1 | /* Copyright (c) 2002 Red Hat Incorporated. |
---|
2 | All rights reserved. |
---|
3 | |
---|
4 | Redistribution and use in source and binary forms, with or without |
---|
5 | modification, are permitted provided that the following conditions are met: |
---|
6 | |
---|
7 | Redistributions of source code must retain the above copyright |
---|
8 | notice, this list of conditions and the following disclaimer. |
---|
9 | |
---|
10 | Redistributions in binary form must reproduce the above copyright |
---|
11 | notice, this list of conditions and the following disclaimer in the |
---|
12 | documentation and/or other materials provided with the distribution. |
---|
13 | |
---|
14 | The name of Red Hat Incorporated may not be used to endorse |
---|
15 | or promote products derived from this software without specific |
---|
16 | prior written permission. |
---|
17 | |
---|
18 | THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" |
---|
19 | AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE |
---|
20 | IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE |
---|
21 | ARE DISCLAIMED. IN NO EVENT SHALL RED HAT INCORPORATED BE LIABLE FOR ANY |
---|
22 | DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES |
---|
23 | (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; |
---|
24 | LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND |
---|
25 | ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
---|
26 | (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS |
---|
27 | SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
---|
28 | */ |
---|
29 | |
---|
30 | /* |
---|
31 | FUNCTION |
---|
32 | <<iswalpha>>, <<iswalpha_l>>---alphabetic wide character test |
---|
33 | |
---|
34 | INDEX |
---|
35 | iswalpha |
---|
36 | |
---|
37 | INDEX |
---|
38 | iswalpha_l |
---|
39 | |
---|
40 | SYNOPSIS |
---|
41 | #include <wctype.h> |
---|
42 | int iswalpha(wint_t <[c]>); |
---|
43 | |
---|
44 | #include <wctype.h> |
---|
45 | int iswalpha_l(wint_t <[c]>, locale_t <[locale]>); |
---|
46 | |
---|
47 | DESCRIPTION |
---|
48 | <<iswalpha>> is a function which classifies wide-character values that |
---|
49 | are alphabetic. |
---|
50 | |
---|
51 | <<iswalpha_l>> is like <<iswalpha>> but performs the check based on the |
---|
52 | locale specified by the locale object locale. If <[locale]> is |
---|
53 | LC_GLOBAL_LOCALE or not a valid locale object, the behaviour is undefined. |
---|
54 | |
---|
55 | RETURNS |
---|
56 | <<iswalpha>>, <<iswalpha_l>> return non-zero if <[c]> is an alphabetic |
---|
57 | wide character. |
---|
58 | |
---|
59 | PORTABILITY |
---|
60 | <<iswalpha>> is C99. |
---|
61 | <<iswalpha_l>> is POSIX-1.2008. |
---|
62 | |
---|
63 | No supporting OS subroutines are required. |
---|
64 | */ |
---|
65 | #include <_ansi.h> |
---|
66 | #include <newlib.h> |
---|
67 | #include <wctype.h> |
---|
68 | #include <string.h> |
---|
69 | #include <ctype.h> |
---|
70 | #include "local.h" |
---|
71 | |
---|
72 | #ifdef _MB_CAPABLE |
---|
73 | #include "utf8alpha.h" |
---|
74 | #endif /* _MB_CAPABLE */ |
---|
75 | |
---|
76 | int |
---|
77 | iswalpha (wint_t c) |
---|
78 | { |
---|
79 | #ifdef _MB_CAPABLE |
---|
80 | unsigned const char *table; |
---|
81 | unsigned char *ptr; |
---|
82 | unsigned char ctmp; |
---|
83 | int size; |
---|
84 | wint_t x; |
---|
85 | |
---|
86 | c = _jp2uc (c); |
---|
87 | |
---|
88 | /* Based on and tested against Unicode 5.2 |
---|
89 | See utf8alpha.h for a description how to fetch the data. */ |
---|
90 | x = (c >> 8); |
---|
91 | /* for some large sections, all characters are alphabetic so handle them here */ |
---|
92 | if ((x >= 0x34 && x <= 0x4c) || |
---|
93 | (x >= 0x4e && x <= 0x9e) || |
---|
94 | (x >= 0xac && x <= 0xd6) || |
---|
95 | (x >= 0x120 && x <= 0x122) || |
---|
96 | (x >= 0x130 && x <= 0x133) || |
---|
97 | (x >= 0x200 && x <= 0x2a5) || |
---|
98 | (x >= 0x2a7 && x <= 0x2b6)) |
---|
99 | return 1; |
---|
100 | |
---|
101 | switch (x) |
---|
102 | { |
---|
103 | case 0x00: |
---|
104 | table = u0; |
---|
105 | size = sizeof(u0); |
---|
106 | break; |
---|
107 | case 0x01: |
---|
108 | case 0x11: |
---|
109 | case 0x15: |
---|
110 | case 0x1e: |
---|
111 | case 0xa0: |
---|
112 | case 0xa1: |
---|
113 | case 0xa2: |
---|
114 | case 0xa3: |
---|
115 | case 0xa5: |
---|
116 | case 0xf9: |
---|
117 | case 0xfc: |
---|
118 | case 0x2f8: |
---|
119 | case 0x2f9: |
---|
120 | return 1; |
---|
121 | case 0x02: |
---|
122 | table = u2; |
---|
123 | size = sizeof(u2); |
---|
124 | break; |
---|
125 | case 0x03: |
---|
126 | table = u3; |
---|
127 | size = sizeof(u3); |
---|
128 | break; |
---|
129 | case 0x04: |
---|
130 | table = u4; |
---|
131 | size = sizeof(u4); |
---|
132 | break; |
---|
133 | case 0x05: |
---|
134 | table = u5; |
---|
135 | size = sizeof(u5); |
---|
136 | break; |
---|
137 | case 0x06: |
---|
138 | table = u6; |
---|
139 | size = sizeof(u6); |
---|
140 | break; |
---|
141 | case 0x07: |
---|
142 | table = u7; |
---|
143 | size = sizeof(u7); |
---|
144 | break; |
---|
145 | case 0x08: |
---|
146 | table = u8; |
---|
147 | size = sizeof(u8); |
---|
148 | break; |
---|
149 | case 0x09: |
---|
150 | table = u9; |
---|
151 | size = sizeof(u9); |
---|
152 | break; |
---|
153 | case 0x0a: |
---|
154 | table = ua; |
---|
155 | size = sizeof(ua); |
---|
156 | break; |
---|
157 | case 0x0b: |
---|
158 | table = ub; |
---|
159 | size = sizeof(ub); |
---|
160 | break; |
---|
161 | case 0x0c: |
---|
162 | table = uc; |
---|
163 | size = sizeof(uc); |
---|
164 | break; |
---|
165 | case 0x0d: |
---|
166 | table = ud; |
---|
167 | size = sizeof(ud); |
---|
168 | break; |
---|
169 | case 0x0e: |
---|
170 | table = ue; |
---|
171 | size = sizeof(ue); |
---|
172 | break; |
---|
173 | case 0x0f: |
---|
174 | table = uf; |
---|
175 | size = sizeof(uf); |
---|
176 | break; |
---|
177 | case 0x10: |
---|
178 | table = u10; |
---|
179 | size = sizeof(u10); |
---|
180 | break; |
---|
181 | case 0x12: |
---|
182 | table = u12; |
---|
183 | size = sizeof(u12); |
---|
184 | break; |
---|
185 | case 0x13: |
---|
186 | table = u13; |
---|
187 | size = sizeof(u13); |
---|
188 | break; |
---|
189 | case 0x14: |
---|
190 | table = u14; |
---|
191 | size = sizeof(u14); |
---|
192 | break; |
---|
193 | case 0x16: |
---|
194 | table = u16; |
---|
195 | size = sizeof(u16); |
---|
196 | break; |
---|
197 | case 0x17: |
---|
198 | table = u17; |
---|
199 | size = sizeof(u17); |
---|
200 | break; |
---|
201 | case 0x18: |
---|
202 | table = u18; |
---|
203 | size = sizeof(u18); |
---|
204 | break; |
---|
205 | case 0x19: |
---|
206 | table = u19; |
---|
207 | size = sizeof(u19); |
---|
208 | break; |
---|
209 | case 0x1a: |
---|
210 | table = u1a; |
---|
211 | size = sizeof(u1a); |
---|
212 | break; |
---|
213 | case 0x1b: |
---|
214 | table = u1b; |
---|
215 | size = sizeof(u1b); |
---|
216 | break; |
---|
217 | case 0x1c: |
---|
218 | table = u1c; |
---|
219 | size = sizeof(u1c); |
---|
220 | break; |
---|
221 | case 0x1d: |
---|
222 | table = u1d; |
---|
223 | size = sizeof(u1d); |
---|
224 | break; |
---|
225 | case 0x1f: |
---|
226 | table = u1f; |
---|
227 | size = sizeof(u1f); |
---|
228 | break; |
---|
229 | case 0x20: |
---|
230 | table = u20; |
---|
231 | size = sizeof(u20); |
---|
232 | break; |
---|
233 | case 0x21: |
---|
234 | table = u21; |
---|
235 | size = sizeof(u21); |
---|
236 | break; |
---|
237 | case 0x24: |
---|
238 | table = u24; |
---|
239 | size = sizeof(u24); |
---|
240 | break; |
---|
241 | case 0x2c: |
---|
242 | table = u2c; |
---|
243 | size = sizeof(u2c); |
---|
244 | break; |
---|
245 | case 0x2d: |
---|
246 | table = u2d; |
---|
247 | size = sizeof(u2d); |
---|
248 | break; |
---|
249 | case 0x2e: |
---|
250 | table = u2e; |
---|
251 | size = sizeof(u2e); |
---|
252 | break; |
---|
253 | case 0x30: |
---|
254 | table = u30; |
---|
255 | size = sizeof(u30); |
---|
256 | break; |
---|
257 | case 0x31: |
---|
258 | table = u31; |
---|
259 | size = sizeof(u31); |
---|
260 | break; |
---|
261 | case 0x4d: |
---|
262 | table = u4d; |
---|
263 | size = sizeof(u4d); |
---|
264 | break; |
---|
265 | case 0x9f: |
---|
266 | table = u9f; |
---|
267 | size = sizeof(u9f); |
---|
268 | break; |
---|
269 | case 0xa4: |
---|
270 | table = ua4; |
---|
271 | size = sizeof(ua4); |
---|
272 | break; |
---|
273 | case 0xa6: |
---|
274 | table = ua6; |
---|
275 | size = sizeof(ua6); |
---|
276 | break; |
---|
277 | case 0xa7: |
---|
278 | table = ua7; |
---|
279 | size = sizeof(ua7); |
---|
280 | break; |
---|
281 | case 0xa8: |
---|
282 | table = ua8; |
---|
283 | size = sizeof(ua8); |
---|
284 | break; |
---|
285 | case 0xa9: |
---|
286 | table = ua9; |
---|
287 | size = sizeof(ua9); |
---|
288 | break; |
---|
289 | case 0xaa: |
---|
290 | table = uaa; |
---|
291 | size = sizeof(uaa); |
---|
292 | break; |
---|
293 | case 0xab: |
---|
294 | table = uab; |
---|
295 | size = sizeof(uab); |
---|
296 | break; |
---|
297 | case 0xd7: |
---|
298 | table = ud7; |
---|
299 | size = sizeof(ud7); |
---|
300 | break; |
---|
301 | case 0xfa: |
---|
302 | table = ufa; |
---|
303 | size = sizeof(ufa); |
---|
304 | break; |
---|
305 | case 0xfb: |
---|
306 | table = ufb; |
---|
307 | size = sizeof(ufb); |
---|
308 | break; |
---|
309 | case 0xfd: |
---|
310 | table = ufd; |
---|
311 | size = sizeof(ufd); |
---|
312 | break; |
---|
313 | case 0xfe: |
---|
314 | table = ufe; |
---|
315 | size = sizeof(ufe); |
---|
316 | break; |
---|
317 | case 0xff: |
---|
318 | table = uff; |
---|
319 | size = sizeof(uff); |
---|
320 | break; |
---|
321 | case 0x100: |
---|
322 | table = u100; |
---|
323 | size = sizeof(u100); |
---|
324 | break; |
---|
325 | case 0x101: |
---|
326 | table = u101; |
---|
327 | size = sizeof(u101); |
---|
328 | break; |
---|
329 | case 0x102: |
---|
330 | table = u102; |
---|
331 | size = sizeof(u102); |
---|
332 | break; |
---|
333 | case 0x103: |
---|
334 | table = u103; |
---|
335 | size = sizeof(u103); |
---|
336 | break; |
---|
337 | case 0x104: |
---|
338 | table = u104; |
---|
339 | size = sizeof(u104); |
---|
340 | break; |
---|
341 | case 0x108: |
---|
342 | table = u108; |
---|
343 | size = sizeof(u108); |
---|
344 | break; |
---|
345 | case 0x109: |
---|
346 | table = u109; |
---|
347 | size = sizeof(u109); |
---|
348 | break; |
---|
349 | case 0x10a: |
---|
350 | table = u10a; |
---|
351 | size = sizeof(u10a); |
---|
352 | break; |
---|
353 | case 0x10b: |
---|
354 | table = u10b; |
---|
355 | size = sizeof(u10b); |
---|
356 | break; |
---|
357 | case 0x10c: |
---|
358 | table = u10c; |
---|
359 | size = sizeof(u10c); |
---|
360 | break; |
---|
361 | case 0x110: |
---|
362 | table = u110; |
---|
363 | size = sizeof(u110); |
---|
364 | break; |
---|
365 | case 0x123: |
---|
366 | table = u123; |
---|
367 | size = sizeof(u123); |
---|
368 | break; |
---|
369 | case 0x124: |
---|
370 | table = u124; |
---|
371 | size = sizeof(u124); |
---|
372 | break; |
---|
373 | case 0x134: |
---|
374 | table = u134; |
---|
375 | size = sizeof(u134); |
---|
376 | break; |
---|
377 | case 0x1d4: |
---|
378 | table = u1d4; |
---|
379 | size = sizeof(u1d4); |
---|
380 | break; |
---|
381 | case 0x1d5: |
---|
382 | table = u1d5; |
---|
383 | size = sizeof(u1d5); |
---|
384 | break; |
---|
385 | case 0x1d6: |
---|
386 | table = u1d6; |
---|
387 | size = sizeof(u1d6); |
---|
388 | break; |
---|
389 | case 0x1d7: |
---|
390 | table = u1d7; |
---|
391 | size = sizeof(u1d7); |
---|
392 | break; |
---|
393 | case 0x1f1: |
---|
394 | table = u1f1; |
---|
395 | size = sizeof(u1f1); |
---|
396 | break; |
---|
397 | case 0x2a6: |
---|
398 | table = u2a6; |
---|
399 | size = sizeof(u2a6); |
---|
400 | break; |
---|
401 | case 0x2b7: |
---|
402 | table = u2b7; |
---|
403 | size = sizeof(u2b7); |
---|
404 | break; |
---|
405 | case 0x2fa: |
---|
406 | table = u2fa; |
---|
407 | size = sizeof(u2fa); |
---|
408 | break; |
---|
409 | default: |
---|
410 | return 0; |
---|
411 | } |
---|
412 | /* we have narrowed down to a section of 256 characters to check */ |
---|
413 | /* now check if c matches the alphabetic wide-chars within that section */ |
---|
414 | ptr = (unsigned char *)table; |
---|
415 | ctmp = (unsigned char)c; |
---|
416 | while (ptr < table + size) |
---|
417 | { |
---|
418 | if (ctmp == *ptr) |
---|
419 | return 1; |
---|
420 | if (ctmp < *ptr) |
---|
421 | return 0; |
---|
422 | /* otherwise c > *ptr */ |
---|
423 | /* look for 0x0 as next element which indicates a range */ |
---|
424 | ++ptr; |
---|
425 | if (ptr < table + size - 1 && *ptr == 0x0) |
---|
426 | { |
---|
427 | /* we have a range..see if c falls within range */ |
---|
428 | ++ptr; |
---|
429 | if (ctmp <= *ptr) |
---|
430 | return 1; |
---|
431 | ++ptr; |
---|
432 | } |
---|
433 | } |
---|
434 | /* not in table */ |
---|
435 | return 0; |
---|
436 | #else |
---|
437 | return (c < (wint_t)0x100 ? isalpha (c) : 0); |
---|
438 | #endif /* _MB_CAPABLE */ |
---|
439 | } |
---|