Commit | Line | Data |
---|---|---|
0168c3d8 KH |
1 | /* Basic character support. |
2 | Copyright (C) 1995, 1997, 1998, 2001 Electrotechnical Laboratory, JAPAN. | |
8f924df7 | 3 | Licensed to the Free Software Foundation. |
ec62e0ac GM |
4 | Copyright (C) 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008 |
5 | Free Software Foundation, Inc. | |
6 | Copyright (C) 2003, 2004, 2005, 2006, 2007, 2008 | |
0168c3d8 KH |
7 | National Institute of Advanced Industrial Science and Technology (AIST) |
8 | Registration Number H13PRO009 | |
9 | ||
10 | This file is part of GNU Emacs. | |
11 | ||
9ec0b715 | 12 | GNU Emacs is free software: you can redistribute it and/or modify |
0168c3d8 | 13 | it under the terms of the GNU General Public License as published by |
9ec0b715 GM |
14 | the Free Software Foundation, either version 3 of the License, or |
15 | (at your option) any later version. | |
0168c3d8 KH |
16 | |
17 | GNU Emacs is distributed in the hope that it will be useful, | |
18 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
19 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
20 | GNU General Public License for more details. | |
21 | ||
22 | You should have received a copy of the GNU General Public License | |
9ec0b715 | 23 | along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>. */ |
0168c3d8 KH |
24 | |
25 | /* At first, see the document in `character.h' to understand the code | |
26 | in this file. */ | |
27 | ||
28 | #ifdef emacs | |
29 | #include <config.h> | |
30 | #endif | |
31 | ||
32 | #include <stdio.h> | |
33 | ||
34 | #ifdef emacs | |
35 | ||
36 | #include <sys/types.h> | |
37 | #include "lisp.h" | |
38 | #include "character.h" | |
39 | #include "buffer.h" | |
40 | #include "charset.h" | |
41 | #include "composite.h" | |
42 | #include "disptab.h" | |
43 | ||
44 | #else /* not emacs */ | |
45 | ||
46 | #include "mulelib.h" | |
47 | ||
48 | #endif /* emacs */ | |
49 | ||
50 | Lisp_Object Qcharacterp; | |
51 | ||
52 | /* Vector of translation table ever defined. | |
53 | ID of a translation table is used to index this vector. */ | |
54 | Lisp_Object Vtranslation_table_vector; | |
55 | ||
56 | /* A char-table for characters which may invoke auto-filling. */ | |
57 | Lisp_Object Vauto_fill_chars; | |
58 | ||
59 | Lisp_Object Qauto_fill_chars; | |
60 | ||
33f91981 KH |
61 | /* Char-table of information about which character to unify to which |
62 | Unicode character. */ | |
0168c3d8 KH |
63 | Lisp_Object Vchar_unify_table; |
64 | ||
65 | /* A char-table. An element is non-nil iff the corresponding | |
66 | character has a printable glyph. */ | |
67 | Lisp_Object Vprintable_chars; | |
68 | ||
69 | /* A char-table. An elemnent is a column-width of the corresponding | |
70 | character. */ | |
71 | Lisp_Object Vchar_width_table; | |
72 | ||
73 | /* A char-table. An element is a symbol indicating the direction | |
74 | property of corresponding character. */ | |
75 | Lisp_Object Vchar_direction_table; | |
76 | ||
8973478b | 77 | /* Variable used locally in the macro FETCH_MULTIBYTE_CHAR. */ |
0168c3d8 | 78 | unsigned char *_fetch_multibyte_char_p; |
0168c3d8 | 79 | |
c57f3328 KH |
80 | /* Char table of scripts. */ |
81 | Lisp_Object Vchar_script_table; | |
82 | ||
c7e14352 KH |
83 | /* Alist of scripts vs representative characters. */ |
84 | Lisp_Object Vscript_representative_chars; | |
85 | ||
c57f3328 KH |
86 | static Lisp_Object Qchar_script_table; |
87 | ||
b672c5ae KH |
88 | /* Mapping table from unibyte chars to multibyte chars. */ |
89 | int unibyte_to_multibyte_table[256]; | |
15843e6f | 90 | |
90c9d035 KH |
91 | /* Nth element is 1 iff unibyte char N can be mapped to a multibyte |
92 | char. */ | |
93 | char unibyte_has_multibyte_table[256]; | |
94 | ||
0168c3d8 KH |
95 | \f |
96 | ||
2bde7652 KH |
97 | /* If character code C has modifier masks, reflect them to the |
98 | character code if possible. Return the resulting code. */ | |
99 | ||
100 | int | |
101 | char_resolve_modifier_mask (c) | |
102 | int c; | |
103 | { | |
d0363d44 | 104 | /* A non-ASCII character can't reflect modifier bits to the code. */ |
2bde7652 KH |
105 | if (! ASCII_CHAR_P ((c & ~CHAR_MODIFIER_MASK))) |
106 | return c; | |
107 | ||
108 | /* For Meta, Shift, and Control modifiers, we need special care. */ | |
2bde7652 KH |
109 | if (c & CHAR_SHIFT) |
110 | { | |
111 | /* Shift modifier is valid only with [A-Za-z]. */ | |
112 | if ((c & 0377) >= 'A' && (c & 0377) <= 'Z') | |
113 | c &= ~CHAR_SHIFT; | |
114 | else if ((c & 0377) >= 'a' && (c & 0377) <= 'z') | |
115 | c = (c & ~CHAR_SHIFT) - ('a' - 'A'); | |
03365d0e KH |
116 | /* Shift modifier for control characters and SPC is ignored. */ |
117 | else if ((c & ~CHAR_MODIFIER_MASK) <= 0x20) | |
d0363d44 KH |
118 | c &= ~CHAR_SHIFT; |
119 | } | |
2bde7652 KH |
120 | if (c & CHAR_CTL) |
121 | { | |
122 | /* Simulate the code in lread.c. */ | |
123 | /* Allow `\C- ' and `\C-?'. */ | |
03365d0e KH |
124 | if ((c & 0377) == ' ') |
125 | c &= ~0177 & ~ CHAR_CTL; | |
126 | else if ((c & 0377) == '?') | |
127 | c = 0177 | (c & ~0177 & ~CHAR_CTL); | |
2bde7652 KH |
128 | /* ASCII control chars are made from letters (both cases), |
129 | as well as the non-letters within 0100...0137. */ | |
130 | else if ((c & 0137) >= 0101 && (c & 0137) <= 0132) | |
131 | c &= (037 | (~0177 & ~CHAR_CTL)); | |
132 | else if ((c & 0177) >= 0100 && (c & 0177) <= 0137) | |
133 | c &= (037 | (~0177 & ~CHAR_CTL)); | |
134 | } | |
03365d0e KH |
135 | if (c & CHAR_META) |
136 | { | |
137 | /* Move the meta bit to the right place for a string. */ | |
138 | c = (c & ~CHAR_META) | 0x80; | |
139 | } | |
2bde7652 KH |
140 | |
141 | return c; | |
142 | } | |
143 | ||
144 | ||
33f91981 KH |
145 | /* Store multibyte form of character C at P. If C has modifier bits, |
146 | handle them appropriately. */ | |
147 | ||
0168c3d8 | 148 | int |
e3d8eb8c | 149 | char_string (c, p) |
5aa91c9b | 150 | unsigned c; |
1889b238 | 151 | unsigned char *p; |
0168c3d8 KH |
152 | { |
153 | int bytes; | |
154 | ||
e3d8eb8c KH |
155 | if (c & CHAR_MODIFIER_MASK) |
156 | { | |
2bde7652 | 157 | c = (unsigned) char_resolve_modifier_mask ((int) c); |
e3d8eb8c KH |
158 | /* If C still has any modifier bits, just ignore it. */ |
159 | c &= ~CHAR_MODIFIER_MASK; | |
160 | } | |
161 | ||
0168c3d8 KH |
162 | MAYBE_UNIFY_CHAR (c); |
163 | ||
e3d8eb8c | 164 | if (c <= MAX_3_BYTE_CHAR) |
0168c3d8 KH |
165 | { |
166 | bytes = CHAR_STRING (c, p); | |
167 | } | |
168 | else if (c <= MAX_4_BYTE_CHAR) | |
169 | { | |
170 | p[0] = (0xF0 | (c >> 18)); | |
171 | p[1] = (0x80 | ((c >> 12) & 0x3F)); | |
172 | p[2] = (0x80 | ((c >> 6) & 0x3F)); | |
173 | p[3] = (0x80 | (c & 0x3F)); | |
174 | bytes = 4; | |
175 | } | |
e3d8eb8c | 176 | else if (c <= MAX_5_BYTE_CHAR) |
0168c3d8 KH |
177 | { |
178 | p[0] = 0xF8; | |
179 | p[1] = (0x80 | ((c >> 18) & 0x0F)); | |
180 | p[2] = (0x80 | ((c >> 12) & 0x3F)); | |
181 | p[3] = (0x80 | ((c >> 6) & 0x3F)); | |
182 | p[4] = (0x80 | (c & 0x3F)); | |
183 | bytes = 5; | |
184 | } | |
5aa91c9b | 185 | else if (c <= MAX_CHAR) |
e3d8eb8c KH |
186 | { |
187 | c = CHAR_TO_BYTE8 (c); | |
188 | bytes = BYTE8_STRING (c, p); | |
189 | } | |
5aa91c9b KH |
190 | else |
191 | error ("Invalid character: %d", c); | |
1889b238 | 192 | |
0168c3d8 KH |
193 | return bytes; |
194 | } | |
195 | ||
196 | ||
33f91981 KH |
197 | /* Return a character whose multibyte form is at P. Set LEN is not |
198 | NULL, it must be a pointer to integer. In that case, set *LEN to | |
199 | the byte length of the multibyte form. If ADVANCED is not NULL, is | |
200 | must be a pointer to unsigned char. In that case, set *ADVANCED to | |
201 | the ending address (i.e. the starting address of the next | |
202 | character) of the multibyte form. */ | |
203 | ||
0168c3d8 | 204 | int |
e3d8eb8c | 205 | string_char (p, advanced, len) |
15843e6f KH |
206 | const unsigned char *p; |
207 | const unsigned char **advanced; | |
0168c3d8 KH |
208 | int *len; |
209 | { | |
1889b238 | 210 | int c; |
15843e6f | 211 | const unsigned char *saved_p = p; |
0168c3d8 KH |
212 | |
213 | if (*p < 0x80 || ! (*p & 0x20) || ! (*p & 0x10)) | |
214 | { | |
215 | c = STRING_CHAR_ADVANCE (p); | |
216 | } | |
217 | else if (! (*p & 0x08)) | |
218 | { | |
219 | c = ((((p)[0] & 0xF) << 18) | |
220 | | (((p)[1] & 0x3F) << 12) | |
221 | | (((p)[2] & 0x3F) << 6) | |
222 | | ((p)[3] & 0x3F)); | |
223 | p += 4; | |
224 | } | |
225 | else | |
226 | { | |
227 | c = ((((p)[1] & 0x3F) << 18) | |
228 | | (((p)[2] & 0x3F) << 12) | |
229 | | (((p)[3] & 0x3F) << 6) | |
230 | | ((p)[4] & 0x3F)); | |
231 | p += 5; | |
232 | } | |
233 | ||
234 | MAYBE_UNIFY_CHAR (c); | |
235 | ||
236 | if (len) | |
237 | *len = p - saved_p; | |
238 | if (advanced) | |
239 | *advanced = p; | |
240 | return c; | |
241 | } | |
242 | ||
243 | ||
244 | /* Translate character C by translation table TABLE. If C is | |
245 | negative, translate a character specified by CHARSET and CODE. If | |
246 | no translation is found in TABLE, return the untranslated | |
10453be9 KH |
247 | character. If TABLE is a list, elements are char tables. In this |
248 | case, translace C by all tables. */ | |
0168c3d8 KH |
249 | |
250 | int | |
251 | translate_char (table, c) | |
252 | Lisp_Object table; | |
253 | int c; | |
254 | { | |
10453be9 KH |
255 | if (CHAR_TABLE_P (table)) |
256 | { | |
257 | Lisp_Object ch; | |
258 | ||
259 | ch = CHAR_TABLE_REF (table, c); | |
260 | if (CHARACTERP (ch)) | |
261 | c = XINT (ch); | |
262 | } | |
263 | else | |
264 | { | |
265 | for (; CONSP (table); table = XCDR (table)) | |
266 | c = translate_char (XCAR (table), c); | |
267 | } | |
268 | return c; | |
0168c3d8 KH |
269 | } |
270 | ||
0168c3d8 | 271 | /* Convert the multibyte character C to unibyte 8-bit character based |
ac86488b KH |
272 | on the current value of charset_unibyte. If dimension of |
273 | charset_unibyte is more than one, return (C & 0xFF). | |
0168c3d8 KH |
274 | |
275 | The argument REV_TBL is now ignored. It will be removed in the | |
276 | future. */ | |
277 | ||
278 | int | |
279 | multibyte_char_to_unibyte (c, rev_tbl) | |
280 | int c; | |
281 | Lisp_Object rev_tbl; | |
282 | { | |
b672c5ae KH |
283 | struct charset *charset; |
284 | unsigned c1; | |
0168c3d8 | 285 | |
b672c5ae KH |
286 | if (CHAR_BYTE8_P (c)) |
287 | return CHAR_TO_BYTE8 (c); | |
288 | charset = CHARSET_FROM_ID (charset_unibyte); | |
289 | c1 = ENCODE_CHAR (charset, c); | |
0168c3d8 KH |
290 | return ((c1 != CHARSET_INVALID_CODE (charset)) ? c1 : c & 0xFF); |
291 | } | |
292 | ||
935d5b02 KH |
293 | /* Like multibyte_char_to_unibyte, but return -1 if C is not supported |
294 | by charset_unibyte. */ | |
295 | ||
296 | int | |
297 | multibyte_char_to_unibyte_safe (c) | |
298 | int c; | |
299 | { | |
300 | struct charset *charset; | |
301 | unsigned c1; | |
302 | ||
303 | if (CHAR_BYTE8_P (c)) | |
304 | return CHAR_TO_BYTE8 (c); | |
305 | charset = CHARSET_FROM_ID (charset_unibyte); | |
306 | c1 = ENCODE_CHAR (charset, c); | |
307 | return ((c1 != CHARSET_INVALID_CODE (charset)) ? c1 : -1); | |
308 | } | |
0168c3d8 KH |
309 | |
310 | DEFUN ("characterp", Fcharacterp, Scharacterp, 1, 2, 0, | |
311 | doc: /* Return non-nil if OBJECT is a character. */) | |
312 | (object, ignore) | |
313 | Lisp_Object object, ignore; | |
314 | { | |
315 | return (CHARACTERP (object) ? Qt : Qnil); | |
316 | } | |
317 | ||
318 | DEFUN ("max-char", Fmax_char, Smax_char, 0, 0, 0, | |
319 | doc: /* Return the character of the maximum code. */) | |
320 | () | |
321 | { | |
322 | return make_number (MAX_CHAR); | |
323 | } | |
324 | ||
325 | DEFUN ("unibyte-char-to-multibyte", Funibyte_char_to_multibyte, | |
326 | Sunibyte_char_to_multibyte, 1, 1, 0, | |
5556875b | 327 | doc: /* Convert the byte CH to multibyte character. */) |
0168c3d8 KH |
328 | (ch) |
329 | Lisp_Object ch; | |
330 | { | |
331 | int c; | |
332 | struct charset *charset; | |
333 | ||
334 | CHECK_CHARACTER (ch); | |
335 | c = XFASTINT (ch); | |
336 | if (c >= 0400) | |
337 | error ("Invalid unibyte character: %d", c); | |
ac86488b | 338 | charset = CHARSET_FROM_ID (charset_unibyte); |
0168c3d8 KH |
339 | c = DECODE_CHAR (charset, c); |
340 | if (c < 0) | |
3c5a53bd | 341 | c = BYTE8_TO_CHAR (XFASTINT (ch)); |
0168c3d8 KH |
342 | return make_number (c); |
343 | } | |
344 | ||
345 | DEFUN ("multibyte-char-to-unibyte", Fmultibyte_char_to_unibyte, | |
346 | Smultibyte_char_to_unibyte, 1, 1, 0, | |
5556875b SM |
347 | doc: /* Convert the multibyte character CH to a byte. |
348 | If the multibyte character does not represent a byte, return -1. */) | |
0168c3d8 KH |
349 | (ch) |
350 | Lisp_Object ch; | |
351 | { | |
5556875b | 352 | int cm; |
0168c3d8 KH |
353 | |
354 | CHECK_CHARACTER (ch); | |
5556875b SM |
355 | cm = XFASTINT (ch); |
356 | if (cm < 256) | |
357 | /* Can't distinguish a byte read from a unibyte buffer from | |
358 | a latin1 char, so let's let it slide. */ | |
359 | return ch; | |
360 | else | |
361 | { | |
2afc21f5 | 362 | int cu = CHAR_TO_BYTE_SAFE (cm); |
5556875b SM |
363 | return make_number (cu); |
364 | } | |
0168c3d8 KH |
365 | } |
366 | ||
367 | DEFUN ("char-bytes", Fchar_bytes, Schar_bytes, 1, 1, 0, | |
368 | doc: /* Return 1 regardless of the argument CHAR. | |
add553ac JB |
369 | This is now an obsolete function. We keep it just for backward compatibility. |
370 | usage: (char-bytes CHAR) */) | |
0168c3d8 KH |
371 | (ch) |
372 | Lisp_Object ch; | |
373 | { | |
374 | CHECK_CHARACTER (ch); | |
375 | return make_number (1); | |
376 | } | |
377 | ||
378 | DEFUN ("char-width", Fchar_width, Schar_width, 1, 1, 0, | |
379 | doc: /* Return width of CHAR when displayed in the current buffer. | |
380 | The width is measured by how many columns it occupies on the screen. | |
add553ac JB |
381 | Tab is taken to occupy `tab-width' columns. |
382 | usage: (char-width CHAR) */) | |
0168c3d8 KH |
383 | (ch) |
384 | Lisp_Object ch; | |
385 | { | |
386 | Lisp_Object disp; | |
387 | int c, width; | |
388 | struct Lisp_Char_Table *dp = buffer_display_table (); | |
389 | ||
390 | CHECK_CHARACTER (ch); | |
391 | c = XINT (ch); | |
392 | ||
393 | /* Get the way the display table would display it. */ | |
394 | disp = dp ? DISP_CHAR_VECTOR (dp, c) : Qnil; | |
395 | ||
396 | if (VECTORP (disp)) | |
397 | width = ASIZE (disp); | |
398 | else | |
399 | width = CHAR_WIDTH (c); | |
400 | ||
401 | return make_number (width); | |
402 | } | |
403 | ||
0168c3d8 KH |
404 | /* Return width of string STR of length LEN when displayed in the |
405 | current buffer. The width is measured by how many columns it | |
406 | occupies on the screen. If PRECISION > 0, return the width of | |
407 | longest substring that doesn't exceed PRECISION, and set number of | |
408 | characters and bytes of the substring in *NCHARS and *NBYTES | |
409 | respectively. */ | |
410 | ||
1889b238 | 411 | int |
0168c3d8 | 412 | c_string_width (str, len, precision, nchars, nbytes) |
8f924df7 | 413 | const unsigned char *str; |
0168c3d8 KH |
414 | int precision, *nchars, *nbytes; |
415 | { | |
416 | int i = 0, i_byte = 0; | |
417 | int width = 0; | |
418 | struct Lisp_Char_Table *dp = buffer_display_table (); | |
419 | ||
420 | while (i_byte < len) | |
421 | { | |
422 | int bytes, thiswidth; | |
423 | Lisp_Object val; | |
424 | int c = STRING_CHAR_AND_LENGTH (str + i_byte, len - i_byte, bytes); | |
425 | ||
426 | if (dp) | |
427 | { | |
428 | val = DISP_CHAR_VECTOR (dp, c); | |
429 | if (VECTORP (val)) | |
430 | thiswidth = XVECTOR (val)->size; | |
431 | else | |
432 | thiswidth = CHAR_WIDTH (c); | |
433 | } | |
434 | else | |
435 | { | |
436 | thiswidth = CHAR_WIDTH (c); | |
437 | } | |
438 | ||
439 | if (precision > 0 | |
440 | && (width + thiswidth > precision)) | |
441 | { | |
442 | *nchars = i; | |
443 | *nbytes = i_byte; | |
444 | return width; | |
445 | } | |
446 | i++; | |
447 | i_byte += bytes; | |
448 | width += thiswidth; | |
449 | } | |
450 | ||
451 | if (precision > 0) | |
452 | { | |
453 | *nchars = i; | |
454 | *nbytes = i_byte; | |
455 | } | |
456 | ||
457 | return width; | |
458 | } | |
459 | ||
1889b238 KH |
460 | /* Return width of string STR of length LEN when displayed in the |
461 | current buffer. The width is measured by how many columns it | |
462 | occupies on the screen. */ | |
463 | ||
464 | int | |
465 | strwidth (str, len) | |
466 | unsigned char *str; | |
467 | int len; | |
468 | { | |
469 | return c_string_width (str, len, -1, NULL, NULL); | |
470 | } | |
471 | ||
0168c3d8 KH |
472 | /* Return width of Lisp string STRING when displayed in the current |
473 | buffer. The width is measured by how many columns it occupies on | |
474 | the screen while paying attention to compositions. If PRECISION > | |
475 | 0, return the width of longest substring that doesn't exceed | |
476 | PRECISION, and set number of characters and bytes of the substring | |
477 | in *NCHARS and *NBYTES respectively. */ | |
478 | ||
479 | int | |
480 | lisp_string_width (string, precision, nchars, nbytes) | |
481 | Lisp_Object string; | |
482 | int precision, *nchars, *nbytes; | |
483 | { | |
8f924df7 | 484 | int len = SCHARS (string); |
0aee65b9 KH |
485 | /* This set multibyte to 0 even if STRING is multibyte when it |
486 | contains only ascii and eight-bit-graphic, but that's | |
487 | intentional. */ | |
488 | int multibyte = len < SBYTES (string); | |
8f924df7 | 489 | unsigned char *str = SDATA (string); |
0168c3d8 KH |
490 | int i = 0, i_byte = 0; |
491 | int width = 0; | |
492 | struct Lisp_Char_Table *dp = buffer_display_table (); | |
493 | ||
494 | while (i < len) | |
495 | { | |
496 | int chars, bytes, thiswidth; | |
497 | Lisp_Object val; | |
498 | int cmp_id; | |
f4bc0685 | 499 | EMACS_INT ignore, end; |
0168c3d8 KH |
500 | |
501 | if (find_composition (i, -1, &ignore, &end, &val, string) | |
502 | && ((cmp_id = get_composition_id (i, i_byte, end - i, val, string)) | |
503 | >= 0)) | |
504 | { | |
505 | thiswidth = composition_table[cmp_id]->width; | |
506 | chars = end - i; | |
507 | bytes = string_char_to_byte (string, end) - i_byte; | |
508 | } | |
0168c3d8 KH |
509 | else |
510 | { | |
0aee65b9 | 511 | int c; |
0168c3d8 | 512 | |
0aee65b9 KH |
513 | if (multibyte) |
514 | c = STRING_CHAR_AND_LENGTH (str + i_byte, len - i_byte, bytes); | |
515 | else | |
516 | c = str[i_byte], bytes = 1; | |
0168c3d8 | 517 | chars = 1; |
0aee65b9 KH |
518 | if (dp) |
519 | { | |
520 | val = DISP_CHAR_VECTOR (dp, c); | |
521 | if (VECTORP (val)) | |
522 | thiswidth = XVECTOR (val)->size; | |
523 | else | |
524 | thiswidth = CHAR_WIDTH (c); | |
525 | } | |
526 | else | |
527 | { | |
528 | thiswidth = CHAR_WIDTH (c); | |
529 | } | |
0168c3d8 KH |
530 | } |
531 | ||
532 | if (precision > 0 | |
533 | && (width + thiswidth > precision)) | |
534 | { | |
535 | *nchars = i; | |
536 | *nbytes = i_byte; | |
537 | return width; | |
538 | } | |
539 | i += chars; | |
540 | i_byte += bytes; | |
541 | width += thiswidth; | |
542 | } | |
543 | ||
544 | if (precision > 0) | |
545 | { | |
546 | *nchars = i; | |
547 | *nbytes = i_byte; | |
548 | } | |
549 | ||
550 | return width; | |
551 | } | |
552 | ||
553 | DEFUN ("string-width", Fstring_width, Sstring_width, 1, 1, 0, | |
554 | doc: /* Return width of STRING when displayed in the current buffer. | |
555 | Width is measured by how many columns it occupies on the screen. | |
556 | When calculating width of a multibyte character in STRING, | |
557 | only the base leading-code is considered; the validity of | |
558 | the following bytes is not checked. Tabs in STRING are always | |
add553ac JB |
559 | taken to occupy `tab-width' columns. |
560 | usage: (string-width STRING) */) | |
0168c3d8 KH |
561 | (str) |
562 | Lisp_Object str; | |
563 | { | |
564 | Lisp_Object val; | |
565 | ||
566 | CHECK_STRING (str); | |
567 | XSETFASTINT (val, lisp_string_width (str, -1, NULL, NULL)); | |
568 | return val; | |
569 | } | |
570 | ||
571 | DEFUN ("char-direction", Fchar_direction, Schar_direction, 1, 1, 0, | |
572 | doc: /* Return the direction of CHAR. | |
add553ac JB |
573 | The returned value is 0 for left-to-right and 1 for right-to-left. |
574 | usage: (char-direction CHAR) */) | |
0168c3d8 KH |
575 | (ch) |
576 | Lisp_Object ch; | |
577 | { | |
578 | int c; | |
579 | ||
580 | CHECK_CHARACTER (ch); | |
581 | c = XINT (ch); | |
582 | return CHAR_TABLE_REF (Vchar_direction_table, c); | |
583 | } | |
584 | ||
0168c3d8 KH |
585 | /* Return the number of characters in the NBYTES bytes at PTR. |
586 | This works by looking at the contents and checking for multibyte | |
587 | sequences while assuming that there's no invalid sequence. | |
588 | However, if the current buffer has enable-multibyte-characters = | |
589 | nil, we treat each byte as a character. */ | |
590 | ||
13818c30 | 591 | EMACS_INT |
0168c3d8 | 592 | chars_in_text (ptr, nbytes) |
8f924df7 | 593 | const unsigned char *ptr; |
13818c30 | 594 | EMACS_INT nbytes; |
0168c3d8 KH |
595 | { |
596 | /* current_buffer is null at early stages of Emacs initialization. */ | |
597 | if (current_buffer == 0 | |
598 | || NILP (current_buffer->enable_multibyte_characters)) | |
599 | return nbytes; | |
600 | ||
601 | return multibyte_chars_in_text (ptr, nbytes); | |
602 | } | |
603 | ||
604 | /* Return the number of characters in the NBYTES bytes at PTR. | |
605 | This works by looking at the contents and checking for multibyte | |
606 | sequences while assuming that there's no invalid sequence. It | |
607 | ignores enable-multibyte-characters. */ | |
608 | ||
13818c30 | 609 | EMACS_INT |
0168c3d8 | 610 | multibyte_chars_in_text (ptr, nbytes) |
8f924df7 | 611 | const unsigned char *ptr; |
13818c30 | 612 | EMACS_INT nbytes; |
0168c3d8 | 613 | { |
8f924df7 | 614 | const unsigned char *endp = ptr + nbytes; |
0168c3d8 KH |
615 | int chars = 0; |
616 | ||
617 | while (ptr < endp) | |
618 | { | |
619 | int len = MULTIBYTE_LENGTH (ptr, endp); | |
620 | ||
621 | if (len == 0) | |
622 | abort (); | |
623 | ptr += len; | |
624 | chars++; | |
625 | } | |
626 | ||
627 | return chars; | |
628 | } | |
629 | ||
630 | /* Parse unibyte text at STR of LEN bytes as a multibyte text, count | |
631 | characters and bytes in it, and store them in *NCHARS and *NBYTES | |
632 | respectively. On counting bytes, pay attention to that 8-bit | |
633 | characters not constructing a valid multibyte sequence are | |
634 | represented by 2-byte in a multibyte text. */ | |
635 | ||
636 | void | |
637 | parse_str_as_multibyte (str, len, nchars, nbytes) | |
8f924df7 | 638 | const unsigned char *str; |
0168c3d8 KH |
639 | int len, *nchars, *nbytes; |
640 | { | |
8f924df7 | 641 | const unsigned char *endp = str + len; |
0168c3d8 KH |
642 | int n, chars = 0, bytes = 0; |
643 | ||
644 | if (len >= MAX_MULTIBYTE_LENGTH) | |
645 | { | |
8f924df7 | 646 | const unsigned char *adjusted_endp = endp - MAX_MULTIBYTE_LENGTH; |
0168c3d8 KH |
647 | while (str < adjusted_endp) |
648 | { | |
649 | if ((n = MULTIBYTE_LENGTH_NO_CHECK (str)) > 0) | |
650 | str += n, bytes += n; | |
651 | else | |
652 | str++, bytes += 2; | |
653 | chars++; | |
654 | } | |
655 | } | |
656 | while (str < endp) | |
657 | { | |
658 | if ((n = MULTIBYTE_LENGTH (str, endp)) > 0) | |
659 | str += n, bytes += n; | |
660 | else | |
661 | str++, bytes += 2; | |
662 | chars++; | |
663 | } | |
664 | ||
665 | *nchars = chars; | |
666 | *nbytes = bytes; | |
667 | return; | |
668 | } | |
669 | ||
670 | /* Arrange unibyte text at STR of NBYTES bytes as a multibyte text. | |
671 | It actually converts only such 8-bit characters that don't contruct | |
672 | a multibyte sequence to multibyte forms of Latin-1 characters. If | |
673 | NCHARS is nonzero, set *NCHARS to the number of characters in the | |
674 | text. It is assured that we can use LEN bytes at STR as a work | |
675 | area and that is enough. Return the number of bytes of the | |
676 | resulting text. */ | |
677 | ||
678 | int | |
679 | str_as_multibyte (str, len, nbytes, nchars) | |
680 | unsigned char *str; | |
681 | int len, nbytes, *nchars; | |
682 | { | |
683 | unsigned char *p = str, *endp = str + nbytes; | |
684 | unsigned char *to; | |
685 | int chars = 0; | |
686 | int n; | |
687 | ||
688 | if (nbytes >= MAX_MULTIBYTE_LENGTH) | |
689 | { | |
690 | unsigned char *adjusted_endp = endp - MAX_MULTIBYTE_LENGTH; | |
691 | while (p < adjusted_endp | |
692 | && (n = MULTIBYTE_LENGTH_NO_CHECK (p)) > 0) | |
693 | p += n, chars++; | |
694 | } | |
695 | while ((n = MULTIBYTE_LENGTH (p, endp)) > 0) | |
696 | p += n, chars++; | |
697 | if (nchars) | |
698 | *nchars = chars; | |
699 | if (p == endp) | |
700 | return nbytes; | |
701 | ||
702 | to = p; | |
703 | nbytes = endp - p; | |
704 | endp = str + len; | |
705 | safe_bcopy ((char *) p, (char *) (endp - nbytes), nbytes); | |
706 | p = endp - nbytes; | |
707 | ||
708 | if (nbytes >= MAX_MULTIBYTE_LENGTH) | |
709 | { | |
710 | unsigned char *adjusted_endp = endp - MAX_MULTIBYTE_LENGTH; | |
711 | while (p < adjusted_endp) | |
712 | { | |
713 | if ((n = MULTIBYTE_LENGTH_NO_CHECK (p)) > 0) | |
714 | { | |
715 | while (n--) | |
716 | *to++ = *p++; | |
717 | } | |
718 | else | |
719 | { | |
720 | int c = *p++; | |
721 | c = BYTE8_TO_CHAR (c); | |
722 | to += CHAR_STRING (c, to); | |
723 | } | |
724 | } | |
725 | chars++; | |
726 | } | |
727 | while (p < endp) | |
728 | { | |
729 | if ((n = MULTIBYTE_LENGTH (p, endp)) > 0) | |
730 | { | |
731 | while (n--) | |
732 | *to++ = *p++; | |
8f924df7 | 733 | } |
0168c3d8 KH |
734 | else |
735 | { | |
736 | int c = *p++; | |
737 | c = BYTE8_TO_CHAR (c); | |
738 | to += CHAR_STRING (c, to); | |
739 | } | |
740 | chars++; | |
741 | } | |
742 | if (nchars) | |
743 | *nchars = chars; | |
744 | return (to - str); | |
745 | } | |
746 | ||
747 | /* Parse unibyte string at STR of LEN bytes, and return the number of | |
748 | bytes it may ocupy when converted to multibyte string by | |
749 | `str_to_multibyte'. */ | |
750 | ||
751 | int | |
752 | parse_str_to_multibyte (str, len) | |
753 | unsigned char *str; | |
754 | int len; | |
755 | { | |
756 | unsigned char *endp = str + len; | |
757 | int bytes; | |
758 | ||
759 | for (bytes = 0; str < endp; str++) | |
760 | bytes += (*str < 0x80) ? 1 : 2; | |
761 | return bytes; | |
762 | } | |
763 | ||
764 | ||
765 | /* Convert unibyte text at STR of NBYTES bytes to a multibyte text | |
766 | that contains the same single-byte characters. It actually | |
767 | converts all 8-bit characters to multibyte forms. It is assured | |
768 | that we can use LEN bytes at STR as a work area and that is | |
769 | enough. */ | |
770 | ||
771 | int | |
772 | str_to_multibyte (str, len, bytes) | |
773 | unsigned char *str; | |
774 | int len, bytes; | |
775 | { | |
776 | unsigned char *p = str, *endp = str + bytes; | |
777 | unsigned char *to; | |
778 | ||
779 | while (p < endp && *p < 0x80) p++; | |
780 | if (p == endp) | |
781 | return bytes; | |
782 | to = p; | |
783 | bytes = endp - p; | |
784 | endp = str + len; | |
785 | safe_bcopy ((char *) p, (char *) (endp - bytes), bytes); | |
786 | p = endp - bytes; | |
8f924df7 | 787 | while (p < endp) |
0168c3d8 KH |
788 | { |
789 | int c = *p++; | |
790 | ||
791 | if (c >= 0x80) | |
792 | c = BYTE8_TO_CHAR (c); | |
793 | to += CHAR_STRING (c, to); | |
794 | } | |
795 | return (to - str); | |
796 | } | |
797 | ||
798 | /* Arrange multibyte text at STR of LEN bytes as a unibyte text. It | |
799 | actually converts characters in the range 0x80..0xFF to | |
800 | unibyte. */ | |
801 | ||
802 | int | |
803 | str_as_unibyte (str, bytes) | |
804 | unsigned char *str; | |
805 | int bytes; | |
806 | { | |
15843e6f KH |
807 | const unsigned char *p = str, *endp = str + bytes; |
808 | unsigned char *to; | |
0168c3d8 KH |
809 | int c, len; |
810 | ||
811 | while (p < endp) | |
812 | { | |
813 | c = *p; | |
814 | len = BYTES_BY_CHAR_HEAD (c); | |
815 | if (CHAR_BYTE8_HEAD_P (c)) | |
816 | break; | |
817 | p += len; | |
818 | } | |
15843e6f | 819 | to = str + (p - str); |
8f924df7 | 820 | while (p < endp) |
0168c3d8 KH |
821 | { |
822 | c = *p; | |
823 | len = BYTES_BY_CHAR_HEAD (c); | |
824 | if (CHAR_BYTE8_HEAD_P (c)) | |
825 | { | |
826 | c = STRING_CHAR_ADVANCE (p); | |
827 | *to++ = CHAR_TO_BYTE8 (c); | |
828 | } | |
829 | else | |
830 | { | |
831 | while (len--) *to++ = *p++; | |
832 | } | |
833 | } | |
834 | return (to - str); | |
835 | } | |
836 | ||
4aa40bb8 KH |
837 | /* Convert eight-bit chars in SRC (in multibyte form) to the |
838 | corresponding byte and store in DST. CHARS is the number of | |
839 | characters in SRC. The value is the number of bytes stored in DST. | |
840 | Usually, the value is the same as CHARS, but is less than it if SRC | |
841 | contains a non-ASCII, non-eight-bit characater. If ACCEPT_LATIN_1 | |
842 | is nonzero, a Latin-1 character is accepted and converted to a byte | |
f27f70ec KH |
843 | of that character code. |
844 | Note: Currently the arg ACCEPT_LATIN_1 is not used. */ | |
4aa40bb8 KH |
845 | |
846 | EMACS_INT | |
847 | str_to_unibyte (src, dst, chars, accept_latin_1) | |
848 | const unsigned char *src; | |
849 | unsigned char *dst; | |
850 | EMACS_INT chars; | |
851 | int accept_latin_1; | |
852 | { | |
853 | EMACS_INT i; | |
854 | ||
855 | for (i = 0; i < chars; i++) | |
856 | { | |
857 | int c = STRING_CHAR_ADVANCE (src); | |
858 | ||
859 | if (CHAR_BYTE8_P (c)) | |
860 | c = CHAR_TO_BYTE8 (c); | |
861 | else if (! ASCII_CHAR_P (c) | |
862 | && (! accept_latin_1 || c >= 0x100)) | |
863 | return i; | |
864 | *dst++ = c; | |
865 | } | |
866 | return i; | |
867 | } | |
868 | ||
869 | ||
0168c3d8 KH |
870 | int |
871 | string_count_byte8 (string) | |
872 | Lisp_Object string; | |
873 | { | |
874 | int multibyte = STRING_MULTIBYTE (string); | |
8f924df7 KH |
875 | int nbytes = SBYTES (string); |
876 | unsigned char *p = SDATA (string); | |
0168c3d8 KH |
877 | unsigned char *pend = p + nbytes; |
878 | int count = 0; | |
879 | int c, len; | |
880 | ||
881 | if (multibyte) | |
882 | while (p < pend) | |
883 | { | |
884 | c = *p; | |
885 | len = BYTES_BY_CHAR_HEAD (c); | |
886 | ||
887 | if (CHAR_BYTE8_HEAD_P (c)) | |
888 | count++; | |
889 | p += len; | |
890 | } | |
891 | else | |
892 | while (p < pend) | |
893 | { | |
894 | if (*p++ >= 0x80) | |
895 | count++; | |
896 | } | |
897 | return count; | |
898 | } | |
899 | ||
900 | ||
901 | Lisp_Object | |
902 | string_escape_byte8 (string) | |
903 | Lisp_Object string; | |
904 | { | |
8f924df7 KH |
905 | int nchars = SCHARS (string); |
906 | int nbytes = SBYTES (string); | |
0168c3d8 KH |
907 | int multibyte = STRING_MULTIBYTE (string); |
908 | int byte8_count; | |
15843e6f KH |
909 | const unsigned char *src, *src_end; |
910 | unsigned char *dst; | |
0168c3d8 KH |
911 | Lisp_Object val; |
912 | int c, len; | |
913 | ||
914 | if (multibyte && nchars == nbytes) | |
915 | return string; | |
916 | ||
917 | byte8_count = string_count_byte8 (string); | |
918 | ||
919 | if (byte8_count == 0) | |
920 | return string; | |
921 | ||
922 | if (multibyte) | |
923 | /* Convert 2-byte sequence of byte8 chars to 4-byte octal. */ | |
7b40ebaf | 924 | val = make_uninit_multibyte_string (nchars + byte8_count * 3, |
0168c3d8 KH |
925 | nbytes + byte8_count * 2); |
926 | else | |
927 | /* Convert 1-byte sequence of byte8 chars to 4-byte octal. */ | |
928 | val = make_uninit_string (nbytes + byte8_count * 3); | |
929 | ||
8f924df7 | 930 | src = SDATA (string); |
0168c3d8 | 931 | src_end = src + nbytes; |
8f924df7 | 932 | dst = SDATA (val); |
0168c3d8 KH |
933 | if (multibyte) |
934 | while (src < src_end) | |
935 | { | |
936 | c = *src; | |
937 | len = BYTES_BY_CHAR_HEAD (c); | |
938 | ||
939 | if (CHAR_BYTE8_HEAD_P (c)) | |
940 | { | |
941 | c = STRING_CHAR_ADVANCE (src); | |
942 | c = CHAR_TO_BYTE8 (c); | |
1889b238 | 943 | sprintf ((char *) dst, "\\%03o", c); |
0168c3d8 KH |
944 | dst += 4; |
945 | } | |
946 | else | |
947 | while (len--) *dst++ = *src++; | |
948 | } | |
949 | else | |
950 | while (src < src_end) | |
951 | { | |
952 | c = *src++; | |
953 | if (c >= 0x80) | |
954 | { | |
1889b238 | 955 | sprintf ((char *) dst, "\\%03o", c); |
0168c3d8 KH |
956 | dst += 4; |
957 | } | |
958 | else | |
959 | *dst++ = c; | |
960 | } | |
961 | return val; | |
962 | } | |
963 | ||
964 | \f | |
8f924df7 | 965 | DEFUN ("string", Fstring, Sstring, 0, MANY, 0, |
0168c3d8 | 966 | doc: /* |
d2e83296 DL |
967 | Concatenate all the argument characters and make the result a string. |
968 | usage: (string &rest CHARACTERS) */) | |
0168c3d8 KH |
969 | (n, args) |
970 | int n; | |
971 | Lisp_Object *args; | |
972 | { | |
973 | int i; | |
974 | unsigned char *buf = (unsigned char *) alloca (MAX_MULTIBYTE_LENGTH * n); | |
975 | unsigned char *p = buf; | |
976 | int c; | |
977 | ||
978 | for (i = 0; i < n; i++) | |
979 | { | |
980 | CHECK_CHARACTER (args[i]); | |
981 | c = XINT (args[i]); | |
982 | p += CHAR_STRING (c, p); | |
983 | } | |
984 | ||
985 | return make_string_from_bytes ((char *) buf, n, p - buf); | |
986 | } | |
987 | ||
70b4969d | 988 | DEFUN ("unibyte-string", Funibyte_string, Sunibyte_string, 0, MANY, 0, |
87d6f965 KH |
989 | doc: /* Concatenate all the argument bytes and make the result a unibyte string. |
990 | usage: (unibyte-string &rest BYTES) */) | |
70b4969d KH |
991 | (n, args) |
992 | int n; | |
993 | Lisp_Object *args; | |
994 | { | |
995 | int i; | |
996 | unsigned char *buf = (unsigned char *) alloca (n); | |
997 | unsigned char *p = buf; | |
998 | unsigned c; | |
999 | ||
1000 | for (i = 0; i < n; i++) | |
1001 | { | |
1002 | CHECK_NATNUM (args[i]); | |
1003 | c = XFASTINT (args[i]); | |
1004 | if (c >= 256) | |
1005 | args_out_of_range_3 (args[i], make_number (0), make_number (255)); | |
1006 | *p++ = c; | |
1007 | } | |
1008 | ||
1009 | return make_string_from_bytes ((char *) buf, n, p - buf); | |
1010 | } | |
1011 | ||
d0363d44 KH |
1012 | DEFUN ("char-resolve-modifers", Fchar_resolve_modifiers, |
1013 | Schar_resolve_modifiers, 1, 1, 0, | |
1014 | doc: /* Resolve modifiers in the character CHAR. | |
1015 | The value is a character with modifiers resolved into the character | |
1016 | code. Unresolved modifiers are kept in the value. | |
1017 | usage: (char-resolve-modifers CHAR) */) | |
1018 | (character) | |
1019 | Lisp_Object character; | |
1020 | { | |
1021 | int c; | |
1022 | ||
1023 | CHECK_NUMBER (character); | |
1024 | c = XINT (character); | |
1025 | return make_number (char_resolve_modifier_mask (c)); | |
1026 | } | |
1027 | ||
0168c3d8 KH |
1028 | void |
1029 | init_character_once () | |
1030 | { | |
1031 | } | |
1032 | ||
1033 | #ifdef emacs | |
1034 | ||
1035 | void | |
1036 | syms_of_character () | |
1037 | { | |
1038 | DEFSYM (Qcharacterp, "characterp"); | |
1039 | DEFSYM (Qauto_fill_chars, "auto-fill-chars"); | |
1040 | ||
1041 | staticpro (&Vchar_unify_table); | |
1042 | Vchar_unify_table = Qnil; | |
1043 | ||
1044 | defsubr (&Smax_char); | |
1045 | defsubr (&Scharacterp); | |
1046 | defsubr (&Sunibyte_char_to_multibyte); | |
1047 | defsubr (&Smultibyte_char_to_unibyte); | |
1048 | defsubr (&Schar_bytes); | |
1049 | defsubr (&Schar_width); | |
1050 | defsubr (&Sstring_width); | |
1051 | defsubr (&Schar_direction); | |
0168c3d8 | 1052 | defsubr (&Sstring); |
70b4969d | 1053 | defsubr (&Sunibyte_string); |
d0363d44 | 1054 | defsubr (&Schar_resolve_modifiers); |
0168c3d8 KH |
1055 | |
1056 | DEFVAR_LISP ("translation-table-vector", &Vtranslation_table_vector, | |
1057 | doc: /* | |
68978cf0 DL |
1058 | Vector recording all translation tables ever defined. |
1059 | Each element is a pair (SYMBOL . TABLE) relating the table to the | |
1060 | symbol naming it. The ID of a translation table is an index into this vector. */); | |
0168c3d8 KH |
1061 | Vtranslation_table_vector = Fmake_vector (make_number (16), Qnil); |
1062 | ||
1063 | DEFVAR_LISP ("auto-fill-chars", &Vauto_fill_chars, | |
1064 | doc: /* | |
1065 | A char-table for characters which invoke auto-filling. | |
1066 | Such characters have value t in this table. */); | |
1067 | Vauto_fill_chars = Fmake_char_table (Qauto_fill_chars, Qnil); | |
6cc0e1ca DL |
1068 | CHAR_TABLE_SET (Vauto_fill_chars, ' ', Qt); |
1069 | CHAR_TABLE_SET (Vauto_fill_chars, '\n', Qt); | |
0168c3d8 KH |
1070 | |
1071 | DEFVAR_LISP ("char-width-table", &Vchar_width_table, | |
1072 | doc: /* | |
1073 | A char-table for width (columns) of each character. */); | |
1074 | Vchar_width_table = Fmake_char_table (Qnil, make_number (1)); | |
be8b50bc KH |
1075 | char_table_set_range (Vchar_width_table, 0x80, 0x9F, make_number (4)); |
1076 | char_table_set_range (Vchar_width_table, MAX_5_BYTE_CHAR + 1, MAX_CHAR, | |
1077 | make_number (4)); | |
0168c3d8 KH |
1078 | |
1079 | DEFVAR_LISP ("char-direction-table", &Vchar_direction_table, | |
1080 | doc: /* A char-table for direction of each character. */); | |
1081 | Vchar_direction_table = Fmake_char_table (Qnil, make_number (1)); | |
1082 | ||
1083 | DEFVAR_LISP ("printable-chars", &Vprintable_chars, | |
1084 | doc: /* A char-table for each printable character. */); | |
db6d4189 | 1085 | Vprintable_chars = Fmake_char_table (Qnil, Qnil); |
67dde660 KH |
1086 | Fset_char_table_range (Vprintable_chars, |
1087 | Fcons (make_number (32), make_number (126)), Qt); | |
1088 | Fset_char_table_range (Vprintable_chars, | |
1089 | Fcons (make_number (160), | |
1090 | make_number (MAX_5_BYTE_CHAR)), Qt); | |
15843e6f | 1091 | |
c57f3328 KH |
1092 | DEFVAR_LISP ("char-script-table", &Vchar_script_table, |
1093 | doc: /* Char table of script symbols. | |
1094 | It has one extra slot whose value is a list of script symbols. */); | |
1095 | ||
1096 | /* Intern this now in case it isn't already done. | |
1097 | Setting this variable twice is harmless. | |
1098 | But don't staticpro it here--that is done in alloc.c. */ | |
1099 | Qchar_table_extra_slots = intern ("char-table-extra-slots"); | |
1100 | DEFSYM (Qchar_script_table, "char-script-table"); | |
1101 | Fput (Qchar_script_table, Qchar_table_extra_slots, make_number (1)); | |
1102 | Vchar_script_table = Fmake_char_table (Qchar_script_table, Qnil); | |
c7e14352 KH |
1103 | |
1104 | DEFVAR_LISP ("script-representative-chars", &Vscript_representative_chars, | |
1105 | doc: /* Alist of scripts vs the representative characters. */); | |
1106 | Vscript_representative_chars = Qnil; | |
0168c3d8 KH |
1107 | } |
1108 | ||
1109 | #endif /* emacs */ | |
fbaf0946 MB |
1110 | |
1111 | /* arch-tag: b6665960-3c3d-4184-85cd-af4318197999 | |
1112 | (do not change this comment) */ |