (Lispy glyph-string), and the value is a body of a
   lgstring.  */

static Lisp_Object gstring_hash_table;

static Lisp_Object gstring_lookup_cache P_ ((Lisp_Object));

static Lisp_Object
gstring_lookup_cache (header)
     Lisp_Object header;
{
  struct Lisp_Hash_Table *h = XHASH_TABLE (gstring_hash_table);
  int i = hash_lookup (h, header, NULL);

  return (i >= 0 ? HASH_VALUE (h, i) : Qnil);
}

Lisp_Object
composition_gstring_put_cache (gstring, len)
     Lisp_Object gstring;
     int len;
{
  struct Lisp_Hash_Table *h = XHASH_TABLE (gstring_hash_table);
  unsigned hash;
  Lisp_Object header, copy;
  int i;

  header = LGSTRING_HEADER (gstring);
  hash = h->hashfn (h, header);
  if (len < 0)
    {
      len = LGSTRING_GLYPH_LEN (gstring);
      for (i = 0; i < len; i++)
	if (NILP (LGSTRING_GLYPH (gstring, i)))
	  break;
      len = i;
    }
      
  copy = Fmake_vector (make_number (len + 2), Qnil);
  LGSTRING_SET_HEADER (copy, Fcopy_sequence (header));
  for (i = 0; i < len; i++)
    LGSTRING_SET_GLYPH (copy, i, Fcopy_sequence (LGSTRING_GLYPH (gstring, i)));
  i = hash_put (h, LGSTRING_HEADER (copy), copy, hash);
  LGSTRING_SET_ID (copy, make_number (i));
  return copy;
}

Lisp_Object
composition_gstring_from_id (id)
     int id;
{
  struct Lisp_Hash_Table *h = XHASH_TABLE (gstring_hash_table);

  return HASH_VALUE (h, id);
}

static Lisp_Object fill_gstring_header P_ ((Lisp_Object, Lisp_Object,
					    Lisp_Object, Lisp_Object,
					    Lisp_Object));

int
composition_gstring_p (gstring)
     Lisp_Object gstring;
{
  Lisp_Object header;
  int i;

  if (! VECTORP (gstring) || ASIZE (gstring) < 2)
    return 0;
  header = LGSTRING_HEADER (gstring);
  if (! VECTORP (header) || ASIZE (header) < 2)
    return 0;
  if (! NILP (LGSTRING_FONT (gstring))
      && ! FONT_OBJECT_P (LGSTRING_FONT (gstring)))
    return 0;
  for (i = 1; i < ASIZE (LGSTRING_HEADER (gstring)); i++)
    if (! NATNUMP (AREF (LGSTRING_HEADER (gstring), i)))
      return 0;
  if (! NILP (LGSTRING_ID (gstring)) && ! NATNUMP (LGSTRING_ID (gstring)))
    return 0;
  for (i = 0; i < LGSTRING_GLYPH_LEN (gstring); i++)
    {
      Lisp_Object glyph = LGSTRING_GLYPH (gstring, i);
      if (NILP (glyph))
	break;
      if (! VECTORP (glyph) || ASIZE (glyph) != LGLYPH_SIZE)
	return 0;
    }
  return 1;
}

int
composition_gstring_width (gstring, from, to, metrics)
     Lisp_Object gstring;
     int from, to;
     struct font_metrics *metrics;
{
  Lisp_Object *glyph;
  int width = 0;

  if (metrics)
    {
      Lisp_Object font_object = LGSTRING_FONT (gstring);
      struct font *font = XFONT_OBJECT (font_object);

      metrics->ascent = font->ascent;
      metrics->descent = font->descent;
      metrics->width = metrics->lbearing = metrics->rbearing = 0;
    }
  for (glyph = &LGSTRING_GLYPH (gstring, from); from < to; from++, glyph++)
    {
      int x;

      if (NILP (LGLYPH_ADJUSTMENT (*glyph)))
	width += LGLYPH_WIDTH (*glyph);
      else
	width += LGLYPH_WADJUST (*glyph);
      if (metrics)
	{
	  x = metrics->width + LGLYPH_LBEARING (*glyph) + LGLYPH_XOFF (*glyph);
	  if (metrics->lbearing > x)
	    metrics->lbearing = x;
	  x = metrics->width + LGLYPH_RBEARING (*glyph) + LGLYPH_XOFF (*glyph);
	  if (metrics->rbearing < x)
	    metrics->rbearing = x;
	  metrics->width = width;
	  x = LGLYPH_ASCENT (*glyph) - LGLYPH_YOFF (*glyph);
	  if (metrics->ascent < x)
	    metrics->ascent = x;
	  x = LGLYPH_DESCENT (*glyph) - LGLYPH_YOFF (*glyph);
	  if (metrics->descent < x)
	    metrics->descent = x;
	}
    }
  return width;
}


static Lisp_Object gstring_work;
static Lisp_Object gstring_work_headers;

static Lisp_Object
fill_gstring_header (header, start, end, font_object, string)
     Lisp_Object header, start, end, font_object, string;
{
  EMACS_INT from, to, from_byte;
  EMACS_INT len, i;

  if (NILP (string))
    {
      if (NILP (current_buffer->enable_multibyte_characters))
	error ("Attempt to shape unibyte text");
      validate_region (&start, &end);
      from = XFASTINT (start);
      to = XFASTINT (end);
      from_byte = CHAR_TO_BYTE (from);
    }
  else
    {
      CHECK_STRING (string);
      if (! STRING_MULTIBYTE (current_buffer->enable_multibyte_characters))
	error ("Attempt to shape unibyte text");
      CHECK_NATNUM (start);
      from = XINT (start);
      CHECK_NATNUM (end);
      to = XINT (end);
      if (from < 0 || from > to || to > SCHARS (string))
	args_out_of_range_3 (string, start, end);
      from_byte = string_char_to_byte (string, from);
    }

  len = to - from;
  if (len == 0)
    error ("Attempt to shape zero-length text");
  if (VECTORP (header))
    {
      if (ASIZE (header) != len + 1)
	args_out_of_range (header, make_number (len + 1));
    }
  else
    {
      if (len <= 8)
	header = AREF (gstring_work_headers, len - 1);
      else
	header = Fmake_vector (make_number (len + 1), Qnil);
    }

  ASET (header, 0, font_object);
  for (i = 0; i < len; i++)
    {
      int c;

      if (NILP (string))
	FETCH_CHAR_ADVANCE_NO_CHECK (c, from, from_byte);
      else
	FETCH_STRING_CHAR_ADVANCE_NO_CHECK (c, string, from, from_byte);
      ASET (header, i + 1, make_number (c));
    }
  return header;
}

extern void font_fill_lglyph_metrics P_ ((Lisp_Object, Lisp_Object));

static void
fill_gstring_body (gstring)
     Lisp_Object gstring;
{
  Lisp_Object font_object = LGSTRING_FONT (gstring);
  Lisp_Object header = AREF (gstring, 0);
  EMACS_INT len = LGSTRING_CHAR_LEN (gstring);
  EMACS_INT i;

  for (i = 0; i < len; i++)
    {
      Lisp_Object g = LGSTRING_GLYPH (gstring, i);
868
      EMACS_INT c = XINT (AREF (header, i + 1));
869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918

      if (NILP (g))
	{
	  g = LGLYPH_NEW ();
	  LGSTRING_SET_GLYPH (gstring, i, g);
	}
      LGLYPH_SET_FROM (g, i);
      LGLYPH_SET_TO (g, i);
      LGLYPH_SET_CHAR (g, c);
      if (! NILP (font_object))
	{
	  font_fill_lglyph_metrics (g, font_object);
	}
      else
	{
	  int width = XFASTINT (CHAR_TABLE_REF (Vchar_width_table, c));

	  LGLYPH_SET_CODE (g, c);
	  LGLYPH_SET_LBEARING (g, 0);
	  LGLYPH_SET_RBEARING (g, width);
	  LGLYPH_SET_WIDTH (g, width);
	  LGLYPH_SET_ASCENT (g, 1);
	  LGLYPH_SET_DESCENT (g, 0);
	}
      LGLYPH_SET_ADJUSTMENT (g, Qnil);
    }
  if (i < LGSTRING_GLYPH_LEN (gstring))
    LGSTRING_SET_GLYPH (gstring, i, Qnil);
}

EXFUN (Fre_search_forward, 4);

/* Try to compose the characters at CHARPOS according to CFT_ELEMENT
   which is an element of composition-fucntion-table (which see).
   LIMIT limits the characters to compose.  STRING, if not nil, is a
   target string.  WIN is a window where the characters are being
   displayed.  */

static Lisp_Object
autocmp_chars (cft_element, charpos, bytepos, limit, win, face, string)
     Lisp_Object cft_element;
     EMACS_INT charpos, bytepos, limit;
     struct window *win;
     struct face *face;
     Lisp_Object string;
{
  int count = SPECPDL_INDEX ();
  FRAME_PTR f = XFRAME (win->frame);
  Lisp_Object pos = make_number (charpos);
  EMACS_INT pt = PT, pt_byte = PT_BYTE;
919
  int lookback;
920 921
  
  record_unwind_save_match_data ();
922
  for (lookback = -1; CONSP (cft_element); cft_element = XCDR (cft_element))
923 924 925 926 927 928 929 930
    {
      Lisp_Object elt = XCAR (cft_element);
      Lisp_Object re;
      Lisp_Object font_object = Qnil, gstring;
      EMACS_INT to;

      if (! VECTORP (elt) || ASIZE (elt) != 3)
	continue;
931 932 933 934
      if (lookback < 0)
	lookback = XFASTINT (AREF (elt, 1));
      else if (lookback != XFASTINT (AREF (elt, 1)))
	break;
935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996
      re = AREF (elt, 0);
      if (NILP (string))
	TEMP_SET_PT_BOTH (charpos, bytepos);
      if (NILP (re)
	  || (STRINGP (re)
	      && (STRINGP (string)
		  ? EQ (Fstring_match (re, string, pos), pos)
		  : (! NILP (Fre_search_forward (re, make_number (limit), Qt, Qnil))
		     && EQ (Fmatch_beginning (make_number (0)), pos)))))
	{
	  to = (NILP (re) ? charpos + 1 : XINT (Fmatch_end (make_number (0))));
#ifdef HAVE_WINDOW_SYSTEM
	  if (FRAME_WINDOW_P (f))
	    {
	      font_object = font_range (charpos, &to, win, face, string);
	      if (! FONT_OBJECT_P (font_object))
		{
		  if (NILP (string))
		    TEMP_SET_PT_BOTH (pt, pt_byte);
		  return unbind_to (count, Qnil);
		}
	    }
#endif	/* not HAVE_WINDOW_SYSTEM */
	  gstring = Fcomposition_get_gstring (pos, make_number (to),
					      font_object, string);
	  if (NILP (LGSTRING_ID (gstring)))
	    {
	      Lisp_Object args[6];

	      args[0] = Vauto_composition_function;
	      args[1] = AREF (elt, 2);
	      args[2] = pos;
	      args[3] = make_number (to);
	      args[4] = font_object;
	      args[5] = string;
	      gstring = safe_call (6, args);
	    }
	  if (NILP (string))
	    TEMP_SET_PT_BOTH (pt, pt_byte);
	  return unbind_to (count, gstring);
	}
    }
  if (NILP (string))
    TEMP_SET_PT_BOTH (pt, pt_byte);
  return unbind_to (count, Qnil);
}


/* Update cmp_it->stop_pos to the next position after CHARPOS (and
   BYTEPOS) where character composition may happen.  If BYTEPOS is
   negative, compoute it.  If it is a static composition, set
   cmp_it->ch to -1.  Otherwise, set cmp_it->ch to the character that
   triggers a automatic composition.  */

void
composition_compute_stop_pos (cmp_it, charpos, bytepos, endpos, string)
     struct composition_it *cmp_it;
     EMACS_INT charpos, bytepos, endpos;
     Lisp_Object string;
{
  EMACS_INT start, end, c;
  Lisp_Object prop, val;
997 998
  /* This is from forward_to_next_line_start in xdisp.c.  */
  const int MAX_NEWLINE_DISTANCE = 500;
999

1000 1001
  if (endpos > charpos + MAX_NEWLINE_DISTANCE)
    endpos = charpos + MAX_NEWLINE_DISTANCE;
1002
  cmp_it->stop_pos = endpos;
1003
  cmp_it->id = -1;
1004
  cmp_it->ch = -2;
1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028
  if (find_composition (charpos, endpos, &start, &end, &prop, string)
      && COMPOSITION_VALID_P (start, end, prop))
    {
      cmp_it->stop_pos = endpos = start;
      cmp_it->ch = -1;
    }
  if (NILP (current_buffer->enable_multibyte_characters)
      || ! FUNCTIONP (Vauto_composition_function))
    return;
  if (bytepos < 0)
    {
      if (STRINGP (string))
	bytepos = string_char_to_byte (string, charpos);
      else
	bytepos = CHAR_TO_BYTE (charpos);
    }

  start = charpos;
  while (charpos < endpos)
    {
      if (STRINGP (string))
	FETCH_STRING_CHAR_ADVANCE (c, string, charpos, bytepos);
      else
	FETCH_CHAR_ADVANCE (c, charpos, bytepos);
1029 1030
      if (c == '\n')
	break;
1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044
      val = CHAR_TABLE_REF (Vcomposition_function_table, c);
      if (! NILP (val))
	{
	  Lisp_Object elt;

	  for (; CONSP (val); val = XCDR (val))
	    {
	      elt = XCAR (val);
	      if (VECTORP (elt) && ASIZE (elt) == 3 && NATNUMP (AREF (elt, 1))
		  && charpos - 1 - XFASTINT (AREF (elt, 1)) >= start)
		break;
	    }
	  if (CONSP (val))
	    {
1045 1046
	      cmp_it->lookback = XFASTINT (AREF (elt, 1));
	      cmp_it->stop_pos = charpos - 1 - cmp_it->lookback;
1047
	      cmp_it->ch = c;
1048
	      return;
1049 1050 1051
	    }
	}
    }
1052 1053
  cmp_it->stop_pos = charpos;
  cmp_it->ch = -2;
1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074
}

/* Check if the character at CHARPOS (and BYTEPOS) is composed
   (possibly with the following charaters) on window W.  ENDPOS limits
   characters to be composed.  FACE, in non-NULL, is a base face of
   the character.  If STRING is not nil, it is a string containing the
   character to check, and CHARPOS and BYTEPOS are indices in the
   string.  In that case, FACE must not be NULL.

   If the character is composed, setup members of CMP_IT (id, nglyphs,
   and from), and return 1.  Otherwise, update CMP_IT->stop_pos, and
   return 0.  */

int
composition_reseat_it (cmp_it, charpos, bytepos, endpos, w, face, string)
     struct composition_it *cmp_it;
     EMACS_INT charpos, bytepos, endpos;
     struct window *w;
     struct face *face;
     Lisp_Object string;
{
1075 1076 1077 1078 1079 1080 1081
  if (cmp_it->ch == -2)
    {
      composition_compute_stop_pos (cmp_it, charpos, bytepos, endpos, string);
      if (cmp_it->ch == -2)
	return 0;
    }

1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095
  if (cmp_it->ch < 0)
    {
      /* We are looking at a static composition.  */
      EMACS_INT start, end;
      Lisp_Object prop;

      find_composition (charpos, -1, &start, &end, &prop, string);
      cmp_it->id = get_composition_id (charpos, bytepos, end - start,
				       prop, string);
      if (cmp_it->id < 0)
	goto no_composition;
      cmp_it->nchars = end - start;
      cmp_it->nglyphs = composition_table[cmp_it->id]->glyph_len;
    }
1096
  else if (w)
1097
    {
1098
      Lisp_Object val, elt;
1099 1100 1101
      int i;

      val = CHAR_TABLE_REF (Vcomposition_function_table, cmp_it->ch);
1102 1103 1104 1105 1106 1107
      for (; CONSP (val); val = XCDR (val))
	{
	  elt = XCAR (val);
	  if (cmp_it->lookback == XFASTINT (AREF (elt, 1)))
	    break;
	}
1108 1109
      if (NILP (val))
	goto no_composition;
1110

1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121
      val = autocmp_chars (val, charpos, bytepos, endpos, w, face, string);
      if (! composition_gstring_p (val))
	goto no_composition;
      if (NILP (LGSTRING_ID (val)))
	val = composition_gstring_put_cache (val, -1);
      cmp_it->id = XINT (LGSTRING_ID (val));
      for (i = 0; i < LGSTRING_GLYPH_LEN (val); i++)
	if (NILP (LGSTRING_GLYPH (val, i)))
	  break;
      cmp_it->nglyphs = i;
    }
1122 1123
  else
    goto no_composition;
1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176
  cmp_it->from = 0;
  return 1;

 no_composition:
  charpos++;
  if (STRINGP (string))
    bytepos += MULTIBYTE_LENGTH_NO_CHECK (SDATA (string) + bytepos);
  else
    INC_POS (bytepos);
  composition_compute_stop_pos (cmp_it, charpos, bytepos, endpos, string);
  return 0;
}

int
composition_update_it (cmp_it, charpos, bytepos, string)
     struct composition_it *cmp_it;
     EMACS_INT charpos, bytepos;
     Lisp_Object string;
{
  int i, c;

  if (cmp_it->ch < 0)
    {
      struct composition *cmp = composition_table[cmp_it->id];

      cmp_it->to = cmp_it->nglyphs;
      if (cmp_it->nglyphs == 0)
	c = -1;
      else
	{
	  for (i = 0; i < cmp->glyph_len; i++)
	    if ((c = COMPOSITION_GLYPH (cmp, i)) != '\t')
	      break;
	  if (c == '\t')
	    c = ' ';
	}
      cmp_it->width = cmp->width;
    }
  else
    {
      Lisp_Object gstring = composition_gstring_from_id (cmp_it->id);

      if (cmp_it->nglyphs == 0)
	{
	  c = -1;
	  cmp_it->nchars = LGSTRING_CHAR_LEN (gstring);
	  cmp_it->width = 0;
	}
      else
	{
	  Lisp_Object glyph = LGSTRING_GLYPH (gstring, cmp_it->from);
	  int from = LGLYPH_FROM (glyph);

Stefan Monnier's avatar
Stefan Monnier committed
1177
	  c = XINT (LGSTRING_CHAR (gstring, from));
1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201
	  cmp_it->nchars = LGLYPH_TO (glyph) - from + 1;
	  cmp_it->width = (LGLYPH_WIDTH (glyph) > 0
			   ? CHAR_WIDTH (LGLYPH_CHAR (glyph)) : 0);
	  for (cmp_it->to = cmp_it->from + 1; cmp_it->to < cmp_it->nglyphs;
	       cmp_it->to++)
	    {
	      glyph = LGSTRING_GLYPH (gstring, cmp_it->to);
	      if (LGLYPH_FROM (glyph) != from)
		break;
	      if (LGLYPH_WIDTH (glyph) > 0)
		cmp_it->width += CHAR_WIDTH (LGLYPH_CHAR (glyph));
	    }
	}
    }

  charpos += cmp_it->nchars;
  if (STRINGP (string))
    cmp_it->nbytes = string_char_to_byte (string, charpos) - bytepos;
  else
    cmp_it->nbytes = CHAR_TO_BYTE (charpos) - bytepos;
  return c;
}


1202 1203 1204 1205 1206 1207 1208 1209 1210 1211
struct position_record
{
  EMACS_INT pos, pos_byte;
  unsigned char *p;
};

/* Update the members of POSTION to the next character boundary.  */
#define FORWARD_CHAR(POSITION, STOP)					\
  do {									\
    (POSITION).pos++;							\
1212 1213 1214 1215 1216 1217 1218 1219
    if ((POSITION).pos == (STOP))					\
      {									\
	(POSITION).p = GAP_END_ADDR;					\
	(POSITION).pos_byte = GPT_BYTE;					\
      }									\
    else								\
      {									\
	(POSITION).pos_byte += BYTES_BY_CHAR_HEAD (*((POSITION).p));	\
1220
	(POSITION).p += BYTES_BY_CHAR_HEAD (*((POSITION).p));		\
1221
      }									\
1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259
  } while (0)

/* Update the members of POSTION to the previous character boundary.  */
#define BACKWARD_CHAR(POSITION, STOP)		\
  do {						\
    if ((POSITION).pos == STOP)			\
      (POSITION).p = GPT_ADDR;			\
    do {					\
      (POSITION).pos_byte--;			\
      (POSITION).p--;				\
    } while (! CHAR_HEAD_P (*((POSITION).p)));	\
    (POSITION).pos--;				\
  } while (0)

static Lisp_Object _work_val;
static int _work_char;

/* 1 iff the character C is composable.  */
#define CHAR_COMPOSABLE_P(C)					\
  (_work_val = CHAR_TABLE_REF (Vunicode_category_table, (C)),	\
   (SYMBOLP (_work_val)						\
    && (_work_char = SDATA (SYMBOL_NAME (_work_val))[0]) != 'C'	\
    && _work_char != 'Z'))

/* This is like find_composition, but find an automatic composition
   instead.  If found, set *GSTRING to the glyph-string representing
   the composition, and return 1.  Otherwise, return 0.  */

static int
find_automatic_composition (pos, limit, start, end, gstring, string)
     EMACS_INT pos, limit, *start, *end;
     Lisp_Object *gstring, string;
{
  EMACS_INT head, tail, stop;
  struct position_record orig, cur, check, prev;
  Lisp_Object check_val, val, elt;
  int check_lookback;
  int c;
1260
  Lisp_Object window;
1261 1262
  struct window *w;

1263
  window = Fget_buffer_window (Fcurrent_buffer (), Qnil);
1264 1265 1266 1267
  if (NILP (window))
    return 0;
  w = XWINDOW (window);

1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408
  orig.pos = pos;
  if (NILP (string))
    {
      head = BEGV, tail = ZV, stop = GPT;
      orig.pos_byte = CHAR_TO_BYTE (orig.pos);
      orig.p = BYTE_POS_ADDR (orig.pos_byte);
    }
  else
    {
      head = 0, tail = SCHARS (string), stop = -1;
      orig.pos_byte = string_char_to_byte (string, orig.pos);
      orig.p = SDATA (string) + orig.pos_byte;
    }
  if (limit < pos)
    {
      head = max (head, limit);
      tail = min (tail, pos + 3);
    }
  else
    {
      tail = min (tail, limit + 3);
    }
  cur = orig;

 retry:
  check_val = Qnil;
  /* At first, check if POS is compoable.  */
  c = STRING_CHAR (cur.p, 0);
  if (! CHAR_COMPOSABLE_P (c))
    {
      if (limit < 0)
	return 0;
      if (limit >= cur.pos)
	goto search_forward;
    }
  else
    {
      val = CHAR_TABLE_REF (Vcomposition_function_table, c);
      if (! NILP (val))
	check_val = val, check = cur;
      else
	while (cur.pos + 1 < tail)
	  {
	    FORWARD_CHAR (cur, stop);
	    c = STRING_CHAR (cur.p, 0);
	    if (! CHAR_COMPOSABLE_P (c))
	      break;
	    val = CHAR_TABLE_REF (Vcomposition_function_table, c);
	    if (NILP (val))
	      continue;
	    check_val = val, check = cur;
	    break;
	  }
      cur = orig;
    }
  /* Rewind back to the position where we can safely search forward
     for compositions.  */
  while (cur.pos > head)
    {
      BACKWARD_CHAR (cur, stop);
      c = STRING_CHAR (cur.p, 0);
      if (! CHAR_COMPOSABLE_P (c))
	break;
      val = CHAR_TABLE_REF (Vcomposition_function_table, c);
      if (! NILP (val))
	check_val = val, check = cur;
    }
  prev = cur;
  /* Now search forward.  */
 search_forward:  
  *gstring = Qnil;
  if (! NILP (check_val) || limit >= orig.pos)
    {
      if (NILP (check_val))
	cur = orig;
      else
	cur = check;
      while (cur.pos < tail)
	{
	  int need_adjustment = 0;

	  if (NILP (check_val))
	    {
	      c = STRING_CHAR (cur.p, 0);
	      check_val = CHAR_TABLE_REF (Vcomposition_function_table, c);
	    }
	  for (; CONSP (check_val); check_val = XCDR (check_val))
	    {
	      elt = XCAR (check_val);
	      if (VECTORP (elt) && ASIZE (elt) == 3 && NATNUMP (AREF (elt, 1))
		  && cur.pos - XFASTINT (AREF (elt, 1)) >= head)
		{
		  check.pos = cur.pos - XFASTINT (AREF (elt, 1));
		  if (check.pos == cur.pos)
		    check.pos_byte = cur.pos_byte;
		  else
		    check.pos_byte = CHAR_TO_BYTE (check.pos);
		  val = autocmp_chars (check_val, check.pos, check.pos_byte,
				       tail, w, NULL, string);
		  need_adjustment = 1;
		  if (! NILP (val))
		    {
		      *gstring = val;
		      *start = check.pos;
		      *end = check.pos + LGSTRING_CHAR_LEN (*gstring);
		      if (*start <= orig.pos ? *end > orig.pos
			  : limit >= orig.pos)
			return 1;
		      cur.pos = *end;
		      cur.pos_byte = CHAR_TO_BYTE (cur.pos);
		      break;
		    }
		}
	    }
	  if (need_adjustment)
	    {
	      /* As we have called Lisp, there's a possibilily that
		 buffer/string is relocated.  */
	      if (NILP (string))
		cur.p  = BYTE_POS_ADDR (cur.pos_byte);
	      else
		cur.p = SDATA (string) + cur.pos_byte;
	    }
	  if (! CONSP (check_val))
	    FORWARD_CHAR (cur, stop);
	  check_val = Qnil;
	}
    }
  if (! NILP (*gstring))
    return (limit >= 0 || (*start <= orig.pos && *end > orig.pos));
  if (limit >= 0 && limit < orig.pos && prev.pos > head)
    {
      cur = prev;
      BACKWARD_CHAR (cur, stop);
      orig = cur;
      tail = orig.pos;
      goto retry;
    }
  return 0;
}

1409 1410 1411 1412 1413
int
composition_adjust_point (last_pt)
     EMACS_INT last_pt;
{
  EMACS_INT charpos, bytepos, startpos, beg, end, pos;
1414 1415
  Lisp_Object val;
  int i;
1416 1417 1418 1419

  if (PT == BEGV || PT == ZV)
    return PT;

1420
  /* At first check the static composition. */
1421 1422 1423 1424 1425 1426 1427 1428 1429 1430
  if (get_property_and_range (PT, Qcomposition, &val, &beg, &end, Qnil)
      && COMPOSITION_VALID_P (beg, end, val)
      && beg < PT /* && end > PT   <- It's always the case.  */
      && (last_pt <= beg || last_pt >= end))
    return (PT < last_pt ? beg : end);

  if (NILP (current_buffer->enable_multibyte_characters)
      || ! FUNCTIONP (Vauto_composition_function))
    return PT;

1431 1432 1433
  /* Next check the automatic composition.  */
  if (! find_automatic_composition (PT, -1, &beg, &end, &val, Qnil)
      || beg == PT)
1434
    return PT;
1435
  for (i = 0; i < LGSTRING_GLYPH_LEN (val); i++)
1436
    {
1437
      Lisp_Object glyph = LGSTRING_GLYPH (val, i);
1438

1439 1440 1441 1442 1443 1444 1445 1446
      if (NILP (glyph))
	break;
      if (beg + LGLYPH_FROM (glyph) == PT)
	return PT;
      if (beg + LGLYPH_TO (glyph) >= PT)
	return (PT < last_pt
		? beg + LGLYPH_FROM (glyph)
		: beg + LGLYPH_TO (glyph) + 1);
1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498
    }
  return PT;
}

DEFUN ("composition-get-gstring", Fcomposition_get_gstring,
       Scomposition_get_gstring, 4, 4, 0,
       doc: /* Return a glyph-string for characters between FROM and TO.
If the glhph string is for graphic display, FONT-OBJECT must be
a font-object to use for those characters.
Otherwise (for terminal display), FONT-OBJECT must be nil.

If the optional 4th argument STRING is not nil, it is a string
containing the target characters between indices FROM and TO.

A glhph-string is a vector containing information about how to display
specific character sequence.  The format is:
   [HEADER ID GLYPH ...]

HEADER is a vector of this form:
    [FONT-OBJECT CHAR ...]
where
    FONT-OBJECT is a font-object for all glyphs in the glyph-string,
    or nil if not yet decided.
    CHARs are characters to be composed by GLYPHs.

ID is an identification number of the glyph-string.  It may be nil if
not yet shaped.

GLYPH is a vector whose elements has this form:
    [ FROM-IDX TO-IDX C CODE WIDTH LBEARING RBEARING ASCENT DESCENT
      [ [X-OFF Y-OFF WADJUST] | nil] ]
where
    FROM-IDX and TO-IDX are used internally and should not be touched.
    C is the character of the glyph.
    CODE is the glyph-code of C in FONT-OBJECT.
    WIDTH thru DESCENT are the metrics (in pixels) of the glyph.
    X-OFF and Y-OFF are offests to the base position for the glyph.
    WADJUST is the adjustment to the normal width of the glyph.

If GLYPH is nil, the remaining elements of the glhph-string vector
must be ignore.  */)
     (from, to, font_object, string)
     Lisp_Object font_object, from, to, string;
{
  Lisp_Object gstring, header;

  if (! NILP (font_object))
    CHECK_FONT_OBJECT (font_object);
  header = fill_gstring_header (Qnil, from, to, font_object, string);
  gstring = gstring_lookup_cache (header);
  if (! NILP (gstring))
    return gstring;
1499 1500
  if (LGSTRING_GLYPH_LEN (gstring_work) < to - from)
    gstring_work = Fmake_vector (make_number (to - from + 2), Qnil);
1501 1502 1503 1504 1505 1506
  LGSTRING_SET_HEADER (gstring_work, header);
  LGSTRING_SET_ID (gstring_work, Qnil);
  fill_gstring_body (gstring_work);
  return gstring_work;
}

Kenichi Handa's avatar
Kenichi Handa committed
1507 1508 1509 1510 1511

/* Emacs Lisp APIs.  */

DEFUN ("compose-region-internal", Fcompose_region_internal,
       Scompose_region_internal, 2, 4, 0,
1512 1513 1514 1515
       doc: /* Internal use only.

Compose text in the region between START and END.
Optional 3rd and 4th arguments are COMPONENTS and MODIFICATION-FUNC
1516
for the composition.  See `compose-region' for more detail.  */)
1517 1518
     (start, end, components, modification_func)
     Lisp_Object start, end, components, modification_func;
Kenichi Handa's avatar
Kenichi Handa committed
1519 1520 1521 1522 1523 1524
{
  validate_region (&start, &end);
  if (!NILP (components)
      && !INTEGERP (components)
      && !CONSP (components)
      && !STRINGP (components))
1525
    CHECK_VECTOR (components);
Kenichi Handa's avatar
Kenichi Handa committed
1526

1527
  compose_text (XINT (start), XINT (end), components, modification_func, Qnil);
Kenichi Handa's avatar
Kenichi Handa committed
1528 1529 1530 1531 1532
  return Qnil;
}

DEFUN ("compose-string-internal", Fcompose_string_internal,
       Scompose_string_internal, 3, 5, 0,
1533 1534 1535 1536
       doc: /* Internal use only.

Compose text between indices START and END of STRING.
Optional 4th and 5th arguments are COMPONENTS and MODIFICATION-FUNC
1537
for the composition.  See `compose-string' for more detail.  */)
Juanma Barranquero's avatar