Debian Codegraph @

podofo/3rdparty/utf8proc/utf8proc.c

raw
1 /* -*- mode: c; c-basic-offset: 2; tab-width: 2; indent-tabs-mode: nil -*- */
2 /*
3 * Copyright (c) 2014-2021 Steven G. Johnson, Jiahao Chen, Peter Colberg, Tony Kelman, Scott P. Jones, and other contributors.
4 * Copyright (c) 2009 Public Software Group e. V., Berlin, Germany
5 *
6 * Permission is hereby granted, free of charge, to any person obtaining a
7 * copy of this software and associated documentation files (the "Software"),
8 * to deal in the Software without restriction, including without limitation
9 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
10 * and/or sell copies of the Software, and to permit persons to whom the
11 * Software is furnished to do so, subject to the following conditions:
12 *
13 * The above copyright notice and this permission notice shall be included in
14 * all copies or substantial portions of the Software.
15 *
16 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
17 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
18 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
19 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
20 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
21 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
22 * DEALINGS IN THE SOFTWARE.
23 */
24
25 /*
26 * This library contains derived data from a modified version of the
27 * Unicode data files.
28 *
29 * The original data files are available at
30 * https://www.unicode.org/Public/UNIDATA/
31 *
32 * Please notice the copyright statement in the file "utf8proc_data.c".
33 */
34
35
36 /*
37 * File name: utf8proc.c
38 *
39 * Description:
40 * Implementation of libutf8proc.
41 */
42
43
44 #include "utf8proc.h"
45
46 #ifndef SSIZE_MAX
47 #define SSIZE_MAX ((size_t)SIZE_MAX/2)
48 #endif
49 #ifndef UINT16_MAX
50 # define UINT16_MAX 65535U
51 #endif
52
53 #include "utf8proc_data.c"
54
55
56 UTF8PROC_DLLEXPORT const utf8proc_int8_t utf8proc_utf8class[256] = {
57 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
58 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
59 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
60 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
61 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
62 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
63 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
64 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
65 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
66 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
67 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
68 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
69 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
70 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
71 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
72 4, 4, 4, 4, 4, 4, 4, 4, 0, 0, 0, 0, 0, 0, 0, 0 };
73
74 #define UTF8PROC_HANGUL_SBASE 0xAC00
75 #define UTF8PROC_HANGUL_LBASE 0x1100
76 #define UTF8PROC_HANGUL_VBASE 0x1161
77 #define UTF8PROC_HANGUL_TBASE 0x11A7
78 #define UTF8PROC_HANGUL_LCOUNT 19
79 #define UTF8PROC_HANGUL_VCOUNT 21
80 #define UTF8PROC_HANGUL_TCOUNT 28
81 #define UTF8PROC_HANGUL_NCOUNT 588
82 #define UTF8PROC_HANGUL_SCOUNT 11172
83 /* END is exclusive */
84 #define UTF8PROC_HANGUL_L_START 0x1100
85 #define UTF8PROC_HANGUL_L_END 0x115A
86 #define UTF8PROC_HANGUL_L_FILLER 0x115F
87 #define UTF8PROC_HANGUL_V_START 0x1160
88 #define UTF8PROC_HANGUL_V_END 0x11A3
89 #define UTF8PROC_HANGUL_T_START 0x11A8
90 #define UTF8PROC_HANGUL_T_END 0x11FA
91 #define UTF8PROC_HANGUL_S_START 0xAC00
92 #define UTF8PROC_HANGUL_S_END 0xD7A4
93
94 /* Should follow semantic-versioning rules (semver.org) based on API
95 compatibility. (Note that the shared-library version number will
96 be different, being based on ABI compatibility.): */
97 #define STRINGIZEx(x) #x
98 #define STRINGIZE(x) STRINGIZEx(x)
99 UTF8PROC_DLLEXPORT const char *utf8proc_version(void) {
100 return STRINGIZE(UTF8PROC_VERSION_MAJOR) "." STRINGIZE(UTF8PROC_VERSION_MINOR) "." STRINGIZE(UTF8PROC_VERSION_PATCH) "";
101 }
102
103 UTF8PROC_DLLEXPORT const char *utf8proc_unicode_version(void) {
104 return "17.0.0";
105 }
106
107 UTF8PROC_DLLEXPORT const char *utf8proc_errmsg(utf8proc_ssize_t errcode) {
108 switch (errcode) {
109 case UTF8PROC_ERROR_NOMEM:
110 return "Memory for processing UTF-8 data could not be allocated.";
111 case UTF8PROC_ERROR_OVERFLOW:
112 return "UTF-8 string is too long to be processed.";
113 case UTF8PROC_ERROR_INVALIDUTF8:
114 return "Invalid UTF-8 string";
115 case UTF8PROC_ERROR_NOTASSIGNED:
116 return "Unassigned Unicode code point found in UTF-8 string.";
117 case UTF8PROC_ERROR_INVALIDOPTS:
118 return "Invalid options for UTF-8 processing chosen.";
119 default:
120 return "An unknown error occurred while processing UTF-8 data.";
121 }
122 }
123
124 #define utf_cont(ch) (((ch) & 0xc0) == 0x80)
125 UTF8PROC_DLLEXPORT utf8proc_ssize_t utf8proc_iterate(
126 const utf8proc_uint8_t *str, utf8proc_ssize_t strlen, utf8proc_int32_t *dst
127 ) {
128 utf8proc_int32_t uc;
129 const utf8proc_uint8_t *end;
130
131 *dst = -1;
132 if (!strlen) return 0;
133 end = str + ((strlen < 0) ? 4 : strlen);
134 uc = *str++;
135 if (uc < 0x80) {
136 *dst = uc;
137 return 1;
138 }
139 // Must be between 0xc2 and 0xf4 inclusive to be valid
140 if ((utf8proc_uint32_t)(uc - 0xc2) > (0xf4-0xc2)) return UTF8PROC_ERROR_INVALIDUTF8;
141 if (uc < 0xe0) { // 2-byte sequence
142 // Must have valid continuation character
143 if (str >= end || !utf_cont(*str)) return UTF8PROC_ERROR_INVALIDUTF8;
144 *dst = ((uc & 0x1f)<<6) | (*str & 0x3f);
145 return 2;
146 }
147 if (uc < 0xf0) { // 3-byte sequence
148 if ((str + 1 >= end) || !utf_cont(*str) || !utf_cont(str[1]))
149 return UTF8PROC_ERROR_INVALIDUTF8;
150 // Check for surrogate chars
151 if (uc == 0xed && *str > 0x9f)
152 return UTF8PROC_ERROR_INVALIDUTF8;
153 uc = ((uc & 0xf)<<12) | ((*str & 0x3f)<<6) | (str[1] & 0x3f);
154 if (uc < 0x800)
155 return UTF8PROC_ERROR_INVALIDUTF8;
156 *dst = uc;
157 return 3;
158 }
159 // 4-byte sequence
160 // Must have 3 valid continuation characters
161 if ((str + 2 >= end) || !utf_cont(*str) || !utf_cont(str[1]) || !utf_cont(str[2]))
162 return UTF8PROC_ERROR_INVALIDUTF8;
163 // Make sure in correct range (0x10000 - 0x10ffff)
164 if (uc == 0xf0) {
165 if (*str < 0x90) return UTF8PROC_ERROR_INVALIDUTF8;
166 } else if (uc == 0xf4) {
167 if (*str > 0x8f) return UTF8PROC_ERROR_INVALIDUTF8;
168 }
169 *dst = ((uc & 7)<<18) | ((*str & 0x3f)<<12) | ((str[1] & 0x3f)<<6) | (str[2] & 0x3f);
170 return 4;
171 }
172
173 UTF8PROC_DLLEXPORT utf8proc_bool utf8proc_codepoint_valid(utf8proc_int32_t uc) {
174 return (((utf8proc_uint32_t)uc)-0xd800 > 0x07ff) && ((utf8proc_uint32_t)uc < 0x110000);
175 }
176
177 UTF8PROC_DLLEXPORT utf8proc_ssize_t utf8proc_encode_char(utf8proc_int32_t uc, utf8proc_uint8_t *dst) {
178 if (uc < 0x00) {
179 return 0;
180 } else if (uc < 0x80) {
181 dst[0] = (utf8proc_uint8_t) uc;
182 return 1;
183 } else if (uc < 0x800) {
184 dst[0] = (utf8proc_uint8_t)(0xC0 + (uc >> 6));
185 dst[1] = (utf8proc_uint8_t)(0x80 + (uc & 0x3F));
186 return 2;
187 // Note: we allow encoding 0xd800-0xdfff here, so as not to change
188 // the API, however, these are actually invalid in UTF-8
189 } else if (uc < 0x10000) {
190 dst[0] = (utf8proc_uint8_t)(0xE0 + (uc >> 12));
191 dst[1] = (utf8proc_uint8_t)(0x80 + ((uc >> 6) & 0x3F));
192 dst[2] = (utf8proc_uint8_t)(0x80 + (uc & 0x3F));
193 return 3;
194 } else if (uc < 0x110000) {
195 dst[0] = (utf8proc_uint8_t)(0xF0 + (uc >> 18));
196 dst[1] = (utf8proc_uint8_t)(0x80 + ((uc >> 12) & 0x3F));
197 dst[2] = (utf8proc_uint8_t)(0x80 + ((uc >> 6) & 0x3F));
198 dst[3] = (utf8proc_uint8_t)(0x80 + (uc & 0x3F));
199 return 4;
200 } else return 0;
201 }
202
203 /* internal version used for inserting 0xff bytes between graphemes */
204 static utf8proc_ssize_t charbound_encode_char(utf8proc_int32_t uc, utf8proc_uint8_t *dst) {
205 if (uc < 0x00) {
206 if (uc == -1) { /* internal value used for grapheme breaks */
207 dst[0] = (utf8proc_uint8_t)0xFF;
208 return 1;
209 }
210 return 0;
211 } else if (uc < 0x80) {
212 dst[0] = (utf8proc_uint8_t)uc;
213 return 1;
214 } else if (uc < 0x800) {
215 dst[0] = (utf8proc_uint8_t)(0xC0 + (uc >> 6));
216 dst[1] = (utf8proc_uint8_t)(0x80 + (uc & 0x3F));
217 return 2;
218 } else if (uc < 0x10000) {
219 dst[0] = (utf8proc_uint8_t)(0xE0 + (uc >> 12));
220 dst[1] = (utf8proc_uint8_t)(0x80 + ((uc >> 6) & 0x3F));
221 dst[2] = (utf8proc_uint8_t)(0x80 + (uc & 0x3F));
222 return 3;
223 } else if (uc < 0x110000) {
224 dst[0] = (utf8proc_uint8_t)(0xF0 + (uc >> 18));
225 dst[1] = (utf8proc_uint8_t)(0x80 + ((uc >> 12) & 0x3F));
226 dst[2] = (utf8proc_uint8_t)(0x80 + ((uc >> 6) & 0x3F));
227 dst[3] = (utf8proc_uint8_t)(0x80 + (uc & 0x3F));
228 return 4;
229 } else return 0;
230 }
231
232 /* internal "unsafe" version that does not check whether uc is in range */
233 static const utf8proc_property_t *unsafe_get_property(utf8proc_int32_t uc) {
234 /* ASSERT: uc >= 0 && uc < 0x110000 */
235 return utf8proc_properties + (
236 utf8proc_stage2table[
237 utf8proc_stage1table[uc >> 8] + (uc & 0xFF)
238 ]
239 );
240 }
241
242 UTF8PROC_DLLEXPORT const utf8proc_property_t *utf8proc_get_property(utf8proc_int32_t uc) {
243 return uc < 0 || uc >= 0x110000 ? utf8proc_properties : unsafe_get_property(uc);
244 }
245
246 /* return whether there is a grapheme break between boundclasses lbc and tbc
247 (according to the definition of extended grapheme clusters)
248
249 Rule numbering refers to TR29 Version 29 (Unicode 9.0.0):
250 http://www.unicode.org/reports/tr29/tr29-29.html
251
252 CAVEATS:
253 Please note that evaluation of GB10 (grapheme breaks between emoji zwj sequences)
254 and GB 12/13 (regional indicator code points) require knowledge of previous characters
255 and are thus not handled by this function. This may result in an incorrect break before
256 an E_Modifier class codepoint and an incorrectly missing break between two
257 REGIONAL_INDICATOR class code points if such support does not exist in the caller.
258
259 See the special support in grapheme_break_extended, for required bookkeeping by the caller.
260 */
261 static utf8proc_bool grapheme_break_simple(int lbc, int tbc) {
262 return
263 (lbc == UTF8PROC_BOUNDCLASS_START) ? true : // GB1
264 (lbc == UTF8PROC_BOUNDCLASS_CR && // GB3
265 tbc == UTF8PROC_BOUNDCLASS_LF) ? false : // ---
266 (lbc >= UTF8PROC_BOUNDCLASS_CR && lbc <= UTF8PROC_BOUNDCLASS_CONTROL) ? true : // GB4
267 (tbc >= UTF8PROC_BOUNDCLASS_CR && tbc <= UTF8PROC_BOUNDCLASS_CONTROL) ? true : // GB5
268 (lbc == UTF8PROC_BOUNDCLASS_L && // GB6
269 (tbc == UTF8PROC_BOUNDCLASS_L || // ---
270 tbc == UTF8PROC_BOUNDCLASS_V || // ---
271 tbc == UTF8PROC_BOUNDCLASS_LV || // ---
272 tbc == UTF8PROC_BOUNDCLASS_LVT)) ? false : // ---
273 ((lbc == UTF8PROC_BOUNDCLASS_LV || // GB7
274 lbc == UTF8PROC_BOUNDCLASS_V) && // ---
275 (tbc == UTF8PROC_BOUNDCLASS_V || // ---
276 tbc == UTF8PROC_BOUNDCLASS_T)) ? false : // ---
277 ((lbc == UTF8PROC_BOUNDCLASS_LVT || // GB8
278 lbc == UTF8PROC_BOUNDCLASS_T) && // ---
279 tbc == UTF8PROC_BOUNDCLASS_T) ? false : // ---
280 (tbc == UTF8PROC_BOUNDCLASS_EXTEND || // GB9
281 tbc == UTF8PROC_BOUNDCLASS_ZWJ || // ---
282 tbc == UTF8PROC_BOUNDCLASS_SPACINGMARK || // GB9a
283 lbc == UTF8PROC_BOUNDCLASS_PREPEND) ? false : // GB9b
284 (lbc == UTF8PROC_BOUNDCLASS_E_ZWG && // GB11 (requires additional handling below)
285 tbc == UTF8PROC_BOUNDCLASS_EXTENDED_PICTOGRAPHIC) ? false : // ----
286 (lbc == UTF8PROC_BOUNDCLASS_REGIONAL_INDICATOR && // GB12/13 (requires additional handling below)
287 tbc == UTF8PROC_BOUNDCLASS_REGIONAL_INDICATOR) ? false : // ----
288 true; // GB999
289 }
290
291 static utf8proc_bool grapheme_break_extended(int lbc, int tbc, int licb, int ticb, utf8proc_int32_t *state)
292 {
293 if (state) {
294 int state_bc, state_icb; /* boundclass and indic_conjunct_break state */
295 if (*state == 0) { /* state initialization */
296 state_bc = lbc;
297 state_icb = licb == UTF8PROC_INDIC_CONJUNCT_BREAK_CONSONANT ? licb : UTF8PROC_INDIC_CONJUNCT_BREAK_NONE;
298 }
299 else { /* lbc and licb are already encoded in *state */
300 state_bc = *state & 0xff; // 1st byte of state is bound class
301 state_icb = *state >> 8; // 2nd byte of state is indic conjunct break
302 }
303
304 utf8proc_bool break_permitted = grapheme_break_simple(state_bc, tbc) &&
305 !(state_icb == UTF8PROC_INDIC_CONJUNCT_BREAK_LINKER
306 && ticb == UTF8PROC_INDIC_CONJUNCT_BREAK_CONSONANT); // GB9c
307
308 // Special support for GB9c. Don't break between two consonants
309 // separated 1+ linker characters and 0+ extend characters in any order.
310 // After a consonant, we enter LINKER state after at least one linker.
311 if (ticb == UTF8PROC_INDIC_CONJUNCT_BREAK_CONSONANT
312 || state_icb == UTF8PROC_INDIC_CONJUNCT_BREAK_CONSONANT
313 || state_icb == UTF8PROC_INDIC_CONJUNCT_BREAK_EXTEND)
314 state_icb = ticb;
315 else if (state_icb == UTF8PROC_INDIC_CONJUNCT_BREAK_LINKER)
316 state_icb = ticb == UTF8PROC_INDIC_CONJUNCT_BREAK_EXTEND ?
317 UTF8PROC_INDIC_CONJUNCT_BREAK_LINKER : ticb;
318
319 // Special support for GB 12/13 made possible by GB999. After two RI
320 // class codepoints we want to force a break. Do this by resetting the
321 // second RI's bound class to UTF8PROC_BOUNDCLASS_OTHER, to force a break
322 // after that character according to GB999 (unless of course such a break is
323 // forbidden by a different rule such as GB9).
324 if (state_bc == tbc && tbc == UTF8PROC_BOUNDCLASS_REGIONAL_INDICATOR)
325 state_bc = UTF8PROC_BOUNDCLASS_OTHER;
326 // Special support for GB11 (emoji extend* zwj / emoji)
327 else if (state_bc == UTF8PROC_BOUNDCLASS_EXTENDED_PICTOGRAPHIC) {
328 if (tbc == UTF8PROC_BOUNDCLASS_EXTEND) // fold EXTEND codepoints into emoji
329 state_bc = UTF8PROC_BOUNDCLASS_EXTENDED_PICTOGRAPHIC;
330 else if (tbc == UTF8PROC_BOUNDCLASS_ZWJ)
331 state_bc = UTF8PROC_BOUNDCLASS_E_ZWG; // state to record emoji+zwg combo
332 else
333 state_bc = tbc;
334 }
335 else
336 state_bc = tbc;
337
338 *state = state_bc + (state_icb << 8);
339 return break_permitted;
340 }
341 else
342 return grapheme_break_simple(lbc, tbc);
343 }
344
345 UTF8PROC_DLLEXPORT utf8proc_bool utf8proc_grapheme_break_stateful(
346 utf8proc_int32_t c1, utf8proc_int32_t c2, utf8proc_int32_t *state) {
347
348 const utf8proc_property_t *p1 = utf8proc_get_property(c1);
349 const utf8proc_property_t *p2 = utf8proc_get_property(c2);
350 return grapheme_break_extended(p1->boundclass,
351 p2->boundclass,
352 p1->indic_conjunct_break,
353 p2->indic_conjunct_break,
354 state);
355 }
356
357
358 UTF8PROC_DLLEXPORT utf8proc_bool utf8proc_grapheme_break(
359 utf8proc_int32_t c1, utf8proc_int32_t c2) {
360 return utf8proc_grapheme_break_stateful(c1, c2, NULL);
361 }
362
363 static utf8proc_int32_t seqindex_decode_entry(const utf8proc_uint16_t **entry)
364 {
365 utf8proc_int32_t entry_cp = **entry;
366 if ((entry_cp & 0xF800) == 0xD800) {
367 *entry = *entry + 1;
368 entry_cp = ((entry_cp & 0x03FF) << 10) | (**entry & 0x03FF);
369 entry_cp += 0x10000;
370 }
371 return entry_cp;
372 }
373
374 static utf8proc_int32_t seqindex_decode_index(const utf8proc_uint32_t seqindex)
375 {
376 const utf8proc_uint16_t *entry = &utf8proc_sequences[seqindex];
377 return seqindex_decode_entry(&entry);
378 }
379
380 static utf8proc_ssize_t seqindex_write_char_decomposed(utf8proc_uint16_t seqindex, utf8proc_int32_t *dst, utf8proc_ssize_t bufsize, utf8proc_option_t options, int *last_boundclass) {
381 utf8proc_ssize_t written = 0;
382 const utf8proc_uint16_t *entry = &utf8proc_sequences[seqindex & 0x3FFF];
383 int len = seqindex >> 14;
384 if (len >= 3) {
385 len = *entry;
386 entry++;
387 }
388 for (; len >= 0; entry++, len--) {
389 utf8proc_int32_t entry_cp = seqindex_decode_entry(&entry);
390
391 written += utf8proc_decompose_char(entry_cp, dst ? dst+written : dst,
392 (bufsize > written) ? (bufsize - written) : 0, options,
393 last_boundclass);
394 if (written < 0) return UTF8PROC_ERROR_OVERFLOW;
395 }
396 return written;
397 }
398
399 UTF8PROC_DLLEXPORT utf8proc_int32_t utf8proc_tolower(utf8proc_int32_t c)
400 {
401 utf8proc_int32_t cl = utf8proc_get_property(c)->lowercase_seqindex;
402 return cl != UINT16_MAX ? seqindex_decode_index((utf8proc_uint32_t)cl) : c;
403 }
404
405 UTF8PROC_DLLEXPORT utf8proc_int32_t utf8proc_toupper(utf8proc_int32_t c)
406 {
407 utf8proc_int32_t cu = utf8proc_get_property(c)->uppercase_seqindex;
408 return cu != UINT16_MAX ? seqindex_decode_index((utf8proc_uint32_t)cu) : c;
409 }
410
411 UTF8PROC_DLLEXPORT utf8proc_int32_t utf8proc_totitle(utf8proc_int32_t c)
412 {
413 utf8proc_int32_t cu = utf8proc_get_property(c)->titlecase_seqindex;
414 return cu != UINT16_MAX ? seqindex_decode_index((utf8proc_uint32_t)cu) : c;
415 }
416
417 UTF8PROC_DLLEXPORT int utf8proc_islower(utf8proc_int32_t c)
418 {
419 const utf8proc_property_t *p = utf8proc_get_property(c);
420 return p->lowercase_seqindex != p->uppercase_seqindex && p->lowercase_seqindex == UINT16_MAX;
421 }
422
423 UTF8PROC_DLLEXPORT int utf8proc_isupper(utf8proc_int32_t c)
424 {
425 const utf8proc_property_t *p = utf8proc_get_property(c);
426 return p->lowercase_seqindex != p->uppercase_seqindex && p->uppercase_seqindex == UINT16_MAX && p->category != UTF8PROC_CATEGORY_LT;
427 }
428
429 /* return a character width analogous to wcwidth (except portable and
430 hopefully less buggy than most system wcwidth functions). */
431 UTF8PROC_DLLEXPORT int utf8proc_charwidth(utf8proc_int32_t c) {
432 return utf8proc_get_property(c)->charwidth;
433 }
434
435 UTF8PROC_DLLEXPORT utf8proc_bool utf8proc_charwidth_ambiguous(utf8proc_int32_t c) {
436 return utf8proc_get_property(c)->ambiguous_width;
437 }
438
439 UTF8PROC_DLLEXPORT utf8proc_category_t utf8proc_category(utf8proc_int32_t c) {
440 return (utf8proc_category_t) utf8proc_get_property(c)->category;
441 }
442
443 UTF8PROC_DLLEXPORT const char *utf8proc_category_string(utf8proc_int32_t c) {
444 static const char s[][3] = {"Cn","Lu","Ll","Lt","Lm","Lo","Mn","Mc","Me","Nd","Nl","No","Pc","Pd","Ps","Pe","Pi","Pf","Po","Sm","Sc","Sk","So","Zs","Zl","Zp","Cc","Cf","Cs","Co"};
445 return s[utf8proc_category(c)];
446 }
447
448 #define utf8proc_decompose_lump(replacement_uc) \
449 return utf8proc_decompose_char((replacement_uc), dst, bufsize, \
450 (utf8proc_option_t)(options & ~(unsigned int)UTF8PROC_LUMP), last_boundclass)
451
452 UTF8PROC_DLLEXPORT utf8proc_ssize_t utf8proc_decompose_char(utf8proc_int32_t uc, utf8proc_int32_t *dst, utf8proc_ssize_t bufsize, utf8proc_option_t options, int *last_boundclass) {
453 const utf8proc_property_t *property;
454 utf8proc_propval_t category;
455 utf8proc_int32_t hangul_sindex;
456 if (uc < 0 || uc >= 0x110000) return UTF8PROC_ERROR_NOTASSIGNED;
457 property = unsafe_get_property(uc);
458 category = property->category;
459 hangul_sindex = uc - UTF8PROC_HANGUL_SBASE;
460 if (options & (UTF8PROC_COMPOSE|UTF8PROC_DECOMPOSE)) {
461 if (hangul_sindex >= 0 && hangul_sindex < UTF8PROC_HANGUL_SCOUNT) {
462 utf8proc_int32_t hangul_tindex;
463 if (bufsize >= 1) {
464 dst[0] = UTF8PROC_HANGUL_LBASE +
465 hangul_sindex / UTF8PROC_HANGUL_NCOUNT;
466 if (bufsize >= 2) dst[1] = UTF8PROC_HANGUL_VBASE +
467 (hangul_sindex % UTF8PROC_HANGUL_NCOUNT) / UTF8PROC_HANGUL_TCOUNT;
468 }
469 hangul_tindex = hangul_sindex % UTF8PROC_HANGUL_TCOUNT;
470 if (!hangul_tindex) return 2;
471 if (bufsize >= 3) dst[2] = UTF8PROC_HANGUL_TBASE + hangul_tindex;
472 return 3;
473 }
474 }
475 if (options & UTF8PROC_REJECTNA) {
476 if (!category) return UTF8PROC_ERROR_NOTASSIGNED;
477 }
478 if (options & UTF8PROC_IGNORE) {
479 if (property->ignorable) return 0;
480 }
481 if (options & UTF8PROC_STRIPNA) {
482 if (!category) return 0;
483 }
484 if (options & UTF8PROC_LUMP) {
485 if (category == UTF8PROC_CATEGORY_ZS) utf8proc_decompose_lump(0x0020);
486 if (uc == 0x2018 || uc == 0x2019 || uc == 0x02BC || uc == 0x02C8)
487 utf8proc_decompose_lump(0x0027);
488 if (category == UTF8PROC_CATEGORY_PD || uc == 0x2212)
489 utf8proc_decompose_lump(0x002D);
490 if (uc == 0x2044 || uc == 0x2215) utf8proc_decompose_lump(0x002F);
491 if (uc == 0x2236) utf8proc_decompose_lump(0x003A);
492 if (uc == 0x2039 || uc == 0x2329 || uc == 0x3008)
493 utf8proc_decompose_lump(0x003C);
494 if (uc == 0x203A || uc == 0x232A || uc == 0x3009)
495 utf8proc_decompose_lump(0x003E);
496 if (uc == 0x2216) utf8proc_decompose_lump(0x005C);
497 if (uc == 0x02C4 || uc == 0x02C6 || uc == 0x2038 || uc == 0x2303)
498 utf8proc_decompose_lump(0x005E);
499 if (category == UTF8PROC_CATEGORY_PC || uc == 0x02CD)
500 utf8proc_decompose_lump(0x005F);
501 if (uc == 0x02CB) utf8proc_decompose_lump(0x0060);
502 if (uc == 0x2223) utf8proc_decompose_lump(0x007C);
503 if (uc == 0x223C) utf8proc_decompose_lump(0x007E);
504 if ((options & UTF8PROC_NLF2LS) && (options & UTF8PROC_NLF2PS)) {
505 if (category == UTF8PROC_CATEGORY_ZL ||
506 category == UTF8PROC_CATEGORY_ZP)
507 utf8proc_decompose_lump(0x000A);
508 }
509 }
510 if (options & UTF8PROC_STRIPMARK) {
511 if (category == UTF8PROC_CATEGORY_MN ||
512 category == UTF8PROC_CATEGORY_MC ||
513 category == UTF8PROC_CATEGORY_ME) return 0;
514 }
515 if (options & UTF8PROC_CASEFOLD) {
516 if (property->casefold_seqindex != UINT16_MAX) {
517 return seqindex_write_char_decomposed(property->casefold_seqindex, dst, bufsize, options, last_boundclass);
518 }
519 }
520 if (options & (UTF8PROC_COMPOSE|UTF8PROC_DECOMPOSE)) {
521 if (property->decomp_seqindex != UINT16_MAX &&
522 (!property->decomp_type || (options & UTF8PROC_COMPAT))) {
523 return seqindex_write_char_decomposed(property->decomp_seqindex, dst, bufsize, options, last_boundclass);
524 }
525 }
526 if (options & UTF8PROC_CHARBOUND) {
527 utf8proc_bool boundary;
528 boundary = grapheme_break_extended(0, property->boundclass, 0, property->indic_conjunct_break,
529 last_boundclass);
530 if (boundary) {
531 if (bufsize >= 1) dst[0] = -1; /* sentinel value for grapheme break */
532 if (bufsize >= 2) dst[1] = uc;
533 return 2;
534 }
535 }
536 if (bufsize >= 1) *dst = uc;
537 return 1;
538 }
539
540 UTF8PROC_DLLEXPORT utf8proc_ssize_t utf8proc_decompose(
541 const utf8proc_uint8_t *str, utf8proc_ssize_t strlen,
542 utf8proc_int32_t *buffer, utf8proc_ssize_t bufsize, utf8proc_option_t options
543 ) {
544 return utf8proc_decompose_custom(str, strlen, buffer, bufsize, options, NULL, NULL);
545 }
546
547 UTF8PROC_DLLEXPORT utf8proc_ssize_t utf8proc_decompose_custom(
548 const utf8proc_uint8_t *str, utf8proc_ssize_t strlen,
549 utf8proc_int32_t *buffer, utf8proc_ssize_t bufsize, utf8proc_option_t options,
550 utf8proc_custom_func custom_func, void *custom_data
551 ) {
552 /* strlen will be ignored, if UTF8PROC_NULLTERM is set in options */
553 utf8proc_ssize_t wpos = 0;
554 if ((options & UTF8PROC_COMPOSE) && (options & UTF8PROC_DECOMPOSE))
555 return UTF8PROC_ERROR_INVALIDOPTS;
556 if ((options & UTF8PROC_STRIPMARK) &&
557 !(options & UTF8PROC_COMPOSE) && !(options & UTF8PROC_DECOMPOSE))
558 return UTF8PROC_ERROR_INVALIDOPTS;
559 {
560 utf8proc_int32_t uc;
561 utf8proc_ssize_t rpos = 0;
562 utf8proc_ssize_t decomp_result;
563 int boundclass = UTF8PROC_BOUNDCLASS_START;
564 while (1) {
565 if (options & UTF8PROC_NULLTERM) {
566 rpos += utf8proc_iterate(str + rpos, -1, &uc);
567 /* checking of return value is not necessary,
568 as 'uc' is < 0 in case of error */
569 if (uc < 0) return UTF8PROC_ERROR_INVALIDUTF8;
570 if (rpos < 0) return UTF8PROC_ERROR_OVERFLOW;
571 if (uc == 0) break;
572 } else {
573 if (rpos >= strlen) break;
574 rpos += utf8proc_iterate(str + rpos, strlen - rpos, &uc);
575 if (uc < 0) return UTF8PROC_ERROR_INVALIDUTF8;
576 }
577 if (custom_func != NULL) {
578 uc = custom_func(uc, custom_data); /* user-specified custom mapping */
579 }
580 decomp_result = utf8proc_decompose_char(
581 uc, buffer ? buffer+wpos : buffer, (bufsize > wpos) ? (bufsize - wpos) : 0, options,
582 &boundclass
583 );
584 if (decomp_result < 0) return decomp_result;
585 wpos += decomp_result;
586 /* prohibiting integer overflows due to too long strings: */
587 if (wpos < 0 ||
588 wpos > (utf8proc_ssize_t)(SSIZE_MAX/sizeof(utf8proc_int32_t)/2))
589 return UTF8PROC_ERROR_OVERFLOW;
590 }
591 }
592 if ((options & (UTF8PROC_COMPOSE|UTF8PROC_DECOMPOSE)) && bufsize >= wpos) {
593 utf8proc_ssize_t pos = 0;
594 while (pos < wpos-1) {
595 utf8proc_int32_t uc1, uc2;
596 const utf8proc_property_t *property1, *property2;
597 uc1 = buffer[pos];
598 if (uc1 < 0) {
599 /* skip grapheme break */
600 pos++;
601 continue;
602 }
603 uc2 = buffer[pos+1];
604 if (uc2 < 0) {
605 /* cannot recombine; skip grapheme break */
606 pos+=2;
607 continue;
608 }
609 property1 = unsafe_get_property(uc1);
610 property2 = unsafe_get_property(uc2);
611 if (property1->combining_class > property2->combining_class &&
612 property2->combining_class > 0) {
613 buffer[pos] = uc2;
614 buffer[pos+1] = uc1;
615 if (pos > 0) pos--; else pos++;
616 } else {
617 pos++;
618 }
619 }
620 }
621 return wpos;
622 }
623
624 UTF8PROC_DLLEXPORT utf8proc_ssize_t utf8proc_normalize_utf32(utf8proc_int32_t *buffer, utf8proc_ssize_t length, utf8proc_option_t options) {
625 /* UTF8PROC_NULLTERM option will be ignored, 'length' is never ignored */
626 if (options & (UTF8PROC_NLF2LS | UTF8PROC_NLF2PS | UTF8PROC_STRIPCC)) {
627 utf8proc_ssize_t rpos;
628 utf8proc_ssize_t wpos = 0;
629 utf8proc_int32_t uc;
630 for (rpos = 0; rpos < length; rpos++) {
631 uc = buffer[rpos];
632 if (uc == 0x000D && rpos < length-1 && buffer[rpos+1] == 0x000A) rpos++;
633 if (uc == 0x000A || uc == 0x000D || uc == 0x0085 ||
634 ((options & UTF8PROC_STRIPCC) && (uc == 0x000B || uc == 0x000C))) {
635 if (options & UTF8PROC_NLF2LS) {
636 if (options & UTF8PROC_NLF2PS) {
637 buffer[wpos++] = 0x000A;
638 } else {
639 buffer[wpos++] = 0x2028;
640 }
641 } else {
642 if (options & UTF8PROC_NLF2PS) {
643 buffer[wpos++] = 0x2029;
644 } else {
645 buffer[wpos++] = 0x0020;
646 }
647 }
648 } else if ((options & UTF8PROC_STRIPCC) &&
649 (uc < 0x0020 || (uc >= 0x007F && uc < 0x00A0))) {
650 if (uc == 0x0009) buffer[wpos++] = 0x0020;
651 } else {
652 buffer[wpos++] = uc;
653 }
654 }
655 length = wpos;
656 }
657 if (options & UTF8PROC_COMPOSE) {
658 utf8proc_int32_t *starter = NULL;
659 const utf8proc_property_t *starter_property = NULL;
660 utf8proc_propval_t max_combining_class = -1;
661 utf8proc_ssize_t rpos;
662 utf8proc_ssize_t wpos = 0;
663 for (rpos = 0; rpos < length; rpos++) {
664 utf8proc_int32_t current_char = buffer[rpos];
665 if (current_char < 0) {
666 /* skip grapheme break */
667 continue;
668 }
669 const utf8proc_property_t *current_property = unsafe_get_property(current_char);
670 if (starter && current_property->combining_class > max_combining_class) {
671 /* combination perhaps possible */
672 utf8proc_int32_t hangul_lindex;
673 utf8proc_int32_t hangul_sindex;
674 hangul_lindex = *starter - UTF8PROC_HANGUL_LBASE;
675 if (hangul_lindex >= 0 && hangul_lindex < UTF8PROC_HANGUL_LCOUNT) {
676 utf8proc_int32_t hangul_vindex;
677 hangul_vindex = current_char - UTF8PROC_HANGUL_VBASE;
678 if (hangul_vindex >= 0 && hangul_vindex < UTF8PROC_HANGUL_VCOUNT) {
679 *starter = UTF8PROC_HANGUL_SBASE +
680 (hangul_lindex * UTF8PROC_HANGUL_VCOUNT + hangul_vindex) *
681 UTF8PROC_HANGUL_TCOUNT;
682 starter_property = NULL;
683 continue;
684 }
685 }
686 hangul_sindex = *starter - UTF8PROC_HANGUL_SBASE;
687 if (hangul_sindex >= 0 && hangul_sindex < UTF8PROC_HANGUL_SCOUNT &&
688 (hangul_sindex % UTF8PROC_HANGUL_TCOUNT) == 0) {
689 utf8proc_int32_t hangul_tindex;
690 hangul_tindex = current_char - UTF8PROC_HANGUL_TBASE;
691 if (hangul_tindex > 0 && hangul_tindex < UTF8PROC_HANGUL_TCOUNT) {
692 *starter += hangul_tindex;
693 starter_property = NULL;
694 continue;
695 }
696 }
697 if (!starter_property) {
698 starter_property = unsafe_get_property(*starter);
699 }
700 int idx = starter_property->comb_index;
701 if (idx < 0x3FF && current_property->comb_issecond) {
702 int len = starter_property->comb_length;
703 utf8proc_int32_t max_second = utf8proc_combinations_second[idx + len - 1];
704 if (current_char <= max_second) {
705 int off;
706 // TODO: binary search? arithmetic search?
707 for (off = 0; off < len; ++off) {
708 utf8proc_int32_t second = utf8proc_combinations_second[idx + off];
709 if (current_char < second) {
710 /* not found */
711 break;
712 }
713 if (current_char == second) {
714 /* found */
715 utf8proc_int32_t composition = utf8proc_combinations_combined[idx + off];
716 *starter = composition;
717 starter_property = NULL;
718 break;
719 }
720 }
721 if (starter_property == NULL) {
722 /* found */
723 continue;
724 }
725 }
726 }
727 }
728 buffer[wpos] = current_char;
729 if (current_property->combining_class) {
730 if (current_property->combining_class > max_combining_class) {
731 max_combining_class = current_property->combining_class;
732 }
733 } else {
734 starter = buffer + wpos;
735 starter_property = NULL;
736 max_combining_class = -1;
737 }
738 wpos++;
739 }
740 length = wpos;
741 }
742 return length;
743 }
744
745 UTF8PROC_DLLEXPORT utf8proc_ssize_t utf8proc_reencode(utf8proc_int32_t *buffer, utf8proc_ssize_t length, utf8proc_option_t options) {
746 /* UTF8PROC_NULLTERM option will be ignored, 'length' is never ignored
747 ASSERT: 'buffer' has one spare byte of free space at the end! */
748 length = utf8proc_normalize_utf32(buffer, length, options);
749 if (length < 0) return length;
750 {
751 utf8proc_ssize_t rpos, wpos = 0;
752 utf8proc_int32_t uc;
753 if (options & UTF8PROC_CHARBOUND) {
754 for (rpos = 0; rpos < length; rpos++) {
755 uc = buffer[rpos];
756 wpos += charbound_encode_char(uc, ((utf8proc_uint8_t *)buffer) + wpos);
757 }
758 } else {
759 for (rpos = 0; rpos < length; rpos++) {
760 uc = buffer[rpos];
761 wpos += utf8proc_encode_char(uc, ((utf8proc_uint8_t *)buffer) + wpos);
762 }
763 }
764 ((utf8proc_uint8_t *)buffer)[wpos] = 0;
765 return wpos;
766 }
767 }
768
769 UTF8PROC_DLLEXPORT utf8proc_ssize_t utf8proc_map(
770 const utf8proc_uint8_t *str, utf8proc_ssize_t strlen, utf8proc_uint8_t **dstptr, utf8proc_option_t options
771 ) {
772 return utf8proc_map_custom(str, strlen, dstptr, options, NULL, NULL);
773 }
774
775 UTF8PROC_DLLEXPORT utf8proc_ssize_t utf8proc_map_custom(
776 const utf8proc_uint8_t *str, utf8proc_ssize_t strlen, utf8proc_uint8_t **dstptr, utf8proc_option_t options,
777 utf8proc_custom_func custom_func, void *custom_data
778 ) {
779 utf8proc_int32_t *buffer;
780 utf8proc_ssize_t result;
781 *dstptr = NULL;
782 result = utf8proc_decompose_custom(str, strlen, NULL, 0, options, custom_func, custom_data);
783 if (result < 0) return result;
784 buffer = (utf8proc_int32_t *) malloc(((utf8proc_size_t)result) * sizeof(utf8proc_int32_t) + 1);
785 if (!buffer) return UTF8PROC_ERROR_NOMEM;
786 result = utf8proc_decompose_custom(str, strlen, buffer, result, options, custom_func, custom_data);
787 if (result < 0) {
788 free(buffer);
789 return result;
790 }
791 result = utf8proc_reencode(buffer, result, options);
792 if (result < 0) {
793 free(buffer);
794 return result;
795 }
796 {
797 utf8proc_int32_t *newptr;
798 newptr = (utf8proc_int32_t *) realloc(buffer, (size_t)result+1);
799 if (newptr) buffer = newptr;
800 }
801 *dstptr = (utf8proc_uint8_t *)buffer;
802 return result;
803 }
804
805 UTF8PROC_DLLEXPORT utf8proc_uint8_t *utf8proc_NFD(const utf8proc_uint8_t *str) {
806 utf8proc_uint8_t *retval;
807 utf8proc_map(str, 0, &retval, (utf8proc_option_t)(UTF8PROC_NULLTERM | UTF8PROC_STABLE |
808 UTF8PROC_DECOMPOSE));
809 return retval;
810 }
811
812 UTF8PROC_DLLEXPORT utf8proc_uint8_t *utf8proc_NFC(const utf8proc_uint8_t *str) {
813 utf8proc_uint8_t *retval;
814 utf8proc_map(str, 0, &retval, (utf8proc_option_t)(UTF8PROC_NULLTERM | UTF8PROC_STABLE |
815 UTF8PROC_COMPOSE));
816 return retval;
817 }
818
819 UTF8PROC_DLLEXPORT utf8proc_uint8_t *utf8proc_NFKD(const utf8proc_uint8_t *str) {
820 utf8proc_uint8_t *retval;
821 utf8proc_map(str, 0, &retval, (utf8proc_option_t)(UTF8PROC_NULLTERM | UTF8PROC_STABLE |
822 UTF8PROC_DECOMPOSE | UTF8PROC_COMPAT));
823 return retval;
824 }
825
826 UTF8PROC_DLLEXPORT utf8proc_uint8_t *utf8proc_NFKC(const utf8proc_uint8_t *str) {
827 utf8proc_uint8_t *retval;
828 utf8proc_map(str, 0, &retval, (utf8proc_option_t)(UTF8PROC_NULLTERM | UTF8PROC_STABLE |
829 UTF8PROC_COMPOSE | UTF8PROC_COMPAT));
830 return retval;
831 }
832
833 UTF8PROC_DLLEXPORT utf8proc_uint8_t *utf8proc_NFKC_Casefold(const utf8proc_uint8_t *str) {
834 utf8proc_uint8_t *retval;
835 utf8proc_map(str, 0, &retval, (utf8proc_option_t)(UTF8PROC_NULLTERM | UTF8PROC_STABLE |
836 UTF8PROC_COMPOSE | UTF8PROC_COMPAT | UTF8PROC_CASEFOLD | UTF8PROC_IGNORE));
837 return retval;
838 }