1 | /*
|
---|
2 | * Copyright (C) 2000, 2001, 2002 Loic Dachary <loic@senga.org>
|
---|
3 | *
|
---|
4 | * This program is free software; you can redistribute it and/or modify
|
---|
5 | * it under the terms of the GNU General Public License as published by
|
---|
6 | * the Free Software Foundation; either version 2 of the License, or
|
---|
7 | * (at your option) any later version.
|
---|
8 | *
|
---|
9 | * This program is distributed in the hope that it will be useful,
|
---|
10 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
11 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
12 | * GNU General Public License for more details.
|
---|
13 | *
|
---|
14 | * You should have received a copy of the GNU General Public License
|
---|
15 | * along with this program; if not, write to the Free Software
|
---|
16 | * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
|
---|
17 | */
|
---|
18 |
|
---|
19 | /*
|
---|
20 | * Provides functions to strip accents from a string in all the
|
---|
21 | * charset supported by iconv(3).
|
---|
22 | *
|
---|
23 | * See the unac(3) manual page for more information.
|
---|
24 | *
|
---|
25 | */
|
---|
26 |
|
---|
27 | #ifndef _unac_h
|
---|
28 | #define _unac_h
|
---|
29 |
|
---|
30 | #ifdef __cplusplus
|
---|
31 | extern "C" {
|
---|
32 | #endif
|
---|
33 |
|
---|
34 | /* Generated by builder. Do not modify. Start defines */
|
---|
35 | #define UNAC_BLOCK_SHIFT 5
|
---|
36 | #define UNAC_BLOCK_MASK ((1 << UNAC_BLOCK_SHIFT) - 1)
|
---|
37 | #define UNAC_BLOCK_SIZE (1 << UNAC_BLOCK_SHIFT)
|
---|
38 | #define UNAC_BLOCK_COUNT 178
|
---|
39 | #define UNAC_INDEXES_SIZE (0x10000 >> UNAC_BLOCK_SHIFT)
|
---|
40 | /* Generated by builder. Do not modify. End defines */
|
---|
41 |
|
---|
42 | /*
|
---|
43 | * Return the unaccented equivalent of the UTF-16 character <c>
|
---|
44 | * in the pointer <p>. The length of the unsigned short array pointed
|
---|
45 | * by <p> is returned in the <l> argument.
|
---|
46 | * The C++ prototype of this macro would be:
|
---|
47 | *
|
---|
48 | * void unac_char(const unsigned short c, unsigned short*& p, int& l)
|
---|
49 | *
|
---|
50 | * See unac(3) in IMPLEMENTATION NOTES for more information about the
|
---|
51 | * tables (unac_data_table, unac_positions) layout.
|
---|
52 | */
|
---|
53 | #define unac_char_utf16(c,p,l) \
|
---|
54 | { \
|
---|
55 | unsigned short index = unac_indexes[(c) >> UNAC_BLOCK_SHIFT]; \
|
---|
56 | unsigned char position = (c) & UNAC_BLOCK_MASK; \
|
---|
57 | (p) = &(unac_data_table[index][unac_positions[index][position]]); \
|
---|
58 | (l) = unac_positions[index][position + 1] - unac_positions[index][position]; \
|
---|
59 | if((l) == 1 && *(p) == 0xFFFF) { \
|
---|
60 | (p) = 0; \
|
---|
61 | (l) = 0; \
|
---|
62 | } \
|
---|
63 | }
|
---|
64 |
|
---|
65 | /*
|
---|
66 | * Return the unaccented equivalent of the UTF-16 string <in> of
|
---|
67 | * length <in_length> in the pointer <out>. The length of the UTF-16
|
---|
68 | * string returned in <out> is stored in <out_length>. If the pointer
|
---|
69 | * *out is null, a new string is allocated using malloc(3). If the
|
---|
70 | * pointer *out is not null, the available length must also be given
|
---|
71 | * in the *out_length argument. The pointer passed to *out must have
|
---|
72 | * been allocated by malloc(3) and may be reallocated by realloc(3) if
|
---|
73 | * needs be. It is the responsibility of the caller to free the
|
---|
74 | * pointer returned in *out. The return value is 0 on success and -1
|
---|
75 | * on error, in which case the errno variable is set to the
|
---|
76 | * corresponding error code.
|
---|
77 | */
|
---|
78 | int unac_string_utf16(const char* in, size_t in_length,
|
---|
79 | char** out, size_t* out_length);
|
---|
80 |
|
---|
81 | /*
|
---|
82 | * The semantic of this function is stricly equal to the function
|
---|
83 | * unac_string_utf16. The <charset> argument applies to the content of the
|
---|
84 | * input string. It is converted to UTF-16 using iconv(3) before calling
|
---|
85 | * the unac_string function and the result is converted from UTF-16 to
|
---|
86 | * the specified <charset> before returning it in the <out> pointer.
|
---|
87 | * For efficiency purpose it is recommended that the caller uses
|
---|
88 | * unac_string and iconv(3) to save buffer allocations overhead.
|
---|
89 | * The return value is 0 on success and -1 on error, in which case
|
---|
90 | * the errno variable is set to the corresponding error code.
|
---|
91 | */
|
---|
92 | int unac_string(const char* charset,
|
---|
93 | const char* in, size_t in_length,
|
---|
94 | char** out, size_t* out_length);
|
---|
95 |
|
---|
96 | /*
|
---|
97 | * Return unac version number.
|
---|
98 | */
|
---|
99 | const char* unac_version(void);
|
---|
100 |
|
---|
101 | #define UNAC_DEBUG_NONE 0x00
|
---|
102 | #define UNAC_DEBUG_LOW 0x01
|
---|
103 | #define UNAC_DEBUG_HIGH 0x02
|
---|
104 |
|
---|
105 | #ifdef HAVE_VSNPRINTF
|
---|
106 | #define UNAC_DEBUG_AVAILABLE 1
|
---|
107 | /*
|
---|
108 | * Set the unac debug level. <l> is one of:
|
---|
109 | * UNAC_DEBUG_NONE for no debug messages at all
|
---|
110 | * UNAC_DEBUG_LOW for minimal information
|
---|
111 | * UNAC_DEBUG_HIGH for extremely verbose information,
|
---|
112 | * only usable when translating a few short strings.
|
---|
113 | *
|
---|
114 | * unac_debug with anything but UNAC_DEBUG_NONE is not
|
---|
115 | * thread safe.
|
---|
116 | */
|
---|
117 | #define unac_debug(l) unac_debug_callback((l), 0, (void*)0);
|
---|
118 |
|
---|
119 | /*
|
---|
120 | * Set the debug level and define a printing function callback.
|
---|
121 | * The <level> debug level is the same as in unac_debug. The
|
---|
122 | * <function> is in charge of dealing with the debug messages,
|
---|
123 | * presumably to print them to the user. The <data> is an opaque
|
---|
124 | * pointer that is passed along to <function>, should it
|
---|
125 | * need to manage a persistent context.
|
---|
126 | *
|
---|
127 | * The prototype of <function> allows two arguments. The first
|
---|
128 | * is the debug message (const char*), the second is the opaque
|
---|
129 | * pointer given as <data> argument to unac_debug_callback.
|
---|
130 | *
|
---|
131 | * If <function> is NULL, messages are printed on the standard
|
---|
132 | * error output using fprintf(stderr...).
|
---|
133 | *
|
---|
134 | * unac_debug_callback with anything but UNAC_DEBUG_NONE is not
|
---|
135 | * thread safe.
|
---|
136 | *
|
---|
137 | */
|
---|
138 | typedef void (*unac_debug_print_t)(const char* message, void* data);
|
---|
139 | void unac_debug_callback(int level, unac_debug_print_t function, void* data);
|
---|
140 | #endif /* HAVE_VSNPRINTF */
|
---|
141 |
|
---|
142 | /* Generated by builder. Do not modify. Start declarations */
|
---|
143 | extern unsigned short unac_indexes[UNAC_INDEXES_SIZE];
|
---|
144 | extern unsigned char unac_positions[UNAC_BLOCK_COUNT][UNAC_BLOCK_SIZE + 1];
|
---|
145 | extern unsigned short* unac_data_table[UNAC_BLOCK_COUNT];
|
---|
146 | extern unsigned short unac_data0[];
|
---|
147 | extern unsigned short unac_data1[];
|
---|
148 | extern unsigned short unac_data2[];
|
---|
149 | extern unsigned short unac_data3[];
|
---|
150 | extern unsigned short unac_data4[];
|
---|
151 | extern unsigned short unac_data5[];
|
---|
152 | extern unsigned short unac_data6[];
|
---|
153 | extern unsigned short unac_data7[];
|
---|
154 | extern unsigned short unac_data8[];
|
---|
155 | extern unsigned short unac_data9[];
|
---|
156 | extern unsigned short unac_data10[];
|
---|
157 | extern unsigned short unac_data11[];
|
---|
158 | extern unsigned short unac_data12[];
|
---|
159 | extern unsigned short unac_data13[];
|
---|
160 | extern unsigned short unac_data14[];
|
---|
161 | extern unsigned short unac_data15[];
|
---|
162 | extern unsigned short unac_data16[];
|
---|
163 | extern unsigned short unac_data17[];
|
---|
164 | extern unsigned short unac_data18[];
|
---|
165 | extern unsigned short unac_data19[];
|
---|
166 | extern unsigned short unac_data20[];
|
---|
167 | extern unsigned short unac_data21[];
|
---|
168 | extern unsigned short unac_data22[];
|
---|
169 | extern unsigned short unac_data23[];
|
---|
170 | extern unsigned short unac_data24[];
|
---|
171 | extern unsigned short unac_data25[];
|
---|
172 | extern unsigned short unac_data26[];
|
---|
173 | extern unsigned short unac_data27[];
|
---|
174 | extern unsigned short unac_data28[];
|
---|
175 | extern unsigned short unac_data29[];
|
---|
176 | extern unsigned short unac_data30[];
|
---|
177 | extern unsigned short unac_data31[];
|
---|
178 | extern unsigned short unac_data32[];
|
---|
179 | extern unsigned short unac_data33[];
|
---|
180 | extern unsigned short unac_data34[];
|
---|
181 | extern unsigned short unac_data35[];
|
---|
182 | extern unsigned short unac_data36[];
|
---|
183 | extern unsigned short unac_data37[];
|
---|
184 | extern unsigned short unac_data38[];
|
---|
185 | extern unsigned short unac_data39[];
|
---|
186 | extern unsigned short unac_data40[];
|
---|
187 | extern unsigned short unac_data41[];
|
---|
188 | extern unsigned short unac_data42[];
|
---|
189 | extern unsigned short unac_data43[];
|
---|
190 | extern unsigned short unac_data44[];
|
---|
191 | extern unsigned short unac_data45[];
|
---|
192 | extern unsigned short unac_data46[];
|
---|
193 | extern unsigned short unac_data47[];
|
---|
194 | extern unsigned short unac_data48[];
|
---|
195 | extern unsigned short unac_data49[];
|
---|
196 | extern unsigned short unac_data50[];
|
---|
197 | extern unsigned short unac_data51[];
|
---|
198 | extern unsigned short unac_data52[];
|
---|
199 | extern unsigned short unac_data53[];
|
---|
200 | extern unsigned short unac_data54[];
|
---|
201 | extern unsigned short unac_data55[];
|
---|
202 | extern unsigned short unac_data56[];
|
---|
203 | extern unsigned short unac_data57[];
|
---|
204 | extern unsigned short unac_data58[];
|
---|
205 | extern unsigned short unac_data59[];
|
---|
206 | extern unsigned short unac_data60[];
|
---|
207 | extern unsigned short unac_data61[];
|
---|
208 | extern unsigned short unac_data62[];
|
---|
209 | extern unsigned short unac_data63[];
|
---|
210 | extern unsigned short unac_data64[];
|
---|
211 | extern unsigned short unac_data65[];
|
---|
212 | extern unsigned short unac_data66[];
|
---|
213 | extern unsigned short unac_data67[];
|
---|
214 | extern unsigned short unac_data68[];
|
---|
215 | extern unsigned short unac_data69[];
|
---|
216 | extern unsigned short unac_data70[];
|
---|
217 | extern unsigned short unac_data71[];
|
---|
218 | extern unsigned short unac_data72[];
|
---|
219 | extern unsigned short unac_data73[];
|
---|
220 | extern unsigned short unac_data74[];
|
---|
221 | extern unsigned short unac_data75[];
|
---|
222 | extern unsigned short unac_data76[];
|
---|
223 | extern unsigned short unac_data77[];
|
---|
224 | extern unsigned short unac_data78[];
|
---|
225 | extern unsigned short unac_data79[];
|
---|
226 | extern unsigned short unac_data80[];
|
---|
227 | extern unsigned short unac_data81[];
|
---|
228 | extern unsigned short unac_data82[];
|
---|
229 | extern unsigned short unac_data83[];
|
---|
230 | extern unsigned short unac_data84[];
|
---|
231 | extern unsigned short unac_data85[];
|
---|
232 | extern unsigned short unac_data86[];
|
---|
233 | extern unsigned short unac_data87[];
|
---|
234 | extern unsigned short unac_data88[];
|
---|
235 | extern unsigned short unac_data89[];
|
---|
236 | extern unsigned short unac_data90[];
|
---|
237 | extern unsigned short unac_data91[];
|
---|
238 | extern unsigned short unac_data92[];
|
---|
239 | extern unsigned short unac_data93[];
|
---|
240 | extern unsigned short unac_data94[];
|
---|
241 | extern unsigned short unac_data95[];
|
---|
242 | extern unsigned short unac_data96[];
|
---|
243 | extern unsigned short unac_data97[];
|
---|
244 | extern unsigned short unac_data98[];
|
---|
245 | extern unsigned short unac_data99[];
|
---|
246 | extern unsigned short unac_data100[];
|
---|
247 | extern unsigned short unac_data101[];
|
---|
248 | extern unsigned short unac_data102[];
|
---|
249 | extern unsigned short unac_data103[];
|
---|
250 | extern unsigned short unac_data104[];
|
---|
251 | extern unsigned short unac_data105[];
|
---|
252 | extern unsigned short unac_data106[];
|
---|
253 | extern unsigned short unac_data107[];
|
---|
254 | extern unsigned short unac_data108[];
|
---|
255 | extern unsigned short unac_data109[];
|
---|
256 | extern unsigned short unac_data110[];
|
---|
257 | extern unsigned short unac_data111[];
|
---|
258 | extern unsigned short unac_data112[];
|
---|
259 | extern unsigned short unac_data113[];
|
---|
260 | extern unsigned short unac_data114[];
|
---|
261 | extern unsigned short unac_data115[];
|
---|
262 | extern unsigned short unac_data116[];
|
---|
263 | extern unsigned short unac_data117[];
|
---|
264 | extern unsigned short unac_data118[];
|
---|
265 | extern unsigned short unac_data119[];
|
---|
266 | extern unsigned short unac_data120[];
|
---|
267 | extern unsigned short unac_data121[];
|
---|
268 | extern unsigned short unac_data122[];
|
---|
269 | extern unsigned short unac_data123[];
|
---|
270 | extern unsigned short unac_data124[];
|
---|
271 | extern unsigned short unac_data125[];
|
---|
272 | extern unsigned short unac_data126[];
|
---|
273 | extern unsigned short unac_data127[];
|
---|
274 | extern unsigned short unac_data128[];
|
---|
275 | extern unsigned short unac_data129[];
|
---|
276 | extern unsigned short unac_data130[];
|
---|
277 | extern unsigned short unac_data131[];
|
---|
278 | extern unsigned short unac_data132[];
|
---|
279 | extern unsigned short unac_data133[];
|
---|
280 | extern unsigned short unac_data134[];
|
---|
281 | extern unsigned short unac_data135[];
|
---|
282 | extern unsigned short unac_data136[];
|
---|
283 | extern unsigned short unac_data137[];
|
---|
284 | extern unsigned short unac_data138[];
|
---|
285 | extern unsigned short unac_data139[];
|
---|
286 | extern unsigned short unac_data140[];
|
---|
287 | extern unsigned short unac_data141[];
|
---|
288 | extern unsigned short unac_data142[];
|
---|
289 | extern unsigned short unac_data143[];
|
---|
290 | extern unsigned short unac_data144[];
|
---|
291 | extern unsigned short unac_data145[];
|
---|
292 | extern unsigned short unac_data146[];
|
---|
293 | extern unsigned short unac_data147[];
|
---|
294 | extern unsigned short unac_data148[];
|
---|
295 | extern unsigned short unac_data149[];
|
---|
296 | extern unsigned short unac_data150[];
|
---|
297 | extern unsigned short unac_data151[];
|
---|
298 | extern unsigned short unac_data152[];
|
---|
299 | extern unsigned short unac_data153[];
|
---|
300 | extern unsigned short unac_data154[];
|
---|
301 | extern unsigned short unac_data155[];
|
---|
302 | extern unsigned short unac_data156[];
|
---|
303 | extern unsigned short unac_data157[];
|
---|
304 | extern unsigned short unac_data158[];
|
---|
305 | extern unsigned short unac_data159[];
|
---|
306 | extern unsigned short unac_data160[];
|
---|
307 | extern unsigned short unac_data161[];
|
---|
308 | extern unsigned short unac_data162[];
|
---|
309 | extern unsigned short unac_data163[];
|
---|
310 | extern unsigned short unac_data164[];
|
---|
311 | extern unsigned short unac_data165[];
|
---|
312 | extern unsigned short unac_data166[];
|
---|
313 | extern unsigned short unac_data167[];
|
---|
314 | extern unsigned short unac_data168[];
|
---|
315 | extern unsigned short unac_data169[];
|
---|
316 | extern unsigned short unac_data170[];
|
---|
317 | extern unsigned short unac_data171[];
|
---|
318 | extern unsigned short unac_data172[];
|
---|
319 | extern unsigned short unac_data173[];
|
---|
320 | extern unsigned short unac_data174[];
|
---|
321 | extern unsigned short unac_data175[];
|
---|
322 | extern unsigned short unac_data176[];
|
---|
323 | extern unsigned short unac_data177[];
|
---|
324 | /* Generated by builder. Do not modify. End declarations */
|
---|
325 |
|
---|
326 | #ifdef __cplusplus
|
---|
327 | }
|
---|
328 | #endif
|
---|
329 |
|
---|
330 | #endif /* _unac_h */
|
---|