]> code.delx.au - gnu-emacs/blob - src/charset.h
(concat): Handle bool-vectors correctly.
[gnu-emacs] / src / charset.h
1 /* Header for multilingual character handler.
2 Copyright (C) 1995, 1997, 1998 Electrotechnical Laboratory, JAPAN.
3 Licensed to the Free Software Foundation.
4
5 This file is part of GNU Emacs.
6
7 GNU Emacs is free software; you can redistribute it and/or modify
8 it under the terms of the GNU General Public License as published by
9 the Free Software Foundation; either version 2, or (at your option)
10 any later version.
11
12 GNU Emacs is distributed in the hope that it will be useful,
13 but WITHOUT ANY WARRANTY; without even the implied warranty of
14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 GNU General Public License for more details.
16
17 You should have received a copy of the GNU General Public License
18 along with GNU Emacs; see the file COPYING. If not, write to
19 the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
20 Boston, MA 02111-1307, USA. */
21
22 #ifndef _CHARSET_H
23 #define _CHARSET_H
24
25 /*** GENERAL NOTE on CHARACTER SET (CHARSET) ***
26
27 A character set ("charset" hereafter) is a meaningful collection
28 (i.e. language, culture, functionality, etc) of characters. Emacs
29 handles multiple charsets at once. Each charset corresponds to one
30 of ISO charsets (except for a special charset for composition
31 characters). Emacs identifies a charset by a unique identification
32 number, whereas ISO identifies a charset by a triplet of DIMENSION,
33 CHARS and FINAL-CHAR. So, hereafter, just saying "charset" means an
34 identification number (integer value).
35
36 The value range of charset is 0x00, 0x80..0xFE. There are four
37 kinds of charset depending on DIMENSION (1 or 2) and CHARS (94 or
38 96). For instance, a charset of DIMENSION2_CHARS94 contains 94x94
39
40
41 Within Emacs Lisp, a charset is treated as a symbol which has a
42 property `charset'. The property value is a vector containing
43 various information about the charset. For readability of C codes,
44 we use the following convention on C variable names:
45 charset_symbol: Emacs Lisp symbol of a charset
46 charset_id: Emacs Lisp integer of an identification number of a charset
47 charset: C integer of an identification number of a charset
48
49 Each charset (except for ASCII) is assigned a base leading-code
50 (range 0x80..0x9D). In addition, a charset of greater than 0xA0
51 (whose base leading-code is 0x9A..0x9D) is assigned an extended
52 leading-code (range 0xA0..0xFE). In this case, each base
53 leading-code specify the allowable range of extended leading-code as
54 shown in the table below. A leading-code is used to represent a
55 character in Emacs' buffer and string.
56
57 We call a charset which has extended leading-code as "private
58 charset" because those are mainly for a charset which is not
59 registered by ISO. On the contrary, we call a charset which does
60 not have extended leading-code as "official charset".
61
62 ---------------------------------------------------------------------------
63 charset dimension base leading-code extended leading-code
64 ---------------------------------------------------------------------------
65 0x00 official dim1 -- none -- -- none --
66 (ASCII)
67 0x01..0x7F --never used--
68 0x80 COMPOSITION same as charset -- none --
69 0x81..0x8F official dim1 same as charset -- none --
70 0x90..0x99 official dim2 same as charset -- none --
71 0x9A..0x9F --never used--
72 0xA0..0xDF private dim1 0x9A same as charset
73 of 1-column width
74 0xE0..0xEF private dim1 0x9B same as charset
75 of 2-column width
76 0xF0..0xF4 private dim2 0x9C same as charset
77 of 1-column width
78 0xF5..0xFE private dim2 0x9D same as charset
79 of 2-column width
80 0xFF --never used--
81 ---------------------------------------------------------------------------
82
83 In the table, "COMPOSITION" means a charset for a composite
84 character which is a character composed from several (up to 16)
85 non-composite characters (components). Although a composite
86 character can contain components of many charsets, a composite
87 character itself belongs to the charset CHARSET-COMPOSITION. See
88 the document "GENERAL NOTE on COMPOSITE CHARACTER" below for more
89 detail.
90
91 */
92
93 /* Definition of special leading-codes. */
94 /* Base leading-code. */
95 /* Special leading-code followed by components of a composite character. */
96 #define LEADING_CODE_COMPOSITION 0x80
97 /* Leading-code followed by extended leading-code. */
98 #define LEADING_CODE_PRIVATE_11 0x9A /* for private DIMENSION1 of 1-column */
99 #define LEADING_CODE_PRIVATE_12 0x9B /* for private DIMENSION1 of 2-column */
100 #define LEADING_CODE_PRIVATE_21 0x9C /* for private DIMENSION2 of 1-column */
101 #define LEADING_CODE_PRIVATE_22 0x9D /* for private DIMENSION2 of 2-column */
102
103 /* Extended leading-code. */
104 /* Start of each extended leading-codes. */
105 #define LEADING_CODE_EXT_11 0xA0 /* follows LEADING_CODE_PRIVATE_11 */
106 #define LEADING_CODE_EXT_12 0xE0 /* follows LEADING_CODE_PRIVATE_12 */
107 #define LEADING_CODE_EXT_21 0xF0 /* follows LEADING_CODE_PRIVATE_21 */
108 #define LEADING_CODE_EXT_22 0xF5 /* follows LEADING_CODE_PRIVATE_22 */
109 /* Maximum value of extended leading-codes. */
110 #define LEADING_CODE_EXT_MAX 0xFE
111
112 /* Definition of minimum/maximum charset of each DIMENSION. */
113 #define MIN_CHARSET_OFFICIAL_DIMENSION1 0x81
114 #define MAX_CHARSET_OFFICIAL_DIMENSION1 0x8F
115 #define MIN_CHARSET_OFFICIAL_DIMENSION2 0x90
116 #define MAX_CHARSET_OFFICIAL_DIMENSION2 0x99
117 #define MIN_CHARSET_PRIVATE_DIMENSION1 LEADING_CODE_EXT_11
118 #define MIN_CHARSET_PRIVATE_DIMENSION2 LEADING_CODE_EXT_21
119
120 /* Maximum value of overall charset identification number. */
121 #define MAX_CHARSET 0xFE
122
123 /* Definition of special charsets. */
124 #define CHARSET_ASCII 0
125 #define CHARSET_COMPOSITION 0x80
126
127 extern int charset_ascii; /* ASCII */
128 extern int charset_composition; /* for a composite character */
129 extern int charset_latin_iso8859_1; /* ISO8859-1 (Latin-1) */
130 extern int charset_jisx0208_1978; /* JISX0208.1978 (Japanese Kanji old set) */
131 extern int charset_jisx0208; /* JISX0208.1983 (Japanese Kanji) */
132 extern int charset_katakana_jisx0201; /* JISX0201.Kana (Japanese Katakana) */
133 extern int charset_latin_jisx0201; /* JISX0201.Roman (Japanese Roman) */
134 extern int charset_big5_1; /* Big5 Level 1 (Chinese Traditional) */
135 extern int charset_big5_2; /* Big5 Level 2 (Chinese Traditional) */
136
137 /* Check if CH is the head of multi-byte form, i.e.,
138 an ASCII character or a base leading-code. */
139 #define CHAR_HEAD_P(ch) ((unsigned char) (ch) < 0xA0)
140
141 /*** GENERAL NOTE on CHARACTER REPRESENTATION ***
142
143 At first, the term "character" or "char" is used for a multilingual
144 character (of course, including ASCII character), not for a byte in
145 computer memory. We use the term "code" or "byte" for the latter
146 case.
147
148 A character is identified by charset and one or two POSITION-CODEs.
149 POSITION-CODE is the position of the character in the charset. A
150 character of DIMENSION1 charset has one POSITION-CODE: POSITION-CODE-1.
151 A character of DIMENSION2 charset has two POSITION-CODE:
152 POSITION-CODE-1 and POSITION-CODE-2. The code range of
153 POSITION-CODE is 0x20..0x7F.
154
155 Emacs has two kinds of representation of a character: multi-byte
156 form (for buffer and string) and single-word form (for character
157 object in Emacs Lisp). The latter is called "character code" here
158 after. Both representation encode the information of charset and
159 POSITION-CODE but in a different way (for instance, MSB of
160 POSITION-CODE is set in multi-byte form).
161
162 For details of multi-byte form, see the section "2. Emacs internal
163 format handlers" of `coding.c'.
164
165 Emacs uses 19 bits for a character code. The bits are divided into
166 3 fields: FIELD1(5bits):FIELD2(7bits):FIELD3(7bits).
167
168 A character code of DIMENSION1 character uses FIELD2 to hold charset
169 and FIELD3 to hold POSITION-CODE-1. A character code of DIMENSION2
170 character uses FIELD1 to hold charset, FIELD2 and FIELD3 to hold
171 POSITION-CODE-1 and POSITION-CODE-2 respectively.
172
173 More precisely...
174
175 FIELD2 of DIMENSION1 character (except for ASCII) is "charset - 0x70".
176 This is to make all character codes except for ASCII greater than
177 256 (ASCII's FIELD2 is 0). So, the range of FIELD2 of DIMENSION1
178 character is 0 or 0x11..0x7F.
179
180 FIELD1 of DIMENSION2 character is "charset - 0x8F" for official
181 charset and "charset - 0xE0" for private charset. So, the range of
182 FIELD1 of DIMENSION2 character is 0x01..0x1E.
183
184 -----------------------------------------------------------------------
185 charset FIELD1 (5-bit) FIELD2 (7-bit) FIELD3 (7-bit)
186 -----------------------------------------------------------------------
187 ASCII 0 0 POSITION-CODE-1
188 DIMENSION1 0 charset - 0x70 POSITION-CODE-1
189 DIMENSION2(o) charset - 0x8F POSITION-CODE-1 POSITION-CODE-2
190 DIMENSION2(p) charset - 0xE0 POSITION-CODE-1 POSITION-CODE-2
191 -----------------------------------------------------------------------
192 "(o)": official, "(p)": private
193 -----------------------------------------------------------------------
194
195 */
196
197 /*** GENERAL NOTE on COMPOSITE CHARACTER ***
198
199 A composite character is a character composed from several (up to
200 16) non-composite characters (components). Although each components
201 can belong to any charset, a composite character itself belongs to
202 the charset `charset-composition' and is assigned a special
203 leading-code `LEADING_CODE_COMPOSITION' for multi-byte form. See
204 the document "2. Emacs internal format handlers" in `coding.c' for
205 more detail about multi-byte form.
206
207 A character code of composite character has special format. In the
208 above document, FIELD1 of a composite character is 0x1F. Each
209 composite character is assigned a sequential number CMPCHAR-ID.
210 FIELD2 and FIELD3 are combined to make 14bits field for holding
211 CMPCHAR-ID, which means that Emacs can handle at most 2^14 (= 16384)
212 composite characters at once.
213
214 -----------------------------------------------------------------------
215 charset FIELD1 (5-bit) FIELD2&3 (14-bit)
216 -----------------------------------------------------------------------
217 CHARSET-COMPOSITION 0x1F CMPCHAR-ID
218 -----------------------------------------------------------------------
219
220 Emacs assigns CMPCHAR-ID to a composite character only when it
221 requires the character code of the composite character (e.g. while
222 displaying the composite character).
223
224 */
225
226 /* Masks of each field of character code. */
227 #define CHAR_FIELD1_MASK (0x1F << 14)
228 #define CHAR_FIELD2_MASK (0x7F << 7)
229 #define CHAR_FIELD3_MASK 0x7F
230
231 /* Macros to access each field of character C. */
232 #define CHAR_FIELD1(c) (((c) & CHAR_FIELD1_MASK) >> 14)
233 #define CHAR_FIELD2(c) (((c) & CHAR_FIELD2_MASK) >> 7)
234 #define CHAR_FIELD3(c) ((c) & CHAR_FIELD3_MASK)
235
236 /* Minimum character code of character of each DIMENSION. */
237 #define MIN_CHAR_OFFICIAL_DIMENSION1 \
238 ((MIN_CHARSET_OFFICIAL_DIMENSION1 - 0x70) << 7)
239 #define MIN_CHAR_PRIVATE_DIMENSION1 \
240 ((MIN_CHARSET_PRIVATE_DIMENSION1 - 0x70) << 7)
241 #define MIN_CHAR_OFFICIAL_DIMENSION2 \
242 ((MIN_CHARSET_OFFICIAL_DIMENSION2 - 0x8F) << 14)
243 #define MIN_CHAR_PRIVATE_DIMENSION2 \
244 ((MIN_CHARSET_PRIVATE_DIMENSION2 - 0xE0) << 14)
245 #define MIN_CHAR_COMPOSITION \
246 (0x1F << 14)
247 #define MAX_CHAR_COMPOSITION GLYPH_MASK_CHAR
248
249 /* 1 if C is an ASCII character, else 0. */
250 #define SINGLE_BYTE_CHAR_P(c) ((c) < 0x100)
251 /* 1 if C is an composite character, else 0. */
252 #define COMPOSITE_CHAR_P(c) ((c) >= MIN_CHAR_COMPOSITION)
253
254 /* 1 if BYTE is a character in itself, in multibyte mode. */
255 #define ASCII_BYTE_P(byte) ((byte) < 0x80)
256
257 /* A char-table containing information of each character set.
258
259 Unlike ordinary char-tables, this doesn't contain any nested table.
260 Only the top level elements are used. Each element is a vector of
261 the following information:
262 CHARSET-ID, BYTES, DIMENSION, CHARS, WIDTH, DIRECTION,
263 LEADING-CODE-BASE, LEADING-CODE-EXT,
264 ISO-FINAL-CHAR, ISO-GRAPHIC-PLANE,
265 REVERSE-CHARSET, SHORT-NAME, LONG-NAME, DESCRIPTION,
266 PLIST.
267
268 CHARSET-ID (integer) is the identification number of the charset.
269
270 BYTE (integer) is the length of multi-byte form of a character in
271 the charset: one of 1, 2, 3, and 4.
272
273 DIMENSION (integer) is the number of bytes to represent a character: 1 or 2.
274
275 CHARS (integer) is the number of characters in a dimension: 94 or 96.
276
277 WIDTH (integer) is the number of columns a character in the charset
278 occupies on the screen: one of 0, 1, and 2.
279
280 DIRECTION (integer) is the rendering direction of characters in the
281 charset when rendering. If 0, render from right to left, else
282 render from left to right.
283
284 LEADING-CODE-BASE (integer) is the base leading-code for the
285 charset.
286
287 LEADING-CODE-EXT (integer) is the extended leading-code for the
288 charset. All charsets of less than 0xA0 has the value 0.
289
290 ISO-FINAL-CHAR (character) is the final character of the
291 corresponding ISO 2022 charset.
292
293 ISO-GRAPHIC-PLANE (integer) is the graphic plane to be invoked
294 while encoding to variants of ISO 2022 coding system, one of the
295 following: 0/graphic-plane-left(GL), 1/graphic-plane-right(GR).
296
297 REVERSE-CHARSET (integer) is the charset which differs only in
298 LEFT-TO-RIGHT value from the charset. If there's no such a
299 charset, the value is -1.
300
301 SHORT-NAME (string) is the short name to refer to the charset.
302
303 LONG-NAME (string) is the long name to refer to the charset.
304
305 DESCRIPTION (string) is the description string of the charset.
306
307 PLIST (property list) may contain any type of information a user
308 want to put and get by functions `put-charset-property' and
309 `get-charset-property' respectively. */
310 extern Lisp_Object Vcharset_table;
311
312 /* Macros to access various information of CHARSET in Vcharset_table.
313 We provide these macros for efficiency. No range check of CHARSET. */
314
315 /* Return entry of CHARSET (lisp integer) in Vcharset_table. */
316 #define CHARSET_TABLE_ENTRY(charset) \
317 XCHAR_TABLE (Vcharset_table)->contents[((charset) == CHARSET_ASCII \
318 ? 0 : (charset) + 128)]
319
320 /* Return information INFO-IDX of CHARSET. */
321 #define CHARSET_TABLE_INFO(charset, info_idx) \
322 XVECTOR (CHARSET_TABLE_ENTRY (charset))->contents[info_idx]
323
324 #define CHARSET_ID_IDX (0)
325 #define CHARSET_BYTES_IDX (1)
326 #define CHARSET_DIMENSION_IDX (2)
327 #define CHARSET_CHARS_IDX (3)
328 #define CHARSET_WIDTH_IDX (4)
329 #define CHARSET_DIRECTION_IDX (5)
330 #define CHARSET_LEADING_CODE_BASE_IDX (6)
331 #define CHARSET_LEADING_CODE_EXT_IDX (7)
332 #define CHARSET_ISO_FINAL_CHAR_IDX (8)
333 #define CHARSET_ISO_GRAPHIC_PLANE_IDX (9)
334 #define CHARSET_REVERSE_CHARSET_IDX (10)
335 #define CHARSET_SHORT_NAME_IDX (11)
336 #define CHARSET_LONG_NAME_IDX (12)
337 #define CHARSET_DESCRIPTION_IDX (13)
338 #define CHARSET_PLIST_IDX (14)
339 /* Size of a vector of each entry of Vcharset_table. */
340 #define CHARSET_MAX_IDX (15)
341
342 /* And several more macros to be used frequently. */
343 #define CHARSET_BYTES(charset) \
344 XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_BYTES_IDX))
345 #define CHARSET_DIMENSION(charset) \
346 XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_DIMENSION_IDX))
347 #define CHARSET_CHARS(charset) \
348 XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_CHARS_IDX))
349 #define CHARSET_WIDTH(charset) \
350 XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_WIDTH_IDX))
351 #define CHARSET_DIRECTION(charset) \
352 XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_DIRECTION_IDX))
353 #define CHARSET_LEADING_CODE_BASE(charset) \
354 XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_LEADING_CODE_BASE_IDX))
355 #define CHARSET_LEADING_CODE_EXT(charset) \
356 XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_LEADING_CODE_EXT_IDX))
357 #define CHARSET_ISO_FINAL_CHAR(charset) \
358 XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_ISO_FINAL_CHAR_IDX))
359 #define CHARSET_ISO_GRAPHIC_PLANE(charset) \
360 XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_ISO_GRAPHIC_PLANE_IDX))
361 #define CHARSET_REVERSE_CHARSET(charset) \
362 XINT (CHARSET_TABLE_INFO (charset, CHARSET_REVERSE_CHARSET_IDX))
363
364 /* Macros to specify direction of a charset. */
365 #define CHARSET_DIRECTION_LEFT_TO_RIGHT 0
366 #define CHARSET_DIRECTION_RIGHT_TO_LEFT 1
367
368 /* A vector of charset symbol indexed by charset-id. This is used
369 only for returning charset symbol from C functions. */
370 extern Lisp_Object Vcharset_symbol_table;
371
372 /* Return symbol of CHARSET. */
373 #define CHARSET_SYMBOL(charset) \
374 XVECTOR (Vcharset_symbol_table)->contents[charset]
375
376 /* 1 if CHARSET is valid, else 0. */
377 #define CHARSET_VALID_P(charset) \
378 ((charset) == 0 \
379 || ((charset) >= 0x80 && (charset) <= MAX_CHARSET_OFFICIAL_DIMENSION2) \
380 || ((charset) >= MIN_CHARSET_PRIVATE_DIMENSION1 && (charset) <= MAX_CHARSET))
381
382 /* 1 if CHARSET is already defined, else 0. */
383 #define CHARSET_DEFINED_P(charset) \
384 (((charset) >= 0) && ((charset) <= MAX_CHARSET) \
385 && !NILP (CHARSET_TABLE_ENTRY (charset)))
386
387 /* Since the information CHARSET-BYTES and CHARSET-WIDTH of
388 Vcharset_table can be retrieved only from the first byte of
389 multi-byte form (an ASCII code or a base leading-code), we provide
390 here tables to be used by macros BYTES_BY_CHAR_HEAD and
391 WIDTH_BY_CHAR_HEAD for faster information retrieval. */
392 extern int bytes_by_char_head[256];
393 extern int width_by_char_head[256];
394
395 #define BYTES_BY_CHAR_HEAD(char_head) bytes_by_char_head[char_head]
396 #define WIDTH_BY_CHAR_HEAD(char_head) width_by_char_head[char_head]
397
398 /* Charset of the character C. */
399 #define CHAR_CHARSET(c) \
400 (SINGLE_BYTE_CHAR_P (c) \
401 ? CHARSET_ASCII \
402 : ((c) < MIN_CHAR_OFFICIAL_DIMENSION2 \
403 ? CHAR_FIELD2 (c) + 0x70 \
404 : ((c) < MIN_CHAR_PRIVATE_DIMENSION2 \
405 ? CHAR_FIELD1 (c) + 0x8F \
406 : ((c) < MIN_CHAR_COMPOSITION \
407 ? CHAR_FIELD1 (c) + 0xE0 \
408 : ((c) <= MAX_CHAR_COMPOSITION \
409 ? CHARSET_COMPOSITION \
410 : CHARSET_ASCII)))))
411
412 /* Return charset at the place pointed by P. */
413 #define CHARSET_AT(p) \
414 (*(p) < 0x80 \
415 ? CHARSET_ASCII \
416 : (*(p) == LEADING_CODE_COMPOSITION \
417 ? CHARSET_COMPOSITION \
418 : (*(p) < LEADING_CODE_PRIVATE_11 \
419 ? (int)*(p) \
420 : (*(p) <= LEADING_CODE_PRIVATE_22 \
421 ? (int)*((p) + 1) \
422 : -1))))
423
424 /* Same as `CHARSET_AT ()' but perhaps runs faster because of an
425 additional argument C which is the code (byte) at P. */
426 #define FIRST_CHARSET_AT(p, c) \
427 ((c) < 0x80 \
428 ? CHARSET_ASCII \
429 : ((c) == LEADING_CODE_COMPOSITION \
430 ? CHARSET_COMPOSITION \
431 : ((c) < LEADING_CODE_PRIVATE_11 \
432 ? (int)(c) \
433 : ((c) <= LEADING_CODE_PRIVATE_22 \
434 ? (int)*((p) + 1) \
435 : -1))))
436
437 /* Check if two characters C1 and C2 belong to the same charset.
438 Always return 0 for composite characters. */
439 #define SAME_CHARSET_P(c1, c2) \
440 (c1 < MIN_CHAR_COMPOSITION \
441 && (SINGLE_BYTE_CHAR_P (c1) \
442 ? SINGLE_BYTE_CHAR_P (c2) \
443 : (c1 < MIN_CHAR_OFFICIAL_DIMENSION2 \
444 ? (c1 & CHAR_FIELD2_MASK) == (c2 & CHAR_FIELD2_MASK) \
445 : (c1 & CHAR_FIELD1_MASK) == (c2 & CHAR_FIELD1_MASK))))
446
447 /* Return a non-ASCII character of which charset is CHARSET and
448 position-codes are C1 and C2. DIMENSION1 character ignores C2. */
449 #define MAKE_NON_ASCII_CHAR(charset, c1, c2) \
450 ((charset) == CHARSET_COMPOSITION \
451 ? MAKE_COMPOSITE_CHAR (((c1) << 7) + (c2)) \
452 : (CHARSET_DIMENSION (charset) == 1 \
453 ? (((charset) - 0x70) << 7) | (c1) \
454 : ((charset) < MIN_CHARSET_PRIVATE_DIMENSION2 \
455 ? (((charset) - 0x8F) << 14) | ((c1) << 7) | (c2) \
456 : (((charset) - 0xE0) << 14) | ((c1) << 7) | (c2))))
457
458 /* Return a composite character of which CMPCHAR-ID is ID. */
459 #define MAKE_COMPOSITE_CHAR(id) (MIN_CHAR_COMPOSITION + (id))
460
461 /* Return CMPCHAR-ID of a composite character C. */
462 #define COMPOSITE_CHAR_ID(c) ((c) - MIN_CHAR_COMPOSITION)
463
464 /* Return a character of which charset is CHARSET and position-codes
465 are C1 and C2. DIMENSION1 character ignores C2. */
466 #define MAKE_CHAR(charset, c1, c2) \
467 ((charset) == CHARSET_ASCII \
468 ? (c1) \
469 : MAKE_NON_ASCII_CHAR ((charset), (c1) & 0x7F, (c2) & 0x7F))
470
471 /* If GENERICP is nonzero, return nonzero iff C is a valid normal or
472 generic character. If GENERICP is zero, return nonzero iff C is a
473 valid normal character. */
474 #define CHAR_VALID_P(c, genericp) \
475 ((c) >= 0 \
476 && (SINGLE_BYTE_CHAR_P (c) || char_valid_p (c, genericp)))
477
478 /* The charset of non-ASCII character C is stored in CHARSET, and the
479 position-codes of C are stored in C1 and C2.
480 We store -1 in C2 if the character is just 2 bytes.
481
482 Do not use this macro for an ASCII character. */
483
484 #define SPLIT_NON_ASCII_CHAR(c, charset, c1, c2) \
485 ((c) < MIN_CHAR_OFFICIAL_DIMENSION2 \
486 ? (charset = CHAR_FIELD2 (c) + 0x70, \
487 c1 = CHAR_FIELD3 (c), \
488 c2 = -1) \
489 : (charset = ((c) < MIN_CHAR_COMPOSITION \
490 ? (CHAR_FIELD1 (c) \
491 + ((c) < MIN_CHAR_PRIVATE_DIMENSION2 ? 0x8F : 0xE0)) \
492 : CHARSET_COMPOSITION), \
493 c1 = CHAR_FIELD2 (c), \
494 c2 = CHAR_FIELD3 (c)))
495
496 /* The charset of character C is stored in CHARSET, and the
497 position-codes of C are stored in C1 and C2.
498 We store -1 in C2 if the character is just 2 bytes. */
499
500 #define SPLIT_CHAR(c, charset, c1, c2) \
501 (SINGLE_BYTE_CHAR_P (c) \
502 ? charset = CHARSET_ASCII, c1 = (c), c2 = -1 \
503 : SPLIT_NON_ASCII_CHAR (c, charset, c1, c2))
504
505 /* The charset of the character at STR is stored in CHARSET, and the
506 position-codes are stored in C1 and C2.
507 We store -1 in C2 if the character is just 2 bytes.
508
509 If the character is a composite character, the upper 7-bit and
510 lower 7-bit of CMPCHAR-ID are set in C1 and C2 respectively. No
511 range checking. */
512
513 #define SPLIT_STRING(str, len, charset, c1, c2) \
514 ((BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) < 2 \
515 || BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) > len \
516 || split_non_ascii_string (str, len, &charset, &c1, &c2) < 0) \
517 ? c1 = *(str), charset = CHARSET_ASCII \
518 : charset)
519
520 /* Mapping table from ISO2022's charset (specified by DIMENSION,
521 CHARS, and FINAL_CHAR) to Emacs' charset. Should be accessed by
522 macro ISO_CHARSET_TABLE (DIMENSION, CHARS, FINAL_CHAR). */
523 extern int iso_charset_table[2][2][128];
524
525 #define ISO_CHARSET_TABLE(dimension, chars, final_char) \
526 iso_charset_table[XINT (dimension) - 1][XINT (chars) > 94][XINT (final_char)]
527
528 #define BASE_LEADING_CODE_P(c) (BYTES_BY_CHAR_HEAD ((unsigned char) (c)) > 1)
529
530 /* The following two macros CHAR_STRING and STRING_CHAR are the main
531 entry points to convert between Emacs two types of character
532 representations: multi-byte form and single-word form (character
533 code). */
534
535 /* Set STR a pointer to the multi-byte form of the character C. If C
536 is not a composite character, the multi-byte form is set in WORKBUF
537 and STR points WORKBUF. The caller should allocate at least 4-byte
538 area at WORKBUF in advance. Returns the length of the multi-byte
539 form. If C is an invalid character code, signal an error. */
540
541 #define CHAR_STRING(c, workbuf, str) \
542 (SINGLE_BYTE_CHAR_P (c) \
543 ? *(str = workbuf) = (unsigned char)(c), 1 \
544 : non_ascii_char_to_string (c, workbuf, &str))
545
546 /* Return a character code of the character of which multi-byte form
547 is at STR and the length is LEN. If STR doesn't contain valid
548 multi-byte form, only the first byte in STR is returned. */
549
550 #define STRING_CHAR(str, len) \
551 ((BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) == 1 \
552 || BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) > (len)) \
553 ? (unsigned char) *(str) \
554 : string_to_non_ascii_char (str, len, 0))
555
556 /* This is like STRING_CHAR but the third arg ACTUAL_LEN is set to
557 the length of the multi-byte form. Just to know the length, use
558 MULTIBYTE_FORM_LENGTH. */
559
560 #define STRING_CHAR_AND_LENGTH(str, len, actual_len) \
561 ((BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) == 1 \
562 || BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) > (len)) \
563 ? (actual_len = 1), (unsigned char) *(str) \
564 : string_to_non_ascii_char (str, len, &actual_len))
565
566 /* Fetch the "next" multibyte character from Lisp string STRING
567 at byte position BYTEIDX, character position CHARIDX.
568 Store it into OUTPUT.
569
570 All the args must be side-effect-free.
571 BYTEIDX and CHARIDX must be lvalues;
572 we increment them past the character fetched. */
573
574 #define FETCH_STRING_CHAR_ADVANCE(OUTPUT, STRING, CHARIDX, BYTEIDX) \
575 if (1) \
576 { \
577 unsigned char *fetch_string_char_ptr = &XSTRING (STRING)->data[BYTEIDX]; \
578 int fetch_string_char_space_left = XSTRING (STRING)->size_byte - BYTEIDX; \
579 int actual_len; \
580 \
581 OUTPUT \
582 = STRING_CHAR_AND_LENGTH (fetch_string_char_ptr, \
583 fetch_string_char_space_left, actual_len); \
584 \
585 BYTEIDX += actual_len; \
586 CHARIDX++; \
587 } \
588 else
589
590 /* Return the length of the multi-byte form at string STR of length LEN. */
591
592 #define MULTIBYTE_FORM_LENGTH(str, len) \
593 ((BYTES_BY_CHAR_HEAD (*(unsigned char *)(str)) == 1 \
594 || BYTES_BY_CHAR_HEAD (*(unsigned char *)(str)) > (len)) \
595 ? 1 \
596 : multibyte_form_length (str, len))
597
598 /* Set C a (possibly multibyte) character at P. P points into a
599 string which is the virtual concatenation of STR1 (which ends at
600 END1) or STR2 (which ends at END2). */
601
602 #define GET_CHAR_AFTER_2(c, p, str1, end1, str2, end2) \
603 do { \
604 const char *dtemp = (p) == (end1) ? (str2) : (p); \
605 const char *dlimit = ((p) >= (str1) && (p) < (end1)) ? (end1) : (end2); \
606 c = STRING_CHAR (dtemp, dlimit - dtemp); \
607 } while (0)
608
609 /* Set C a (possibly multibyte) character before P. P points into a
610 string which is the virtual concatenation of STR1 (which ends at
611 END1) or STR2 (which ends at END2). */
612
613 #define GET_CHAR_BEFORE_2(c, p, str1, end1, str2, end2) \
614 do { \
615 const char *dtemp = (p); \
616 const char *dlimit = ((p) > (str2) && (p) <= (end2)) ? (str2) : (str1); \
617 while (dtemp-- > dlimit && (unsigned char) *dtemp >= 0xA0); \
618 c = STRING_CHAR (dtemp, p - dtemp); \
619 } while (0)
620
621 #ifdef emacs
622
623 /* Increase the buffer point POS of the current buffer to the next
624 character boundary. This macro relies on the fact that *GPT_ADDR
625 and *Z_ADDR are always accessible and the values are '\0'. No
626 range checking of POS. */
627 #define INC_POS(pos) \
628 do { \
629 unsigned char *p = BYTE_POS_ADDR (pos); \
630 pos++; \
631 if (*p++ >= 0x80) \
632 while (!CHAR_HEAD_P (*p)) p++, pos++; \
633 } while (0)
634
635 /* Decrease the buffer point POS of the current buffer to the previous
636 character boundary. No range checking of POS. */
637 #define DEC_POS(pos) \
638 do { \
639 unsigned char *p, *p_min; \
640 int pos_saved = --pos; \
641 if (pos < GPT_BYTE) \
642 p = BEG_ADDR + pos - 1, p_min = BEG_ADDR; \
643 else \
644 p = BEG_ADDR + GAP_SIZE + pos - 1, p_min = GAP_END_ADDR; \
645 while (p > p_min && !CHAR_HEAD_P (*p)) p--, pos--; \
646 if (*p < 0x80 && pos != pos_saved) pos = pos_saved; \
647 } while (0)
648
649 /* Increment both CHARPOS and BYTEPOS, each in the appropriate way. */
650
651 #define INC_BOTH(charpos, bytepos) \
652 do \
653 { \
654 (charpos)++; \
655 INC_POS ((bytepos)); \
656 } \
657 while (0)
658
659 /* Decrement both CHARPOS and BYTEPOS, each in the appropriate way. */
660
661 #define DEC_BOTH(charpos, bytepos) \
662 do \
663 { \
664 (charpos)--; \
665 DEC_POS ((bytepos)); \
666 } \
667 while (0)
668
669 /* Increase the buffer point POS of the current buffer to the next
670 character boundary. This macro relies on the fact that *GPT_ADDR
671 and *Z_ADDR are always accessible and the values are '\0'. No
672 range checking of POS. */
673 #define BUF_INC_POS(buf, pos) \
674 do { \
675 unsigned char *p = BUF_BYTE_ADDRESS (buf, pos); \
676 pos++; \
677 if (*p++ >= 0x80) \
678 while (!CHAR_HEAD_P (*p)) p++, pos++; \
679 } while (0)
680
681 /* Decrease the buffer point POS of the current buffer to the previous
682 character boundary. No range checking of POS. */
683 #define BUF_DEC_POS(buf, pos) \
684 do { \
685 unsigned char *p, *p_min; \
686 int pos_saved = --pos; \
687 if (pos < BUF_GPT_BYTE (buf)) \
688 { \
689 p = BUF_BEG_ADDR (buf) + pos - 1; \
690 p_min = BUF_BEG_ADDR (buf); \
691 } \
692 else \
693 { \
694 p = BUF_BEG_ADDR (buf) + BUF_GAP_SIZE (buf) + pos - 1; \
695 p_min = BUF_GAP_END_ADDR (buf); \
696 } \
697 while (p > p_min && !CHAR_HEAD_P (*p)) p--, pos--; \
698 if (*p < 0x80 && pos != pos_saved) pos = pos_saved; \
699 } while (0)
700
701 #endif /* emacs */
702
703 /* Maximum counts of components in one composite character. */
704 #define MAX_COMPONENT_COUNT 16
705
706 /* Structure to hold information of a composite character. */
707 struct cmpchar_info {
708 /* Byte length of the composite character. */
709 int len;
710
711 /* Multi-byte form of the composite character. */
712 unsigned char *data;
713
714 /* Length of glyph codes. */
715 int glyph_len;
716
717 /* Width of the overall glyph of the composite character. */
718 int width;
719
720 /* Pointer to an array of glyph codes of the composite character.
721 This actually contains only character code, no face. */
722 GLYPH *glyph;
723
724 /* Pointer to an array of composition rules. The value has the form:
725 (0xA0 + ((GLOBAL-REF-POINT << 2) | NEW-REF-POINT))
726 where each XXX-REF-POINT is 0..8. */
727 unsigned char *cmp_rule;
728
729 /* Pointer to an array of x-axis offset of left edge of glyphs
730 relative to the left of of glyph[0] except for the first element
731 which is the absolute offset from the left edge of overall glyph.
732 The actual pixel offset should be calculated by multiplying each
733 frame's one column width by this value:
734 (i.e. FONT_WIDTH (f->output_data.x->font) * col_offset[N]). */
735 float *col_offset;
736
737 /* Work slot used by `dumpglyphs' (xterm.c). */
738 int face_work;
739 };
740
741 /* Table of pointers to the structure `cmpchar_info' indexed by
742 CMPCHAR-ID. */
743 extern struct cmpchar_info **cmpchar_table;
744 /* Number of the current composite characters. */
745 extern int n_cmpchars;
746
747 /* This is the maximum length of multi-byte form. */
748 #define MAX_LENGTH_OF_MULTI_BYTE_FORM (MAX_COMPONENT_COUNT * 6)
749
750 /* Maximum character code currently used. */
751 #define MAX_CHAR (MIN_CHAR_COMPOSITION + n_cmpchars)
752
753 extern int unify_char P_ ((Lisp_Object, int, int, int, int));
754 extern int split_non_ascii_string P_ ((unsigned char *, int, int *,
755 unsigned char *, unsigned char *));
756 extern int string_to_non_ascii_char P_ ((unsigned char *, int, int *));
757 extern int non_ascii_char_to_string P_ ((int, unsigned char *, unsigned char **));
758 extern int multibyte_form_length P_ ((unsigned char *, int));
759 extern int str_cmpchar_id P_ ((unsigned char *, int));
760 extern int get_charset_id P_ ((Lisp_Object));
761 extern int cmpchar_component P_ ((unsigned int, unsigned int));
762 extern int find_charset_in_str P_ ((unsigned char *, int, int *, Lisp_Object));
763 extern int strwidth P_ ((unsigned char *, int));
764
765 extern Lisp_Object Vcharacter_unification_table_vector;
766 #define UNIFICATION_ID_TABLE(id) \
767 (XCONS(XVECTOR(Vcharacter_unification_table_vector)->contents[(id)])->cdr)
768
769 /* Copy LEN bytes from FROM to TO. This macro should be used only
770 when a caller knows that LEN is short and the obvious copy loop is
771 faster than calling bcopy which has some overhead. */
772
773 #define BCOPY_SHORT(from, to, len) \
774 do { \
775 int i = len; \
776 unsigined char *from_p = from, *to_p = to; \
777 while (i--) *from_p++ = *to_p++; \
778 } while (0)
779
780 #endif /* _CHARSET_H */