Commit | Line | Data |
---|---|---|
75c8c592 | 1 | /* Basic multilingual character support. |
35e623fb | 2 | Copyright (C) 1995, 1997, 1998 Electrotechnical Laboratory, JAPAN. |
75c8c592 | 3 | Licensed to the Free Software Foundation. |
4ed46869 | 4 | |
369314dc KH |
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. | |
4ed46869 | 11 | |
369314dc KH |
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. | |
4ed46869 | 16 | |
369314dc KH |
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. */ | |
4ed46869 KH |
21 | |
22 | /* At first, see the document in `charset.h' to understand the code in | |
23 | this file. */ | |
24 | ||
25 | #include <stdio.h> | |
26 | ||
27 | #ifdef emacs | |
28 | ||
29 | #include <sys/types.h> | |
30 | #include <config.h> | |
31 | #include "lisp.h" | |
32 | #include "buffer.h" | |
33 | #include "charset.h" | |
34 | #include "coding.h" | |
fc6b09bf | 35 | #include "disptab.h" |
4ed46869 KH |
36 | |
37 | #else /* not emacs */ | |
38 | ||
39 | #include "mulelib.h" | |
40 | ||
41 | #endif /* emacs */ | |
42 | ||
43 | Lisp_Object Qcharset, Qascii, Qcomposition; | |
44 | ||
45 | /* Declaration of special leading-codes. */ | |
46 | int leading_code_composition; /* for composite characters */ | |
47 | int leading_code_private_11; /* for private DIMENSION1 of 1-column */ | |
48 | int leading_code_private_12; /* for private DIMENSION1 of 2-column */ | |
49 | int leading_code_private_21; /* for private DIMENSION2 of 1-column */ | |
50 | int leading_code_private_22; /* for private DIMENSION2 of 2-column */ | |
51 | ||
52 | /* Declaration of special charsets. */ | |
53 | int charset_ascii; /* ASCII */ | |
54 | int charset_composition; /* for a composite character */ | |
55 | int charset_latin_iso8859_1; /* ISO8859-1 (Latin-1) */ | |
56 | int charset_jisx0208_1978; /* JISX0208.1978 (Japanese Kanji old set) */ | |
57 | int charset_jisx0208; /* JISX0208.1983 (Japanese Kanji) */ | |
58 | int charset_katakana_jisx0201; /* JISX0201.Kana (Japanese Katakana) */ | |
59 | int charset_latin_jisx0201; /* JISX0201.Roman (Japanese Roman) */ | |
60 | int charset_big5_1; /* Big5 Level 1 (Chinese Traditional) */ | |
61 | int charset_big5_2; /* Big5 Level 2 (Chinese Traditional) */ | |
62 | ||
b0e3cf2b KH |
63 | int min_composite_char; |
64 | ||
4ed46869 KH |
65 | Lisp_Object Qcharset_table; |
66 | ||
67 | /* A char-table containing information of each character set. */ | |
68 | Lisp_Object Vcharset_table; | |
69 | ||
70 | /* A vector of charset symbol indexed by charset-id. This is used | |
71 | only for returning charset symbol from C functions. */ | |
72 | Lisp_Object Vcharset_symbol_table; | |
73 | ||
74 | /* A list of charset symbols ever defined. */ | |
75 | Lisp_Object Vcharset_list; | |
76 | ||
b0e3cf2b KH |
77 | /* Vector of unification table ever defined. |
78 | An ID of a unification table is an index of this vector. */ | |
79 | Lisp_Object Vcharacter_unification_table_vector; | |
80 | ||
4ed46869 KH |
81 | /* Tables used by macros BYTES_BY_CHAR_HEAD and WIDTH_BY_CHAR_HEAD. */ |
82 | int bytes_by_char_head[256]; | |
83 | int width_by_char_head[256]; | |
84 | ||
85 | /* Mapping table from ISO2022's charset (specified by DIMENSION, | |
86 | CHARS, and FINAL-CHAR) to Emacs' charset. */ | |
87 | int iso_charset_table[2][2][128]; | |
88 | ||
513ee442 KH |
89 | /* Table of pointers to the structure `cmpchar_info' indexed by |
90 | CMPCHAR-ID. */ | |
91 | struct cmpchar_info **cmpchar_table; | |
92 | /* The current size of `cmpchar_table'. */ | |
93 | static int cmpchar_table_size; | |
94 | /* Number of the current composite characters. */ | |
95 | int n_cmpchars; | |
96 | ||
4ed46869 KH |
97 | /* Variables used locally in the macro FETCH_MULTIBYTE_CHAR. */ |
98 | unsigned char *_fetch_multibyte_char_p; | |
99 | int _fetch_multibyte_char_len; | |
100 | ||
35e623fb RS |
101 | /* Offset to add to a non-ASCII value when inserting it. */ |
102 | int nonascii_insert_offset; | |
103 | ||
4cf9710d RS |
104 | /* Translation table for converting non-ASCII unibyte characters |
105 | to multibyte codes, or nil. */ | |
106 | Lisp_Object Vnonascii_translate_table; | |
107 | ||
046b1f03 RS |
108 | #define min(X, Y) ((X) < (Y) ? (X) : (Y)) |
109 | #define max(X, Y) ((X) > (Y) ? (X) : (Y)) | |
110 | \f | |
93bcb785 KH |
111 | void |
112 | invalid_character (c) | |
113 | int c; | |
114 | { | |
115 | error ("Invalid character: %o, %d, 0x%x", c); | |
116 | } | |
117 | ||
118 | ||
4ed46869 KH |
119 | /* Set STR a pointer to the multi-byte form of the character C. If C |
120 | is not a composite character, the multi-byte form is set in WORKBUF | |
121 | and STR points WORKBUF. The caller should allocate at least 4-byte | |
122 | area at WORKBUF in advance. Returns the length of the multi-byte | |
bd4c6dd0 KH |
123 | form. If C is an invalid character to have a multi-byte form, |
124 | signal an error. | |
4ed46869 KH |
125 | |
126 | Use macro `CHAR_STRING (C, WORKBUF, STR)' instead of calling this | |
127 | function directly if C can be an ASCII character. */ | |
128 | ||
129 | int | |
130 | non_ascii_char_to_string (c, workbuf, str) | |
131 | int c; | |
132 | unsigned char *workbuf, **str; | |
133 | { | |
6dc0722d | 134 | int charset, c1, c2; |
4ed46869 KH |
135 | |
136 | if (COMPOSITE_CHAR_P (c)) | |
137 | { | |
138 | int cmpchar_id = COMPOSITE_CHAR_ID (c); | |
139 | ||
140 | if (cmpchar_id < n_cmpchars) | |
141 | { | |
142 | *str = cmpchar_table[cmpchar_id]->data; | |
143 | return cmpchar_table[cmpchar_id]->len; | |
144 | } | |
145 | else | |
146 | { | |
93bcb785 | 147 | invalid_character (c); |
4ed46869 KH |
148 | } |
149 | } | |
150 | ||
151 | SPLIT_NON_ASCII_CHAR (c, charset, c1, c2); | |
bd4c6dd0 KH |
152 | if (!charset |
153 | || ! CHARSET_DEFINED_P (charset) | |
154 | || c1 >= 0 && c1 < 32 | |
155 | || c2 >= 0 && c2 < 32) | |
93bcb785 | 156 | invalid_character (c); |
4ed46869 KH |
157 | |
158 | *str = workbuf; | |
159 | *workbuf++ = CHARSET_LEADING_CODE_BASE (charset); | |
160 | if (*workbuf = CHARSET_LEADING_CODE_EXT (charset)) | |
161 | workbuf++; | |
162 | *workbuf++ = c1 | 0x80; | |
6dc0722d | 163 | if (c2 >= 0) |
4ed46869 KH |
164 | *workbuf++ = c2 | 0x80; |
165 | ||
166 | return (workbuf - *str); | |
167 | } | |
168 | ||
169 | /* Return a non-ASCII character of which multi-byte form is at STR of | |
170 | length LEN. If ACTUAL_LEN is not NULL, the actual length of the | |
171 | character is set to the address ACTUAL_LEN. | |
172 | ||
173 | Use macro `STRING_CHAR (STR, LEN)' instead of calling this function | |
174 | directly if STR can hold an ASCII character. */ | |
175 | ||
176 | string_to_non_ascii_char (str, len, actual_len) | |
8867de67 | 177 | const unsigned char *str; |
4ed46869 KH |
178 | int len, *actual_len; |
179 | { | |
180 | int charset; | |
181 | unsigned char c1, c2; | |
182 | register int c; | |
183 | ||
184 | if (SPLIT_STRING (str, len, charset, c1, c2) == CHARSET_ASCII) | |
185 | { | |
186 | if (actual_len) | |
187 | *actual_len = 1; | |
188 | return (int) *str; | |
189 | } | |
190 | ||
191 | c = MAKE_NON_ASCII_CHAR (charset, c1, c2); | |
192 | ||
193 | if (actual_len) | |
194 | *actual_len = (charset == CHARSET_COMPOSITION | |
195 | ? cmpchar_table[COMPOSITE_CHAR_ID (c)]->len | |
196 | : BYTES_BY_CHAR_HEAD (*str)); | |
197 | return c; | |
198 | } | |
199 | ||
200 | /* Return the length of the multi-byte form at string STR of length LEN. */ | |
201 | int | |
202 | multibyte_form_length (str, len) | |
8867de67 | 203 | const unsigned char *str; |
4ed46869 KH |
204 | int len; |
205 | { | |
206 | int charset; | |
207 | unsigned char c1, c2; | |
208 | register int c; | |
209 | ||
210 | if (SPLIT_STRING (str, len, charset, c1, c2) == CHARSET_ASCII) | |
211 | return 1; | |
212 | ||
213 | return (charset == CHARSET_COMPOSITION | |
214 | ? cmpchar_table[(c1 << 7) | c2]->len | |
215 | : BYTES_BY_CHAR_HEAD (*str)); | |
216 | } | |
217 | ||
218 | /* Check if string STR of length LEN contains valid multi-byte form of | |
219 | a character. If valid, charset and position codes of the character | |
220 | is set at *CHARSET, *C1, and *C2, and return 0. If not valid, | |
221 | return -1. This should be used only in the macro SPLIT_STRING | |
222 | which checks range of STR in advance. */ | |
223 | ||
224 | split_non_ascii_string (str, len, charset, c1, c2) | |
8867de67 KH |
225 | register const unsigned char *str; |
226 | register unsigned char *c1, *c2; | |
4ed46869 KH |
227 | register int len, *charset; |
228 | { | |
229 | register unsigned int cs = *str++; | |
230 | ||
231 | if (cs == LEADING_CODE_COMPOSITION) | |
232 | { | |
233 | int cmpchar_id = str_cmpchar_id (str - 1, len); | |
234 | ||
235 | if (cmpchar_id < 0) | |
236 | return -1; | |
237 | *charset = cs, *c1 = cmpchar_id >> 7, *c2 = cmpchar_id & 0x7F; | |
238 | } | |
239 | else if ((cs < LEADING_CODE_PRIVATE_11 || (cs = *str++) >= 0xA0) | |
240 | && CHARSET_DEFINED_P (cs)) | |
241 | { | |
242 | *charset = cs; | |
243 | if (*str < 0xA0) | |
244 | return -1; | |
245 | *c1 = (*str++) & 0x7F; | |
246 | if (CHARSET_DIMENSION (cs) == 2) | |
247 | { | |
248 | if (*str < 0xA0) | |
249 | return -1; | |
250 | *c2 = (*str++) & 0x7F; | |
251 | } | |
252 | } | |
253 | else | |
254 | return -1; | |
255 | return 0; | |
256 | } | |
257 | ||
23d2a7f1 KH |
258 | /* Return a character unified with C (or a character made of CHARSET, |
259 | C1, and C2) in unification table TABLE. If no unification is found | |
260 | in TABLE, return C. */ | |
261 | unify_char (table, c, charset, c1, c2) | |
262 | Lisp_Object table; | |
263 | int c, charset, c1, c2; | |
264 | { | |
265 | Lisp_Object ch; | |
266 | int alt_charset, alt_c1, alt_c2, dimension; | |
267 | ||
268 | if (c < 0) c = MAKE_CHAR (charset, c1, c2); | |
269 | if (!CHAR_TABLE_P (table) | |
270 | || (ch = Faref (table, make_number (c)), !INTEGERP (ch)) | |
271 | || XINT (ch) < 0) | |
272 | return c; | |
273 | ||
274 | SPLIT_CHAR (XFASTINT (ch), alt_charset, alt_c1, alt_c2); | |
275 | dimension = CHARSET_DIMENSION (alt_charset); | |
276 | if (dimension == 1 && alt_c1 > 0 || dimension == 2 && alt_c2 > 0) | |
277 | /* CH is not a generic character, just return it. */ | |
278 | return XFASTINT (ch); | |
279 | ||
280 | /* Since CH is a generic character, we must return a specific | |
281 | charater which has the same position codes as C from CH. */ | |
282 | if (charset < 0) | |
283 | SPLIT_CHAR (c, charset, c1, c2); | |
284 | if (dimension != CHARSET_DIMENSION (charset)) | |
285 | /* We can't make such a character because of dimension mismatch. */ | |
286 | return c; | |
23d2a7f1 KH |
287 | return MAKE_CHAR (alt_charset, c1, c2); |
288 | } | |
289 | ||
d2665018 KH |
290 | /* Convert the unibyte character C to multibyte based on |
291 | Vnonascii_translate_table or nonascii_insert_offset. If they can't | |
292 | convert C to a valid multibyte character, convert it based on | |
293 | DEFAULT_NONASCII_INSERT_OFFSET which makes C a Latin-1 character. */ | |
35e623fb | 294 | |
35e623fb RS |
295 | unibyte_char_to_multibyte (c) |
296 | int c; | |
297 | { | |
d2665018 | 298 | if (c >= 0240 && c < 0400) |
35e623fb | 299 | { |
d2665018 KH |
300 | int c_save = c; |
301 | ||
4cf9710d RS |
302 | if (! NILP (Vnonascii_translate_table)) |
303 | c = XINT (Faref (Vnonascii_translate_table, make_number (c))); | |
304 | else if (nonascii_insert_offset > 0) | |
35e623fb | 305 | c += nonascii_insert_offset; |
d2665018 KH |
306 | if (c >= 0240 && (c < 0400 || ! VALID_MULTIBYTE_CHAR_P (c))) |
307 | c = c_save + DEFAULT_NONASCII_INSERT_OFFSET; | |
35e623fb RS |
308 | } |
309 | return c; | |
310 | } | |
311 | \f | |
4ed46869 KH |
312 | /* Update the table Vcharset_table with the given arguments (see the |
313 | document of `define-charset' for the meaning of each argument). | |
314 | Several other table contents are also updated. The caller should | |
315 | check the validity of CHARSET-ID and the remaining arguments in | |
316 | advance. */ | |
317 | ||
318 | void | |
319 | update_charset_table (charset_id, dimension, chars, width, direction, | |
320 | iso_final_char, iso_graphic_plane, | |
321 | short_name, long_name, description) | |
322 | Lisp_Object charset_id, dimension, chars, width, direction; | |
323 | Lisp_Object iso_final_char, iso_graphic_plane; | |
324 | Lisp_Object short_name, long_name, description; | |
325 | { | |
326 | int charset = XINT (charset_id); | |
327 | int bytes; | |
328 | unsigned char leading_code_base, leading_code_ext; | |
329 | ||
6dc0722d KH |
330 | if (NILP (CHARSET_TABLE_ENTRY (charset))) |
331 | CHARSET_TABLE_ENTRY (charset) | |
332 | = Fmake_vector (make_number (CHARSET_MAX_IDX), Qnil); | |
4ed46869 KH |
333 | |
334 | /* Get byte length of multibyte form, base leading-code, and | |
335 | extended leading-code of the charset. See the comment under the | |
336 | title "GENERAL NOTE on CHARACTER SET (CHARSET)" in charset.h. */ | |
337 | bytes = XINT (dimension); | |
338 | if (charset < MIN_CHARSET_PRIVATE_DIMENSION1) | |
339 | { | |
340 | /* Official charset, it doesn't have an extended leading-code. */ | |
341 | if (charset != CHARSET_ASCII) | |
342 | bytes += 1; /* For a base leading-code. */ | |
343 | leading_code_base = charset; | |
344 | leading_code_ext = 0; | |
345 | } | |
346 | else | |
347 | { | |
348 | /* Private charset. */ | |
349 | bytes += 2; /* For base and extended leading-codes. */ | |
350 | leading_code_base | |
351 | = (charset < LEADING_CODE_EXT_12 | |
352 | ? LEADING_CODE_PRIVATE_11 | |
353 | : (charset < LEADING_CODE_EXT_21 | |
354 | ? LEADING_CODE_PRIVATE_12 | |
355 | : (charset < LEADING_CODE_EXT_22 | |
356 | ? LEADING_CODE_PRIVATE_21 | |
357 | : LEADING_CODE_PRIVATE_22))); | |
358 | leading_code_ext = charset; | |
359 | } | |
360 | ||
361 | CHARSET_TABLE_INFO (charset, CHARSET_ID_IDX) = charset_id; | |
362 | CHARSET_TABLE_INFO (charset, CHARSET_BYTES_IDX) = make_number (bytes); | |
363 | CHARSET_TABLE_INFO (charset, CHARSET_DIMENSION_IDX) = dimension; | |
364 | CHARSET_TABLE_INFO (charset, CHARSET_CHARS_IDX) = chars; | |
365 | CHARSET_TABLE_INFO (charset, CHARSET_WIDTH_IDX) = width; | |
366 | CHARSET_TABLE_INFO (charset, CHARSET_DIRECTION_IDX) = direction; | |
367 | CHARSET_TABLE_INFO (charset, CHARSET_LEADING_CODE_BASE_IDX) | |
368 | = make_number (leading_code_base); | |
369 | CHARSET_TABLE_INFO (charset, CHARSET_LEADING_CODE_EXT_IDX) | |
370 | = make_number (leading_code_ext); | |
371 | CHARSET_TABLE_INFO (charset, CHARSET_ISO_FINAL_CHAR_IDX) = iso_final_char; | |
372 | CHARSET_TABLE_INFO (charset, CHARSET_ISO_GRAPHIC_PLANE_IDX) | |
373 | = iso_graphic_plane; | |
374 | CHARSET_TABLE_INFO (charset, CHARSET_SHORT_NAME_IDX) = short_name; | |
375 | CHARSET_TABLE_INFO (charset, CHARSET_LONG_NAME_IDX) = long_name; | |
376 | CHARSET_TABLE_INFO (charset, CHARSET_DESCRIPTION_IDX) = description; | |
377 | CHARSET_TABLE_INFO (charset, CHARSET_PLIST_IDX) = Qnil; | |
378 | ||
379 | { | |
380 | /* If we have already defined a charset which has the same | |
381 | DIMENSION, CHARS and ISO-FINAL-CHAR but the different | |
382 | DIRECTION, we must update the entry REVERSE-CHARSET of both | |
383 | charsets. If there's no such charset, the value of the entry | |
384 | is set to nil. */ | |
385 | int i; | |
386 | ||
513ee442 | 387 | for (i = 0; i <= MAX_CHARSET; i++) |
4ed46869 KH |
388 | if (!NILP (CHARSET_TABLE_ENTRY (i))) |
389 | { | |
390 | if (CHARSET_DIMENSION (i) == XINT (dimension) | |
391 | && CHARSET_CHARS (i) == XINT (chars) | |
392 | && CHARSET_ISO_FINAL_CHAR (i) == XINT (iso_final_char) | |
393 | && CHARSET_DIRECTION (i) != XINT (direction)) | |
394 | { | |
395 | CHARSET_TABLE_INFO (charset, CHARSET_REVERSE_CHARSET_IDX) | |
396 | = make_number (i); | |
397 | CHARSET_TABLE_INFO (i, CHARSET_REVERSE_CHARSET_IDX) = charset_id; | |
398 | break; | |
399 | } | |
400 | } | |
513ee442 | 401 | if (i > MAX_CHARSET) |
4ed46869 KH |
402 | /* No such a charset. */ |
403 | CHARSET_TABLE_INFO (charset, CHARSET_REVERSE_CHARSET_IDX) | |
404 | = make_number (-1); | |
405 | } | |
406 | ||
407 | if (charset != CHARSET_ASCII | |
408 | && charset < MIN_CHARSET_PRIVATE_DIMENSION1) | |
409 | { | |
410 | /* Update tables bytes_by_char_head and width_by_char_head. */ | |
411 | bytes_by_char_head[leading_code_base] = bytes; | |
412 | width_by_char_head[leading_code_base] = XINT (width); | |
413 | ||
414 | /* Update table emacs_code_class. */ | |
415 | emacs_code_class[charset] = (bytes == 2 | |
416 | ? EMACS_leading_code_2 | |
417 | : (bytes == 3 | |
418 | ? EMACS_leading_code_3 | |
419 | : EMACS_leading_code_4)); | |
420 | } | |
421 | ||
422 | /* Update table iso_charset_table. */ | |
423 | if (ISO_CHARSET_TABLE (dimension, chars, iso_final_char) < 0) | |
424 | ISO_CHARSET_TABLE (dimension, chars, iso_final_char) = charset; | |
425 | } | |
426 | ||
427 | #ifdef emacs | |
428 | ||
429 | /* Return charset id of CHARSET_SYMBOL, or return -1 if CHARSET_SYMBOL | |
430 | is invalid. */ | |
431 | int | |
432 | get_charset_id (charset_symbol) | |
433 | Lisp_Object charset_symbol; | |
434 | { | |
435 | Lisp_Object val; | |
436 | int charset; | |
437 | ||
438 | return ((SYMBOLP (charset_symbol) | |
439 | && (val = Fget (charset_symbol, Qcharset), VECTORP (val)) | |
440 | && (charset = XINT (XVECTOR (val)->contents[CHARSET_ID_IDX]), | |
441 | CHARSET_VALID_P (charset))) | |
442 | ? charset : -1); | |
443 | } | |
444 | ||
445 | /* Return an identification number for a new private charset of | |
446 | DIMENSION and WIDTH. If there's no more room for the new charset, | |
447 | return 0. */ | |
448 | Lisp_Object | |
449 | get_new_private_charset_id (dimension, width) | |
450 | int dimension, width; | |
451 | { | |
452 | int charset, from, to; | |
453 | ||
454 | if (dimension == 1) | |
455 | { | |
456 | if (width == 1) | |
457 | from = LEADING_CODE_EXT_11, to = LEADING_CODE_EXT_12; | |
458 | else | |
459 | from = LEADING_CODE_EXT_12, to = LEADING_CODE_EXT_21; | |
460 | } | |
461 | else | |
462 | { | |
463 | if (width == 1) | |
464 | from = LEADING_CODE_EXT_21, to = LEADING_CODE_EXT_22; | |
465 | else | |
b0e3cf2b | 466 | from = LEADING_CODE_EXT_22, to = LEADING_CODE_EXT_MAX + 1; |
4ed46869 KH |
467 | } |
468 | ||
469 | for (charset = from; charset < to; charset++) | |
470 | if (!CHARSET_DEFINED_P (charset)) break; | |
471 | ||
472 | return make_number (charset < to ? charset : 0); | |
473 | } | |
474 | ||
475 | DEFUN ("define-charset", Fdefine_charset, Sdefine_charset, 3, 3, 0, | |
476 | "Define CHARSET-ID as the identification number of CHARSET with INFO-VECTOR.\n\ | |
23d2a7f1 | 477 | If CHARSET-ID is nil, it is decided automatically, which means CHARSET is\n\ |
4ed46869 KH |
478 | treated as a private charset.\n\ |
479 | INFO-VECTOR is a vector of the format:\n\ | |
480 | [DIMENSION CHARS WIDTH DIRECTION ISO-FINAL-CHAR ISO-GRAPHIC-PLANE\n\ | |
481 | SHORT-NAME LONG-NAME DESCRIPTION]\n\ | |
482 | The meanings of each elements is as follows:\n\ | |
483 | DIMENSION (integer) is the number of bytes to represent a character: 1 or 2.\n\ | |
484 | CHARS (integer) is the number of characters in a dimension: 94 or 96.\n\ | |
485 | WIDTH (integer) is the number of columns a character in the charset\n\ | |
486 | occupies on the screen: one of 0, 1, and 2.\n\ | |
487 | \n\ | |
488 | DIRECTION (integer) is the rendering direction of characters in the\n\ | |
489 | charset when rendering. If 0, render from right to left, else\n\ | |
490 | render from left to right.\n\ | |
491 | \n\ | |
492 | ISO-FINAL-CHAR (character) is the final character of the\n\ | |
493 | corresponding ISO 2022 charset.\n\ | |
494 | \n\ | |
495 | ISO-GRAPHIC-PLANE (integer) is the graphic plane to be invoked\n\ | |
496 | while encoding to variants of ISO 2022 coding system, one of the\n\ | |
497 | following: 0/graphic-plane-left(GL), 1/graphic-plane-right(GR).\n\ | |
498 | \n\ | |
499 | SHORT-NAME (string) is the short name to refer to the charset.\n\ | |
500 | \n\ | |
501 | LONG-NAME (string) is the long name to refer to the charset.\n\ | |
502 | \n\ | |
503 | DESCRIPTION (string) is the description string of the charset.") | |
504 | (charset_id, charset_symbol, info_vector) | |
505 | Lisp_Object charset_id, charset_symbol, info_vector; | |
506 | { | |
507 | Lisp_Object *vec; | |
508 | ||
509 | if (!NILP (charset_id)) | |
510 | CHECK_NUMBER (charset_id, 0); | |
511 | CHECK_SYMBOL (charset_symbol, 1); | |
512 | CHECK_VECTOR (info_vector, 2); | |
513 | ||
514 | if (! NILP (charset_id)) | |
515 | { | |
516 | if (! CHARSET_VALID_P (XINT (charset_id))) | |
517 | error ("Invalid CHARSET: %d", XINT (charset_id)); | |
518 | else if (CHARSET_DEFINED_P (XINT (charset_id))) | |
519 | error ("Already defined charset: %d", XINT (charset_id)); | |
520 | } | |
521 | ||
522 | vec = XVECTOR (info_vector)->contents; | |
523 | if (XVECTOR (info_vector)->size != 9 | |
524 | || !INTEGERP (vec[0]) || !(XINT (vec[0]) == 1 || XINT (vec[0]) == 2) | |
525 | || !INTEGERP (vec[1]) || !(XINT (vec[1]) == 94 || XINT (vec[1]) == 96) | |
526 | || !INTEGERP (vec[2]) || !(XINT (vec[2]) == 1 || XINT (vec[2]) == 2) | |
527 | || !INTEGERP (vec[3]) || !(XINT (vec[3]) == 0 || XINT (vec[3]) == 1) | |
528 | || !INTEGERP (vec[4]) || !(XINT (vec[4]) >= '0' && XINT (vec[4]) <= '~') | |
529 | || !INTEGERP (vec[5]) || !(XINT (vec[5]) == 0 || XINT (vec[5]) == 1) | |
530 | || !STRINGP (vec[6]) | |
531 | || !STRINGP (vec[7]) | |
532 | || !STRINGP (vec[8])) | |
533 | error ("Invalid info-vector argument for defining charset %s", | |
534 | XSYMBOL (charset_symbol)->name->data); | |
535 | ||
536 | if (NILP (charset_id)) | |
537 | { | |
538 | charset_id = get_new_private_charset_id (XINT (vec[0]), XINT (vec[2])); | |
539 | if (XINT (charset_id) == 0) | |
540 | error ("There's no room for a new private charset %s", | |
541 | XSYMBOL (charset_symbol)->name->data); | |
542 | } | |
543 | ||
544 | update_charset_table (charset_id, vec[0], vec[1], vec[2], vec[3], | |
545 | vec[4], vec[5], vec[6], vec[7], vec[8]); | |
6dc0722d | 546 | Fput (charset_symbol, Qcharset, CHARSET_TABLE_ENTRY (XINT (charset_id))); |
4ed46869 KH |
547 | CHARSET_SYMBOL (XINT (charset_id)) = charset_symbol; |
548 | Vcharset_list = Fcons (charset_symbol, Vcharset_list); | |
549 | return Qnil; | |
550 | } | |
551 | ||
3fac5a51 KH |
552 | DEFUN ("get-unused-iso-final-char", Fget_unused_iso_final_char, |
553 | Sget_unused_iso_final_char, 2, 2, 0, | |
554 | "Return an unsed ISO's final char for a charset of DIMENISION and CHARS.\n\ | |
555 | DIMENSION is the number of bytes to represent a character: 1 or 2.\n\ | |
556 | CHARS is the number of characters in a dimension: 94 or 96.\n\ | |
557 | \n\ | |
558 | This final char is for private use, thus the range is `0' (48) .. `?' (63).\n\ | |
559 | If there's no unused final char for the specified kind of charset,\n\ | |
560 | return nil.") | |
561 | (dimension, chars) | |
562 | Lisp_Object dimension, chars; | |
563 | { | |
564 | int final_char; | |
565 | ||
566 | CHECK_NUMBER (dimension, 0); | |
567 | CHECK_NUMBER (chars, 1); | |
568 | if (XINT (dimension) != 1 && XINT (dimension) != 2) | |
569 | error ("Invalid charset dimension %d, it should be 1 or 2", | |
570 | XINT (dimension)); | |
571 | if (XINT (chars) != 94 && XINT (chars) != 96) | |
572 | error ("Invalid charset chars %d, it should be 94 or 96", | |
573 | XINT (chars)); | |
574 | for (final_char = '0'; final_char <= '?'; final_char++) | |
575 | { | |
576 | if (ISO_CHARSET_TABLE (dimension, chars, make_number (final_char)) < 0) | |
577 | break; | |
578 | } | |
579 | return (final_char <= '?' ? make_number (final_char) : Qnil); | |
580 | } | |
581 | ||
4ed46869 KH |
582 | DEFUN ("declare-equiv-charset", Fdeclare_equiv_charset, Sdeclare_equiv_charset, |
583 | 4, 4, 0, | |
584 | "Declare a charset of DIMENSION, CHARS, FINAL-CHAR is the same as CHARSET.\n\ | |
585 | CHARSET should be defined by `defined-charset' in advance.") | |
586 | (dimension, chars, final_char, charset_symbol) | |
587 | Lisp_Object dimension, chars, final_char, charset_symbol; | |
588 | { | |
589 | int charset; | |
590 | ||
591 | CHECK_NUMBER (dimension, 0); | |
592 | CHECK_NUMBER (chars, 1); | |
593 | CHECK_NUMBER (final_char, 2); | |
594 | CHECK_SYMBOL (charset_symbol, 3); | |
595 | ||
596 | if (XINT (dimension) != 1 && XINT (dimension) != 2) | |
597 | error ("Invalid DIMENSION %d, it should be 1 or 2", XINT (dimension)); | |
598 | if (XINT (chars) != 94 && XINT (chars) != 96) | |
599 | error ("Invalid CHARS %d, it should be 94 or 96", XINT (chars)); | |
600 | if (XINT (final_char) < '0' || XFASTINT (final_char) > '~') | |
601 | error ("Invalid FINAL-CHAR %c, it should be `0'..`~'", XINT (chars)); | |
602 | if ((charset = get_charset_id (charset_symbol)) < 0) | |
603 | error ("Invalid charset %s", XSYMBOL (charset_symbol)->name->data); | |
604 | ||
605 | ISO_CHARSET_TABLE (dimension, chars, final_char) = charset; | |
606 | return Qnil; | |
607 | } | |
608 | ||
609 | /* Return number of different charsets in STR of length LEN. In | |
610 | addition, for each found charset N, CHARSETS[N] is set 1. The | |
a29e3b1b | 611 | caller should allocate CHARSETS (MAX_CHARSET + 1 elements) in advance. |
23d2a7f1 | 612 | It may lookup a unification table TABLE if supplied. */ |
4ed46869 KH |
613 | |
614 | int | |
23d2a7f1 | 615 | find_charset_in_str (str, len, charsets, table) |
028d516b KH |
616 | unsigned char *str; |
617 | int len, *charsets; | |
23d2a7f1 | 618 | Lisp_Object table; |
4ed46869 | 619 | { |
733eafd8 | 620 | register int num = 0, c; |
4ed46869 | 621 | |
23d2a7f1 KH |
622 | if (! CHAR_TABLE_P (table)) |
623 | table = Qnil; | |
624 | ||
4ed46869 KH |
625 | while (len > 0) |
626 | { | |
05505664 | 627 | int bytes, charset; |
733eafd8 | 628 | c = *str; |
23d2a7f1 | 629 | |
733eafd8 | 630 | if (c == LEADING_CODE_COMPOSITION) |
05505664 | 631 | { |
733eafd8 KH |
632 | int cmpchar_id = str_cmpchar_id (str, len); |
633 | GLYPH *glyph; | |
05505664 | 634 | |
733eafd8 | 635 | if (cmpchar_id > 0) |
05505664 | 636 | { |
733eafd8 KH |
637 | struct cmpchar_info *cmpcharp = cmpchar_table[cmpchar_id]; |
638 | int i; | |
639 | ||
640 | for (i = 0; i < cmpcharp->glyph_len; i++) | |
641 | { | |
642 | c = cmpcharp->glyph[i]; | |
643 | if (!NILP (table)) | |
644 | { | |
645 | if ((c = unify_char (table, c, 0, 0, 0)) < 0) | |
646 | c = cmpcharp->glyph[i]; | |
647 | } | |
648 | if ((charset = CHAR_CHARSET (c)) < 0) | |
649 | charset = CHARSET_ASCII; | |
650 | if (!charsets[charset]) | |
651 | { | |
652 | charsets[charset] = 1; | |
653 | num += 1; | |
654 | } | |
655 | } | |
656 | str += cmpcharp->len; | |
657 | len -= cmpcharp->len; | |
658 | continue; | |
05505664 | 659 | } |
05505664 | 660 | |
733eafd8 KH |
661 | charset = CHARSET_ASCII; |
662 | bytes = 1; | |
663 | } | |
23d2a7f1 KH |
664 | else |
665 | { | |
733eafd8 KH |
666 | c = STRING_CHAR_AND_LENGTH (str, len, bytes); |
667 | if (! NILP (table)) | |
668 | { | |
669 | int c1 = unify_char (table, c, 0, 0, 0); | |
670 | if (c1 >= 0) | |
671 | c = c1; | |
672 | } | |
673 | charset = CHAR_CHARSET (c); | |
23d2a7f1 | 674 | } |
4ed46869 KH |
675 | |
676 | if (!charsets[charset]) | |
677 | { | |
678 | charsets[charset] = 1; | |
679 | num += 1; | |
680 | } | |
681 | str += bytes; | |
682 | len -= bytes; | |
683 | } | |
684 | return num; | |
685 | } | |
686 | ||
687 | DEFUN ("find-charset-region", Ffind_charset_region, Sfind_charset_region, | |
23d2a7f1 | 688 | 2, 3, 0, |
4ed46869 | 689 | "Return a list of charsets in the region between BEG and END.\n\ |
23d2a7f1 KH |
690 | BEG and END are buffer positions.\n\ |
691 | Optional arg TABLE if non-nil is a unification table to look up.") | |
692 | (beg, end, table) | |
693 | Lisp_Object beg, end, table; | |
4ed46869 | 694 | { |
028d516b | 695 | int charsets[MAX_CHARSET + 1]; |
6ae1f27e | 696 | int from, from_byte, to, stop, stop_byte, i; |
4ed46869 KH |
697 | Lisp_Object val; |
698 | ||
699 | validate_region (&beg, &end); | |
700 | from = XFASTINT (beg); | |
701 | stop = to = XFASTINT (end); | |
6ae1f27e | 702 | |
4ed46869 | 703 | if (from < GPT && GPT < to) |
6ae1f27e RS |
704 | { |
705 | stop = GPT; | |
706 | stop_byte = GPT_BYTE; | |
707 | } | |
708 | else | |
709 | stop_byte = CHAR_TO_BYTE (stop); | |
710 | ||
711 | from_byte = CHAR_TO_BYTE (from); | |
712 | ||
028d516b | 713 | bzero (charsets, (MAX_CHARSET + 1) * sizeof (int)); |
4ed46869 KH |
714 | while (1) |
715 | { | |
6ae1f27e RS |
716 | find_charset_in_str (BYTE_POS_ADDR (from_byte), stop_byte - from_byte, |
717 | charsets, table); | |
4ed46869 | 718 | if (stop < to) |
6ae1f27e RS |
719 | { |
720 | from = stop, from_byte = stop_byte; | |
721 | stop = to, stop_byte = CHAR_TO_BYTE (stop); | |
722 | } | |
4ed46869 KH |
723 | else |
724 | break; | |
725 | } | |
6ae1f27e | 726 | |
4ed46869 | 727 | val = Qnil; |
513ee442 | 728 | for (i = MAX_CHARSET; i >= 0; i--) |
4ed46869 KH |
729 | if (charsets[i]) |
730 | val = Fcons (CHARSET_SYMBOL (i), val); | |
731 | return val; | |
732 | } | |
733 | ||
734 | DEFUN ("find-charset-string", Ffind_charset_string, Sfind_charset_string, | |
23d2a7f1 KH |
735 | 1, 2, 0, |
736 | "Return a list of charsets in STR.\n\ | |
737 | Optional arg TABLE if non-nil is a unification table to look up.") | |
738 | (str, table) | |
739 | Lisp_Object str, table; | |
4ed46869 | 740 | { |
a29e3b1b | 741 | int charsets[MAX_CHARSET + 1]; |
4ed46869 KH |
742 | int i; |
743 | Lisp_Object val; | |
744 | ||
745 | CHECK_STRING (str, 0); | |
87b089ad RS |
746 | |
747 | if (! STRING_MULTIBYTE (str)) | |
748 | return Qnil; | |
749 | ||
a29e3b1b | 750 | bzero (charsets, (MAX_CHARSET + 1) * sizeof (int)); |
fc932ac6 | 751 | find_charset_in_str (XSTRING (str)->data, STRING_BYTES (XSTRING (str)), |
23d2a7f1 | 752 | charsets, table); |
4ed46869 | 753 | val = Qnil; |
513ee442 | 754 | for (i = MAX_CHARSET; i >= 0; i--) |
4ed46869 KH |
755 | if (charsets[i]) |
756 | val = Fcons (CHARSET_SYMBOL (i), val); | |
757 | return val; | |
758 | } | |
759 | \f | |
760 | DEFUN ("make-char-internal", Fmake_char_internal, Smake_char_internal, 1, 3, 0, | |
513ee442 | 761 | "") |
4ed46869 KH |
762 | (charset, code1, code2) |
763 | Lisp_Object charset, code1, code2; | |
764 | { | |
765 | CHECK_NUMBER (charset, 0); | |
766 | ||
767 | if (NILP (code1)) | |
768 | XSETFASTINT (code1, 0); | |
769 | else | |
770 | CHECK_NUMBER (code1, 1); | |
771 | if (NILP (code2)) | |
772 | XSETFASTINT (code2, 0); | |
773 | else | |
774 | CHECK_NUMBER (code2, 2); | |
775 | ||
776 | if (!CHARSET_DEFINED_P (XINT (charset))) | |
777 | error ("Invalid charset: %d", XINT (charset)); | |
778 | ||
779 | return make_number (MAKE_CHAR (XINT (charset), XINT (code1), XINT (code2))); | |
780 | } | |
781 | ||
782 | DEFUN ("split-char", Fsplit_char, Ssplit_char, 1, 1, 0, | |
783 | "Return list of charset and one or two position-codes of CHAR.") | |
784 | (ch) | |
785 | Lisp_Object ch; | |
786 | { | |
787 | Lisp_Object val; | |
6dc0722d | 788 | int charset, c1, c2; |
4ed46869 KH |
789 | |
790 | CHECK_NUMBER (ch, 0); | |
791 | SPLIT_CHAR (XFASTINT (ch), charset, c1, c2); | |
6dc0722d | 792 | return (c2 >= 0 |
4ed46869 KH |
793 | ? Fcons (CHARSET_SYMBOL (charset), |
794 | Fcons (make_number (c1), Fcons (make_number (c2), Qnil))) | |
795 | : Fcons (CHARSET_SYMBOL (charset), Fcons (make_number (c1), Qnil))); | |
796 | } | |
797 | ||
798 | DEFUN ("char-charset", Fchar_charset, Schar_charset, 1, 1, 0, | |
799 | "Return charset of CHAR.") | |
800 | (ch) | |
801 | Lisp_Object ch; | |
802 | { | |
803 | CHECK_NUMBER (ch, 0); | |
804 | ||
805 | return CHARSET_SYMBOL (CHAR_CHARSET (XINT (ch))); | |
806 | } | |
807 | ||
808 | DEFUN ("iso-charset", Fiso_charset, Siso_charset, 3, 3, 0, | |
2b71bb78 KH |
809 | "Return charset of ISO's specification DIMENSION, CHARS, and FINAL-CHAR.\n\ |
810 | \n\ | |
811 | ISO 2022's designation sequence (escape sequence) distinguishes charsets\n\ | |
812 | by their DIMENSION, CHARS, and FINAL-CHAR,\n\ | |
813 | where as Emacs distinguishes them by charset symbol.\n\ | |
814 | See the documentation of the function `charset-info' for the meanings of\n\ | |
815 | DIMENSION, CHARS, and FINAL-CHAR.") | |
4ed46869 KH |
816 | (dimension, chars, final_char) |
817 | Lisp_Object dimension, chars, final_char; | |
818 | { | |
819 | int charset; | |
820 | ||
821 | CHECK_NUMBER (dimension, 0); | |
822 | CHECK_NUMBER (chars, 1); | |
823 | CHECK_NUMBER (final_char, 2); | |
824 | ||
825 | if ((charset = ISO_CHARSET_TABLE (dimension, chars, final_char)) < 0) | |
826 | return Qnil; | |
827 | return CHARSET_SYMBOL (charset); | |
828 | } | |
829 | ||
9d3d8cba KH |
830 | /* If GENERICP is nonzero, return nonzero iff C is a valid normal or |
831 | generic character. If GENERICP is zero, return nonzero iff C is a | |
832 | valid normal character. Do not call this function directly, | |
833 | instead use macro CHAR_VALID_P. */ | |
834 | int | |
835 | char_valid_p (c, genericp) | |
836 | int c, genericp; | |
837 | { | |
838 | int charset, c1, c2; | |
839 | ||
840 | if (c < 0) | |
841 | return 0; | |
842 | if (SINGLE_BYTE_CHAR_P (c)) | |
843 | return 1; | |
844 | SPLIT_NON_ASCII_CHAR (c, charset, c1, c2); | |
845 | if (!CHARSET_VALID_P (charset)) | |
846 | return 0; | |
847 | return (c < MIN_CHAR_COMPOSITION | |
848 | ? ((c & CHAR_FIELD1_MASK) /* i.e. dimension of C is two. */ | |
849 | ? (genericp && c1 == 0 && c2 == 0 | |
850 | || c1 >= 32 && c2 >= 32) | |
851 | : (genericp && c1 == 0 | |
852 | || c1 >= 32)) | |
853 | : c < MIN_CHAR_COMPOSITION + n_cmpchars); | |
854 | } | |
855 | ||
856 | DEFUN ("char-valid-p", Fchar_valid_p, Schar_valid_p, 1, 2, 0, | |
a9d02884 DL |
857 | "Return t if OBJECT is a valid normal character.\n\ |
858 | If optional arg GENERICP is non-nil, also return t if OBJECT is\n\ | |
9d3d8cba KH |
859 | a valid generic character.") |
860 | (object, genericp) | |
861 | Lisp_Object object, genericp; | |
862 | { | |
863 | if (! NATNUMP (object)) | |
864 | return Qnil; | |
865 | return (CHAR_VALID_P (XFASTINT (object), !NILP (genericp)) ? Qt : Qnil); | |
866 | } | |
867 | ||
d2665018 KH |
868 | DEFUN ("unibyte-char-to-multibyte", Funibyte_char_to_multibyte, |
869 | Sunibyte_char_to_multibyte, 1, 1, 0, | |
870 | "Convert the unibyte character CH to multibyte character.\n\ | |
871 | The conversion is done based on nonascii-translate-table (which see)\n\ | |
872 | or nonascii-insert-offset (which see).") | |
873 | (ch) | |
874 | Lisp_Object ch; | |
875 | { | |
876 | int c; | |
877 | ||
878 | CHECK_NUMBER (ch, 0); | |
879 | c = XINT (ch); | |
880 | if (c < 0 || c >= 0400) | |
881 | error ("Invalid unibyte character: %d", c); | |
882 | c = unibyte_char_to_multibyte (c); | |
883 | if (c < 0) | |
884 | error ("Can't convert to multibyte character: %d", XINT (ch)); | |
885 | return make_number (c); | |
886 | } | |
887 | ||
4ed46869 KH |
888 | DEFUN ("char-bytes", Fchar_bytes, Schar_bytes, 1, 1, 0, |
889 | "Return byte length of multi-byte form of CHAR.") | |
890 | (ch) | |
891 | Lisp_Object ch; | |
892 | { | |
893 | Lisp_Object val; | |
894 | int bytes; | |
895 | ||
896 | CHECK_NUMBER (ch, 0); | |
897 | if (COMPOSITE_CHAR_P (XFASTINT (ch))) | |
898 | { | |
899 | unsigned int id = COMPOSITE_CHAR_ID (XFASTINT (ch)); | |
900 | ||
901 | bytes = (id < n_cmpchars ? cmpchar_table[id]->len : 1); | |
902 | } | |
903 | else | |
904 | { | |
905 | int charset = CHAR_CHARSET (XFASTINT (ch)); | |
906 | ||
907 | bytes = CHARSET_DEFINED_P (charset) ? CHARSET_BYTES (charset) : 1; | |
908 | } | |
909 | ||
910 | XSETFASTINT (val, bytes); | |
911 | return val; | |
912 | } | |
913 | ||
914 | /* Return the width of character of which multi-byte form starts with | |
915 | C. The width is measured by how many columns occupied on the | |
916 | screen when displayed in the current buffer. */ | |
917 | ||
918 | #define ONE_BYTE_CHAR_WIDTH(c) \ | |
919 | (c < 0x20 \ | |
920 | ? (c == '\t' \ | |
53316e55 | 921 | ? XFASTINT (current_buffer->tab_width) \ |
4ed46869 KH |
922 | : (c == '\n' ? 0 : (NILP (current_buffer->ctl_arrow) ? 4 : 2))) \ |
923 | : (c < 0x7f \ | |
924 | ? 1 \ | |
925 | : (c == 0x7F \ | |
926 | ? (NILP (current_buffer->ctl_arrow) ? 4 : 2) \ | |
927 | : ((! NILP (current_buffer->enable_multibyte_characters) \ | |
928 | && BASE_LEADING_CODE_P (c)) \ | |
929 | ? WIDTH_BY_CHAR_HEAD (c) \ | |
930 | : 4)))) \ | |
931 | ||
932 | ||
933 | DEFUN ("char-width", Fchar_width, Schar_width, 1, 1, 0, | |
934 | "Return width of CHAR when displayed in the current buffer.\n\ | |
935 | The width is measured by how many columns it occupies on the screen.") | |
936 | (ch) | |
937 | Lisp_Object ch; | |
938 | { | |
859f2b3c | 939 | Lisp_Object val, disp; |
4ed46869 | 940 | int c; |
51c4025f | 941 | struct Lisp_Char_Table *dp = buffer_display_table (); |
4ed46869 KH |
942 | |
943 | CHECK_NUMBER (ch, 0); | |
944 | ||
859f2b3c RS |
945 | c = XINT (ch); |
946 | ||
947 | /* Get the way the display table would display it. */ | |
51c4025f | 948 | disp = dp ? DISP_CHAR_VECTOR (dp, c) : Qnil; |
859f2b3c RS |
949 | |
950 | if (VECTORP (disp)) | |
951 | XSETINT (val, XVECTOR (disp)->size); | |
952 | else if (SINGLE_BYTE_CHAR_P (c)) | |
953 | XSETINT (val, ONE_BYTE_CHAR_WIDTH (c)); | |
4ed46869 KH |
954 | else if (COMPOSITE_CHAR_P (c)) |
955 | { | |
956 | int id = COMPOSITE_CHAR_ID (XFASTINT (ch)); | |
957 | XSETFASTINT (val, (id < n_cmpchars ? cmpchar_table[id]->width : 0)); | |
958 | } | |
959 | else | |
960 | { | |
961 | int charset = CHAR_CHARSET (c); | |
962 | ||
963 | XSETFASTINT (val, CHARSET_WIDTH (charset)); | |
964 | } | |
965 | return val; | |
966 | } | |
967 | ||
968 | /* Return width of string STR of length LEN when displayed in the | |
969 | current buffer. The width is measured by how many columns it | |
970 | occupies on the screen. */ | |
859f2b3c | 971 | |
4ed46869 KH |
972 | int |
973 | strwidth (str, len) | |
974 | unsigned char *str; | |
975 | int len; | |
976 | { | |
977 | unsigned char *endp = str + len; | |
978 | int width = 0; | |
c4a4e28f | 979 | struct Lisp_Char_Table *dp = buffer_display_table (); |
4ed46869 | 980 | |
859f2b3c RS |
981 | while (str < endp) |
982 | { | |
983 | if (*str == LEADING_CODE_COMPOSITION) | |
984 | { | |
985 | int id = str_cmpchar_id (str, endp - str); | |
986 | ||
987 | if (id < 0) | |
988 | { | |
989 | width += 4; | |
990 | str++; | |
991 | } | |
992 | else | |
993 | { | |
994 | width += cmpchar_table[id]->width; | |
995 | str += cmpchar_table[id]->len; | |
996 | } | |
997 | } | |
998 | else | |
999 | { | |
1000 | Lisp_Object disp; | |
e515b0a9 KH |
1001 | int thislen; |
1002 | int c = STRING_CHAR_AND_LENGTH (str, endp - str, thislen); | |
859f2b3c RS |
1003 | |
1004 | /* Get the way the display table would display it. */ | |
acc35c36 RS |
1005 | if (dp) |
1006 | disp = DISP_CHAR_VECTOR (dp, c); | |
1007 | else | |
1008 | disp = Qnil; | |
859f2b3c RS |
1009 | |
1010 | if (VECTORP (disp)) | |
e515b0a9 | 1011 | width += XVECTOR (disp)->size; |
859f2b3c | 1012 | else |
e515b0a9 | 1013 | width += ONE_BYTE_CHAR_WIDTH (*str); |
859f2b3c | 1014 | |
e515b0a9 | 1015 | str += thislen; |
859f2b3c RS |
1016 | } |
1017 | } | |
4ed46869 KH |
1018 | return width; |
1019 | } | |
1020 | ||
1021 | DEFUN ("string-width", Fstring_width, Sstring_width, 1, 1, 0, | |
1022 | "Return width of STRING when displayed in the current buffer.\n\ | |
1023 | Width is measured by how many columns it occupies on the screen.\n\ | |
046b1f03 RS |
1024 | When calculating width of a multibyte character in STRING,\n\ |
1025 | only the base leading-code is considered; the validity of\n\ | |
1026 | the following bytes is not checked.") | |
4ed46869 KH |
1027 | (str) |
1028 | Lisp_Object str; | |
1029 | { | |
1030 | Lisp_Object val; | |
1031 | ||
1032 | CHECK_STRING (str, 0); | |
fc932ac6 RS |
1033 | XSETFASTINT (val, strwidth (XSTRING (str)->data, |
1034 | STRING_BYTES (XSTRING (str)))); | |
4ed46869 KH |
1035 | return val; |
1036 | } | |
1037 | ||
1038 | DEFUN ("char-direction", Fchar_direction, Schar_direction, 1, 1, 0, | |
1039 | "Return the direction of CHAR.\n\ | |
1040 | The returned value is 0 for left-to-right and 1 for right-to-left.") | |
1041 | (ch) | |
1042 | Lisp_Object ch; | |
1043 | { | |
1044 | int charset; | |
1045 | ||
1046 | CHECK_NUMBER (ch, 0); | |
1047 | charset = CHAR_CHARSET (XFASTINT (ch)); | |
1048 | if (!CHARSET_DEFINED_P (charset)) | |
93bcb785 | 1049 | invalid_character (XINT (ch)); |
4ed46869 KH |
1050 | return CHARSET_TABLE_INFO (charset, CHARSET_DIRECTION_IDX); |
1051 | } | |
1052 | ||
af4fecb4 | 1053 | DEFUN ("chars-in-region", Fchars_in_region, Schars_in_region, 2, 2, 0, |
6ae1f27e | 1054 | "Return number of characters between BEG and END.") |
046b1f03 RS |
1055 | (beg, end) |
1056 | Lisp_Object beg, end; | |
1057 | { | |
6ae1f27e | 1058 | int from, to; |
046b1f03 RS |
1059 | |
1060 | from = min (XFASTINT (beg), XFASTINT (end)); | |
a8a35e61 | 1061 | to = max (XFASTINT (beg), XFASTINT (end)); |
046b1f03 | 1062 | |
6ae1f27e RS |
1063 | return to - from; |
1064 | } | |
9036eb45 | 1065 | |
87b089ad RS |
1066 | /* Return the number of characters in the NBYTES bytes at PTR. |
1067 | This works by looking at the contents and checking for multibyte sequences. | |
1068 | However, if the current buffer has enable-multibyte-characters = nil, | |
1069 | we treat each byte as a character. */ | |
1070 | ||
6ae1f27e RS |
1071 | int |
1072 | chars_in_text (ptr, nbytes) | |
1073 | unsigned char *ptr; | |
1074 | int nbytes; | |
1075 | { | |
93bcb785 | 1076 | unsigned char *endp, c; |
6ae1f27e | 1077 | int chars; |
046b1f03 | 1078 | |
87b089ad RS |
1079 | /* current_buffer is null at early stages of Emacs initialization. */ |
1080 | if (current_buffer == 0 | |
1081 | || NILP (current_buffer->enable_multibyte_characters)) | |
6ae1f27e | 1082 | return nbytes; |
a8a35e61 | 1083 | |
6ae1f27e RS |
1084 | endp = ptr + nbytes; |
1085 | chars = 0; | |
046b1f03 | 1086 | |
6ae1f27e RS |
1087 | while (ptr < endp) |
1088 | { | |
93bcb785 KH |
1089 | c = *ptr++; |
1090 | ||
1091 | if (BASE_LEADING_CODE_P (c)) | |
1092 | while (ptr < endp && ! CHAR_HEAD_P (*ptr)) ptr++; | |
046b1f03 RS |
1093 | chars++; |
1094 | } | |
1095 | ||
6ae1f27e | 1096 | return chars; |
046b1f03 RS |
1097 | } |
1098 | ||
87b089ad RS |
1099 | /* Return the number of characters in the NBYTES bytes at PTR. |
1100 | This works by looking at the contents and checking for multibyte sequences. | |
1101 | It ignores enable-multibyte-characters. */ | |
1102 | ||
1103 | int | |
1104 | multibyte_chars_in_text (ptr, nbytes) | |
1105 | unsigned char *ptr; | |
1106 | int nbytes; | |
1107 | { | |
93bcb785 | 1108 | unsigned char *endp, c; |
87b089ad RS |
1109 | int chars; |
1110 | ||
1111 | endp = ptr + nbytes; | |
1112 | chars = 0; | |
1113 | ||
1114 | while (ptr < endp) | |
1115 | { | |
93bcb785 KH |
1116 | c = *ptr++; |
1117 | ||
1118 | if (BASE_LEADING_CODE_P (c)) | |
1119 | while (ptr < endp && ! CHAR_HEAD_P (*ptr)) ptr++; | |
87b089ad RS |
1120 | chars++; |
1121 | } | |
1122 | ||
1123 | return chars; | |
1124 | } | |
1125 | ||
1126 | DEFUN ("string", Fstring, Sstring, 1, MANY, 0, | |
4ed46869 | 1127 | "Concatenate all the argument characters and make the result a string.") |
53316e55 KH |
1128 | (n, args) |
1129 | int n; | |
4ed46869 KH |
1130 | Lisp_Object *args; |
1131 | { | |
53316e55 | 1132 | int i; |
4ed46869 | 1133 | unsigned char *buf |
bd4c6dd0 | 1134 | = (unsigned char *) alloca (MAX_LENGTH_OF_MULTI_BYTE_FORM * n); |
4ed46869 KH |
1135 | unsigned char *p = buf; |
1136 | Lisp_Object val; | |
1137 | ||
1138 | for (i = 0; i < n; i++) | |
1139 | { | |
1140 | int c, len; | |
1141 | unsigned char *str; | |
1142 | ||
1143 | if (!INTEGERP (args[i])) | |
b0e3cf2b | 1144 | CHECK_NUMBER (args[i], 0); |
4ed46869 KH |
1145 | c = XINT (args[i]); |
1146 | len = CHAR_STRING (c, p, str); | |
1147 | if (p != str) | |
1148 | /* C is a composite character. */ | |
1149 | bcopy (str, p, len); | |
1150 | p += len; | |
1151 | } | |
1152 | ||
27802600 | 1153 | val = make_string_from_bytes (buf, n, p - buf); |
4ed46869 KH |
1154 | return val; |
1155 | } | |
1156 | ||
1157 | #endif /* emacs */ | |
1158 | \f | |
1159 | /*** Composite characters staffs ***/ | |
1160 | ||
1161 | /* Each composite character is identified by CMPCHAR-ID which is | |
1162 | assigned when Emacs needs the character code of the composite | |
1163 | character (e.g. when displaying it on the screen). See the | |
1164 | document "GENERAL NOTE on COMPOSITE CHARACTER" in `charset.h' how a | |
1165 | composite character is represented in Emacs. */ | |
1166 | ||
1167 | /* If `static' is defined, it means that it is defined to null string. */ | |
1168 | #ifndef static | |
1169 | /* The following function is copied from lread.c. */ | |
1170 | static int | |
1171 | hash_string (ptr, len) | |
1172 | unsigned char *ptr; | |
1173 | int len; | |
1174 | { | |
1175 | register unsigned char *p = ptr; | |
1176 | register unsigned char *end = p + len; | |
1177 | register unsigned char c; | |
1178 | register int hash = 0; | |
1179 | ||
1180 | while (p != end) | |
1181 | { | |
1182 | c = *p++; | |
1183 | if (c >= 0140) c -= 40; | |
1184 | hash = ((hash<<3) + (hash>>28) + c); | |
1185 | } | |
1186 | return hash & 07777777777; | |
1187 | } | |
1188 | #endif | |
1189 | ||
4ed46869 KH |
1190 | #define CMPCHAR_HASH_TABLE_SIZE 0xFFF |
1191 | ||
1192 | static int *cmpchar_hash_table[CMPCHAR_HASH_TABLE_SIZE]; | |
1193 | ||
1194 | /* Each element of `cmpchar_hash_table' is a pointer to an array of | |
1195 | integer, where the 1st element is the size of the array, the 2nd | |
1196 | element is how many elements are actually used in the array, and | |
1197 | the remaining elements are CMPCHAR-IDs of composite characters of | |
1198 | the same hash value. */ | |
1199 | #define CMPCHAR_HASH_SIZE(table) table[0] | |
1200 | #define CMPCHAR_HASH_USED(table) table[1] | |
1201 | #define CMPCHAR_HASH_CMPCHAR_ID(table, i) table[i] | |
1202 | ||
1203 | /* Return CMPCHAR-ID of the composite character in STR of the length | |
1204 | LEN. If the composite character has not yet been registered, | |
1205 | register it in `cmpchar_table' and assign new CMPCHAR-ID. This | |
1206 | is the sole function for assigning CMPCHAR-ID. */ | |
1207 | int | |
1208 | str_cmpchar_id (str, len) | |
8867de67 | 1209 | const unsigned char *str; |
4ed46869 KH |
1210 | int len; |
1211 | { | |
1212 | int hash_idx, *hashp; | |
1213 | unsigned char *buf; | |
1214 | int embedded_rule; /* 1 if composition rule is embedded. */ | |
1215 | int chars; /* number of components. */ | |
1216 | int i; | |
1217 | struct cmpchar_info *cmpcharp; | |
1218 | ||
4ed46869 KH |
1219 | /* The second byte 0xFF means compostion rule is embedded. */ |
1220 | embedded_rule = (str[1] == 0xFF); | |
1221 | ||
1222 | /* At first, get the actual length of the composite character. */ | |
1223 | { | |
8867de67 | 1224 | const unsigned char *p, *endp = str + 1, *lastp = str + len; |
4ed46869 KH |
1225 | int bytes; |
1226 | ||
6ae1f27e | 1227 | while (endp < lastp && ! CHAR_HEAD_P (*endp)) endp++; |
93bcb785 KH |
1228 | if (endp - str < 5) |
1229 | /* Any composite char have at least 5-byte length. */ | |
1230 | return -1; | |
1231 | ||
4ed46869 | 1232 | chars = 0; |
93bcb785 | 1233 | p = str + 1; |
4ed46869 KH |
1234 | while (p < endp) |
1235 | { | |
93bcb785 | 1236 | if (embedded_rule) p++; |
4ed46869 | 1237 | /* No need of checking if *P is 0xA0 because |
93bcb785 KH |
1238 | BYTES_BY_CHAR_HEAD (0x80) surely returns 2. */ |
1239 | p += BYTES_BY_CHAR_HEAD (*p - 0x20); | |
4ed46869 KH |
1240 | chars++; |
1241 | } | |
93bcb785 KH |
1242 | if (p > endp || chars < 2 || chars > MAX_COMPONENT_COUNT) |
1243 | /* Invalid components. */ | |
4ed46869 | 1244 | return -1; |
93bcb785 | 1245 | len = p - str; |
4ed46869 KH |
1246 | } |
1247 | hash_idx = hash_string (str, len) % CMPCHAR_HASH_TABLE_SIZE; | |
1248 | hashp = cmpchar_hash_table[hash_idx]; | |
1249 | ||
1250 | /* Then, look into the hash table. */ | |
1251 | if (hashp != NULL) | |
1252 | /* Find the correct one among composite characters of the same | |
1253 | hash value. */ | |
1254 | for (i = 2; i < CMPCHAR_HASH_USED (hashp); i++) | |
1255 | { | |
1256 | cmpcharp = cmpchar_table[CMPCHAR_HASH_CMPCHAR_ID (hashp, i)]; | |
1257 | if (len == cmpcharp->len | |
1258 | && ! bcmp (str, cmpcharp->data, len)) | |
1259 | return CMPCHAR_HASH_CMPCHAR_ID (hashp, i); | |
1260 | } | |
1261 | ||
1262 | /* We have to register the composite character in cmpchar_table. */ | |
513ee442 KH |
1263 | if (n_cmpchars > (CHAR_FIELD2_MASK | CHAR_FIELD3_MASK)) |
1264 | /* No, we have no more room for a new composite character. */ | |
1265 | return -1; | |
1266 | ||
4ed46869 KH |
1267 | /* Make the entry in hash table. */ |
1268 | if (hashp == NULL) | |
1269 | { | |
1270 | /* Make a table for 8 composite characters initially. */ | |
1271 | hashp = (cmpchar_hash_table[hash_idx] | |
1272 | = (int *) xmalloc (sizeof (int) * (2 + 8))); | |
1273 | CMPCHAR_HASH_SIZE (hashp) = 10; | |
1274 | CMPCHAR_HASH_USED (hashp) = 2; | |
1275 | } | |
1276 | else if (CMPCHAR_HASH_USED (hashp) >= CMPCHAR_HASH_SIZE (hashp)) | |
1277 | { | |
1278 | CMPCHAR_HASH_SIZE (hashp) += 8; | |
1279 | hashp = (cmpchar_hash_table[hash_idx] | |
1280 | = (int *) xrealloc (hashp, | |
1281 | sizeof (int) * CMPCHAR_HASH_SIZE (hashp))); | |
1282 | } | |
1283 | CMPCHAR_HASH_CMPCHAR_ID (hashp, CMPCHAR_HASH_USED (hashp)) = n_cmpchars; | |
1284 | CMPCHAR_HASH_USED (hashp)++; | |
1285 | ||
1286 | /* Set information of the composite character in cmpchar_table. */ | |
1287 | if (cmpchar_table_size == 0) | |
1288 | { | |
1289 | /* This is the first composite character to be registered. */ | |
1290 | cmpchar_table_size = 256; | |
1291 | cmpchar_table | |
1292 | = (struct cmpchar_info **) xmalloc (sizeof (cmpchar_table[0]) | |
1293 | * cmpchar_table_size); | |
1294 | } | |
1295 | else if (cmpchar_table_size <= n_cmpchars) | |
1296 | { | |
1297 | cmpchar_table_size += 256; | |
1298 | cmpchar_table | |
1299 | = (struct cmpchar_info **) xrealloc (cmpchar_table, | |
1300 | sizeof (cmpchar_table[0]) | |
1301 | * cmpchar_table_size); | |
1302 | } | |
1303 | ||
1304 | cmpcharp = (struct cmpchar_info *) xmalloc (sizeof (struct cmpchar_info)); | |
1305 | ||
1306 | cmpcharp->len = len; | |
1307 | cmpcharp->data = (unsigned char *) xmalloc (len + 1); | |
1308 | bcopy (str, cmpcharp->data, len); | |
1309 | cmpcharp->data[len] = 0; | |
1310 | cmpcharp->glyph_len = chars; | |
1311 | cmpcharp->glyph = (GLYPH *) xmalloc (sizeof (GLYPH) * chars); | |
1312 | if (embedded_rule) | |
1313 | { | |
1314 | cmpcharp->cmp_rule = (unsigned char *) xmalloc (chars); | |
1315 | cmpcharp->col_offset = (float *) xmalloc (sizeof (float) * chars); | |
1316 | } | |
1317 | else | |
1318 | { | |
1319 | cmpcharp->cmp_rule = NULL; | |
1320 | cmpcharp->col_offset = NULL; | |
1321 | } | |
1322 | ||
1323 | /* Setup GLYPH data and composition rules (if any) so as not to make | |
1324 | them every time on displaying. */ | |
1325 | { | |
1326 | unsigned char *bufp; | |
1327 | int width; | |
1328 | float leftmost = 0.0, rightmost = 1.0; | |
1329 | ||
1330 | if (embedded_rule) | |
1331 | /* At first, col_offset[N] is set to relative to col_offset[0]. */ | |
1332 | cmpcharp->col_offset[0] = 0; | |
1333 | ||
1334 | for (i = 0, bufp = cmpcharp->data + 1; i < chars; i++) | |
1335 | { | |
1336 | if (embedded_rule) | |
1337 | cmpcharp->cmp_rule[i] = *bufp++; | |
1338 | ||
1339 | if (*bufp == 0xA0) /* This is an ASCII character. */ | |
1340 | { | |
1341 | cmpcharp->glyph[i] = FAST_MAKE_GLYPH ((*++bufp & 0x7F), 0); | |
1342 | width = 1; | |
1343 | bufp++; | |
1344 | } | |
1345 | else /* Multibyte character. */ | |
1346 | { | |
1347 | /* Make `bufp' point normal multi-byte form temporally. */ | |
1348 | *bufp -= 0x20; | |
1349 | cmpcharp->glyph[i] | |
1350 | = FAST_MAKE_GLYPH (string_to_non_ascii_char (bufp, 4, 0), 0); | |
1351 | width = WIDTH_BY_CHAR_HEAD (*bufp); | |
1352 | *bufp += 0x20; | |
1353 | bufp += BYTES_BY_CHAR_HEAD (*bufp - 0x20); | |
1354 | } | |
1355 | ||
1356 | if (embedded_rule && i > 0) | |
1357 | { | |
1358 | /* Reference points (global_ref and new_ref) are | |
1359 | encoded as below: | |
1360 | ||
1361 | 0--1--2 -- ascent | |
1362 | | | | |
1363 | | | | |
1364 | | 4 -+--- center | |
1365 | -- 3 5 -- baseline | |
1366 | | | | |
1367 | 6--7--8 -- descent | |
1368 | ||
1369 | Now, we calculate the column offset of the new glyph | |
1370 | from the left edge of the first glyph. This can avoid | |
1371 | the same calculation everytime displaying this | |
1372 | composite character. */ | |
1373 | ||
1374 | /* Reference points of global glyph and new glyph. */ | |
1375 | int global_ref = (cmpcharp->cmp_rule[i] - 0xA0) / 9; | |
1376 | int new_ref = (cmpcharp->cmp_rule[i] - 0xA0) % 9; | |
1377 | /* Column offset relative to the first glyph. */ | |
1378 | float left = (leftmost | |
1379 | + (global_ref % 3) * (rightmost - leftmost) / 2.0 | |
1380 | - (new_ref % 3) * width / 2.0); | |
1381 | ||
1382 | cmpcharp->col_offset[i] = left; | |
1383 | if (left < leftmost) | |
1384 | leftmost = left; | |
1385 | if (left + width > rightmost) | |
1386 | rightmost = left + width; | |
1387 | } | |
1388 | else | |
1389 | { | |
1390 | if (width > rightmost) | |
1391 | rightmost = width; | |
1392 | } | |
1393 | } | |
1394 | if (embedded_rule) | |
1395 | { | |
1396 | /* Now col_offset[N] are relative to the left edge of the | |
1397 | first component. Make them relative to the left edge of | |
1398 | overall glyph. */ | |
1399 | for (i = 0; i < chars; i++) | |
1400 | cmpcharp->col_offset[i] -= leftmost; | |
1401 | /* Make rightmost holds width of overall glyph. */ | |
1402 | rightmost -= leftmost; | |
1403 | } | |
1404 | ||
1405 | cmpcharp->width = rightmost; | |
1406 | if (cmpcharp->width < rightmost) | |
1407 | /* To get a ceiling integer value. */ | |
1408 | cmpcharp->width++; | |
1409 | } | |
1410 | ||
1411 | cmpchar_table[n_cmpchars] = cmpcharp; | |
1412 | ||
1413 | return n_cmpchars++; | |
1414 | } | |
1415 | ||
1416 | /* Return the Nth element of the composite character C. */ | |
1417 | int | |
1418 | cmpchar_component (c, n) | |
1419 | unsigned int c, n; | |
1420 | { | |
1421 | int id = COMPOSITE_CHAR_ID (c); | |
1422 | ||
1423 | if (id >= n_cmpchars /* C is not a valid composite character. */ | |
1424 | || n >= cmpchar_table[id]->glyph_len) /* No such component. */ | |
1425 | return -1; | |
1426 | /* No face data is stored in glyph code. */ | |
1427 | return ((int) (cmpchar_table[id]->glyph[n])); | |
1428 | } | |
1429 | ||
1430 | DEFUN ("cmpcharp", Fcmpcharp, Scmpcharp, 1, 1, 0, | |
1431 | "T if CHAR is a composite character.") | |
1432 | (ch) | |
1433 | Lisp_Object ch; | |
1434 | { | |
1435 | CHECK_NUMBER (ch, 0); | |
1436 | return (COMPOSITE_CHAR_P (XINT (ch)) ? Qt : Qnil); | |
1437 | } | |
1438 | ||
1439 | DEFUN ("composite-char-component", Fcmpchar_component, Scmpchar_component, | |
1440 | 2, 2, 0, | |
1441 | "Return the IDXth component character of composite character CHARACTER.") | |
1442 | (character, idx) | |
1443 | Lisp_Object character, idx; | |
1444 | { | |
1445 | int c; | |
1446 | ||
1447 | CHECK_NUMBER (character, 0); | |
1448 | CHECK_NUMBER (idx, 1); | |
1449 | ||
1450 | if ((c = cmpchar_component (XINT (character), XINT (idx))) < 0) | |
1451 | args_out_of_range (character, idx); | |
1452 | ||
1453 | return make_number (c); | |
1454 | } | |
1455 | ||
1456 | DEFUN ("composite-char-composition-rule", Fcmpchar_cmp_rule, Scmpchar_cmp_rule, | |
1457 | 2, 2, 0, | |
55001746 KH |
1458 | "Return the Nth composition rule embedded in composite character CHARACTER.\n\ |
1459 | The returned rule is for composing the Nth component\n\ | |
1460 | on the (N-1)th component. If N is 0, the returned value is always 255.") | |
1461 | (character, n) | |
1462 | Lisp_Object character, n; | |
4ed46869 KH |
1463 | { |
1464 | int id, i; | |
1465 | ||
1466 | CHECK_NUMBER (character, 0); | |
55001746 | 1467 | CHECK_NUMBER (n, 1); |
4ed46869 KH |
1468 | |
1469 | id = COMPOSITE_CHAR_ID (XINT (character)); | |
1470 | if (id < 0 || id >= n_cmpchars) | |
1471 | error ("Invalid composite character: %d", XINT (character)); | |
55001746 | 1472 | i = XINT (n); |
4ed46869 | 1473 | if (i > cmpchar_table[id]->glyph_len) |
55001746 | 1474 | args_out_of_range (character, n); |
4ed46869 KH |
1475 | |
1476 | return make_number (cmpchar_table[id]->cmp_rule[i]); | |
1477 | } | |
1478 | ||
1479 | DEFUN ("composite-char-composition-rule-p", Fcmpchar_cmp_rule_p, | |
1480 | Scmpchar_cmp_rule_p, 1, 1, 0, | |
1481 | "Return non-nil if composite character CHARACTER contains a embedded rule.") | |
1482 | (character) | |
1483 | Lisp_Object character; | |
1484 | { | |
1485 | int id; | |
1486 | ||
1487 | CHECK_NUMBER (character, 0); | |
1488 | id = COMPOSITE_CHAR_ID (XINT (character)); | |
1489 | if (id < 0 || id >= n_cmpchars) | |
1490 | error ("Invalid composite character: %d", XINT (character)); | |
1491 | ||
1492 | return (cmpchar_table[id]->cmp_rule ? Qt : Qnil); | |
1493 | } | |
1494 | ||
1495 | DEFUN ("composite-char-component-count", Fcmpchar_cmp_count, | |
1496 | Scmpchar_cmp_count, 1, 1, 0, | |
1497 | "Return number of compoents of composite character CHARACTER.") | |
1498 | (character) | |
1499 | Lisp_Object character; | |
1500 | { | |
1501 | int id; | |
1502 | ||
1503 | CHECK_NUMBER (character, 0); | |
1504 | id = COMPOSITE_CHAR_ID (XINT (character)); | |
1505 | if (id < 0 || id >= n_cmpchars) | |
1506 | error ("Invalid composite character: %d", XINT (character)); | |
1507 | ||
1508 | return (make_number (cmpchar_table[id]->glyph_len)); | |
1509 | } | |
1510 | ||
1511 | DEFUN ("compose-string", Fcompose_string, Scompose_string, | |
1512 | 1, 1, 0, | |
1513 | "Return one char string composed from all characters in STRING.") | |
1514 | (str) | |
1515 | Lisp_Object str; | |
1516 | { | |
1517 | unsigned char buf[MAX_LENGTH_OF_MULTI_BYTE_FORM], *p, *pend, *ptemp; | |
1518 | int len, i; | |
1519 | ||
1520 | CHECK_STRING (str, 0); | |
1521 | ||
1522 | buf[0] = LEADING_CODE_COMPOSITION; | |
1523 | p = XSTRING (str)->data; | |
fc932ac6 | 1524 | pend = p + STRING_BYTES (XSTRING (str)); |
4ed46869 KH |
1525 | i = 1; |
1526 | while (p < pend) | |
1527 | { | |
1528 | if (*p < 0x20 || *p == 127) /* control code */ | |
1529 | error ("Invalid component character: %d", *p); | |
1530 | else if (*p < 0x80) /* ASCII */ | |
1531 | { | |
1532 | if (i + 2 >= MAX_LENGTH_OF_MULTI_BYTE_FORM) | |
1533 | error ("Too long string to be composed: %s", XSTRING (str)->data); | |
1534 | /* Prepend an ASCII charset indicator 0xA0, set MSB of the | |
1535 | code itself. */ | |
1536 | buf[i++] = 0xA0; | |
1537 | buf[i++] = *p++ + 0x80; | |
1538 | } | |
1539 | else if (*p == LEADING_CODE_COMPOSITION) /* composite char */ | |
1540 | { | |
1541 | /* Already composed. Eliminate the heading | |
1542 | LEADING_CODE_COMPOSITION, keep the remaining bytes | |
1543 | unchanged. */ | |
1544 | p++; | |
1545 | ptemp = p; | |
6ae1f27e | 1546 | while (! CHAR_HEAD_P (*p)) p++; |
4ed46869 KH |
1547 | if (i + (p - ptemp) >= MAX_LENGTH_OF_MULTI_BYTE_FORM) |
1548 | error ("Too long string to be composed: %s", XSTRING (str)->data); | |
1549 | bcopy (ptemp, buf + i, p - ptemp); | |
1550 | i += p - ptemp; | |
1551 | } | |
1552 | else /* multibyte char */ | |
1553 | { | |
1554 | /* Add 0x20 to the base leading-code, keep the remaining | |
1555 | bytes unchanged. */ | |
1556 | len = BYTES_BY_CHAR_HEAD (*p); | |
1557 | if (i + len >= MAX_LENGTH_OF_MULTI_BYTE_FORM) | |
1558 | error ("Too long string to be composed: %s", XSTRING (str)->data); | |
1559 | bcopy (p, buf + i, len); | |
1560 | buf[i] += 0x20; | |
1561 | p += len, i += len; | |
1562 | } | |
1563 | } | |
1564 | ||
1565 | if (i < 5) | |
1566 | /* STR contains only one character, which can't be composed. */ | |
1567 | error ("Too short string to be composed: %s", XSTRING (str)->data); | |
1568 | ||
27802600 | 1569 | return make_string_from_bytes (buf, 1, i); |
4ed46869 KH |
1570 | } |
1571 | ||
1572 | \f | |
1573 | charset_id_internal (charset_name) | |
1574 | char *charset_name; | |
1575 | { | |
1576 | Lisp_Object val = Fget (intern (charset_name), Qcharset); | |
1577 | ||
1578 | if (!VECTORP (val)) | |
1579 | error ("Charset %s is not defined", charset_name); | |
1580 | ||
1581 | return (XINT (XVECTOR (val)->contents[0])); | |
1582 | } | |
1583 | ||
1584 | DEFUN ("setup-special-charsets", Fsetup_special_charsets, | |
1585 | Ssetup_special_charsets, 0, 0, 0, "Internal use only.") | |
1586 | () | |
1587 | { | |
1588 | charset_latin_iso8859_1 = charset_id_internal ("latin-iso8859-1"); | |
1589 | charset_jisx0208_1978 = charset_id_internal ("japanese-jisx0208-1978"); | |
1590 | charset_jisx0208 = charset_id_internal ("japanese-jisx0208"); | |
1591 | charset_katakana_jisx0201 = charset_id_internal ("katakana-jisx0201"); | |
1592 | charset_latin_jisx0201 = charset_id_internal ("latin-jisx0201"); | |
1593 | charset_big5_1 = charset_id_internal ("chinese-big5-1"); | |
1594 | charset_big5_2 = charset_id_internal ("chinese-big5-2"); | |
1595 | return Qnil; | |
1596 | } | |
1597 | ||
1598 | init_charset_once () | |
1599 | { | |
1600 | int i, j, k; | |
1601 | ||
1602 | staticpro (&Vcharset_table); | |
1603 | staticpro (&Vcharset_symbol_table); | |
1604 | ||
1605 | /* This has to be done here, before we call Fmake_char_table. */ | |
1606 | Qcharset_table = intern ("charset-table"); | |
1607 | staticpro (&Qcharset_table); | |
1608 | ||
1609 | /* Intern this now in case it isn't already done. | |
1610 | Setting this variable twice is harmless. | |
1611 | But don't staticpro it here--that is done in alloc.c. */ | |
1612 | Qchar_table_extra_slots = intern ("char-table-extra-slots"); | |
1613 | ||
1614 | /* Now we are ready to set up this property, so we can | |
1615 | create the charset table. */ | |
1616 | Fput (Qcharset_table, Qchar_table_extra_slots, make_number (0)); | |
1617 | Vcharset_table = Fmake_char_table (Qcharset_table, Qnil); | |
1618 | ||
513ee442 | 1619 | Vcharset_symbol_table = Fmake_vector (make_number (MAX_CHARSET + 1), Qnil); |
4ed46869 KH |
1620 | |
1621 | /* Setup tables. */ | |
1622 | for (i = 0; i < 2; i++) | |
1623 | for (j = 0; j < 2; j++) | |
1624 | for (k = 0; k < 128; k++) | |
1625 | iso_charset_table [i][j][k] = -1; | |
1626 | ||
1627 | bzero (cmpchar_hash_table, sizeof cmpchar_hash_table); | |
1628 | cmpchar_table_size = n_cmpchars = 0; | |
1629 | ||
1630 | for (i = 0; i < 256; i++) | |
1631 | BYTES_BY_CHAR_HEAD (i) = 1; | |
1632 | BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_11) = 3; | |
1633 | BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_12) = 3; | |
1634 | BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_21) = 4; | |
1635 | BYTES_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_22) = 4; | |
1636 | /* The following doesn't reflect the actual bytes, but just to tell | |
1637 | that it is a start of a multibyte character. */ | |
1638 | BYTES_BY_CHAR_HEAD (LEADING_CODE_COMPOSITION) = 2; | |
1639 | ||
1640 | for (i = 0; i < 128; i++) | |
1641 | WIDTH_BY_CHAR_HEAD (i) = 1; | |
1642 | for (; i < 256; i++) | |
1643 | WIDTH_BY_CHAR_HEAD (i) = 4; | |
1644 | WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_11) = 1; | |
1645 | WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_12) = 2; | |
1646 | WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_21) = 1; | |
1647 | WIDTH_BY_CHAR_HEAD (LEADING_CODE_PRIVATE_22) = 2; | |
1648 | } | |
1649 | ||
1650 | #ifdef emacs | |
1651 | ||
1652 | syms_of_charset () | |
1653 | { | |
1654 | Qascii = intern ("ascii"); | |
1655 | staticpro (&Qascii); | |
1656 | ||
1657 | Qcharset = intern ("charset"); | |
1658 | staticpro (&Qcharset); | |
1659 | ||
1660 | /* Define ASCII charset now. */ | |
1661 | update_charset_table (make_number (CHARSET_ASCII), | |
1662 | make_number (1), make_number (94), | |
1663 | make_number (1), | |
1664 | make_number (0), | |
1665 | make_number ('B'), | |
1666 | make_number (0), | |
1667 | build_string ("ASCII"), | |
1668 | build_string ("ASCII"), | |
1669 | build_string ("ASCII (ISO646 IRV)")); | |
1670 | CHARSET_SYMBOL (CHARSET_ASCII) = Qascii; | |
1671 | Fput (Qascii, Qcharset, CHARSET_TABLE_ENTRY (CHARSET_ASCII)); | |
1672 | ||
1673 | Qcomposition = intern ("composition"); | |
1674 | staticpro (&Qcomposition); | |
1675 | CHARSET_SYMBOL (CHARSET_COMPOSITION) = Qcomposition; | |
1676 | ||
1677 | defsubr (&Sdefine_charset); | |
3fac5a51 | 1678 | defsubr (&Sget_unused_iso_final_char); |
4ed46869 KH |
1679 | defsubr (&Sdeclare_equiv_charset); |
1680 | defsubr (&Sfind_charset_region); | |
1681 | defsubr (&Sfind_charset_string); | |
1682 | defsubr (&Smake_char_internal); | |
1683 | defsubr (&Ssplit_char); | |
1684 | defsubr (&Schar_charset); | |
1685 | defsubr (&Siso_charset); | |
9d3d8cba | 1686 | defsubr (&Schar_valid_p); |
d2665018 | 1687 | defsubr (&Sunibyte_char_to_multibyte); |
4ed46869 KH |
1688 | defsubr (&Schar_bytes); |
1689 | defsubr (&Schar_width); | |
1690 | defsubr (&Sstring_width); | |
1691 | defsubr (&Schar_direction); | |
af4fecb4 | 1692 | defsubr (&Schars_in_region); |
87b089ad | 1693 | defsubr (&Sstring); |
4ed46869 KH |
1694 | defsubr (&Scmpcharp); |
1695 | defsubr (&Scmpchar_component); | |
1696 | defsubr (&Scmpchar_cmp_rule); | |
1697 | defsubr (&Scmpchar_cmp_rule_p); | |
1698 | defsubr (&Scmpchar_cmp_count); | |
1699 | defsubr (&Scompose_string); | |
1700 | defsubr (&Ssetup_special_charsets); | |
1701 | ||
1702 | DEFVAR_LISP ("charset-list", &Vcharset_list, | |
1703 | "List of charsets ever defined."); | |
1704 | Vcharset_list = Fcons (Qascii, Qnil); | |
1705 | ||
b0e3cf2b KH |
1706 | DEFVAR_LISP ("character-unification-table-vector", |
1707 | &Vcharacter_unification_table_vector, | |
1708 | "Vector of cons cell of a symbol and unification table ever defined.\n\ | |
1709 | An ID of a unification table is an index of this vector."); | |
1710 | Vcharacter_unification_table_vector = Fmake_vector (make_number (16), Qnil); | |
1711 | ||
4ed46869 KH |
1712 | DEFVAR_INT ("leading-code-composition", &leading_code_composition, |
1713 | "Leading-code of composite characters."); | |
1714 | leading_code_composition = LEADING_CODE_COMPOSITION; | |
1715 | ||
1716 | DEFVAR_INT ("leading-code-private-11", &leading_code_private_11, | |
1717 | "Leading-code of private TYPE9N charset of column-width 1."); | |
1718 | leading_code_private_11 = LEADING_CODE_PRIVATE_11; | |
1719 | ||
1720 | DEFVAR_INT ("leading-code-private-12", &leading_code_private_12, | |
1721 | "Leading-code of private TYPE9N charset of column-width 2."); | |
1722 | leading_code_private_12 = LEADING_CODE_PRIVATE_12; | |
1723 | ||
1724 | DEFVAR_INT ("leading-code-private-21", &leading_code_private_21, | |
1725 | "Leading-code of private TYPE9Nx9N charset of column-width 1."); | |
1726 | leading_code_private_21 = LEADING_CODE_PRIVATE_21; | |
1727 | ||
1728 | DEFVAR_INT ("leading-code-private-22", &leading_code_private_22, | |
1729 | "Leading-code of private TYPE9Nx9N charset of column-width 2."); | |
1730 | leading_code_private_22 = LEADING_CODE_PRIVATE_22; | |
35e623fb RS |
1731 | |
1732 | DEFVAR_INT ("nonascii-insert-offset", &nonascii_insert_offset, | |
d2665018 | 1733 | "Offset for converting non-ASCII unibyte codes 0240...0377 to multibyte.\n\ |
4cf9710d RS |
1734 | This is used for converting unibyte text to multibyte,\n\ |
1735 | and for inserting character codes specified by number.\n\n\ | |
1736 | Conversion is performed only when multibyte characters are enabled,\n\ | |
1737 | and it serves to convert a Latin-1 or similar 8-bit character code\n\ | |
1738 | to the corresponding Emacs character code.\n\ | |
1739 | If `nonascii-translate-table' is non-nil, it overrides this variable."); | |
35e623fb | 1740 | nonascii_insert_offset = 0; |
b0e3cf2b | 1741 | |
4cf9710d RS |
1742 | DEFVAR_LISP ("nonascii-translate-table", &Vnonascii_translate_table, |
1743 | "Translate table for converting non-ASCII unibyte codes to multibyte.\n\ | |
1744 | This is used for converting unibyte text to multibyte,\n\ | |
1745 | and for inserting character codes specified by number.\n\n\ | |
1746 | Conversion is performed only when multibyte characters are enabled,\n\ | |
1747 | and it serves to convert a Latin-1 or similar 8-bit character code\n\ | |
1748 | to the corresponding Emacs character code.\n\n\ | |
1749 | If this is nil, `nonascii-insert-offset' is used instead."); | |
1750 | Vnonascii_translate_table = Qnil; | |
1751 | ||
b0e3cf2b KH |
1752 | DEFVAR_INT ("min-composite-char", &min_composite_char, |
1753 | "Minimum character code of a composite character."); | |
1754 | min_composite_char = MIN_CHAR_COMPOSITION; | |
4ed46869 KH |
1755 | } |
1756 | ||
1757 | #endif /* emacs */ |