chartab.c 40.3 KB
Newer Older
1
/* chartab.c -- char-table support
2
   Copyright (C) 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011
3 4 5 6 7
     National Institute of Advanced Industrial Science and Technology (AIST)
     Registration Number H13PRO009

This file is part of GNU Emacs.

8
GNU Emacs is free software: you can redistribute it and/or modify
9
it under the terms of the GNU General Public License as published by
10 11
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
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/>.  */
20 21

#include <config.h>
22

23 24 25 26
#include "lisp.h"
#include "character.h"
#include "charset.h"
#include "ccl.h"
27 28 29 30 31 32 33 34 35 36 37 38

/* 64/16/32/128 */

/* Number of elements in Nth level char-table.  */
const int chartab_size[4] =
  { (1 << CHARTAB_SIZE_BITS_0),
    (1 << CHARTAB_SIZE_BITS_1),
    (1 << CHARTAB_SIZE_BITS_2),
    (1 << CHARTAB_SIZE_BITS_3) };

/* Number of characters each element of Nth level char-table
   covers.  */
39
static const int chartab_chars[4] =
40 41 42 43 44 45 46
  { (1 << (CHARTAB_SIZE_BITS_1 + CHARTAB_SIZE_BITS_2 + CHARTAB_SIZE_BITS_3)),
    (1 << (CHARTAB_SIZE_BITS_2 + CHARTAB_SIZE_BITS_3)),
    (1 << CHARTAB_SIZE_BITS_3),
    1 };

/* Number of characters (in bits) each element of Nth level char-table
   covers.  */
47
static const int chartab_bits[4] =
48 49 50 51 52 53 54 55
  { (CHARTAB_SIZE_BITS_1 + CHARTAB_SIZE_BITS_2 + CHARTAB_SIZE_BITS_3),
    (CHARTAB_SIZE_BITS_2 + CHARTAB_SIZE_BITS_3),
    CHARTAB_SIZE_BITS_3,
    0 };

#define CHARTAB_IDX(c, depth, min_char)		\
  (((c) - (min_char)) >> chartab_bits[(depth)])

56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77

/* Preamble for uniprop (Unicode character property) tables.  See the
   comment of "Unicode character property tables".  */

/* Purpose of uniprop tables. */
static Lisp_Object Qchar_code_property_table;

/* Types of decoder and encoder functions for uniprop values.  */
typedef Lisp_Object (*uniprop_decoder_t) (Lisp_Object, Lisp_Object);
typedef Lisp_Object (*uniprop_encoder_t) (Lisp_Object, Lisp_Object);

static Lisp_Object uniprop_table_uncompress (Lisp_Object, int);
static uniprop_decoder_t uniprop_get_decoder (Lisp_Object);

/* 1 iff TABLE is a uniprop table.  */
#define UNIPROP_TABLE_P(TABLE)					\
  (EQ (XCHAR_TABLE (TABLE)->purpose, Qchar_code_property_table)	\
   && CHAR_TABLE_EXTRA_SLOTS (XCHAR_TABLE (TABLE)) == 5)

/* Return a decoder for values in the uniprop table TABLE.  */
#define UNIPROP_GET_DECODER(TABLE)	\
  (UNIPROP_TABLE_P (TABLE) ? uniprop_get_decoder (TABLE) : NULL)
78

79 80 81 82 83 84 85 86 87
/* Nonzero iff OBJ is a string representing uniprop values of 128
   succeeding characters (the bottom level of a char-table) by a
   compressed format.  We are sure that no property value has a string
   starting with '\001' nor '\002'.  */
#define UNIPROP_COMPRESSED_FORM_P(OBJ)	\
  (STRINGP (OBJ) && SCHARS (OBJ) > 0	\
   && ((SREF (OBJ, 0) == 1 || (SREF (OBJ, 0) == 2))))


Paul Eggert's avatar
Paul Eggert committed
88
DEFUN ("make-char-table", Fmake_char_table, Smake_char_table, 1, 2, 0,
89
       doc: /* Return a newly created char-table, with purpose PURPOSE.
90 91
Each element is initialized to INIT, which defaults to nil.

92 93 94 95
PURPOSE should be a symbol.  If it has a `char-table-extra-slots'
property, the property's value should be an integer between 0 and 10
that specifies how many extra slots the char-table has.  Otherwise,
the char-table has no extra slot.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
96
  (register Lisp_Object purpose, Lisp_Object init)
97 98 99
{
  Lisp_Object vector;
  Lisp_Object n;
100
  int n_extras;
101 102 103
  int size;

  CHECK_SYMBOL (purpose);
104 105 106 107
  n = Fget (purpose, Qchar_table_extra_slots);
  if (NILP (n))
    n_extras = 0;
  else
108
    {
109
      CHECK_NATNUM (n);
110
      if (XINT (n) > 10)
111
	args_out_of_range (n, Qnil);
112
      n_extras = XINT (n);
113 114 115 116
    }

  size = VECSIZE (struct Lisp_Char_Table) - 1 + n_extras;
  vector = Fmake_vector (make_number (size), init);
Miles Bader's avatar
Miles Bader committed
117
  XSETPVECTYPE (XVECTOR (vector), PVEC_CHAR_TABLE);
Paul Eggert's avatar
Paul Eggert committed
118 119
  set_char_table_parent (vector, Qnil);
  set_char_table_purpose (vector, purpose);
120 121 122 123 124
  XSETCHAR_TABLE (vector, XCHAR_TABLE (vector));
  return vector;
}

static Lisp_Object
125
make_sub_char_table (int depth, int min_char, Lisp_Object defalt)
126 127 128 129 130
{
  Lisp_Object table;
  int size = VECSIZE (struct Lisp_Sub_Char_Table) - 1 + chartab_size[depth];

  table = Fmake_vector (make_number (size), defalt);
Miles Bader's avatar
Miles Bader committed
131
  XSETPVECTYPE (XVECTOR (table), PVEC_SUB_CHAR_TABLE);
132 133 134 135 136 137 138
  XSUB_CHAR_TABLE (table)->depth = make_number (depth);
  XSUB_CHAR_TABLE (table)->min_char = make_number (min_char);

  return table;
}

static Lisp_Object
139
char_table_ascii (Lisp_Object table)
140
{
141
  Lisp_Object sub, val;
142 143

  sub = XCHAR_TABLE (table)->contents[0];
144 145
  if (! SUB_CHAR_TABLE_P (sub))
    return sub;
146
  sub = XSUB_CHAR_TABLE (sub)->contents[0];
147 148
  if (! SUB_CHAR_TABLE_P (sub))
    return sub;
149 150 151 152
  val = XSUB_CHAR_TABLE (sub)->contents[0];
  if (UNIPROP_TABLE_P (table) && UNIPROP_COMPRESSED_FORM_P (val))
    val = uniprop_table_uncompress (sub, 0);
  return val;
153 154
}

155
static Lisp_Object
156
copy_sub_char_table (Lisp_Object table)
157 158 159
{
  int depth = XINT (XSUB_CHAR_TABLE (table)->depth);
  int min_char = XINT (XSUB_CHAR_TABLE (table)->min_char);
160
  Lisp_Object copy = make_sub_char_table (depth, min_char, Qnil);
161 162 163 164 165
  int i;

  /* Recursively copy any sub char-tables.  */
  for (i = 0; i < chartab_size[depth]; i++)
    {
166 167 168
      Lisp_Object val = XSUB_CHAR_TABLE (table)->contents[i];
      set_sub_char_table_contents
	(copy, i, SUB_CHAR_TABLE_P (val) ? copy_sub_char_table (val) : val);
169 170 171 172 173 174 175
    }

  return copy;
}


Lisp_Object
176
copy_char_table (Lisp_Object table)
177 178
{
  Lisp_Object copy;
179
  int size = XCHAR_TABLE (table)->header.size & PSEUDOVECTOR_SIZE_MASK;
180 181 182
  int i;

  copy = Fmake_vector (make_number (size), Qnil);
Miles Bader's avatar
Miles Bader committed
183
  XSETPVECTYPE (XVECTOR (copy), PVEC_CHAR_TABLE);
Paul Eggert's avatar
Paul Eggert committed
184 185 186
  set_char_table_defalt (copy, XCHAR_TABLE (table)->defalt);
  set_char_table_parent (copy, XCHAR_TABLE (table)->parent);
  set_char_table_purpose (copy, XCHAR_TABLE (table)->purpose);
187
  for (i = 0; i < chartab_size[0]; i++)
188
    set_char_table_contents
Paul Eggert's avatar
Paul Eggert committed
189
      (copy, i,
190 191 192
       (SUB_CHAR_TABLE_P (XCHAR_TABLE (table)->contents[i])
	? copy_sub_char_table (XCHAR_TABLE (table)->contents[i])
	: XCHAR_TABLE (table)->contents[i]));
Paul Eggert's avatar
Paul Eggert committed
193
  set_char_table_ascii (copy, char_table_ascii (copy));
194 195
  size -= VECSIZE (struct Lisp_Char_Table) - 1;
  for (i = 0; i < size; i++)
196
    set_char_table_extras (copy, i, XCHAR_TABLE (table)->extras[i]);
197 198 199 200 201

  XSETCHAR_TABLE (copy, XCHAR_TABLE (copy));
  return copy;
}

Andreas Schwab's avatar
Andreas Schwab committed
202
static Lisp_Object
203
sub_char_table_ref (Lisp_Object table, int c, bool is_uniprop)
204 205 206 207 208
{
  struct Lisp_Sub_Char_Table *tbl = XSUB_CHAR_TABLE (table);
  int depth = XINT (tbl->depth);
  int min_char = XINT (tbl->min_char);
  Lisp_Object val;
209
  int idx = CHARTAB_IDX (c, depth, min_char);
210

211 212 213
  val = tbl->contents[idx];
  if (is_uniprop && UNIPROP_COMPRESSED_FORM_P (val))
    val = uniprop_table_uncompress (table, idx);
214
  if (SUB_CHAR_TABLE_P (val))
215
    val = sub_char_table_ref (val, c, is_uniprop);
216 217 218 219
  return val;
}

Lisp_Object
220
char_table_ref (Lisp_Object table, int c)
221 222 223 224 225 226 227 228 229 230 231 232 233 234
{
  struct Lisp_Char_Table *tbl = XCHAR_TABLE (table);
  Lisp_Object val;

  if (ASCII_CHAR_P (c))
    {
      val = tbl->ascii;
      if (SUB_CHAR_TABLE_P (val))
	val = XSUB_CHAR_TABLE (val)->contents[c];
    }
  else
    {
      val = tbl->contents[CHARTAB_IDX (c, 0, 0)];
      if (SUB_CHAR_TABLE_P (val))
235
	val = sub_char_table_ref (val, c, UNIPROP_TABLE_P (table));
236 237 238 239 240 241 242 243
    }
  if (NILP (val))
    {
      val = tbl->defalt;
      if (NILP (val) && CHAR_TABLE_P (tbl->parent))
	val = char_table_ref (tbl->parent, c);
    }
  return val;
Kenichi Handa's avatar
Kenichi Handa committed
244
}
245 246

static Lisp_Object
247
sub_char_table_ref_and_range (Lisp_Object table, int c, int *from, int *to,
248
			      Lisp_Object defalt, bool is_uniprop)
249 250 251 252
{
  struct Lisp_Sub_Char_Table *tbl = XSUB_CHAR_TABLE (table);
  int depth = XINT (tbl->depth);
  int min_char = XINT (tbl->min_char);
253
  int chartab_idx = CHARTAB_IDX (c, depth, min_char), idx;
254
  Lisp_Object val;
Kenichi Handa's avatar
Kenichi Handa committed
255

256
  val = tbl->contents[chartab_idx];
257 258
  if (is_uniprop && UNIPROP_COMPRESSED_FORM_P (val))
    val = uniprop_table_uncompress (table, chartab_idx);
259
  if (SUB_CHAR_TABLE_P (val))
260
    val = sub_char_table_ref_and_range (val, c, from, to, defalt, is_uniprop);
261 262 263
  else if (NILP (val))
    val = defalt;

264
  idx = chartab_idx;
265
  while (idx > 0 && *from < min_char + idx * chartab_chars[depth])
266
    {
267 268
      Lisp_Object this_val;

269 270 271
      c = min_char + idx * chartab_chars[depth] - 1;
      idx--;
      this_val = tbl->contents[idx];
272 273
      if (is_uniprop && UNIPROP_COMPRESSED_FORM_P (this_val))
	this_val = uniprop_table_uncompress (table, idx);
274
      if (SUB_CHAR_TABLE_P (this_val))
275 276
	this_val = sub_char_table_ref_and_range (this_val, c, from, to, defalt,
						 is_uniprop);
277 278 279 280
      else if (NILP (this_val))
	this_val = defalt;

      if (! EQ (this_val, val))
281 282 283 284
	{
	  *from = c + 1;
	  break;
	}
285
    }
286 287 288
  while (((c = (chartab_idx + 1) * chartab_chars[depth])
	  < chartab_chars[depth - 1])
	 && (c += min_char) <= *to)
289
    {
290 291
      Lisp_Object this_val;

292 293
      chartab_idx++;
      this_val = tbl->contents[chartab_idx];
294 295
      if (is_uniprop && UNIPROP_COMPRESSED_FORM_P (this_val))
	this_val = uniprop_table_uncompress (table, chartab_idx);
296
      if (SUB_CHAR_TABLE_P (this_val))
297 298
	this_val = sub_char_table_ref_and_range (this_val, c, from, to, defalt,
						 is_uniprop);
299 300 301
      else if (NILP (this_val))
	this_val = defalt;
      if (! EQ (this_val, val))
302 303 304 305
	{
	  *to = c - 1;
	  break;
	}
306
    }
307

308 309 310 311
  return val;
}


312 313 314 315
/* Return the value for C in char-table TABLE.  Shrink the range *FROM
   and *TO to cover characters (containing C) that have the same value
   as C.  It is not assured that the values of (*FROM - 1) and (*TO +
   1) are different from that of C.  */
316

317
Lisp_Object
318
char_table_ref_and_range (Lisp_Object table, int c, int *from, int *to)
319 320
{
  struct Lisp_Char_Table *tbl = XCHAR_TABLE (table);
321
  int chartab_idx = CHARTAB_IDX (c, 0, 0), idx;
322
  Lisp_Object val;
323
  bool is_uniprop = UNIPROP_TABLE_P (table);
324

325
  val = tbl->contents[chartab_idx];
326 327 328 329
  if (*from < 0)
    *from = 0;
  if (*to < 0)
    *to = MAX_CHAR;
330 331
  if (is_uniprop && UNIPROP_COMPRESSED_FORM_P (val))
    val = uniprop_table_uncompress (table, chartab_idx);
332
  if (SUB_CHAR_TABLE_P (val))
333 334
    val = sub_char_table_ref_and_range (val, c, from, to, tbl->defalt,
					is_uniprop);
335 336
  else if (NILP (val))
    val = tbl->defalt;
337
  idx = chartab_idx;
338
  while (*from < idx * chartab_chars[0])
339
    {
340 341
      Lisp_Object this_val;

342 343 344
      c = idx * chartab_chars[0] - 1;
      idx--;
      this_val = tbl->contents[idx];
345 346
      if (is_uniprop && UNIPROP_COMPRESSED_FORM_P (this_val))
	this_val = uniprop_table_uncompress (table, idx);
347
      if (SUB_CHAR_TABLE_P (this_val))
348
	this_val = sub_char_table_ref_and_range (this_val, c, from, to,
349
						 tbl->defalt, is_uniprop);
350 351 352 353
      else if (NILP (this_val))
	this_val = tbl->defalt;

      if (! EQ (this_val, val))
354 355 356 357
	{
	  *from = c + 1;
	  break;
	}
358
    }
359
  while (*to >= (chartab_idx + 1) * chartab_chars[0])
360
    {
361 362
      Lisp_Object this_val;

363 364 365
      chartab_idx++;
      c = chartab_idx * chartab_chars[0];
      this_val = tbl->contents[chartab_idx];
366 367
      if (is_uniprop && UNIPROP_COMPRESSED_FORM_P (this_val))
	this_val = uniprop_table_uncompress (table, chartab_idx);
368
      if (SUB_CHAR_TABLE_P (this_val))
369
	this_val = sub_char_table_ref_and_range (this_val, c, from, to,
370
						 tbl->defalt, is_uniprop);
371 372 373
      else if (NILP (this_val))
	this_val = tbl->defalt;
      if (! EQ (this_val, val))
374 375 376 377
	{
	  *to = c - 1;
	  break;
	}
378 379 380
    }

  return val;
381
}
382 383 384


static void
385
sub_char_table_set (Lisp_Object table, int c, Lisp_Object val, bool is_uniprop)
386 387 388 389 390 391
{
  struct Lisp_Sub_Char_Table *tbl = XSUB_CHAR_TABLE (table);
  int depth = XINT ((tbl)->depth);
  int min_char = XINT ((tbl)->min_char);
  int i = CHARTAB_IDX (c, depth, min_char);
  Lisp_Object sub;
Kenichi Handa's avatar
Kenichi Handa committed
392

393
  if (depth == 3)
394
    set_sub_char_table_contents (table, i, val);
395 396 397 398 399
  else
    {
      sub = tbl->contents[i];
      if (! SUB_CHAR_TABLE_P (sub))
	{
400 401 402 403 404 405 406
	  if (is_uniprop && UNIPROP_COMPRESSED_FORM_P (sub))
	    sub = uniprop_table_uncompress (table, i);
	  else
	    {
	      sub = make_sub_char_table (depth + 1,
					 min_char + i * chartab_chars[depth],
					 sub);
407
	      set_sub_char_table_contents (table, i, sub);
408
	    }
409
	}
410
      sub_char_table_set (sub, c, val, is_uniprop);
411 412 413
    }
}

414
void
415
char_table_set (Lisp_Object table, int c, Lisp_Object val)
416 417 418 419 420
{
  struct Lisp_Char_Table *tbl = XCHAR_TABLE (table);

  if (ASCII_CHAR_P (c)
      && SUB_CHAR_TABLE_P (tbl->ascii))
421
    set_sub_char_table_contents (tbl->ascii, c, val);
422 423 424 425 426 427 428 429 430
  else
    {
      int i = CHARTAB_IDX (c, 0, 0);
      Lisp_Object sub;

      sub = tbl->contents[i];
      if (! SUB_CHAR_TABLE_P (sub))
	{
	  sub = make_sub_char_table (1, i * chartab_chars[0], sub);
431
	  set_char_table_contents (table, i, sub);
432
	}
433
      sub_char_table_set (sub, c, val, UNIPROP_TABLE_P (table));
434
      if (ASCII_CHAR_P (c))
Paul Eggert's avatar
Paul Eggert committed
435
	set_char_table_ascii (table, char_table_ascii (table));
436 437 438 439
    }
}

static void
440
sub_char_table_set_range (Lisp_Object table, int from, int to, Lisp_Object val,
441
			  bool is_uniprop)
442
{
443 444 445 446 447 448 449 450 451 452
  struct Lisp_Sub_Char_Table *tbl = XSUB_CHAR_TABLE (table);
  int depth = XINT ((tbl)->depth);
  int min_char = XINT ((tbl)->min_char);
  int chars_in_block = chartab_chars[depth];
  int i, c, lim = chartab_size[depth];

  if (from < min_char)
    from = min_char;
  i = CHARTAB_IDX (from, depth, min_char);
  c = min_char + chars_in_block * i;
453
  for (; i < lim; i++, c += chars_in_block)
454
    {
455 456 457
      if (c > to)
	break;
      if (from <= c && c + chars_in_block - 1 <= to)
458
	set_sub_char_table_contents (table, i, val);
459 460 461 462 463 464 465 466 467 468
      else
	{
	  Lisp_Object sub = tbl->contents[i];
	  if (! SUB_CHAR_TABLE_P (sub))
	    {
	      if (is_uniprop && UNIPROP_COMPRESSED_FORM_P (sub))
		sub = uniprop_table_uncompress (table, i);
	      else
		{
		  sub = make_sub_char_table (depth + 1, c, sub);
469
		  set_sub_char_table_contents (table, i, sub);
470 471 472 473
		}
	    }
	  sub_char_table_set_range (sub, from, to, val, is_uniprop);
	}
474 475 476 477
    }
}


478
void
479
char_table_set_range (Lisp_Object table, int from, int to, Lisp_Object val)
480 481 482 483 484 485 486
{
  struct Lisp_Char_Table *tbl = XCHAR_TABLE (table);

  if (from == to)
    char_table_set (table, from, val);
  else
    {
487
      bool is_uniprop = UNIPROP_TABLE_P (table);
488 489 490 491 492 493 494 495 496
      int lim = CHARTAB_IDX (to, 0, 0);
      int i, c;

      for (i = CHARTAB_IDX (from, 0, 0), c = 0; i <= lim;
	   i++, c += chartab_chars[0])
	{
	  if (c > to)
	    break;
	  if (from <= c && c + chartab_chars[0] - 1 <= to)
497
	    set_char_table_contents (table, i, val);
498 499 500 501 502 503
	  else
	    {
	      Lisp_Object sub = tbl->contents[i];
	      if (! SUB_CHAR_TABLE_P (sub))
		{
		  sub = make_sub_char_table (1, i * chartab_chars[0], sub);
504
		  set_char_table_contents (table, i, sub);
505 506 507 508
		}
	      sub_char_table_set_range (sub, from, to, val, is_uniprop);
	    }
	}
509
      if (ASCII_CHAR_P (from))
Paul Eggert's avatar
Paul Eggert committed
510
	set_char_table_ascii (table, char_table_ascii (table));
511 512 513 514 515 516 517 518
    }
}


DEFUN ("char-table-subtype", Fchar_table_subtype, Schar_table_subtype,
       1, 1, 0,
       doc: /*
Return the subtype of char-table CHAR-TABLE.  The value is a symbol.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
519
  (Lisp_Object char_table)
520 521 522 523 524 525 526 527 528 529 530 531 532
{
  CHECK_CHAR_TABLE (char_table);

  return XCHAR_TABLE (char_table)->purpose;
}

DEFUN ("char-table-parent", Fchar_table_parent, Schar_table_parent,
       1, 1, 0,
       doc: /* Return the parent char-table of CHAR-TABLE.
The value is either nil or another char-table.
If CHAR-TABLE holds nil for a given character,
then the actual applicable value is inherited from the parent char-table
\(or from its parents, if necessary).  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
533
  (Lisp_Object char_table)
534 535 536 537 538 539
{
  CHECK_CHAR_TABLE (char_table);

  return XCHAR_TABLE (char_table)->parent;
}

Paul Eggert's avatar
Paul Eggert committed
540
DEFUN ("set-char-table-parent", Fset_char_table_parent, Sset_char_table_parent,
541 542
       2, 2, 0,
       doc: /* Set the parent char-table of CHAR-TABLE to PARENT.
Kenichi Handa's avatar
Kenichi Handa committed
543
Return PARENT.  PARENT must be either nil or another char-table.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
544
  (Lisp_Object char_table, Lisp_Object parent)
545 546 547 548 549 550 551 552 553 554 555 556 557 558
{
  Lisp_Object temp;

  CHECK_CHAR_TABLE (char_table);

  if (!NILP (parent))
    {
      CHECK_CHAR_TABLE (parent);

      for (temp = parent; !NILP (temp); temp = XCHAR_TABLE (temp)->parent)
	if (EQ (temp, char_table))
	  error ("Attempt to make a chartable be its own parent");
    }

Paul Eggert's avatar
Paul Eggert committed
559
  set_char_table_parent (char_table, parent);
560 561 562 563

  return parent;
}

Paul Eggert's avatar
Paul Eggert committed
564
DEFUN ("char-table-extra-slot", Fchar_table_extra_slot, Schar_table_extra_slot,
565 566
       2, 2, 0,
       doc: /* Return the value of CHAR-TABLE's extra-slot number N.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
567
  (Lisp_Object char_table, Lisp_Object n)
568 569 570 571 572 573 574 575 576 577
{
  CHECK_CHAR_TABLE (char_table);
  CHECK_NUMBER (n);
  if (XINT (n) < 0
      || XINT (n) >= CHAR_TABLE_EXTRA_SLOTS (XCHAR_TABLE (char_table)))
    args_out_of_range (char_table, n);

  return XCHAR_TABLE (char_table)->extras[XINT (n)];
}

Paul Eggert's avatar
Paul Eggert committed
578
DEFUN ("set-char-table-extra-slot", Fset_char_table_extra_slot,
579 580 581
       Sset_char_table_extra_slot,
       3, 3, 0,
       doc: /* Set CHAR-TABLE's extra-slot number N to VALUE.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
582
  (Lisp_Object char_table, Lisp_Object n, Lisp_Object value)
583 584 585 586 587 588 589
{
  CHECK_CHAR_TABLE (char_table);
  CHECK_NUMBER (n);
  if (XINT (n) < 0
      || XINT (n) >= CHAR_TABLE_EXTRA_SLOTS (XCHAR_TABLE (char_table)))
    args_out_of_range (char_table, n);

590
  set_char_table_extras (char_table, XINT (n), value);
591
  return value;
592 593 594 595 596 597 598
}

DEFUN ("char-table-range", Fchar_table_range, Schar_table_range,
       2, 2, 0,
       doc: /* Return the value in CHAR-TABLE for a range of characters RANGE.
RANGE should be nil (for the default value),
a cons of character codes (for characters in the range), or a character code.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
599
  (Lisp_Object char_table, Lisp_Object range)
600 601 602 603 604 605
{
  Lisp_Object val;
  CHECK_CHAR_TABLE (char_table);

  if (EQ (range, Qnil))
    val = XCHAR_TABLE (char_table)->defalt;
606 607
  else if (CHARACTERP (range))
    val = CHAR_TABLE_REF (char_table, XFASTINT (range));
608 609 610 611
  else if (CONSP (range))
    {
      int from, to;

Kenichi Handa's avatar
Kenichi Handa committed
612 613
      CHECK_CHARACTER_CAR (range);
      CHECK_CHARACTER_CDR (range);
614 615 616
      from = XFASTINT (XCAR (range));
      to = XFASTINT (XCDR (range));
      val = char_table_ref_and_range (char_table, from, &from, &to);
617 618 619 620 621 622 623
      /* Not yet implemented. */
    }
  else
    error ("Invalid RANGE argument to `char-table-range'");
  return val;
}

Paul Eggert's avatar
Paul Eggert committed
624
DEFUN ("set-char-table-range", Fset_char_table_range, Sset_char_table_range,
625
       3, 3, 0,
Kenichi Handa's avatar
Kenichi Handa committed
626
       doc: /* Set the value in CHAR-TABLE for a range of characters RANGE to VALUE.
627
RANGE should be t (for all characters), nil (for the default value),
Kenichi Handa's avatar
Kenichi Handa committed
628 629
a cons of character codes (for characters in the range),
or a character code.  Return VALUE.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
630
  (Lisp_Object char_table, Lisp_Object range, Lisp_Object value)
631 632 633 634 635 636
{
  CHECK_CHAR_TABLE (char_table);
  if (EQ (range, Qt))
    {
      int i;

Paul Eggert's avatar
Paul Eggert committed
637
      set_char_table_ascii (char_table, value);
638
      for (i = 0; i < chartab_size[0]; i++)
639
	set_char_table_contents (char_table, i, value);
640 641
    }
  else if (EQ (range, Qnil))
Paul Eggert's avatar
Paul Eggert committed
642
    set_char_table_defalt (char_table, value);
643
  else if (CHARACTERP (range))
644 645 646
    char_table_set (char_table, XINT (range), value);
  else if (CONSP (range))
    {
Kenichi Handa's avatar
Kenichi Handa committed
647 648
      CHECK_CHARACTER_CAR (range);
      CHECK_CHARACTER_CDR (range);
649 650 651 652 653 654 655 656 657 658
      char_table_set_range (char_table,
			    XINT (XCAR (range)), XINT (XCDR (range)), value);
    }
  else
    error ("Invalid RANGE argument to `set-char-table-range'");

  return value;
}

/* Look up the element in TABLE at index CH, and return it as an
659
   integer.  If the element is not a character, return CH itself.  */
660 661

int
662
char_table_translate (Lisp_Object table, int ch)
663 664 665
{
  Lisp_Object value;
  value = Faref (table, make_number (ch));
666
  if (! CHARACTERP (value))
667 668 669 670 671
    return ch;
  return XINT (value);
}

static Lisp_Object
672
optimize_sub_char_table (Lisp_Object table, Lisp_Object test)
673 674 675 676
{
  struct Lisp_Sub_Char_Table *tbl = XSUB_CHAR_TABLE (table);
  int depth = XINT (tbl->depth);
  Lisp_Object elt, this;
677 678
  int i;
  bool optimizable;
679 680 681

  elt = XSUB_CHAR_TABLE (table)->contents[0];
  if (SUB_CHAR_TABLE_P (elt))
682 683
    {
      elt = optimize_sub_char_table (elt, test);
684
      set_sub_char_table_contents (table, 0, elt);
685
    }
686
  optimizable = SUB_CHAR_TABLE_P (elt) ? 0 : 1;
687 688 689 690
  for (i = 1; i < chartab_size[depth]; i++)
    {
      this = XSUB_CHAR_TABLE (table)->contents[i];
      if (SUB_CHAR_TABLE_P (this))
691 692
	{
	  this = optimize_sub_char_table (this, test);
693
	  set_sub_char_table_contents (table, i, this);
694
	}
695 696
      if (optimizable
	  && (NILP (test) ? NILP (Fequal (this, elt)) /* defaults to `equal'. */
697 698
	      : EQ (test, Qeq) ? !EQ (this, elt)      /* Optimize `eq' case.  */
	      : NILP (call2 (test, this, elt))))
699
	optimizable = 0;
700 701
    }

702
  return (optimizable ? elt : table);
703 704
}

Paul Eggert's avatar
Paul Eggert committed
705
DEFUN ("optimize-char-table", Foptimize_char_table, Soptimize_char_table,
706 707 708 709
       1, 2, 0,
       doc: /* Optimize CHAR-TABLE.
TEST is the comparison function used to decide whether two entries are
equivalent and can be merged.  It defaults to `equal'.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
710
  (Lisp_Object char_table, Lisp_Object test)
711 712 713 714 715 716 717 718 719 720
{
  Lisp_Object elt;
  int i;

  CHECK_CHAR_TABLE (char_table);

  for (i = 0; i < chartab_size[0]; i++)
    {
      elt = XCHAR_TABLE (char_table)->contents[i];
      if (SUB_CHAR_TABLE_P (elt))
721
	set_char_table_contents
722
	  (char_table, i, optimize_sub_char_table (elt, test));
723
    }
724
  /* Reset the `ascii' cache, in case it got optimized away.  */
Paul Eggert's avatar
Paul Eggert committed
725
  set_char_table_ascii (char_table, char_table_ascii (char_table));
726

727 728 729 730
  return Qnil;
}


731 732 733
/* Map C_FUNCTION or FUNCTION over TABLE (top or sub char-table),
   calling it for each character or group of characters that share a
   value.  RANGE is a cons (FROM . TO) specifying the range of target
734
   characters, VAL is a value of FROM in TABLE, TOP is the top
735 736 737 738 739
   char-table.

   ARG is passed to C_FUNCTION when that is called.

   It returns the value of last character covered by TABLE (not the
Paul Eggert's avatar
Paul Eggert committed
740
   value inherited from the parent), and by side-effect, the car part
741 742 743
   of RANGE is updated to the minimum character C where C and all the
   following characters in TABLE have the same value.  */

744
static Lisp_Object
745 746
map_sub_char_table (void (*c_function) (Lisp_Object, Lisp_Object, Lisp_Object),
		    Lisp_Object function, Lisp_Object table, Lisp_Object arg, Lisp_Object val,
747
		    Lisp_Object range, Lisp_Object top)
748
{
749 750
  /* Depth of TABLE.  */
  int depth;
Juanma Barranquero's avatar
Juanma Barranquero committed
751
  /* Minimum and maximum characters covered by TABLE. */
752 753 754 755
  int min_char, max_char;
  /* Number of characters covered by one element of TABLE.  */
  int chars_in_block;
  int from = XINT (XCAR (range)), to = XINT (XCDR (range));
756
  int i, c;
757
  bool is_uniprop = UNIPROP_TABLE_P (top);
758
  uniprop_decoder_t decoder = UNIPROP_GET_DECODER (top);
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
  if (SUB_CHAR_TABLE_P (table))
    {
      struct Lisp_Sub_Char_Table *tbl = XSUB_CHAR_TABLE (table);

      depth = XINT (tbl->depth);
      min_char = XINT (tbl->min_char);
      max_char = min_char + chartab_chars[depth - 1] - 1;
    }
  else
    {
      depth = 0;
      min_char = 0;
      max_char = MAX_CHAR;
    }
  chars_in_block = chartab_chars[depth];

  if (to < max_char)
    max_char = to;
  /* Set I to the index of the first element to check.  */
  if (from <= min_char)
    i = 0;
  else
    i = (from - min_char) / chars_in_block;
  for (c = min_char + chars_in_block * i; c <= max_char;
       i++, c += chars_in_block)
785
    {
786 787 788
      Lisp_Object this = (SUB_CHAR_TABLE_P (table)
			  ? XSUB_CHAR_TABLE (table)->contents[i]
			  : XCHAR_TABLE (table)->contents[i]);
789
      int nextc = c + chars_in_block;
790

791 792
      if (is_uniprop && UNIPROP_COMPRESSED_FORM_P (this))
	this = uniprop_table_uncompress (table, i);
793
      if (SUB_CHAR_TABLE_P (this))
794 795 796 797
	{
	  if (to >= nextc)
	    XSETCDR (range, make_number (nextc - 1));
	  val = map_sub_char_table (c_function, function, this, arg,
798
				    val, range, top);
799
	}
800
      else
801
	{
802
	  if (NILP (this))
803
	    this = XCHAR_TABLE (top)->defalt;
804
	  if (!EQ (val, this))
805
	    {
806
	      bool different_value = 1;
807 808 809

	      if (NILP (val))
		{
810
		  if (! NILP (XCHAR_TABLE (top)->parent))
811
		    {
812
		      Lisp_Object parent = XCHAR_TABLE (top)->parent;
813 814 815 816
		      Lisp_Object temp = XCHAR_TABLE (parent)->parent;

		      /* This is to get a value of FROM in PARENT
			 without checking the parent of PARENT.  */
Paul Eggert's avatar
Paul Eggert committed
817
		      set_char_table_parent (parent, Qnil);
818
		      val = CHAR_TABLE_REF (parent, from);
Paul Eggert's avatar
Paul Eggert committed
819
		      set_char_table_parent (parent, temp);
820 821 822
		      XSETCDR (range, make_number (c - 1));
		      val = map_sub_char_table (c_function, function,
						parent, arg, val, range,
823
						parent);
824
		      if (EQ (val, this))
825 826 827 828
			different_value = 0;
		    }
		}
	      if (! NILP (val) && different_value)
829
		{
Kenichi Handa's avatar
Kenichi Handa committed
830
		  XSETCDR (range, make_number (c - 1));
831
		  if (EQ (XCAR (range), XCDR (range)))
832 833 834 835
		    {
		      if (c_function)
			(*c_function) (arg, XCAR (range), val);
		      else
836 837 838 839 840
			{
			  if (decoder)
			    val = decoder (top, val);
			  call2 (function, XCAR (range), val);
			}
841
		    }
842
		  else
843 844 845 846
		    {
		      if (c_function)
			(*c_function) (arg, range, val);
		      else
847 848 849 850 851
			{
			  if (decoder)
			    val = decoder (top, val);
			  call2 (function, range, val);
			}
852
		    }
853
		}
854
	      val = this;
855
	      from = c;
Kenichi Handa's avatar
Kenichi Handa committed
856
	      XSETCAR (range, make_number (c));
857 858
	    }
	}
859
      XSETCDR (range, make_number (to));
860 861 862 863 864 865 866 867
    }
  return val;
}


/* Map C_FUNCTION or FUNCTION over TABLE, calling it for each
   character or group of characters that share a value.

Kenichi Handa's avatar
Kenichi Handa committed
868
   ARG is passed to C_FUNCTION when that is called.  */
869 870

void
871 872
map_char_table (void (*c_function) (Lisp_Object, Lisp_Object, Lisp_Object),
		Lisp_Object function, Lisp_Object table, Lisp_Object arg)
873
{
874 875 876
  Lisp_Object range, val, parent;
  struct gcpro gcpro1, gcpro2, gcpro3, gcpro4;
  uniprop_decoder_t decoder = UNIPROP_GET_DECODER (table);
877

878
  range = Fcons (make_number (0), make_number (MAX_CHAR));
879 880 881
  parent = XCHAR_TABLE (table)->parent;

  GCPRO4 (table, arg, range, parent);
882 883 884
  val = XCHAR_TABLE (table)->ascii;
  if (SUB_CHAR_TABLE_P (val))
    val = XSUB_CHAR_TABLE (val)->contents[0];
885
  val = map_sub_char_table (c_function, function, table, arg, val, range,
886 887
			    table);

888 889 890
  /* If VAL is nil and TABLE has a parent, we must consult the parent
     recursively.  */
  while (NILP (val) && ! NILP (XCHAR_TABLE (table)->parent))
891
    {
892
      Lisp_Object temp;
893 894
      int from = XINT (XCAR (range));

895 896
      parent = XCHAR_TABLE (table)->parent;
      temp = XCHAR_TABLE (parent)->parent;
897 898
      /* This is to get a value of FROM in PARENT without checking the
	 parent of PARENT.  */
Paul Eggert's avatar
Paul Eggert committed
899
      set_char_table_parent (parent, Qnil);
900
      val = CHAR_TABLE_REF (parent, from);
Paul Eggert's avatar
Paul Eggert committed
901
      set_char_table_parent (parent, temp);
902
      val = map_sub_char_table (c_function, function, parent, arg, val, range,
903
				parent);
904
      table = parent;
905
    }
906 907 908

  if (! NILP (val))
    {
909 910 911 912 913
      if (EQ (XCAR (range), XCDR (range)))
	{
	  if (c_function)
	    (*c_function) (arg, XCAR (range), val);
	  else
914 915 916 917 918
	    {
	      if (decoder)
		val = decoder (table, val);
	      call2 (function, XCAR (range), val);
	    }
919
	}
920
      else
921 922 923 924
	{
	  if (c_function)
	    (*c_function) (arg, range, val);
	  else
925 926 927 928 929
	    {
	      if (decoder)
		val = decoder (table, val);
	      call2 (function, range, val);
	    }
930
	}
931
    }
932 933

  UNGCPRO;
934 935 936 937
}

DEFUN ("map-char-table", Fmap_char_table, Smap_char_table,
  2, 2, 0,
Chong Yidong's avatar
Chong Yidong committed
938 939 940 941 942
       doc: /* Call FUNCTION for each character in CHAR-TABLE that has non-nil value.
FUNCTION is called with two arguments, KEY and VALUE.
KEY is a character code or a cons of character codes specifying a
range of characters that have the same value.
VALUE is what (char-table-range CHAR-TABLE KEY) returns.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
943
  (Lisp_Object function, Lisp_Object char_table)
944 945 946
{
  CHECK_CHAR_TABLE (char_table);

Kenichi Handa's avatar
Kenichi Handa committed
947
  map_char_table (NULL, function, char_table, char_table);
948 949 950
  return Qnil;
}

951 952

static void
953 954 955 956
map_sub_char_table_for_charset (void (*c_function) (Lisp_Object, Lisp_Object),
				Lisp_Object function, Lisp_Object table, Lisp_Object arg,
				Lisp_Object range, struct charset *charset,
				unsigned from, unsigned to)
957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000
{
  struct Lisp_Sub_Char_Table *tbl = XSUB_CHAR_TABLE (table);
  int depth = XINT (tbl->depth);
  int c, i;

  if (depth < 3)
    for (i = 0, c = XINT (tbl->min_char); i < chartab_size[depth];
	 i++, c += chartab_chars[depth])
      {
	Lisp_Object this;

	this = tbl->contents[i];
	if (SUB_CHAR_TABLE_P (this))
	  map_sub_char_table_for_charset (c_function, function, this, arg,
					  range, charset, from, to);
	else
	  {
	    if (! NILP (XCAR (range)))
	      {
		XSETCDR (range, make_number (c - 1));
		if (c_function)
		  (*c_function) (arg, range);
		else
		  call2 (function, range, arg);
	      }
	    XSETCAR (range, Qnil);
	  }
      }
  else
    for (i = 0, c = XINT (tbl->min_char); i < chartab_size[depth]; i++, c ++)
      {
	Lisp_Object this;
	unsigned code;

	this = tbl->contents[i];
	if (NILP (this)
	    || (charset
		&& (code = ENCODE_CHAR (charset, c),
		    (code < from || code > to))))
	  {
	    if (! NILP (XCAR (range)))
	      {
		XSETCDR (range, make_number (c - 1));
		if (c_function)
1001
		  (*c_function) (arg, range);
1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015
		else
		  call2 (function, range, arg);
		XSETCAR (range, Qnil);
	      }
	  }
	else
	  {
	    if (NILP (XCAR (range)))
	      XSETCAR (range, make_number (c));
	  }
      }
}


1016
/* Support function for `map-charset-chars'.  Map C_FUNCTION or
Juanma Barranquero's avatar
Juanma Barranquero committed
1017
   FUNCTION over TABLE, calling it for each character or a group of
1018 1019