Commit | Line | Data |
---|---|---|
ca4c9455 | 1 | /* Composite sequence support. |
acaf905b | 2 | Copyright (C) 2001-2012 Free Software Foundation, Inc. |
5df4f04c | 3 | Copyright (C) 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 |
ce03bf76 KH |
4 | National Institute of Advanced Industrial Science and Technology (AIST) |
5 | Registration Number H14PRO021 | |
f30d8d94 | 6 | Copyright (C) 2003, 2006 |
1527c36e KH |
7 | National Institute of Advanced Industrial Science and Technology (AIST) |
8 | Registration Number H13PRO009 | |
ca4c9455 KH |
9 | |
10 | This file is part of GNU Emacs. | |
11 | ||
9ec0b715 | 12 | GNU Emacs is free software: you can redistribute it and/or modify |
ca4c9455 | 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. | |
ca4c9455 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/>. */ |
ca4c9455 KH |
24 | |
25 | #include <config.h> | |
f162bcc3 PE |
26 | |
27 | #define COMPOSITE_INLINE EXTERN_INLINE | |
28 | ||
ca4c9455 | 29 | #include "lisp.h" |
1527c36e | 30 | #include "character.h" |
e5560ff7 | 31 | #include "buffer.h" |
f5199465 | 32 | #include "coding.h" |
ca4c9455 | 33 | #include "intervals.h" |
58753d74 KH |
34 | #include "window.h" |
35 | #include "frame.h" | |
36 | #include "dispextern.h" | |
37 | #include "font.h" | |
f5199465 KH |
38 | #include "termhooks.h" |
39 | ||
ca4c9455 KH |
40 | |
41 | /* Emacs uses special text property `composition' to support character | |
42 | composition. A sequence of characters that have the same (i.e. eq) | |
43 | `composition' property value is treated as a single composite | |
44 | sequence (we call it just `composition' here after). Characters in | |
45 | a composition are all composed somehow on the screen. | |
46 | ||
47 | The property value has this form when the composition is made: | |
48 | ((LENGTH . COMPONENTS) . MODIFICATION-FUNC) | |
49 | then turns to this form: | |
50 | (COMPOSITION-ID . (LENGTH COMPONENTS-VEC . MODIFICATION-FUNC)) | |
51 | when the composition is registered in composition_hash_table and | |
52 | composition_table. These rather peculiar structures were designed | |
53 | to make it easy to distinguish them quickly (we can do that by | |
54 | checking only the first element) and to extract LENGTH (from the | |
55 | former form) and COMPOSITION-ID (from the latter form). | |
56 | ||
57 | We register a composition when it is displayed, or when the width | |
58 | is required (for instance, to calculate columns). | |
59 | ||
60 | LENGTH -- Length of the composition. This information is used to | |
61 | check the validity of the composition. | |
62 | ||
63 | COMPONENTS -- Character, string, vector, list, or nil. | |
64 | ||
65 | If it is nil, characters in the text are composed relatively | |
66 | according to their metrics in font glyphs. | |
67 | ||
68 | If it is a character or a string, the character or characters | |
69 | in the string are composed relatively. | |
70 | ||
71 | If it is a vector or list of integers, the element is a | |
72 | character or an encoded composition rule. The characters are | |
73 | composed according to the rules. (2N)th elements are | |
74 | characters to be composed and (2N+1)th elements are | |
75 | composition rules to tell how to compose (2N+2)th element with | |
76 | the previously composed 2N glyphs. | |
77 | ||
763d7377 EZ |
78 | COMPONENTS-VEC -- Vector of integers. In a relative composition, |
79 | the elements are the characters to be composed. In a rule-base | |
ca4c9455 KH |
80 | composition, the elements are characters or encoded |
81 | composition rules. | |
82 | ||
83 | MODIFICATION-FUNC -- If non nil, it is a function to call when the | |
84 | composition gets invalid after a modification in a buffer. If | |
85 | it is nil, a function in `composition-function-table' of the | |
86 | first character in the sequence is called. | |
87 | ||
88 | COMPOSITION-ID --Identification number of the composition. It is | |
89 | used as an index to composition_table for the composition. | |
90 | ||
91 | When Emacs has to display a composition or has to know its | |
92 | displaying width, the function get_composition_id is called. It | |
93 | returns COMPOSITION-ID so that the caller can access the | |
94 | information about the composition through composition_table. If a | |
95 | COMPOSITION-ID has not yet been assigned to the composition, | |
96 | get_composition_id checks the validity of `composition' property, | |
97 | and, if valid, assigns a new ID, registers the information in | |
98 | composition_hash_table and composition_table, and changes the form | |
763d7377 EZ |
99 | of the property value. If the property is invalid, |
100 | get_composition_id returns -1 without changing the property value. | |
ca4c9455 | 101 | |
763d7377 | 102 | We use two tables to keep the information about composition; |
ca4c9455 KH |
103 | composition_hash_table and composition_table. |
104 | ||
763d7377 | 105 | The former is a hash table whose keys are COMPONENTS-VECs and |
ca4c9455 | 106 | values are the corresponding COMPOSITION-IDs. This hash table is |
4abc7470 | 107 | weak, but as each key (COMPONENTS-VEC) is also kept as a value of the |
ca4c9455 | 108 | `composition' property, it won't be collected as garbage until all |
4abc7470 | 109 | bits of text that have the same COMPONENTS-VEC are deleted. |
ca4c9455 KH |
110 | |
111 | The latter is a table of pointers to `struct composition' indexed | |
4abc7470 | 112 | by COMPOSITION-ID. This structure keeps the other information (see |
ca4c9455 KH |
113 | composite.h). |
114 | ||
115 | In general, a text property holds information about individual | |
116 | characters. But, a `composition' property holds information about | |
4abc7470 | 117 | a sequence of characters (in this sense, it is like the `intangible' |
ca4c9455 | 118 | property). That means that we should not share the property value |
4abc7470 | 119 | in adjacent compositions -- we can't distinguish them if they have the |
ca4c9455 KH |
120 | same property. So, after any changes, we call |
121 | `update_compositions' and change a property of one of adjacent | |
122 | compositions to a copy of it. This function also runs a proper | |
123 | composition modification function to make a composition that gets | |
124 | invalid by the change valid again. | |
125 | ||
4abc7470 | 126 | As the value of the `composition' property holds information about a |
ca4c9455 | 127 | specific range of text, the value gets invalid if we change the |
4abc7470 | 128 | text in the range. We treat the `composition' property as always |
ca4c9455 KH |
129 | rear-nonsticky (currently by setting default-text-properties to |
130 | (rear-nonsticky (composition))) and we never make properties of | |
131 | adjacent compositions identical. Thus, any such changes make the | |
4abc7470 | 132 | range just shorter. So, we can check the validity of the `composition' |
ca4c9455 KH |
133 | property by comparing LENGTH information with the actual length of |
134 | the composition. | |
135 | ||
136 | */ | |
137 | ||
138 | ||
139 | Lisp_Object Qcomposition; | |
140 | ||
141 | /* Table of pointers to the structure `composition' indexed by | |
142 | COMPOSITION-ID. This structure is for storing information about | |
143 | each composition except for COMPONENTS-VEC. */ | |
144 | struct composition **composition_table; | |
145 | ||
146 | /* The current size of `composition_table'. */ | |
ebfa62c0 | 147 | static ptrdiff_t composition_table_size; |
ca4c9455 KH |
148 | |
149 | /* Number of compositions currently made. */ | |
ebfa62c0 | 150 | ptrdiff_t n_compositions; |
ca4c9455 KH |
151 | |
152 | /* Hash table for compositions. The key is COMPONENTS-VEC of | |
153 | `composition' property. The value is the corresponding | |
154 | COMPOSITION-ID. */ | |
155 | Lisp_Object composition_hash_table; | |
156 | ||
955cbe7b PE |
157 | static Lisp_Object Qauto_composed; |
158 | static Lisp_Object Qauto_composition_function; | |
763d7377 EZ |
159 | /* Maximum number of characters to look back for |
160 | auto-compositions. */ | |
895416e3 KH |
161 | #define MAX_AUTO_COMPOSITION_LOOKBACK 3 |
162 | ||
ca4c9455 KH |
163 | /* Temporary variable used in macros COMPOSITION_XXX. */ |
164 | Lisp_Object composition_temp; | |
f30d8d94 | 165 | |
ca4c9455 | 166 | \f |
ca4c9455 KH |
167 | /* Return COMPOSITION-ID of a composition at buffer position |
168 | CHARPOS/BYTEPOS and length NCHARS. The `composition' property of | |
169 | the sequence is PROP. STRING, if non-nil, is a string that | |
170 | contains the composition instead of the current buffer. | |
171 | ||
172 | If the composition is invalid, return -1. */ | |
173 | ||
ebfa62c0 | 174 | ptrdiff_t |
d311d28c | 175 | get_composition_id (ptrdiff_t charpos, ptrdiff_t bytepos, ptrdiff_t nchars, |
579c18d0 | 176 | Lisp_Object prop, Lisp_Object string) |
ca4c9455 KH |
177 | { |
178 | Lisp_Object id, length, components, key, *key_contents; | |
d86d0d74 | 179 | ptrdiff_t glyph_len; |
ca4c9455 | 180 | struct Lisp_Hash_Table *hash_table = XHASH_TABLE (composition_hash_table); |
d3411f89 | 181 | ptrdiff_t hash_index; |
0de4bb68 | 182 | EMACS_UINT hash_code; |
d86d0d74 | 183 | enum composition_method method; |
ca4c9455 | 184 | struct composition *cmp; |
d311d28c | 185 | ptrdiff_t i; |
579c18d0 | 186 | int ch; |
ca4c9455 | 187 | |
0065d054 PE |
188 | /* Maximum length of a string of glyphs. XftGlyphExtents limits |
189 | this to INT_MAX, and Emacs limits it further. Divide INT_MAX - 1 | |
190 | by 2 because x_produce_glyphs computes glyph_len * 2 + 1. Divide | |
191 | the size by MAX_MULTIBYTE_LENGTH because encode_terminal_code | |
192 | multiplies glyph_len by MAX_MULTIBYTE_LENGTH. */ | |
d86d0d74 | 193 | enum { |
0065d054 PE |
194 | GLYPH_LEN_MAX = min ((INT_MAX - 1) / 2, |
195 | min (PTRDIFF_MAX, SIZE_MAX) / MAX_MULTIBYTE_LENGTH) | |
d86d0d74 PE |
196 | }; |
197 | ||
ca4c9455 KH |
198 | /* PROP should be |
199 | Form-A: ((LENGTH . COMPONENTS) . MODIFICATION-FUNC) | |
200 | or | |
201 | Form-B: (COMPOSITION-ID . (LENGTH COMPONENTS-VEC . MODIFICATION-FUNC)) | |
202 | */ | |
203 | if (nchars == 0 || !CONSP (prop)) | |
204 | goto invalid_composition; | |
205 | ||
206 | id = XCAR (prop); | |
207 | if (INTEGERP (id)) | |
208 | { | |
209 | /* PROP should be Form-B. */ | |
210 | if (XINT (id) < 0 || XINT (id) >= n_compositions) | |
211 | goto invalid_composition; | |
212 | return XINT (id); | |
213 | } | |
214 | ||
215 | /* PROP should be Form-A. | |
216 | Thus, ID should be (LENGTH . COMPONENTS). */ | |
217 | if (!CONSP (id)) | |
218 | goto invalid_composition; | |
219 | length = XCAR (id); | |
220 | if (!INTEGERP (length) || XINT (length) != nchars) | |
221 | goto invalid_composition; | |
222 | ||
223 | components = XCDR (id); | |
224 | ||
225 | /* Check if the same composition has already been registered or not | |
226 | by consulting composition_hash_table. The key for this table is | |
227 | COMPONENTS (converted to a vector COMPONENTS-VEC) or, if it is | |
228 | nil, vector of characters in the composition range. */ | |
229 | if (INTEGERP (components)) | |
230 | key = Fmake_vector (make_number (1), components); | |
231 | else if (STRINGP (components) || CONSP (components)) | |
232 | key = Fvconcat (1, &components); | |
233 | else if (VECTORP (components)) | |
234 | key = components; | |
235 | else if (NILP (components)) | |
236 | { | |
237 | key = Fmake_vector (make_number (nchars), Qnil); | |
238 | if (STRINGP (string)) | |
239 | for (i = 0; i < nchars; i++) | |
240 | { | |
241 | FETCH_STRING_CHAR_ADVANCE (ch, string, charpos, bytepos); | |
28be1ada | 242 | ASET (key, i, make_number (ch)); |
ca4c9455 KH |
243 | } |
244 | else | |
245 | for (i = 0; i < nchars; i++) | |
246 | { | |
247 | FETCH_CHAR_ADVANCE (ch, charpos, bytepos); | |
28be1ada | 248 | ASET (key, i, make_number (ch)); |
ca4c9455 KH |
249 | } |
250 | } | |
251 | else | |
252 | goto invalid_composition; | |
253 | ||
254 | hash_index = hash_lookup (hash_table, key, &hash_code); | |
255 | if (hash_index >= 0) | |
256 | { | |
257 | /* We have already registered the same composition. Change PROP | |
258 | from Form-A above to Form-B while replacing COMPONENTS with | |
259 | COMPONENTS-VEC stored in the hash table. We can directly | |
260 | modify the cons cell of PROP because it is not shared. */ | |
261 | key = HASH_KEY (hash_table, hash_index); | |
262 | id = HASH_VALUE (hash_table, hash_index); | |
f3fbd155 KR |
263 | XSETCAR (prop, id); |
264 | XSETCDR (prop, Fcons (make_number (nchars), Fcons (key, XCDR (prop)))); | |
ca4c9455 KH |
265 | return XINT (id); |
266 | } | |
267 | ||
268 | /* This composition is a new one. We must register it. */ | |
177c0ea7 | 269 | |
ca4c9455 | 270 | /* Check if we have sufficient memory to store this information. */ |
0065d054 PE |
271 | if (composition_table_size <= n_compositions) |
272 | composition_table = xpalloc (composition_table, &composition_table_size, | |
273 | 1, -1, sizeof *composition_table); | |
ca4c9455 KH |
274 | |
275 | key_contents = XVECTOR (key)->contents; | |
276 | ||
277 | /* Check if the contents of COMPONENTS are valid if COMPONENTS is a | |
278 | vector or a list. It should be a sequence of: | |
279 | char1 rule1 char2 rule2 char3 ... ruleN charN+1 */ | |
f30d8d94 | 280 | |
dd5e1ed4 | 281 | if (VECTORP (components) |
f30d8d94 KH |
282 | && ASIZE (components) >= 2 |
283 | && VECTORP (AREF (components, 0))) | |
284 | { | |
285 | /* COMPONENTS is a glyph-string. */ | |
d311d28c | 286 | ptrdiff_t len = ASIZE (key); |
f30d8d94 KH |
287 | |
288 | for (i = 1; i < len; i++) | |
289 | if (! VECTORP (AREF (key, i))) | |
290 | goto invalid_composition; | |
291 | } | |
dd5e1ed4 | 292 | else if (VECTORP (components) || CONSP (components)) |
ca4c9455 | 293 | { |
d311d28c | 294 | ptrdiff_t len = ASIZE (key); |
ca4c9455 KH |
295 | |
296 | /* The number of elements should be odd. */ | |
297 | if ((len % 2) == 0) | |
298 | goto invalid_composition; | |
299 | /* All elements should be integers (character or encoded | |
300 | composition rule). */ | |
301 | for (i = 0; i < len; i++) | |
302 | { | |
303 | if (!INTEGERP (key_contents[i])) | |
304 | goto invalid_composition; | |
305 | } | |
306 | } | |
307 | ||
308 | /* Change PROP from Form-A above to Form-B. We can directly modify | |
309 | the cons cell of PROP because it is not shared. */ | |
310 | XSETFASTINT (id, n_compositions); | |
f3fbd155 KR |
311 | XSETCAR (prop, id); |
312 | XSETCDR (prop, Fcons (make_number (nchars), Fcons (key, XCDR (prop)))); | |
ca4c9455 KH |
313 | |
314 | /* Register the composition in composition_hash_table. */ | |
315 | hash_index = hash_put (hash_table, key, id, hash_code); | |
316 | ||
d86d0d74 PE |
317 | method = (NILP (components) |
318 | ? COMPOSITION_RELATIVE | |
319 | : ((INTEGERP (components) || STRINGP (components)) | |
320 | ? COMPOSITION_WITH_ALTCHARS | |
321 | : COMPOSITION_WITH_RULE_ALTCHARS)); | |
322 | ||
323 | glyph_len = (method == COMPOSITION_WITH_RULE_ALTCHARS | |
324 | ? (ASIZE (key) + 1) / 2 | |
325 | : ASIZE (key)); | |
326 | ||
0065d054 | 327 | if (GLYPH_LEN_MAX < glyph_len) |
d86d0d74 PE |
328 | memory_full (SIZE_MAX); |
329 | ||
ca4c9455 | 330 | /* Register the composition in composition_table. */ |
38182d90 | 331 | cmp = xmalloc (sizeof *cmp); |
ca4c9455 | 332 | |
d86d0d74 | 333 | cmp->method = method; |
ca4c9455 | 334 | cmp->hash_index = hash_index; |
ca4c9455 | 335 | cmp->glyph_len = glyph_len; |
0065d054 | 336 | cmp->offsets = xnmalloc (glyph_len, 2 * sizeof *cmp->offsets); |
ca4c9455 KH |
337 | cmp->font = NULL; |
338 | ||
58753d74 | 339 | if (cmp->method != COMPOSITION_WITH_RULE_ALTCHARS) |
ca4c9455 KH |
340 | { |
341 | /* Relative composition. */ | |
342 | cmp->width = 0; | |
343 | for (i = 0; i < glyph_len; i++) | |
344 | { | |
345 | int this_width; | |
346 | ch = XINT (key_contents[i]); | |
bd01620e EZ |
347 | /* TAB in a composition means display glyphs with padding |
348 | space on the left or right. */ | |
8e86803c | 349 | this_width = (ch == '\t' ? 1 : CHAR_WIDTH (ch)); |
ca4c9455 KH |
350 | if (cmp->width < this_width) |
351 | cmp->width = this_width; | |
352 | } | |
353 | } | |
354 | else | |
355 | { | |
356 | /* Rule-base composition. */ | |
cb3c2e3e | 357 | double leftmost = 0.0, rightmost; |
ca4c9455 KH |
358 | |
359 | ch = XINT (key_contents[0]); | |
69df789a | 360 | rightmost = ch != '\t' ? CHAR_WIDTH (ch) : 1; |
ca4c9455 KH |
361 | |
362 | for (i = 1; i < glyph_len; i += 2) | |
363 | { | |
b13a45c6 | 364 | int rule, gref, nref; |
ca4c9455 | 365 | int this_width; |
cb3c2e3e | 366 | double this_left; |
ca4c9455 KH |
367 | |
368 | rule = XINT (key_contents[i]); | |
369 | ch = XINT (key_contents[i + 1]); | |
69df789a | 370 | this_width = ch != '\t' ? CHAR_WIDTH (ch) : 1; |
ca4c9455 KH |
371 | |
372 | /* A composition rule is specified by an integer value | |
373 | that encodes global and new reference points (GREF and | |
374 | NREF). GREF and NREF are specified by numbers as | |
375 | below: | |
376 | 0---1---2 -- ascent | |
377 | | | | |
378 | | | | |
379 | | | | |
380 | 9--10--11 -- center | |
381 | | | | |
382 | ---3---4---5--- baseline | |
383 | | | | |
384 | 6---7---8 -- descent | |
385 | */ | |
b13a45c6 | 386 | COMPOSITION_DECODE_REFS (rule, gref, nref); |
ca4c9455 KH |
387 | this_left = (leftmost |
388 | + (gref % 3) * (rightmost - leftmost) / 2.0 | |
389 | - (nref % 3) * this_width / 2.0); | |
390 | ||
391 | if (this_left < leftmost) | |
392 | leftmost = this_left; | |
393 | if (this_left + this_width > rightmost) | |
394 | rightmost = this_left + this_width; | |
395 | } | |
396 | ||
397 | cmp->width = rightmost - leftmost; | |
398 | if (cmp->width < (rightmost - leftmost)) | |
399 | /* To get a ceiling integer value. */ | |
400 | cmp->width++; | |
401 | } | |
402 | ||
403 | composition_table[n_compositions] = cmp; | |
404 | ||
405 | return n_compositions++; | |
406 | ||
407 | invalid_composition: | |
408 | /* Would it be better to remove this `composition' property? */ | |
409 | return -1; | |
410 | } | |
411 | ||
412 | \f | |
90b3fe91 KH |
413 | /* Find a static composition at or nearest to position POS of OBJECT |
414 | (buffer or string). | |
ca4c9455 KH |
415 | |
416 | OBJECT defaults to the current buffer. If there's a composition at | |
417 | POS, set *START and *END to the start and end of the sequence, | |
418 | *PROP to the `composition' property, and return 1. | |
419 | ||
420 | If there's no composition at POS and LIMIT is negative, return 0. | |
421 | ||
422 | Otherwise, search for a composition forward (LIMIT > POS) or | |
423 | backward (LIMIT < POS). In this case, LIMIT bounds the search. | |
424 | ||
425 | If a composition is found, set *START, *END, and *PROP as above, | |
426 | and return 1, else return 0. | |
427 | ||
428 | This doesn't check the validity of composition. */ | |
429 | ||
de1339b0 | 430 | bool |
d311d28c PE |
431 | find_composition (ptrdiff_t pos, ptrdiff_t limit, |
432 | ptrdiff_t *start, ptrdiff_t *end, | |
579c18d0 | 433 | Lisp_Object *prop, Lisp_Object object) |
ca4c9455 KH |
434 | { |
435 | Lisp_Object val; | |
436 | ||
437 | if (get_property_and_range (pos, Qcomposition, prop, start, end, object)) | |
438 | return 1; | |
439 | ||
440 | if (limit < 0 || limit == pos) | |
441 | return 0; | |
442 | ||
443 | if (limit > pos) /* search forward */ | |
d279f620 KH |
444 | { |
445 | val = Fnext_single_property_change (make_number (pos), Qcomposition, | |
446 | object, make_number (limit)); | |
447 | pos = XINT (val); | |
448 | if (pos == limit) | |
449 | return 0; | |
450 | } | |
ca4c9455 | 451 | else /* search backward */ |
d279f620 KH |
452 | { |
453 | if (get_property_and_range (pos - 1, Qcomposition, prop, start, end, | |
454 | object)) | |
455 | return 1; | |
456 | val = Fprevious_single_property_change (make_number (pos), Qcomposition, | |
457 | object, make_number (limit)); | |
458 | pos = XINT (val); | |
459 | if (pos == limit) | |
460 | return 0; | |
461 | pos--; | |
462 | } | |
ca4c9455 KH |
463 | get_property_and_range (pos, Qcomposition, prop, start, end, object); |
464 | return 1; | |
465 | } | |
466 | ||
467 | /* Run a proper function to adjust the composition sitting between | |
468 | FROM and TO with property PROP. */ | |
469 | ||
470 | static void | |
d311d28c | 471 | run_composition_function (ptrdiff_t from, ptrdiff_t to, Lisp_Object prop) |
ca4c9455 | 472 | { |
7d019510 | 473 | Lisp_Object func; |
d311d28c | 474 | ptrdiff_t start, end; |
ca4c9455 KH |
475 | |
476 | func = COMPOSITION_MODIFICATION_FUNC (prop); | |
477 | /* If an invalid composition precedes or follows, try to make them | |
478 | valid too. */ | |
479 | if (from > BEGV | |
480 | && find_composition (from - 1, -1, &start, &end, &prop, Qnil) | |
481 | && !COMPOSITION_VALID_P (start, end, prop)) | |
482 | from = start; | |
483 | if (to < ZV | |
484 | && find_composition (to, -1, &start, &end, &prop, Qnil) | |
485 | && !COMPOSITION_VALID_P (start, end, prop)) | |
486 | to = end; | |
775b3d2d | 487 | if (!NILP (Ffboundp (func))) |
ca4c9455 | 488 | call2 (func, make_number (from), make_number (to)); |
ca4c9455 KH |
489 | } |
490 | ||
491 | /* Make invalid compositions adjacent to or inside FROM and TO valid. | |
492 | CHECK_MASK is bitwise `or' of mask bits defined by macros | |
493 | CHECK_XXX (see the comment in composite.h). | |
494 | ||
b418f8a6 | 495 | It also resets the text-property `auto-composed' to a proper region |
f96ba4c1 KH |
496 | so that automatic character composition works correctly later while |
497 | displaying the region. | |
8f924df7 | 498 | |
ca4c9455 KH |
499 | This function is called when a buffer text is changed. If the |
500 | change is deletion, FROM == TO. Otherwise, FROM < TO. */ | |
501 | ||
502 | void | |
d311d28c | 503 | update_compositions (ptrdiff_t from, ptrdiff_t to, int check_mask) |
ca4c9455 | 504 | { |
7d019510 | 505 | Lisp_Object prop; |
d311d28c | 506 | ptrdiff_t start, end; |
f96ba4c1 KH |
507 | /* The beginning and end of the region to set the property |
508 | `auto-composed' to nil. */ | |
d311d28c | 509 | ptrdiff_t min_pos = from, max_pos = to; |
ca4c9455 | 510 | |
6c1aa7f1 GM |
511 | if (inhibit_modification_hooks) |
512 | return; | |
177c0ea7 | 513 | |
d3f40cbd KH |
514 | /* If FROM and TO are not in a valid range, do nothing. */ |
515 | if (! (BEGV <= from && from <= to && to <= ZV)) | |
516 | return; | |
517 | ||
ca4c9455 KH |
518 | if (check_mask & CHECK_HEAD) |
519 | { | |
520 | /* FROM should be at composition boundary. But, insertion or | |
521 | deletion will make two compositions adjacent and | |
522 | indistinguishable when they have same (eq) property. To | |
523 | avoid it, in such a case, we change the property of the | |
524 | latter to the copy of it. */ | |
525 | if (from > BEGV | |
553d3164 KH |
526 | && find_composition (from - 1, -1, &start, &end, &prop, Qnil) |
527 | && COMPOSITION_VALID_P (start, end, prop)) | |
ca4c9455 | 528 | { |
f96ba4c1 KH |
529 | min_pos = start; |
530 | if (end > to) | |
531 | max_pos = end; | |
ca4c9455 KH |
532 | if (from < end) |
533 | Fput_text_property (make_number (from), make_number (end), | |
534 | Qcomposition, | |
535 | Fcons (XCAR (prop), XCDR (prop)), Qnil); | |
536 | run_composition_function (start, end, prop); | |
537 | from = end; | |
538 | } | |
dd33cc56 | 539 | else if (from < ZV |
553d3164 | 540 | && find_composition (from, -1, &start, &from, &prop, Qnil) |
9657d668 | 541 | && COMPOSITION_VALID_P (start, from, prop)) |
f96ba4c1 KH |
542 | { |
543 | if (from > to) | |
544 | max_pos = from; | |
545 | run_composition_function (start, from, prop); | |
546 | } | |
ca4c9455 KH |
547 | } |
548 | ||
549 | if (check_mask & CHECK_INSIDE) | |
550 | { | |
551 | /* In this case, we are sure that (check & CHECK_TAIL) is also | |
552 | nonzero. Thus, here we should check only compositions before | |
553 | (to - 1). */ | |
554 | while (from < to - 1 | |
555 | && find_composition (from, to, &start, &from, &prop, Qnil) | |
9657d668 | 556 | && COMPOSITION_VALID_P (start, from, prop) |
ca4c9455 KH |
557 | && from < to - 1) |
558 | run_composition_function (start, from, prop); | |
559 | } | |
560 | ||
561 | if (check_mask & CHECK_TAIL) | |
562 | { | |
563 | if (from < to | |
553d3164 KH |
564 | && find_composition (to - 1, -1, &start, &end, &prop, Qnil) |
565 | && COMPOSITION_VALID_P (start, end, prop)) | |
ca4c9455 KH |
566 | { |
567 | /* TO should be also at composition boundary. But, | |
568 | insertion or deletion will make two compositions adjacent | |
569 | and indistinguishable when they have same (eq) property. | |
570 | To avoid it, in such a case, we change the property of | |
571 | the former to the copy of it. */ | |
572 | if (to < end) | |
f96ba4c1 KH |
573 | { |
574 | Fput_text_property (make_number (start), make_number (to), | |
575 | Qcomposition, | |
576 | Fcons (XCAR (prop), XCDR (prop)), Qnil); | |
577 | max_pos = end; | |
578 | } | |
ca4c9455 KH |
579 | run_composition_function (start, end, prop); |
580 | } | |
581 | else if (to < ZV | |
553d3164 KH |
582 | && find_composition (to, -1, &start, &end, &prop, Qnil) |
583 | && COMPOSITION_VALID_P (start, end, prop)) | |
f96ba4c1 KH |
584 | { |
585 | run_composition_function (start, end, prop); | |
586 | max_pos = end; | |
587 | } | |
ca4c9455 | 588 | } |
f96ba4c1 | 589 | if (min_pos < max_pos) |
9d440521 | 590 | { |
d311d28c | 591 | ptrdiff_t count = SPECPDL_INDEX (); |
9d440521 KH |
592 | |
593 | specbind (Qinhibit_read_only, Qt); | |
594 | specbind (Qinhibit_modification_hooks, Qt); | |
595 | specbind (Qinhibit_point_motion_hooks, Qt); | |
596 | Fremove_list_of_text_properties (make_number (min_pos), | |
597 | make_number (max_pos), | |
598 | Fcons (Qauto_composed, Qnil), Qnil); | |
599 | unbind_to (count, Qnil); | |
600 | } | |
ca4c9455 KH |
601 | } |
602 | ||
c1361885 KH |
603 | |
604 | /* Modify composition property values in LIST destructively. LIST is | |
605 | a list as returned from text_property_list. Change values to the | |
606 | top-level copies of them so that none of them are `eq'. */ | |
607 | ||
608 | void | |
971de7fb | 609 | make_composition_value_copy (Lisp_Object list) |
c1361885 KH |
610 | { |
611 | Lisp_Object plist, val; | |
612 | ||
613 | for (; CONSP (list); list = XCDR (list)) | |
614 | { | |
615 | plist = XCAR (XCDR (XCDR (XCAR (list)))); | |
616 | while (CONSP (plist) && CONSP (XCDR (plist))) | |
617 | { | |
618 | if (EQ (XCAR (plist), Qcomposition) | |
619 | && (val = XCAR (XCDR (plist)), CONSP (val))) | |
f3fbd155 | 620 | XSETCAR (XCDR (plist), Fcons (XCAR (val), XCDR (val))); |
c1361885 KH |
621 | plist = XCDR (XCDR (plist)); |
622 | } | |
623 | } | |
624 | } | |
625 | ||
626 | ||
ca4c9455 KH |
627 | /* Make text in the region between START and END a composition that |
628 | has COMPONENTS and MODIFICATION-FUNC. | |
629 | ||
630 | If STRING is non-nil, then operate on characters contained between | |
631 | indices START and END in STRING. */ | |
632 | ||
633 | void | |
d311d28c | 634 | compose_text (ptrdiff_t start, ptrdiff_t end, Lisp_Object components, |
579c18d0 | 635 | Lisp_Object modification_func, Lisp_Object string) |
ca4c9455 KH |
636 | { |
637 | Lisp_Object prop; | |
638 | ||
639 | prop = Fcons (Fcons (make_number (end - start), components), | |
640 | modification_func); | |
641 | Fput_text_property (make_number (start), make_number (end), | |
642 | Qcomposition, prop, string); | |
643 | } | |
58753d74 KH |
644 | |
645 | ||
d311d28c PE |
646 | static Lisp_Object autocmp_chars (Lisp_Object, ptrdiff_t, ptrdiff_t, |
647 | ptrdiff_t, struct window *, | |
f57e2426 | 648 | struct face *, Lisp_Object); |
58753d74 KH |
649 | |
650 | \f | |
651 | /* Lisp glyph-string handlers */ | |
652 | ||
653 | /* Hash table for automatic composition. The key is a header of a | |
654 | lgstring (Lispy glyph-string), and the value is a body of a | |
655 | lgstring. */ | |
656 | ||
657 | static Lisp_Object gstring_hash_table; | |
658 | ||
f57e2426 | 659 | static Lisp_Object gstring_lookup_cache (Lisp_Object); |
58753d74 KH |
660 | |
661 | static Lisp_Object | |
971de7fb | 662 | gstring_lookup_cache (Lisp_Object header) |
58753d74 KH |
663 | { |
664 | struct Lisp_Hash_Table *h = XHASH_TABLE (gstring_hash_table); | |
d3411f89 | 665 | ptrdiff_t i = hash_lookup (h, header, NULL); |
58753d74 KH |
666 | |
667 | return (i >= 0 ? HASH_VALUE (h, i) : Qnil); | |
668 | } | |
669 | ||
670 | Lisp_Object | |
d311d28c | 671 | composition_gstring_put_cache (Lisp_Object gstring, ptrdiff_t len) |
58753d74 KH |
672 | { |
673 | struct Lisp_Hash_Table *h = XHASH_TABLE (gstring_hash_table); | |
0de4bb68 | 674 | EMACS_UINT hash; |
58753d74 | 675 | Lisp_Object header, copy; |
d311d28c | 676 | ptrdiff_t i; |
58753d74 KH |
677 | |
678 | header = LGSTRING_HEADER (gstring); | |
b7432bb2 | 679 | hash = h->test.hashfn (&h->test, header); |
58753d74 KH |
680 | if (len < 0) |
681 | { | |
d311d28c | 682 | ptrdiff_t j, glyph_len = LGSTRING_GLYPH_LEN (gstring); |
27ccc379 PE |
683 | for (j = 0; j < glyph_len; j++) |
684 | if (NILP (LGSTRING_GLYPH (gstring, j))) | |
58753d74 | 685 | break; |
27ccc379 | 686 | len = j; |
58753d74 | 687 | } |
087e2ea9 | 688 | |
d311d28c | 689 | lint_assume (len <= TYPE_MAXIMUM (ptrdiff_t) - 2); |
58753d74 KH |
690 | copy = Fmake_vector (make_number (len + 2), Qnil); |
691 | LGSTRING_SET_HEADER (copy, Fcopy_sequence (header)); | |
692 | for (i = 0; i < len; i++) | |
693 | LGSTRING_SET_GLYPH (copy, i, Fcopy_sequence (LGSTRING_GLYPH (gstring, i))); | |
694 | i = hash_put (h, LGSTRING_HEADER (copy), copy, hash); | |
695 | LGSTRING_SET_ID (copy, make_number (i)); | |
696 | return copy; | |
697 | } | |
698 | ||
699 | Lisp_Object | |
ebfa62c0 | 700 | composition_gstring_from_id (ptrdiff_t id) |
58753d74 KH |
701 | { |
702 | struct Lisp_Hash_Table *h = XHASH_TABLE (gstring_hash_table); | |
703 | ||
704 | return HASH_VALUE (h, id); | |
705 | } | |
706 | ||
f57e2426 J |
707 | static Lisp_Object fill_gstring_header (Lisp_Object, Lisp_Object, |
708 | Lisp_Object, Lisp_Object, | |
709 | Lisp_Object); | |
58753d74 | 710 | |
de1339b0 | 711 | bool |
971de7fb | 712 | composition_gstring_p (Lisp_Object gstring) |
58753d74 KH |
713 | { |
714 | Lisp_Object header; | |
d311d28c | 715 | ptrdiff_t i; |
58753d74 KH |
716 | |
717 | if (! VECTORP (gstring) || ASIZE (gstring) < 2) | |
718 | return 0; | |
719 | header = LGSTRING_HEADER (gstring); | |
720 | if (! VECTORP (header) || ASIZE (header) < 2) | |
721 | return 0; | |
722 | if (! NILP (LGSTRING_FONT (gstring)) | |
f5199465 KH |
723 | && (! FONT_OBJECT_P (LGSTRING_FONT (gstring)) |
724 | && ! CODING_SYSTEM_P (LGSTRING_FONT (gstring)))) | |
58753d74 KH |
725 | return 0; |
726 | for (i = 1; i < ASIZE (LGSTRING_HEADER (gstring)); i++) | |
727 | if (! NATNUMP (AREF (LGSTRING_HEADER (gstring), i))) | |
728 | return 0; | |
729 | if (! NILP (LGSTRING_ID (gstring)) && ! NATNUMP (LGSTRING_ID (gstring))) | |
730 | return 0; | |
731 | for (i = 0; i < LGSTRING_GLYPH_LEN (gstring); i++) | |
732 | { | |
733 | Lisp_Object glyph = LGSTRING_GLYPH (gstring, i); | |
734 | if (NILP (glyph)) | |
735 | break; | |
736 | if (! VECTORP (glyph) || ASIZE (glyph) != LGLYPH_SIZE) | |
737 | return 0; | |
738 | } | |
739 | return 1; | |
740 | } | |
741 | ||
742 | int | |
d311d28c | 743 | composition_gstring_width (Lisp_Object gstring, ptrdiff_t from, ptrdiff_t to, |
579c18d0 | 744 | struct font_metrics *metrics) |
58753d74 KH |
745 | { |
746 | Lisp_Object *glyph; | |
747 | int width = 0; | |
748 | ||
749 | if (metrics) | |
750 | { | |
751 | Lisp_Object font_object = LGSTRING_FONT (gstring); | |
58753d74 | 752 | |
f5199465 KH |
753 | if (FONT_OBJECT_P (font_object)) |
754 | { | |
755 | struct font *font = XFONT_OBJECT (font_object); | |
756 | ||
757 | metrics->ascent = font->ascent; | |
758 | metrics->descent = font->descent; | |
759 | } | |
760 | else | |
761 | { | |
762 | metrics->ascent = 1; | |
763 | metrics->descent = 0; | |
764 | } | |
58753d74 KH |
765 | metrics->width = metrics->lbearing = metrics->rbearing = 0; |
766 | } | |
4939150c | 767 | for (glyph = lgstring_glyph_addr (gstring, from); from < to; from++, glyph++) |
58753d74 KH |
768 | { |
769 | int x; | |
770 | ||
771 | if (NILP (LGLYPH_ADJUSTMENT (*glyph))) | |
772 | width += LGLYPH_WIDTH (*glyph); | |
773 | else | |
774 | width += LGLYPH_WADJUST (*glyph); | |
775 | if (metrics) | |
776 | { | |
777 | x = metrics->width + LGLYPH_LBEARING (*glyph) + LGLYPH_XOFF (*glyph); | |
778 | if (metrics->lbearing > x) | |
779 | metrics->lbearing = x; | |
780 | x = metrics->width + LGLYPH_RBEARING (*glyph) + LGLYPH_XOFF (*glyph); | |
781 | if (metrics->rbearing < x) | |
782 | metrics->rbearing = x; | |
783 | metrics->width = width; | |
784 | x = LGLYPH_ASCENT (*glyph) - LGLYPH_YOFF (*glyph); | |
785 | if (metrics->ascent < x) | |
786 | metrics->ascent = x; | |
895416e3 | 787 | x = LGLYPH_DESCENT (*glyph) + LGLYPH_YOFF (*glyph); |
58753d74 KH |
788 | if (metrics->descent < x) |
789 | metrics->descent = x; | |
790 | } | |
791 | } | |
792 | return width; | |
793 | } | |
794 | ||
795 | ||
796 | static Lisp_Object gstring_work; | |
797 | static Lisp_Object gstring_work_headers; | |
798 | ||
799 | static Lisp_Object | |
971de7fb | 800 | fill_gstring_header (Lisp_Object header, Lisp_Object start, Lisp_Object end, Lisp_Object font_object, Lisp_Object string) |
58753d74 | 801 | { |
d311d28c PE |
802 | ptrdiff_t from, to, from_byte; |
803 | ptrdiff_t len, i; | |
58753d74 KH |
804 | |
805 | if (NILP (string)) | |
806 | { | |
4b4deea2 | 807 | if (NILP (BVAR (current_buffer, enable_multibyte_characters))) |
58753d74 KH |
808 | error ("Attempt to shape unibyte text"); |
809 | validate_region (&start, &end); | |
810 | from = XFASTINT (start); | |
811 | to = XFASTINT (end); | |
812 | from_byte = CHAR_TO_BYTE (from); | |
813 | } | |
814 | else | |
815 | { | |
816 | CHECK_STRING (string); | |
793ffee8 | 817 | if (! STRING_MULTIBYTE (string)) |
58753d74 | 818 | error ("Attempt to shape unibyte text"); |
d311d28c PE |
819 | /* The caller checks that START and END are nonnegative integers. */ |
820 | if (! (XINT (start) <= XINT (end) && XINT (end) <= SCHARS (string))) | |
821 | args_out_of_range_3 (string, start, end); | |
58753d74 | 822 | from = XINT (start); |
58753d74 | 823 | to = XINT (end); |
58753d74 KH |
824 | from_byte = string_char_to_byte (string, from); |
825 | } | |
826 | ||
827 | len = to - from; | |
828 | if (len == 0) | |
829 | error ("Attempt to shape zero-length text"); | |
830 | if (VECTORP (header)) | |
831 | { | |
832 | if (ASIZE (header) != len + 1) | |
833 | args_out_of_range (header, make_number (len + 1)); | |
834 | } | |
835 | else | |
836 | { | |
837 | if (len <= 8) | |
838 | header = AREF (gstring_work_headers, len - 1); | |
839 | else | |
840 | header = Fmake_vector (make_number (len + 1), Qnil); | |
841 | } | |
842 | ||
843 | ASET (header, 0, font_object); | |
844 | for (i = 0; i < len; i++) | |
845 | { | |
846 | int c; | |
847 | ||
848 | if (NILP (string)) | |
849 | FETCH_CHAR_ADVANCE_NO_CHECK (c, from, from_byte); | |
850 | else | |
851 | FETCH_STRING_CHAR_ADVANCE_NO_CHECK (c, string, from, from_byte); | |
852 | ASET (header, i + 1, make_number (c)); | |
853 | } | |
854 | return header; | |
855 | } | |
856 | ||
58753d74 | 857 | static void |
971de7fb | 858 | fill_gstring_body (Lisp_Object gstring) |
58753d74 KH |
859 | { |
860 | Lisp_Object font_object = LGSTRING_FONT (gstring); | |
861 | Lisp_Object header = AREF (gstring, 0); | |
d311d28c PE |
862 | ptrdiff_t len = LGSTRING_CHAR_LEN (gstring); |
863 | ptrdiff_t i; | |
58753d74 KH |
864 | |
865 | for (i = 0; i < len; i++) | |
866 | { | |
867 | Lisp_Object g = LGSTRING_GLYPH (gstring, i); | |
f8c86b69 | 868 | int c = XFASTINT (AREF (header, i + 1)); |
58753d74 KH |
869 | |
870 | if (NILP (g)) | |
871 | { | |
872 | g = LGLYPH_NEW (); | |
873 | LGSTRING_SET_GLYPH (gstring, i, g); | |
874 | } | |
875 | LGLYPH_SET_FROM (g, i); | |
876 | LGLYPH_SET_TO (g, i); | |
877 | LGLYPH_SET_CHAR (g, c); | |
f5199465 | 878 | if (FONT_OBJECT_P (font_object)) |
58753d74 KH |
879 | { |
880 | font_fill_lglyph_metrics (g, font_object); | |
881 | } | |
882 | else | |
883 | { | |
884 | int width = XFASTINT (CHAR_TABLE_REF (Vchar_width_table, c)); | |
885 | ||
886 | LGLYPH_SET_CODE (g, c); | |
887 | LGLYPH_SET_LBEARING (g, 0); | |
888 | LGLYPH_SET_RBEARING (g, width); | |
889 | LGLYPH_SET_WIDTH (g, width); | |
890 | LGLYPH_SET_ASCENT (g, 1); | |
891 | LGLYPH_SET_DESCENT (g, 0); | |
892 | } | |
893 | LGLYPH_SET_ADJUSTMENT (g, Qnil); | |
894 | } | |
895 | if (i < LGSTRING_GLYPH_LEN (gstring)) | |
896 | LGSTRING_SET_GLYPH (gstring, i, Qnil); | |
897 | } | |
898 | ||
58753d74 | 899 | |
10f72a37 KH |
900 | /* Try to compose the characters at CHARPOS according to composition |
901 | rule RULE ([PATTERN PREV-CHARS FUNC]). LIMIT limits the characters | |
902 | to compose. STRING, if not nil, is a target string. WIN is a | |
903 | window where the characters are being displayed. If characters are | |
904 | successfully composed, return the composition as a glyph-string | |
905 | object. Otherwise return nil. */ | |
58753d74 KH |
906 | |
907 | static Lisp_Object | |
d311d28c | 908 | autocmp_chars (Lisp_Object rule, ptrdiff_t charpos, ptrdiff_t bytepos, ptrdiff_t limit, struct window *win, struct face *face, Lisp_Object string) |
58753d74 | 909 | { |
d311d28c | 910 | ptrdiff_t count = SPECPDL_INDEX (); |
d3d50620 | 911 | FRAME_PTR f = XFRAME (win->frame); |
58753d74 | 912 | Lisp_Object pos = make_number (charpos); |
d311d28c PE |
913 | ptrdiff_t to; |
914 | ptrdiff_t pt = PT, pt_byte = PT_BYTE; | |
10f72a37 | 915 | Lisp_Object re, font_object, lgstring; |
d311d28c | 916 | ptrdiff_t len; |
087e2ea9 | 917 | |
58753d74 | 918 | record_unwind_save_match_data (); |
10f72a37 KH |
919 | re = AREF (rule, 0); |
920 | if (NILP (re)) | |
921 | len = 1; | |
922 | else if (! STRINGP (re)) | |
923 | return unbind_to (count, Qnil); | |
924 | else if ((len = fast_looking_at (re, charpos, bytepos, limit, -1, string)) | |
925 | > 0) | |
58753d74 | 926 | { |
10f72a37 KH |
927 | if (NILP (string)) |
928 | len = BYTE_TO_CHAR (bytepos + len) - charpos; | |
929 | else | |
930 | len = string_byte_to_char (string, bytepos + len) - charpos; | |
931 | } | |
932 | if (len <= 0) | |
933 | return unbind_to (count, Qnil); | |
934 | to = limit = charpos + len; | |
58753d74 | 935 | #ifdef HAVE_WINDOW_SYSTEM |
10f72a37 KH |
936 | if (FRAME_WINDOW_P (f)) |
937 | { | |
938 | font_object = font_range (charpos, &to, win, face, string); | |
939 | if (! FONT_OBJECT_P (font_object) | |
940 | || (! NILP (re) | |
941 | && to < limit | |
942 | && (fast_looking_at (re, charpos, bytepos, to, -1, string) <= 0))) | |
943 | return unbind_to (count, Qnil); | |
944 | } | |
945 | else | |
58753d74 | 946 | #endif /* not HAVE_WINDOW_SYSTEM */ |
d3d50620 | 947 | font_object = win->frame; |
10f72a37 KH |
948 | lgstring = Fcomposition_get_gstring (pos, make_number (to), font_object, |
949 | string); | |
950 | if (NILP (LGSTRING_ID (lgstring))) | |
951 | { | |
10f72a37 KH |
952 | /* Save point as marker before calling out to lisp. */ |
953 | if (NILP (string)) | |
657924ff DA |
954 | record_unwind_protect (restore_point_unwind, |
955 | build_marker (current_buffer, pt, pt_byte)); | |
6cd7a139 DA |
956 | lgstring = safe_call (6, Vauto_composition_function, AREF (rule, 2), |
957 | pos, make_number (to), font_object, string); | |
58753d74 | 958 | } |
10f72a37 | 959 | return unbind_to (count, lgstring); |
58753d74 KH |
960 | } |
961 | ||
82ebc97b | 962 | static Lisp_Object _work_val; |
82ebc97b | 963 | |
f6aa6ec6 KH |
964 | /* 1 iff the character C is composable. Characters of general |
965 | category Z? or C? are not composable except for ZWNJ and ZWJ. */ | |
966 | ||
82ebc97b | 967 | #define CHAR_COMPOSABLE_P(C) \ |
a28d4396 KH |
968 | ((C) > ' ' \ |
969 | && ((C) == 0x200C || (C) == 0x200D \ | |
970 | || (_work_val = CHAR_TABLE_REF (Vunicode_category_table, (C)), \ | |
c805dec0 KH |
971 | (INTEGERP (_work_val) \ |
972 | && (XINT (_work_val) <= UNICODE_CATEGORY_So))))) | |
58753d74 KH |
973 | |
974 | /* Update cmp_it->stop_pos to the next position after CHARPOS (and | |
975 | BYTEPOS) where character composition may happen. If BYTEPOS is | |
82ebc97b KH |
976 | negative, compute it. ENDPOS is a limit of searching. If it is |
977 | less than CHARPOS, search backward to ENDPOS+1 assuming that | |
978 | set_iterator_to_next works in reverse order. In this case, if a | |
979 | composition closest to CHARPOS is found, set cmp_it->stop_pos to | |
980 | the last character of the composition. | |
981 | ||
982 | If no composition is found, set cmp_it->ch to -2. If a static | |
983 | composition is found, set cmp_it->ch to -1. Otherwise, set | |
984 | cmp_it->ch to the character that triggers the automatic | |
985 | composition. */ | |
58753d74 KH |
986 | |
987 | void | |
d311d28c | 988 | composition_compute_stop_pos (struct composition_it *cmp_it, ptrdiff_t charpos, ptrdiff_t bytepos, ptrdiff_t endpos, Lisp_Object string) |
58753d74 | 989 | { |
d311d28c | 990 | ptrdiff_t start, end; |
f8c86b69 | 991 | int c; |
58753d74 | 992 | Lisp_Object prop, val; |
3ffdafce KH |
993 | /* This is from forward_to_next_line_start in xdisp.c. */ |
994 | const int MAX_NEWLINE_DISTANCE = 500; | |
58753d74 | 995 | |
82ebc97b KH |
996 | if (charpos < endpos) |
997 | { | |
998 | if (endpos > charpos + MAX_NEWLINE_DISTANCE) | |
999 | endpos = charpos + MAX_NEWLINE_DISTANCE; | |
1000 | } | |
1001 | else if (endpos < charpos) | |
1002 | { | |
1003 | /* We search backward for a position to check composition. */ | |
1004 | if (endpos < 0) | |
1005 | { | |
1006 | /* But we don't know where to stop the searching. */ | |
1007 | endpos = NILP (string) ? BEGV - 1 : -1; | |
1008 | /* Usually we don't reach ENDPOS because we stop searching | |
1009 | at an uncomposable character (NL, LRE, etc). */ | |
1010 | } | |
1011 | } | |
053ca52b | 1012 | cmp_it->id = -1; |
44566dc7 | 1013 | cmp_it->ch = -2; |
82ebc97b KH |
1014 | cmp_it->reversed_p = 0; |
1015 | cmp_it->stop_pos = endpos; | |
1016 | if (charpos == endpos) | |
1017 | return; | |
1018 | /* FIXME: Bidi is not yet handled well in static composition. */ | |
1019 | if (charpos < endpos | |
1020 | && find_composition (charpos, endpos, &start, &end, &prop, string) | |
15fa4783 | 1021 | && start >= charpos |
58753d74 KH |
1022 | && COMPOSITION_VALID_P (start, end, prop)) |
1023 | { | |
1024 | cmp_it->stop_pos = endpos = start; | |
1025 | cmp_it->ch = -1; | |
1026 | } | |
4b4deea2 | 1027 | if (NILP (BVAR (current_buffer, enable_multibyte_characters)) |
d9a7c140 | 1028 | || NILP (Vauto_composition_mode)) |
58753d74 KH |
1029 | return; |
1030 | if (bytepos < 0) | |
1031 | { | |
82ebc97b | 1032 | if (NILP (string)) |
58753d74 | 1033 | bytepos = CHAR_TO_BYTE (charpos); |
82ebc97b KH |
1034 | else |
1035 | bytepos = string_char_to_byte (string, charpos); | |
58753d74 KH |
1036 | } |
1037 | ||
1038 | start = charpos; | |
82ebc97b | 1039 | if (charpos < endpos) |
58753d74 | 1040 | { |
82ebc97b KH |
1041 | /* Forward search. */ |
1042 | while (charpos < endpos) | |
ea8ba975 | 1043 | { |
82ebc97b KH |
1044 | if (STRINGP (string)) |
1045 | FETCH_STRING_CHAR_ADVANCE (c, string, charpos, bytepos); | |
1046 | else | |
1047 | FETCH_CHAR_ADVANCE (c, charpos, bytepos); | |
1048 | if (c == '\n') | |
1049 | { | |
1050 | cmp_it->ch = -2; | |
1051 | break; | |
1052 | } | |
1053 | val = CHAR_TABLE_REF (Vcomposition_function_table, c); | |
1054 | if (! NILP (val)) | |
1055 | { | |
1056 | Lisp_Object elt; | |
10f72a37 | 1057 | int ridx; |
82ebc97b | 1058 | |
10f72a37 | 1059 | for (ridx = 0; CONSP (val); val = XCDR (val), ridx++) |
82ebc97b KH |
1060 | { |
1061 | elt = XCAR (val); | |
1062 | if (VECTORP (elt) && ASIZE (elt) == 3 | |
1063 | && NATNUMP (AREF (elt, 1)) | |
1064 | && charpos - 1 - XFASTINT (AREF (elt, 1)) >= start) | |
1065 | break; | |
1066 | } | |
1067 | if (CONSP (val)) | |
1068 | { | |
10f72a37 | 1069 | cmp_it->rule_idx = ridx; |
82ebc97b KH |
1070 | cmp_it->lookback = XFASTINT (AREF (elt, 1)); |
1071 | cmp_it->stop_pos = charpos - 1 - cmp_it->lookback; | |
1072 | cmp_it->ch = c; | |
1073 | return; | |
1074 | } | |
1075 | } | |
ea8ba975 | 1076 | } |
ff94e32c KH |
1077 | if (charpos == endpos) |
1078 | { | |
1079 | /* We couldn't find a composition point before ENDPOS. But, | |
1080 | some character after ENDPOS may be composed with | |
1081 | characters before ENDPOS. So, we should stop at the safe | |
1082 | point. */ | |
1083 | charpos = endpos - MAX_AUTO_COMPOSITION_LOOKBACK; | |
1084 | if (charpos < start) | |
1085 | charpos = start; | |
1086 | } | |
82ebc97b | 1087 | } |
b7050583 | 1088 | else if (charpos > endpos) |
82ebc97b KH |
1089 | { |
1090 | /* Search backward for a pattern that may be composed and the | |
1091 | position of (possibly) the last character of the match is | |
1092 | closest to (but not after) START. The reason for the last | |
d07f018c EZ |
1093 | character is that set_iterator_to_next works in reverse order, |
1094 | and thus we must stop at the last character for composition | |
82ebc97b KH |
1095 | check. */ |
1096 | unsigned char *p; | |
1097 | int len; | |
d07f018c EZ |
1098 | /* Limit byte position used in fast_looking_at. This is the |
1099 | byte position of the character after START. */ | |
d311d28c | 1100 | ptrdiff_t limit; |
82ebc97b KH |
1101 | |
1102 | if (NILP (string)) | |
1103 | p = BYTE_POS_ADDR (bytepos); | |
1104 | else | |
1105 | p = SDATA (string) + bytepos; | |
1106 | c = STRING_CHAR_AND_LENGTH (p, len); | |
1107 | limit = bytepos + len; | |
1108 | while (CHAR_COMPOSABLE_P (c)) | |
58753d74 | 1109 | { |
10f72a37 KH |
1110 | val = CHAR_TABLE_REF (Vcomposition_function_table, c); |
1111 | if (! NILP (val)) | |
82ebc97b | 1112 | { |
10f72a37 | 1113 | Lisp_Object elt; |
d311d28c | 1114 | int ridx, blen; |
58753d74 | 1115 | |
10f72a37 | 1116 | for (ridx = 0; CONSP (val); val = XCDR (val), ridx++) |
82ebc97b | 1117 | { |
10f72a37 KH |
1118 | elt = XCAR (val); |
1119 | if (VECTORP (elt) && ASIZE (elt) == 3 | |
1120 | && NATNUMP (AREF (elt, 1)) | |
d311d28c | 1121 | && charpos - XFASTINT (AREF (elt, 1)) > endpos) |
82ebc97b | 1122 | { |
d311d28c PE |
1123 | ptrdiff_t back = XFASTINT (AREF (elt, 1)); |
1124 | ptrdiff_t cpos = charpos - back, bpos; | |
10f72a37 KH |
1125 | |
1126 | if (back == 0) | |
1127 | bpos = bytepos; | |
82ebc97b | 1128 | else |
10f72a37 KH |
1129 | bpos = (NILP (string) ? CHAR_TO_BYTE (cpos) |
1130 | : string_char_to_byte (string, cpos)); | |
1131 | if (STRINGP (AREF (elt, 0))) | |
8f5201ae PE |
1132 | blen = fast_looking_at (AREF (elt, 0), cpos, bpos, |
1133 | start + 1, limit, string); | |
10f72a37 | 1134 | else |
8f5201ae PE |
1135 | blen = 1; |
1136 | if (blen > 0) | |
82ebc97b | 1137 | { |
10f72a37 | 1138 | /* Make CPOS point to the last character of |
8f5201ae PE |
1139 | match. Note that BLEN is byte-length. */ |
1140 | if (blen > 1) | |
10f72a37 | 1141 | { |
8f5201ae | 1142 | bpos += blen; |
10f72a37 KH |
1143 | if (NILP (string)) |
1144 | cpos = BYTE_TO_CHAR (bpos) - 1; | |
1145 | else | |
1146 | cpos = string_byte_to_char (string, bpos) - 1; | |
1147 | } | |
1148 | back = cpos - (charpos - back); | |
1149 | if (cmp_it->stop_pos < cpos | |
1150 | || (cmp_it->stop_pos == cpos | |
1151 | && cmp_it->lookback < back)) | |
1152 | { | |
1153 | cmp_it->rule_idx = ridx; | |
1154 | cmp_it->stop_pos = cpos; | |
1155 | cmp_it->ch = c; | |
1156 | cmp_it->lookback = back; | |
1157 | cmp_it->nchars = back + 1; | |
1158 | } | |
82ebc97b KH |
1159 | } |
1160 | } | |
1161 | } | |
1162 | } | |
1163 | if (charpos - 1 == endpos) | |
1164 | break; | |
1165 | if (STRINGP (string)) | |
58753d74 | 1166 | { |
82ebc97b KH |
1167 | p--, bytepos--; |
1168 | while (! CHAR_HEAD_P (*p)) | |
1169 | p--, bytepos--; | |
1170 | charpos--; | |
1171 | } | |
1172 | else | |
1173 | { | |
1174 | DEC_BOTH (charpos, bytepos); | |
1175 | p = BYTE_POS_ADDR (bytepos); | |
58753d74 | 1176 | } |
82ebc97b KH |
1177 | c = STRING_CHAR (p); |
1178 | } | |
1179 | if (cmp_it->ch >= 0) | |
1180 | /* We found a position to check. */ | |
1181 | return; | |
1182 | /* Skip all uncomposable characters. */ | |
1183 | if (NILP (string)) | |
1184 | { | |
1185 | while (charpos - 1 > endpos && ! CHAR_COMPOSABLE_P (c)) | |
58753d74 | 1186 | { |
82ebc97b KH |
1187 | DEC_BOTH (charpos, bytepos); |
1188 | c = FETCH_MULTIBYTE_CHAR (bytepos); | |
1189 | } | |
1190 | } | |
1191 | else | |
1192 | { | |
1193 | while (charpos - 1 > endpos && ! CHAR_COMPOSABLE_P (c)) | |
1194 | { | |
1195 | p--; | |
1196 | while (! CHAR_HEAD_P (*p)) | |
1197 | p--; | |
1198 | charpos--; | |
1199 | c = STRING_CHAR (p); | |
58753d74 KH |
1200 | } |
1201 | } | |
1202 | } | |
44566dc7 | 1203 | cmp_it->stop_pos = charpos; |
58753d74 KH |
1204 | } |
1205 | ||
1206 | /* Check if the character at CHARPOS (and BYTEPOS) is composed | |
22e33406 | 1207 | (possibly with the following characters) on window W. ENDPOS limits |
58753d74 KH |
1208 | characters to be composed. FACE, in non-NULL, is a base face of |
1209 | the character. If STRING is not nil, it is a string containing the | |
1210 | character to check, and CHARPOS and BYTEPOS are indices in the | |
1211 | string. In that case, FACE must not be NULL. | |
1212 | ||
1213 | If the character is composed, setup members of CMP_IT (id, nglyphs, | |
de1339b0 PE |
1214 | from, to, reversed_p), and return true. Otherwise, update |
1215 | CMP_IT->stop_pos, and return false. */ | |
58753d74 | 1216 | |
de1339b0 PE |
1217 | bool |
1218 | composition_reseat_it (struct composition_it *cmp_it, ptrdiff_t charpos, | |
1219 | ptrdiff_t bytepos, ptrdiff_t endpos, struct window *w, | |
1220 | struct face *face, Lisp_Object string) | |
58753d74 | 1221 | { |
3ffdafce KH |
1222 | if (cmp_it->ch == -2) |
1223 | { | |
1224 | composition_compute_stop_pos (cmp_it, charpos, bytepos, endpos, string); | |
10f72a37 | 1225 | if (cmp_it->ch == -2 || cmp_it->stop_pos != charpos) |
82ebc97b | 1226 | /* The current position is not composed. */ |
3ffdafce KH |
1227 | return 0; |
1228 | } | |
1229 | ||
3cccbd87 KH |
1230 | if (endpos < 0) |
1231 | endpos = NILP (string) ? BEGV : 0; | |
1232 | ||
58753d74 KH |
1233 | if (cmp_it->ch < 0) |
1234 | { | |
1235 | /* We are looking at a static composition. */ | |
d311d28c | 1236 | ptrdiff_t start, end; |
58753d74 KH |
1237 | Lisp_Object prop; |
1238 | ||
1239 | find_composition (charpos, -1, &start, &end, &prop, string); | |
1240 | cmp_it->id = get_composition_id (charpos, bytepos, end - start, | |
1241 | prop, string); | |
1242 | if (cmp_it->id < 0) | |
1243 | goto no_composition; | |
1244 | cmp_it->nchars = end - start; | |
1245 | cmp_it->nglyphs = composition_table[cmp_it->id]->glyph_len; | |
1246 | } | |
29fb7306 | 1247 | else if (w) |
58753d74 | 1248 | { |
10f72a37 | 1249 | Lisp_Object lgstring = Qnil; |
b246f932 | 1250 | Lisp_Object val, elt; |
d311d28c | 1251 | ptrdiff_t i; |
58753d74 KH |
1252 | |
1253 | val = CHAR_TABLE_REF (Vcomposition_function_table, cmp_it->ch); | |
10f72a37 | 1254 | for (i = 0; i < cmp_it->rule_idx; i++, val = XCDR (val)); |
82ebc97b | 1255 | if (charpos < endpos) |
90b3fe91 | 1256 | { |
82ebc97b KH |
1257 | for (; CONSP (val); val = XCDR (val)) |
1258 | { | |
1259 | elt = XCAR (val); | |
10f72a37 KH |
1260 | if (! VECTORP (elt) || ASIZE (elt) != 3 |
1261 | || ! INTEGERP (AREF (elt, 1))) | |
1262 | continue; | |
1263 | if (XFASTINT (AREF (elt, 1)) != cmp_it->lookback) | |
aa3830c4 | 1264 | goto no_composition; |
10f72a37 KH |
1265 | lgstring = autocmp_chars (elt, charpos, bytepos, endpos, |
1266 | w, face, string); | |
1267 | if (composition_gstring_p (lgstring)) | |
82ebc97b | 1268 | break; |
10f72a37 KH |
1269 | lgstring = Qnil; |
1270 | /* Composition failed perhaps because the font doesn't | |
1271 | support sufficient range of characters. Try the | |
1272 | other composition rules if any. */ | |
82ebc97b | 1273 | } |
82ebc97b | 1274 | cmp_it->reversed_p = 0; |
90b3fe91 | 1275 | } |
82ebc97b KH |
1276 | else |
1277 | { | |
d311d28c | 1278 | ptrdiff_t cpos = charpos, bpos = bytepos; |
90b3fe91 | 1279 | |
3cccbd87 KH |
1280 | cmp_it->reversed_p = 1; |
1281 | elt = XCAR (val); | |
1282 | if (cmp_it->lookback > 0) | |
82ebc97b | 1283 | { |
3cccbd87 KH |
1284 | cpos = charpos - cmp_it->lookback; |
1285 | if (STRINGP (string)) | |
1286 | bpos = string_char_to_byte (string, cpos); | |
1287 | else | |
1288 | bpos = CHAR_TO_BYTE (cpos); | |
82ebc97b | 1289 | } |
3cccbd87 KH |
1290 | lgstring = autocmp_chars (elt, cpos, bpos, charpos + 1, w, face, |
1291 | string); | |
1292 | if (! composition_gstring_p (lgstring) | |
1293 | || cpos + LGSTRING_CHAR_LEN (lgstring) - 1 != charpos) | |
1294 | /* Composition failed or didn't cover the current | |
1295 | character. */ | |
1296 | goto no_composition; | |
82ebc97b | 1297 | } |
10f72a37 KH |
1298 | if (NILP (lgstring)) |
1299 | goto no_composition; | |
1300 | if (NILP (LGSTRING_ID (lgstring))) | |
1301 | lgstring = composition_gstring_put_cache (lgstring, -1); | |
1302 | cmp_it->id = XINT (LGSTRING_ID (lgstring)); | |
1303 | for (i = 0; i < LGSTRING_GLYPH_LEN (lgstring); i++) | |
1304 | if (NILP (LGSTRING_GLYPH (lgstring, i))) | |
58753d74 KH |
1305 | break; |
1306 | cmp_it->nglyphs = i; | |
82ebc97b KH |
1307 | cmp_it->from = 0; |
1308 | cmp_it->to = i; | |
58753d74 | 1309 | } |
29fb7306 KH |
1310 | else |
1311 | goto no_composition; | |
58753d74 KH |
1312 | return 1; |
1313 | ||
1314 | no_composition: | |
82ebc97b KH |
1315 | if (charpos == endpos) |
1316 | return 0; | |
1317 | if (charpos < endpos) | |
1318 | { | |
1319 | charpos++; | |
10f72a37 | 1320 | if (NILP (string)) |
82ebc97b | 1321 | INC_POS (bytepos); |
10f72a37 | 1322 | else |
aa3830c4 | 1323 | bytepos += BYTES_BY_CHAR_HEAD (*(SDATA (string) + bytepos)); |
82ebc97b | 1324 | } |
58753d74 | 1325 | else |
82ebc97b KH |
1326 | { |
1327 | charpos--; | |
1328 | /* BYTEPOS is calculated in composition_compute_stop_pos */ | |
1329 | bytepos = -1; | |
1330 | } | |
3cccbd87 KH |
1331 | if (cmp_it->reversed_p) |
1332 | endpos = -1; | |
58753d74 KH |
1333 | composition_compute_stop_pos (cmp_it, charpos, bytepos, endpos, string); |
1334 | return 0; | |
1335 | } | |
1336 | ||
10f72a37 KH |
1337 | /* Update charpos, nchars, nbytes, and width of the current grapheme |
1338 | cluster. | |
1339 | ||
1340 | If the composition is static or automatic in L2R context, the | |
1341 | cluster is identified by CMP_IT->from, and CHARPOS is the position | |
1342 | of the first character of the cluster. In this case, update | |
1343 | CMP_IT->to too. | |
1344 | ||
1345 | If the composition is automatic in R2L context, the cluster is | |
1346 | identified by CMP_IT->to, and CHARPOS is the position of the last | |
1347 | character of the cluster. In this case, update CMP_IT->from too. | |
1348 | ||
1349 | The return value is the character code of the first character of | |
1350 | the cluster, or -1 if the composition is somehow broken. */ | |
82ebc97b | 1351 | |
58753d74 | 1352 | int |
d311d28c | 1353 | composition_update_it (struct composition_it *cmp_it, ptrdiff_t charpos, ptrdiff_t bytepos, Lisp_Object string) |
58753d74 | 1354 | { |
80e079b2 | 1355 | int i, c IF_LINT (= 0); |
58753d74 KH |
1356 | |
1357 | if (cmp_it->ch < 0) | |
1358 | { | |
10f72a37 | 1359 | /* static composition */ |
58753d74 KH |
1360 | struct composition *cmp = composition_table[cmp_it->id]; |
1361 | ||
10f72a37 | 1362 | cmp_it->charpos = charpos; |
58753d74 KH |
1363 | cmp_it->to = cmp_it->nglyphs; |
1364 | if (cmp_it->nglyphs == 0) | |
1365 | c = -1; | |
1366 | else | |
1367 | { | |
1368 | for (i = 0; i < cmp->glyph_len; i++) | |
bd01620e EZ |
1369 | /* TAB in a composition means display glyphs with padding |
1370 | space on the left or right. */ | |
58753d74 KH |
1371 | if ((c = COMPOSITION_GLYPH (cmp, i)) != '\t') |
1372 | break; | |
1373 | if (c == '\t') | |
1374 | c = ' '; | |
1375 | } | |
1376 | cmp_it->width = cmp->width; | |
10f72a37 KH |
1377 | charpos += cmp_it->nchars; |
1378 | if (STRINGP (string)) | |
1379 | cmp_it->nbytes = string_char_to_byte (string, charpos) - bytepos; | |
1380 | else | |
1381 | cmp_it->nbytes = CHAR_TO_BYTE (charpos) - bytepos; | |
58753d74 KH |
1382 | } |
1383 | else | |
1384 | { | |
b7432bb2 | 1385 | /* Automatic composition. */ |
58753d74 | 1386 | Lisp_Object gstring = composition_gstring_from_id (cmp_it->id); |
10f72a37 | 1387 | Lisp_Object glyph; |
d311d28c | 1388 | ptrdiff_t from; |
58753d74 KH |
1389 | |
1390 | if (cmp_it->nglyphs == 0) | |
1391 | { | |
58753d74 KH |
1392 | cmp_it->nchars = LGSTRING_CHAR_LEN (gstring); |
1393 | cmp_it->width = 0; | |
10f72a37 KH |
1394 | cmp_it->from = cmp_it->to = 0; |
1395 | return -1; | |
58753d74 | 1396 | } |
10f72a37 | 1397 | if (! cmp_it->reversed_p) |
58753d74 | 1398 | { |
10f72a37 KH |
1399 | glyph = LGSTRING_GLYPH (gstring, cmp_it->from); |
1400 | from = LGLYPH_FROM (glyph); | |
58753d74 KH |
1401 | for (cmp_it->to = cmp_it->from + 1; cmp_it->to < cmp_it->nglyphs; |
1402 | cmp_it->to++) | |
1403 | { | |
1404 | glyph = LGSTRING_GLYPH (gstring, cmp_it->to); | |
1405 | if (LGLYPH_FROM (glyph) != from) | |
1406 | break; | |
58753d74 | 1407 | } |
10f72a37 | 1408 | cmp_it->charpos = charpos; |
58753d74 | 1409 | } |
82ebc97b KH |
1410 | else |
1411 | { | |
10f72a37 KH |
1412 | glyph = LGSTRING_GLYPH (gstring, cmp_it->to - 1); |
1413 | from = LGLYPH_FROM (glyph); | |
1414 | cmp_it->charpos = charpos - (LGLYPH_TO (glyph) - from); | |
1415 | for (cmp_it->from = cmp_it->to - 1; cmp_it->from > 0; | |
1416 | cmp_it->from--) | |
82ebc97b | 1417 | { |
10f72a37 | 1418 | glyph = LGSTRING_GLYPH (gstring, cmp_it->from - 1); |
82ebc97b KH |
1419 | if (LGLYPH_FROM (glyph) != from) |
1420 | break; | |
82ebc97b | 1421 | } |
10f72a37 KH |
1422 | } |
1423 | glyph = LGSTRING_GLYPH (gstring, cmp_it->from); | |
1424 | cmp_it->nchars = LGLYPH_TO (glyph) + 1 - from; | |
1425 | cmp_it->nbytes = 0; | |
1426 | cmp_it->width = 0; | |
1427 | for (i = cmp_it->nchars - 1; i >= 0; i--) | |
1428 | { | |
1429 | c = XINT (LGSTRING_CHAR (gstring, i)); | |
1430 | cmp_it->nbytes += CHAR_BYTES (c); | |
63257f6c | 1431 | cmp_it->width += CHAR_WIDTH (c); |
82ebc97b | 1432 | } |
58753d74 | 1433 | } |
58753d74 KH |
1434 | return c; |
1435 | } | |
1436 | ||
1437 | ||
90b3fe91 KH |
1438 | struct position_record |
1439 | { | |
d311d28c | 1440 | ptrdiff_t pos, pos_byte; |
90b3fe91 KH |
1441 | unsigned char *p; |
1442 | }; | |
1443 | ||
763d7377 | 1444 | /* Update the members of POSITION to the next character boundary. */ |
90b3fe91 KH |
1445 | #define FORWARD_CHAR(POSITION, STOP) \ |
1446 | do { \ | |
90b3fe91 | 1447 | (POSITION).pos++; \ |
900c4486 KH |
1448 | if ((POSITION).pos == (STOP)) \ |
1449 | { \ | |
1450 | (POSITION).p = GAP_END_ADDR; \ | |
1451 | (POSITION).pos_byte = GPT_BYTE; \ | |
1452 | } \ | |
1453 | else \ | |
1454 | { \ | |
900c4486 | 1455 | (POSITION).pos_byte += BYTES_BY_CHAR_HEAD (*((POSITION).p)); \ |
d45a49e3 | 1456 | (POSITION).p += BYTES_BY_CHAR_HEAD (*((POSITION).p)); \ |
900c4486 | 1457 | } \ |
90b3fe91 KH |
1458 | } while (0) |
1459 | ||
763d7377 | 1460 | /* Update the members of POSITION to the previous character boundary. */ |
90b3fe91 KH |
1461 | #define BACKWARD_CHAR(POSITION, STOP) \ |
1462 | do { \ | |
a28d4396 | 1463 | if ((POSITION).pos == (STOP)) \ |
90b3fe91 KH |
1464 | (POSITION).p = GPT_ADDR; \ |
1465 | do { \ | |
1466 | (POSITION).pos_byte--; \ | |
1467 | (POSITION).p--; \ | |
1468 | } while (! CHAR_HEAD_P (*((POSITION).p))); \ | |
1469 | (POSITION).pos--; \ | |
1470 | } while (0) | |
1471 | ||
90b3fe91 | 1472 | /* This is like find_composition, but find an automatic composition |
a28d4396 KH |
1473 | instead. It is assured that POS is not within a static |
1474 | composition. If found, set *GSTRING to the glyph-string | |
de1339b0 PE |
1475 | representing the composition, and return true. Otherwise, *GSTRING to |
1476 | Qnil, and return false. */ | |
90b3fe91 | 1477 | |
de1339b0 | 1478 | static bool |
d311d28c PE |
1479 | find_automatic_composition (ptrdiff_t pos, ptrdiff_t limit, |
1480 | ptrdiff_t *start, ptrdiff_t *end, | |
a28d4396 | 1481 | Lisp_Object *gstring, Lisp_Object string) |
90b3fe91 | 1482 | { |
d311d28c | 1483 | ptrdiff_t head, tail, stop; |
a28d4396 KH |
1484 | /* Forward limit position of checking a composition taking a |
1485 | looking-back count into account. */ | |
d311d28c | 1486 | ptrdiff_t fore_check_limit; |
a28d4396 | 1487 | struct position_record cur, prev; |
90b3fe91 | 1488 | int c; |
29fb7306 | 1489 | Lisp_Object window; |
90b3fe91 | 1490 | struct window *w; |
de1339b0 | 1491 | bool need_adjustment = 0; |
90b3fe91 | 1492 | |
3eda4b19 | 1493 | window = Fget_buffer_window (Fcurrent_buffer (), Qnil); |
29fb7306 KH |
1494 | if (NILP (window)) |
1495 | return 0; | |
1496 | w = XWINDOW (window); | |
1497 | ||
a28d4396 | 1498 | cur.pos = pos; |
90b3fe91 KH |
1499 | if (NILP (string)) |
1500 | { | |
1501 | head = BEGV, tail = ZV, stop = GPT; | |
a28d4396 KH |
1502 | cur.pos_byte = CHAR_TO_BYTE (cur.pos); |
1503 | cur.p = BYTE_POS_ADDR (cur.pos_byte); | |
90b3fe91 KH |
1504 | } |
1505 | else | |
1506 | { | |
1507 | head = 0, tail = SCHARS (string), stop = -1; | |
a28d4396 KH |
1508 | cur.pos_byte = string_char_to_byte (string, cur.pos); |
1509 | cur.p = SDATA (string) + cur.pos_byte; | |
90b3fe91 | 1510 | } |
a28d4396 KH |
1511 | if (limit < 0) |
1512 | /* Finding a composition covering the character after POS is the | |
1513 | same as setting LIMIT to POS. */ | |
1514 | limit = pos; | |
1515 | if (limit <= pos) | |
1516 | fore_check_limit = min (tail, pos + 1 + MAX_AUTO_COMPOSITION_LOOKBACK); | |
90b3fe91 | 1517 | else |
895416e3 | 1518 | fore_check_limit = min (tail, limit + MAX_AUTO_COMPOSITION_LOOKBACK); |
90b3fe91 | 1519 | |
a28d4396 | 1520 | /* Provided that we have these possible compositions now: |
895416e3 | 1521 | |
a28d4396 KH |
1522 | POS: 1 2 3 4 5 6 7 8 9 |
1523 | |-A-| | |
1524 | |-B-|-C-|--D--| | |
895416e3 | 1525 | |
a28d4396 KH |
1526 | Here, it is known that characters after positions 1 and 9 can |
1527 | never be composed (i.e. ! CHAR_COMPOSABLE_P (CH)), and | |
1528 | composition A is an invalid one because it's partially covered by | |
1529 | the valid composition C. And to know whether a composition is | |
1530 | valid or not, the only way is to start searching forward from a | |
1531 | position that can not be a tail part of composition (it's 2 in | |
1532 | the above case). | |
1533 | ||
1534 | Now we have these cases (1 through 4): | |
1535 | ||
1536 | -- character after POS is ... -- | |
1537 | not composable composable | |
1538 | LIMIT <= POS (1) (3) | |
1539 | POS < LIMIT (2) (4) | |
1540 | ||
1541 | Among them, in case (2), we simply search forward from POS. | |
1542 | ||
1543 | In the other cases, we at first rewind back to the position where | |
1544 | the previous character is not composable or the beginning of | |
1545 | buffer (string), then search compositions forward. In case (1) | |
1546 | and (3) we repeat this process until a composition is found. */ | |
1547 | ||
1548 | while (1) | |
1549 | { | |
62a6e103 | 1550 | c = STRING_CHAR (cur.p); |
90b3fe91 | 1551 | if (! CHAR_COMPOSABLE_P (c)) |
90b3fe91 | 1552 | { |
a28d4396 KH |
1553 | if (limit <= pos) /* case (1) */ |
1554 | { | |
1555 | do { | |
31bfc35c | 1556 | if (cur.pos <= limit) |
a28d4396 KH |
1557 | return 0; |
1558 | BACKWARD_CHAR (cur, stop); | |
1559 | c = STRING_CHAR (cur.p); | |
1560 | } while (! CHAR_COMPOSABLE_P (c)); | |
1561 | fore_check_limit = cur.pos + 1; | |
1562 | } | |
1563 | else /* case (2) */ | |
1564 | /* No need of rewinding back. */ | |
1565 | goto search_forward; | |
1566 | } | |
90b3fe91 | 1567 | |
a28d4396 KH |
1568 | /* Rewind back to the position where we can safely search |
1569 | forward for compositions. It is assured that the character | |
1570 | at cur.pos is composable. */ | |
1571 | while (head < cur.pos) | |
1572 | { | |
1573 | prev = cur; | |
1574 | BACKWARD_CHAR (cur, stop); | |
1575 | c = STRING_CHAR (cur.p); | |
1576 | if (! CHAR_COMPOSABLE_P (c)) | |
90b3fe91 | 1577 | { |
a28d4396 KH |
1578 | cur = prev; |
1579 | break; | |
90b3fe91 | 1580 | } |
a28d4396 KH |
1581 | } |
1582 | ||
1583 | search_forward: | |
1584 | /* Now search forward. */ | |
1585 | *gstring = Qnil; | |
1586 | prev = cur; /* remember the start of searching position. */ | |
1587 | while (cur.pos < fore_check_limit) | |
1588 | { | |
1589 | Lisp_Object val; | |
1590 | ||
1591 | c = STRING_CHAR (cur.p); | |
1592 | for (val = CHAR_TABLE_REF (Vcomposition_function_table, c); | |
1593 | CONSP (val); val = XCDR (val)) | |
90b3fe91 | 1594 | { |
a28d4396 KH |
1595 | Lisp_Object elt = XCAR (val); |
1596 | ||
1597 | if (VECTORP (elt) && ASIZE (elt) == 3 && NATNUMP (AREF (elt, 1))) | |
90b3fe91 | 1598 | { |
a28d4396 KH |
1599 | EMACS_INT check_pos = cur.pos - XFASTINT (AREF (elt, 1)); |
1600 | struct position_record check; | |
1601 | ||
1602 | if (check_pos < head | |
1603 | || (limit <= pos ? pos < check_pos | |
1604 | : limit <= check_pos)) | |
1605 | continue; | |
1606 | for (check = cur; check_pos < check.pos; ) | |
1607 | BACKWARD_CHAR (check, stop); | |
1608 | *gstring = autocmp_chars (elt, check.pos, check.pos_byte, | |
1609 | tail, w, NULL, string); | |
90b3fe91 | 1610 | need_adjustment = 1; |
a28d4396 | 1611 | if (NILP (*gstring)) |
90b3fe91 | 1612 | { |
a28d4396 KH |
1613 | /* As we have called Lisp, there's a possibility |
1614 | that buffer/string is relocated. */ | |
1615 | if (NILP (string)) | |
1616 | cur.p = BYTE_POS_ADDR (cur.pos_byte); | |
1617 | else | |
1618 | cur.p = SDATA (string) + cur.pos_byte; | |
1619 | } | |
1620 | else | |
1621 | { | |
1622 | /* We found a candidate of a target composition. */ | |
90b3fe91 KH |
1623 | *start = check.pos; |
1624 | *end = check.pos + LGSTRING_CHAR_LEN (*gstring); | |
a28d4396 KH |
1625 | if (pos < limit |
1626 | ? pos < *end | |
1627 | : *start <= pos && pos < *end) | |
1628 | /* This is the target composition. */ | |
90b3fe91 KH |
1629 | return 1; |
1630 | cur.pos = *end; | |
a28d4396 KH |
1631 | if (NILP (string)) |
1632 | { | |
1633 | cur.pos_byte = CHAR_TO_BYTE (cur.pos); | |
1634 | cur.p = BYTE_POS_ADDR (cur.pos_byte); | |
1635 | } | |
1636 | else | |
1637 | { | |
1638 | cur.pos_byte = string_char_to_byte (string, cur.pos); | |
1639 | cur.p = SDATA (string) + cur.pos_byte; | |
1640 | } | |
90b3fe91 KH |
1641 | break; |
1642 | } | |
1643 | } | |
1644 | } | |
a28d4396 KH |
1645 | if (! CONSP (val)) |
1646 | /* We found no composition here. */ | |
90b3fe91 | 1647 | FORWARD_CHAR (cur, stop); |
90b3fe91 | 1648 | } |
a28d4396 KH |
1649 | |
1650 | if (pos < limit) /* case (2) and (4)*/ | |
1651 | return 0; | |
1652 | if (! NILP (*gstring)) | |
1653 | return 1; | |
1654 | if (prev.pos == head) | |
1655 | return 0; | |
90b3fe91 | 1656 | cur = prev; |
a28d4396 KH |
1657 | if (need_adjustment) |
1658 | { | |
1659 | if (NILP (string)) | |
1660 | cur.p = BYTE_POS_ADDR (cur.pos_byte); | |
1661 | else | |
1662 | cur.p = SDATA (string) + cur.pos_byte; | |
1663 | } | |
90b3fe91 | 1664 | BACKWARD_CHAR (cur, stop); |
90b3fe91 | 1665 | } |
90b3fe91 KH |
1666 | } |
1667 | ||
e614ea00 KH |
1668 | /* Return the adjusted point provided that point is moved from LAST_PT |
1669 | to NEW_PT. */ | |
1670 | ||
d311d28c PE |
1671 | ptrdiff_t |
1672 | composition_adjust_point (ptrdiff_t last_pt, ptrdiff_t new_pt) | |
58753d74 | 1673 | { |
d311d28c | 1674 | ptrdiff_t i, beg, end; |
90b3fe91 | 1675 | Lisp_Object val; |
58753d74 | 1676 | |
e614ea00 KH |
1677 | if (new_pt == BEGV || new_pt == ZV) |
1678 | return new_pt; | |
58753d74 | 1679 | |
90b3fe91 | 1680 | /* At first check the static composition. */ |
e614ea00 | 1681 | if (get_property_and_range (new_pt, Qcomposition, &val, &beg, &end, Qnil) |
895416e3 KH |
1682 | && COMPOSITION_VALID_P (beg, end, val)) |
1683 | { | |
e614ea00 | 1684 | if (beg < new_pt /* && end > new_pt <- It's always the case. */ |
895416e3 | 1685 | && (last_pt <= beg || last_pt >= end)) |
e614ea00 KH |
1686 | return (new_pt < last_pt ? beg : end); |
1687 | return new_pt; | |
895416e3 | 1688 | } |
58753d74 | 1689 | |
4b4deea2 | 1690 | if (NILP (BVAR (current_buffer, enable_multibyte_characters)) |
d9a7c140 | 1691 | || NILP (Vauto_composition_mode)) |
e614ea00 | 1692 | return new_pt; |
58753d74 | 1693 | |
90b3fe91 | 1694 | /* Next check the automatic composition. */ |
d311d28c | 1695 | if (! find_automatic_composition (new_pt, (ptrdiff_t) -1, &beg, &end, &val, |
e614ea00 KH |
1696 | Qnil) |
1697 | || beg == new_pt) | |
1698 | return new_pt; | |
90b3fe91 | 1699 | for (i = 0; i < LGSTRING_GLYPH_LEN (val); i++) |
58753d74 | 1700 | { |
90b3fe91 | 1701 | Lisp_Object glyph = LGSTRING_GLYPH (val, i); |
58753d74 | 1702 | |
90b3fe91 KH |
1703 | if (NILP (glyph)) |
1704 | break; | |
e614ea00 KH |
1705 | if (beg + LGLYPH_FROM (glyph) == new_pt) |
1706 | return new_pt; | |
1707 | if (beg + LGLYPH_TO (glyph) >= new_pt) | |
1708 | return (new_pt < last_pt | |
90b3fe91 KH |
1709 | ? beg + LGLYPH_FROM (glyph) |
1710 | : beg + LGLYPH_TO (glyph) + 1); | |
58753d74 | 1711 | } |
e614ea00 | 1712 | return new_pt; |
58753d74 KH |
1713 | } |
1714 | ||
1715 | DEFUN ("composition-get-gstring", Fcomposition_get_gstring, | |
1716 | Scomposition_get_gstring, 4, 4, 0, | |
1717 | doc: /* Return a glyph-string for characters between FROM and TO. | |
9d751859 | 1718 | If the glyph string is for graphic display, FONT-OBJECT must be |
58753d74 | 1719 | a font-object to use for those characters. |
f5199465 KH |
1720 | Otherwise (for terminal display), FONT-OBJECT must be a terminal ID, a |
1721 | frame, or nil for the selected frame's terminal device. | |
58753d74 KH |
1722 | |
1723 | If the optional 4th argument STRING is not nil, it is a string | |
1724 | containing the target characters between indices FROM and TO. | |
1725 | ||
9d751859 EZ |
1726 | A glyph-string is a vector containing information about how to display |
1727 | a specific character sequence. The format is: | |
58753d74 KH |
1728 | [HEADER ID GLYPH ...] |
1729 | ||
1730 | HEADER is a vector of this form: | |
1731 | [FONT-OBJECT CHAR ...] | |
1732 | where | |
1733 | FONT-OBJECT is a font-object for all glyphs in the glyph-string, | |
f5199465 | 1734 | or the terminal coding system of the specified terminal. |
58753d74 KH |
1735 | CHARs are characters to be composed by GLYPHs. |
1736 | ||
1737 | ID is an identification number of the glyph-string. It may be nil if | |
1738 | not yet shaped. | |
1739 | ||
9d751859 | 1740 | GLYPH is a vector whose elements have this form: |
58753d74 KH |
1741 | [ FROM-IDX TO-IDX C CODE WIDTH LBEARING RBEARING ASCENT DESCENT |
1742 | [ [X-OFF Y-OFF WADJUST] | nil] ] | |
1743 | where | |
1744 | FROM-IDX and TO-IDX are used internally and should not be touched. | |
1745 | C is the character of the glyph. | |
1746 | CODE is the glyph-code of C in FONT-OBJECT. | |
1747 | WIDTH thru DESCENT are the metrics (in pixels) of the glyph. | |
9d751859 | 1748 | X-OFF and Y-OFF are offsets to the base position for the glyph. |
58753d74 KH |
1749 | WADJUST is the adjustment to the normal width of the glyph. |
1750 | ||
9d751859 EZ |
1751 | If GLYPH is nil, the remaining elements of the glyph-string vector |
1752 | should be ignored. */) | |
5842a27b | 1753 | (Lisp_Object from, Lisp_Object to, Lisp_Object font_object, Lisp_Object string) |
58753d74 KH |
1754 | { |
1755 | Lisp_Object gstring, header; | |
d311d28c | 1756 | ptrdiff_t frompos, topos; |
58753d74 | 1757 | |
ea8ba975 KH |
1758 | CHECK_NATNUM (from); |
1759 | CHECK_NATNUM (to); | |
f5199465 KH |
1760 | if (! FONT_OBJECT_P (font_object)) |
1761 | { | |
1762 | struct coding_system *coding; | |
1763 | struct terminal *terminal = get_terminal (font_object, 1); | |
1764 | ||
1765 | coding = ((TERMINAL_TERMINAL_CODING (terminal)->common_flags | |
1766 | & CODING_REQUIRE_ENCODING_MASK) | |
1767 | ? TERMINAL_TERMINAL_CODING (terminal) : &safe_terminal_coding); | |
1768 | font_object = CODING_ID_NAME (coding->id); | |
1769 | } | |
1770 | ||
58753d74 KH |
1771 | header = fill_gstring_header (Qnil, from, to, font_object, string); |
1772 | gstring = gstring_lookup_cache (header); | |
1773 | if (! NILP (gstring)) | |
1774 | return gstring; | |
46f905e1 | 1775 | |
ea8ba975 KH |
1776 | frompos = XINT (from); |
1777 | topos = XINT (to); | |
46f905e1 SM |
1778 | if (LGSTRING_GLYPH_LEN (gstring_work) < topos - frompos) |
1779 | gstring_work = Fmake_vector (make_number (topos - frompos + 2), Qnil); | |
58753d74 KH |
1780 | LGSTRING_SET_HEADER (gstring_work, header); |
1781 | LGSTRING_SET_ID (gstring_work, Qnil); | |
1782 | fill_gstring_body (gstring_work); | |
1783 | return gstring_work; | |
1784 | } | |
1785 | ||
ca4c9455 KH |
1786 | \f |
1787 | /* Emacs Lisp APIs. */ | |
1788 | ||
1789 | DEFUN ("compose-region-internal", Fcompose_region_internal, | |
1790 | Scompose_region_internal, 2, 4, 0, | |
335c5470 PJ |
1791 | doc: /* Internal use only. |
1792 | ||
1793 | Compose text in the region between START and END. | |
1794 | Optional 3rd and 4th arguments are COMPONENTS and MODIFICATION-FUNC | |
9d751859 | 1795 | for the composition. See `compose-region' for more details. */) |
5842a27b | 1796 | (Lisp_Object start, Lisp_Object end, Lisp_Object components, Lisp_Object modification_func) |
ca4c9455 KH |
1797 | { |
1798 | validate_region (&start, &end); | |
1799 | if (!NILP (components) | |
1800 | && !INTEGERP (components) | |
1801 | && !CONSP (components) | |
1802 | && !STRINGP (components)) | |
b7826503 | 1803 | CHECK_VECTOR (components); |
ca4c9455 | 1804 | |
d562f8ab | 1805 | compose_text (XINT (start), XINT (end), components, modification_func, Qnil); |
ca4c9455 KH |
1806 | return Qnil; |
1807 | } | |
1808 | ||
1809 | DEFUN ("compose-string-internal", Fcompose_string_internal, | |
1810 | Scompose_string_internal, 3, 5, 0, | |
335c5470 PJ |
1811 | doc: /* Internal use only. |
1812 | ||
1813 | Compose text between indices START and END of STRING. | |
1814 | Optional 4th and 5th arguments are COMPONENTS and MODIFICATION-FUNC | |
9d751859 | 1815 | for the composition. See `compose-string' for more details. */) |
5842a27b | 1816 | (Lisp_Object string, Lisp_Object start, Lisp_Object end, Lisp_Object components, Lisp_Object modification_func) |
ca4c9455 | 1817 | { |
b7826503 PJ |
1818 | CHECK_STRING (string); |
1819 | CHECK_NUMBER (start); | |
1820 | CHECK_NUMBER (end); | |
ca4c9455 KH |
1821 | |
1822 | if (XINT (start) < 0 || | |
1823 | XINT (start) > XINT (end) | |
d5db4077 | 1824 | || XINT (end) > SCHARS (string)) |
ca4c9455 KH |
1825 | args_out_of_range (start, end); |
1826 | ||
d562f8ab | 1827 | compose_text (XINT (start), XINT (end), components, modification_func, string); |
ca4c9455 KH |
1828 | return string; |
1829 | } | |
1830 | ||
1831 | DEFUN ("find-composition-internal", Ffind_composition_internal, | |
177c0ea7 | 1832 | Sfind_composition_internal, 4, 4, 0, |
335c5470 PJ |
1833 | doc: /* Internal use only. |
1834 | ||
1835 | Return information about composition at or nearest to position POS. | |
9d751859 | 1836 | See `find-composition' for more details. */) |
5842a27b | 1837 | (Lisp_Object pos, Lisp_Object limit, Lisp_Object string, Lisp_Object detail_p) |
ca4c9455 | 1838 | { |
90b3fe91 | 1839 | Lisp_Object prop, tail, gstring; |
d311d28c | 1840 | ptrdiff_t start, end, from, to; |
ca4c9455 KH |
1841 | int id; |
1842 | ||
b7826503 | 1843 | CHECK_NUMBER_COERCE_MARKER (pos); |
ca4c9455 KH |
1844 | if (!NILP (limit)) |
1845 | { | |
b7826503 | 1846 | CHECK_NUMBER_COERCE_MARKER (limit); |
d311d28c | 1847 | to = min (XINT (limit), ZV); |
ca4c9455 KH |
1848 | } |
1849 | else | |
90b3fe91 | 1850 | to = -1; |
177c0ea7 | 1851 | |
ca4c9455 | 1852 | if (!NILP (string)) |
e3b3e327 | 1853 | { |
b7826503 | 1854 | CHECK_STRING (string); |
d5db4077 | 1855 | if (XINT (pos) < 0 || XINT (pos) > SCHARS (string)) |
e3b3e327 GM |
1856 | args_out_of_range (string, pos); |
1857 | } | |
1858 | else | |
1859 | { | |
fa9090b8 | 1860 | if (XINT (pos) < BEGV || XINT (pos) > ZV) |
e3b3e327 GM |
1861 | args_out_of_range (Fcurrent_buffer (), pos); |
1862 | } | |
d311d28c | 1863 | from = XINT (pos); |
ca4c9455 | 1864 | |
90b3fe91 KH |
1865 | if (!find_composition (from, to, &start, &end, &prop, string)) |
1866 | { | |
4b4deea2 | 1867 | if (!NILP (BVAR (current_buffer, enable_multibyte_characters)) |
d9a7c140 | 1868 | && ! NILP (Vauto_composition_mode) |
90b3fe91 KH |
1869 | && find_automatic_composition (from, to, &start, &end, &gstring, |
1870 | string)) | |
1871 | return list3 (make_number (start), make_number (end), gstring); | |
1872 | return Qnil; | |
1873 | } | |
1874 | if ((end <= XINT (pos) || start > XINT (pos))) | |
1875 | { | |
d311d28c | 1876 | ptrdiff_t s, e; |
90b3fe91 KH |
1877 | |
1878 | if (find_automatic_composition (from, to, &s, &e, &gstring, string) | |
1879 | && (e <= XINT (pos) ? e > end : s < start)) | |
fcebfc6a | 1880 | return list3 (make_number (s), make_number (e), gstring); |
90b3fe91 | 1881 | } |
ca4c9455 KH |
1882 | if (!COMPOSITION_VALID_P (start, end, prop)) |
1883 | return Fcons (make_number (start), Fcons (make_number (end), | |
1884 | Fcons (Qnil, Qnil))); | |
1885 | if (NILP (detail_p)) | |
1886 | return Fcons (make_number (start), Fcons (make_number (end), | |
1887 | Fcons (Qt, Qnil))); | |
1888 | ||
1889 | if (COMPOSITION_REGISTERD_P (prop)) | |
1890 | id = COMPOSITION_ID (prop); | |
1891 | else | |
1892 | { | |
d311d28c | 1893 | ptrdiff_t start_byte = (NILP (string) |
579c18d0 EZ |
1894 | ? CHAR_TO_BYTE (start) |
1895 | : string_char_to_byte (string, start)); | |
ca4c9455 KH |
1896 | id = get_composition_id (start, start_byte, end - start, prop, string); |
1897 | } | |
1898 | ||
1899 | if (id >= 0) | |
1900 | { | |
1901 | Lisp_Object components, relative_p, mod_func; | |
1902 | enum composition_method method = COMPOSITION_METHOD (prop); | |
1903 | int width = composition_table[id]->width; | |
1904 | ||
1905 | components = Fcopy_sequence (COMPOSITION_COMPONENTS (prop)); | |
1906 | relative_p = (method == COMPOSITION_WITH_RULE_ALTCHARS | |
1907 | ? Qnil : Qt); | |
1908 | mod_func = COMPOSITION_MODIFICATION_FUNC (prop); | |
1909 | tail = Fcons (components, | |
1910 | Fcons (relative_p, | |
1911 | Fcons (mod_func, | |
1912 | Fcons (make_number (width), Qnil)))); | |
1913 | } | |
1914 | else | |
1915 | tail = Qnil; | |
1916 | ||
1917 | return Fcons (make_number (start), Fcons (make_number (end), tail)); | |
1918 | } | |
1919 | ||
1920 | \f | |
1921 | void | |
971de7fb | 1922 | syms_of_composite (void) |
ca4c9455 | 1923 | { |
58753d74 KH |
1924 | int i; |
1925 | ||
cd3520a4 | 1926 | DEFSYM (Qcomposition, "composition"); |
ca4c9455 | 1927 | |
58753d74 | 1928 | /* Make a hash table for static composition. */ |
ca4c9455 | 1929 | { |
09654086 | 1930 | Lisp_Object args[6]; |
177c0ea7 | 1931 | |
ca4c9455 KH |
1932 | args[0] = QCtest; |
1933 | args[1] = Qequal; | |
1934 | args[2] = QCweakness; | |
dc47eccc | 1935 | /* We used to make the hash table weak so that unreferenced |
ca101cff | 1936 | compositions can be garbage-collected. But, usually once |
dc47eccc KH |
1937 | created compositions are repeatedly used in an Emacs session, |
1938 | and thus it's not worth to save memory in such a way. So, we | |
1939 | make the table not weak. */ | |
6a83ee8a | 1940 | args[3] = Qnil; |
ca4c9455 KH |
1941 | args[4] = QCsize; |
1942 | args[5] = make_number (311); | |
09654086 | 1943 | composition_hash_table = Fmake_hash_table (6, args); |
ca4c9455 KH |
1944 | staticpro (&composition_hash_table); |
1945 | } | |
1946 | ||
58753d74 KH |
1947 | /* Make a hash table for glyph-string. */ |
1948 | { | |
1949 | Lisp_Object args[6]; | |
58753d74 KH |
1950 | args[0] = QCtest; |
1951 | args[1] = Qequal; | |
1952 | args[2] = QCweakness; | |
1953 | args[3] = Qnil; | |
1954 | args[4] = QCsize; | |
1955 | args[5] = make_number (311); | |
1956 | gstring_hash_table = Fmake_hash_table (6, args); | |
1957 | staticpro (&gstring_hash_table); | |
1958 | } | |
1959 | ||
1960 | staticpro (&gstring_work_headers); | |
1961 | gstring_work_headers = Fmake_vector (make_number (8), Qnil); | |
1962 | for (i = 0; i < 8; i++) | |
1963 | ASET (gstring_work_headers, i, Fmake_vector (make_number (i + 2), Qnil)); | |
1964 | staticpro (&gstring_work); | |
1965 | gstring_work = Fmake_vector (make_number (10), Qnil); | |
1966 | ||
ca4c9455 KH |
1967 | /* Text property `composition' should be nonsticky by default. */ |
1968 | Vtext_property_default_nonsticky | |
1969 | = Fcons (Fcons (Qcomposition, Qt), Vtext_property_default_nonsticky); | |
1970 | ||
29208e82 | 1971 | DEFVAR_LISP ("compose-chars-after-function", Vcompose_chars_after_function, |
335c5470 PJ |
1972 | doc: /* Function to adjust composition of buffer text. |
1973 | ||
9d751859 EZ |
1974 | This function is called with three arguments: FROM, TO, and OBJECT. |
1975 | FROM and TO specify the range of text whose composition should be | |
335c5470 PJ |
1976 | adjusted. OBJECT, if non-nil, is a string that contains the text. |
1977 | ||
1978 | This function is called after a text with `composition' property is | |
1979 | inserted or deleted to keep `composition' property of buffer text | |
1980 | valid. | |
1981 | ||
1982 | The default value is the function `compose-chars-after'. */); | |
d67b4f80 | 1983 | Vcompose_chars_after_function = intern_c_string ("compose-chars-after"); |
ca4c9455 | 1984 | |
cd3520a4 JB |
1985 | DEFSYM (Qauto_composed, "auto-composed"); |
1986 | DEFSYM (Qauto_composition_function, "auto-composition-function"); | |
40add26d | 1987 | |
29208e82 | 1988 | DEFVAR_LISP ("auto-composition-mode", Vauto_composition_mode, |
d9a7c140 KH |
1989 | doc: /* Non-nil if Auto-Composition mode is enabled. |
1990 | Use the command `auto-composition-mode' to change this variable. */); | |
1991 | Vauto_composition_mode = Qt; | |
1992 | ||
29208e82 | 1993 | DEFVAR_LISP ("auto-composition-function", Vauto_composition_function, |
f96ba4c1 | 1994 | doc: /* Function to call to compose characters automatically. |
9d751859 | 1995 | This function is called from the display routine with four arguments: |
ddc872ba | 1996 | FROM, TO, WINDOW, and STRING. |
335c5470 | 1997 | |
ddc872ba KH |
1998 | If STRING is nil, the function must compose characters in the region |
1999 | between FROM and TO in the current buffer. | |
335c5470 | 2000 | |
ddc872ba KH |
2001 | Otherwise, STRING is a string, and FROM and TO are indices into the |
2002 | string. In this case, the function must compose characters in the | |
2003 | string. */); | |
f96ba4c1 | 2004 | Vauto_composition_function = Qnil; |
40add26d | 2005 | |
29208e82 | 2006 | DEFVAR_LISP ("composition-function-table", Vcomposition_function_table, |
9d751859 | 2007 | doc: /* Char-table of functions for automatic character composition. |
58753d74 KH |
2008 | For each character that has to be composed automatically with |
2009 | preceding and/or following characters, this char-table contains | |
2010 | a function to call to compose that character. | |
2011 | ||
2012 | The element at index C in the table, if non-nil, is a list of | |
10f72a37 | 2013 | composition rules of this form: ([PATTERN PREV-CHARS FUNC] ...) |
58753d74 | 2014 | |
9d751859 | 2015 | PATTERN is a regular expression which C and the surrounding |
58753d74 KH |
2016 | characters must match. |
2017 | ||
895416e3 KH |
2018 | PREV-CHARS is a non-negative integer (less than 4) specifying how many |
2019 | characters before C to check the matching with PATTERN. If it is 0, | |
2020 | PATTERN must match C and the following characters. If it is 1, | |
2021 | PATTERN must match a character before C and the following characters. | |
58753d74 KH |
2022 | |
2023 | If PREV-CHARS is 0, PATTERN can be nil, which means that the | |
2024 | single character C should be composed. | |
2025 | ||
2026 | FUNC is a function to return a glyph-string representing a | |
9d751859 | 2027 | composition of the characters that match PATTERN. It is |
58753d74 KH |
2028 | called with one argument GSTRING. |
2029 | ||
2030 | GSTRING is a template of a glyph-string to return. It is already | |
2031 | filled with a proper header for the characters to compose, and | |
2032 | glyphs corresponding to those characters one by one. The | |
9d751859 | 2033 | function must return a new glyph-string with the same header as |
58753d74 KH |
2034 | GSTRING, or modify GSTRING itself and return it. |
2035 | ||
2036 | See also the documentation of `auto-composition-mode'. */); | |
2037 | Vcomposition_function_table = Fmake_char_table (Qnil, Qnil); | |
2038 | ||
ca4c9455 KH |
2039 | defsubr (&Scompose_region_internal); |
2040 | defsubr (&Scompose_string_internal); | |
2041 | defsubr (&Sfind_composition_internal); | |
58753d74 | 2042 | defsubr (&Scomposition_get_gstring); |
ca4c9455 | 2043 | } |