1 | /* |
2 | ******************************************************************************* |
3 | * |
4 | * Copyright (C) 2002-2012, International Business Machines |
5 | * Corporation and others. All Rights Reserved. |
6 | * |
7 | ******************************************************************************* |
8 | * file name: utf_old.h |
9 | * encoding: US-ASCII |
10 | * tab size: 8 (not used) |
11 | * indentation:4 |
12 | * |
13 | * created on: 2002sep21 |
14 | * created by: Markus W. Scherer |
15 | */ |
16 | |
17 | /** |
18 | * \file |
19 | * \brief C API: Deprecated macros for Unicode string handling |
20 | */ |
21 | |
22 | /** |
23 | * |
24 | * The macros in utf_old.h are all deprecated and their use discouraged. |
25 | * Some of the design principles behind the set of UTF macros |
26 | * have changed or proved impractical. |
27 | * Almost all of the old "UTF macros" are at least renamed. |
28 | * If you are looking for a new equivalent to an old macro, please see the |
29 | * comment at the old one. |
30 | * |
31 | * Brief summary of reasons for deprecation: |
32 | * - Switch on UTF_SIZE (selection of UTF-8/16/32 default string processing) |
33 | * was impractical. |
34 | * - Switch on UTF_SAFE etc. (selection of unsafe/safe/strict default string processing) |
35 | * was of little use and impractical. |
36 | * - Whole classes of macros became obsolete outside of the UTF_SIZE/UTF_SAFE |
37 | * selection framework: UTF32_ macros (all trivial) |
38 | * and UTF_ default and intermediate macros (all aliases). |
39 | * - The selection framework also caused many macro aliases. |
40 | * - Change in Unicode standard: "irregular" sequences (3.0) became illegal (3.2). |
41 | * - Change of language in Unicode standard: |
42 | * Growing distinction between internal x-bit Unicode strings and external UTF-x |
43 | * forms, with the former more lenient. |
44 | * Suggests renaming of UTF16_ macros to U16_. |
45 | * - The prefix "UTF_" without a width number confused some users. |
46 | * - "Safe" append macros needed the addition of an error indicator output. |
47 | * - "Safe" UTF-8 macros used legitimate (if rarely used) code point values |
48 | * to indicate error conditions. |
49 | * - The use of the "_CHAR" infix for code point operations confused some users. |
50 | * |
51 | * More details: |
52 | * |
53 | * Until ICU 2.2, utf.h theoretically allowed to choose among UTF-8/16/32 |
54 | * for string processing, and among unsafe/safe/strict default macros for that. |
55 | * |
56 | * It proved nearly impossible to write non-trivial, high-performance code |
57 | * that is UTF-generic. |
58 | * Unsafe default macros would be dangerous for default string processing, |
59 | * and the main reason for the "strict" versions disappeared: |
60 | * Between Unicode 3.0 and 3.2 all "irregular" UTF-8 sequences became illegal. |
61 | * The only other conditions that "strict" checked for were non-characters, |
62 | * which are valid during processing. Only during text input/output should they |
63 | * be checked, and at that time other well-formedness checks may be |
64 | * necessary or useful as well. |
65 | * This can still be done by using U16_NEXT and U_IS_UNICODE_NONCHAR |
66 | * or U_IS_UNICODE_CHAR. |
67 | * |
68 | * The old UTF8_..._SAFE macros also used some normal Unicode code points |
69 | * to indicate malformed sequences. |
70 | * The new UTF8_ macros without suffix use negative values instead. |
71 | * |
72 | * The entire contents of utf32.h was moved here without replacement |
73 | * because all those macros were trivial and |
74 | * were meaningful only in the framework of choosing the UTF size. |
75 | * |
76 | * See Jitterbug 2150 and its discussion on the ICU mailing list |
77 | * in September 2002. |
78 | * |
79 | * <hr> |
80 | * |
81 | * <em>Obsolete part</em> of pre-ICU 2.4 utf.h file documentation: |
82 | * |
83 | * <p>The original concept for these files was for ICU to allow |
84 | * in principle to set which UTF (UTF-8/16/32) is used internally |
85 | * by defining UTF_SIZE to either 8, 16, or 32. utf.h would then define the UChar type |
86 | * accordingly. UTF-16 was the default.</p> |
87 | * |
88 | * <p>This concept has been abandoned. |
89 | * A lot of the ICU source code assumes UChar strings are in UTF-16. |
90 | * This is especially true for low-level code like |
91 | * conversion, normalization, and collation. |
92 | * The utf.h header enforces the default of UTF-16. |
93 | * The UTF-8 and UTF-32 macros remain for now for completeness and backward compatibility.</p> |
94 | * |
95 | * <p>Accordingly, utf.h defines UChar to be an unsigned 16-bit integer. If this matches wchar_t, then |
96 | * UChar is defined to be exactly wchar_t, otherwise uint16_t.</p> |
97 | * |
98 | * <p>UChar32 is defined to be a signed 32-bit integer (int32_t), large enough for a 21-bit |
99 | * Unicode code point (Unicode scalar value, 0..0x10ffff). |
100 | * Before ICU 2.4, the definition of UChar32 was similarly platform-dependent as |
101 | * the definition of UChar. For details see the documentation for UChar32 itself.</p> |
102 | * |
103 | * <p>utf.h also defines a number of C macros for handling single Unicode code points and |
104 | * for using UTF Unicode strings. It includes utf8.h, utf16.h, and utf32.h for the actual |
105 | * implementations of those macros and then aliases one set of them (for UTF-16) for general use. |
106 | * The UTF-specific macros have the UTF size in the macro name prefixes (UTF16_...), while |
107 | * the general alias macros always begin with UTF_...</p> |
108 | * |
109 | * <p>Many string operations can be done with or without error checking. |
110 | * Where such a distinction is useful, there are two versions of the macros, "unsafe" and "safe" |
111 | * ones with ..._UNSAFE and ..._SAFE suffixes. The unsafe macros are fast but may cause |
112 | * program failures if the strings are not well-formed. The safe macros have an additional, boolean |
113 | * parameter "strict". If strict is FALSE, then only illegal sequences are detected. |
114 | * Otherwise, irregular sequences and non-characters are detected as well (like single surrogates). |
115 | * Safe macros return special error code points for illegal/irregular sequences: |
116 | * Typically, U+ffff, or values that would result in a code unit sequence of the same length |
117 | * as the erroneous input sequence.<br> |
118 | * Note that _UNSAFE macros have fewer parameters: They do not have the strictness parameter, and |
119 | * they do not have start/length parameters for boundary checking.</p> |
120 | * |
121 | * <p>Here, the macros are aliased in two steps: |
122 | * In the first step, the UTF-specific macros with UTF16_ prefix and _UNSAFE and _SAFE suffixes are |
123 | * aliased according to the UTF_SIZE to macros with UTF_ prefix and the same suffixes and signatures. |
124 | * Then, in a second step, the default, general alias macros are set to use either the unsafe or |
125 | * the safe/not strict (default) or the safe/strict macro; |
126 | * these general macros do not have a strictness parameter.</p> |
127 | * |
128 | * <p>It is possible to change the default choice for the general alias macros to be unsafe, safe/not strict or safe/strict. |
129 | * The default is safe/not strict. It is not recommended to select the unsafe macros as the basis for |
130 | * Unicode string handling in ICU! To select this, define UTF_SAFE, UTF_STRICT, or UTF_UNSAFE.</p> |
131 | * |
132 | * <p>For general use, one should use the default, general macros with UTF_ prefix and no _SAFE/_UNSAFE suffix. |
133 | * Only in some cases it may be necessary to control the choice of macro directly and use a less generic alias. |
134 | * For example, if it can be assumed that a string is well-formed and the index will stay within the bounds, |
135 | * then the _UNSAFE version may be used. |
136 | * If a UTF-8 string is to be processed, then the macros with UTF8_ prefixes need to be used.</p> |
137 | * |
138 | * <hr> |
139 | * |
140 | * @deprecated ICU 2.4. Use the macros in utf.h, utf16.h, utf8.h instead. |
141 | */ |
142 | |
143 | #ifndef __UTF_OLD_H__ |
144 | #define __UTF_OLD_H__ |
145 | |
146 | #ifndef U_HIDE_DEPRECATED_API |
147 | |
148 | #include "unicode/utf.h" |
149 | #include "unicode/utf8.h" |
150 | #include "unicode/utf16.h" |
151 | |
152 | /* Formerly utf.h, part 1 --------------------------------------------------- */ |
153 | |
154 | #ifdef U_USE_UTF_DEPRECATES |
155 | /** |
156 | * Unicode string and array offset and index type. |
157 | * ICU always counts Unicode code units (UChars) for |
158 | * string offsets, indexes, and lengths, not Unicode code points. |
159 | * |
160 | * @obsolete ICU 2.6. Use int32_t directly instead since this API will be removed in that release. |
161 | */ |
162 | typedef int32_t UTextOffset; |
163 | #endif |
164 | |
165 | /** Number of bits in a Unicode string code unit - ICU uses 16-bit Unicode. @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
166 | #define UTF_SIZE 16 |
167 | |
168 | /** |
169 | * The default choice for general Unicode string macros is to use the ..._SAFE macro implementations |
170 | * with strict=FALSE. |
171 | * |
172 | * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
173 | */ |
174 | #define UTF_SAFE |
175 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
176 | #undef UTF_UNSAFE |
177 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
178 | #undef UTF_STRICT |
179 | |
180 | /** |
181 | * UTF8_ERROR_VALUE_1 and UTF8_ERROR_VALUE_2 are special error values for UTF-8, |
182 | * which need 1 or 2 bytes in UTF-8: |
183 | * \code |
184 | * U+0015 = NAK = Negative Acknowledge, C0 control character |
185 | * U+009f = highest C1 control character |
186 | * \endcode |
187 | * |
188 | * These are used by UTF8_..._SAFE macros so that they can return an error value |
189 | * that needs the same number of code units (bytes) as were seen by |
190 | * a macro. They should be tested with UTF_IS_ERROR() or UTF_IS_VALID(). |
191 | * |
192 | * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
193 | */ |
194 | #define UTF8_ERROR_VALUE_1 0x15 |
195 | |
196 | /** |
197 | * See documentation on UTF8_ERROR_VALUE_1 for details. |
198 | * |
199 | * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
200 | */ |
201 | #define UTF8_ERROR_VALUE_2 0x9f |
202 | |
203 | /** |
204 | * Error value for all UTFs. This code point value will be set by macros with error |
205 | * checking if an error is detected. |
206 | * |
207 | * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
208 | */ |
209 | #define UTF_ERROR_VALUE 0xffff |
210 | |
211 | /** |
212 | * Is a given 32-bit code an error value |
213 | * as returned by one of the macros for any UTF? |
214 | * |
215 | * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
216 | */ |
217 | #define UTF_IS_ERROR(c) \ |
218 | (((c)&0xfffe)==0xfffe || (c)==UTF8_ERROR_VALUE_1 || (c)==UTF8_ERROR_VALUE_2) |
219 | |
220 | /** |
221 | * This is a combined macro: Is c a valid Unicode value _and_ not an error code? |
222 | * |
223 | * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
224 | */ |
225 | #define UTF_IS_VALID(c) \ |
226 | (UTF_IS_UNICODE_CHAR(c) && \ |
227 | (c)!=UTF8_ERROR_VALUE_1 && (c)!=UTF8_ERROR_VALUE_2) |
228 | |
229 | /** |
230 | * Is this code unit or code point a surrogate (U+d800..U+dfff)? |
231 | * @deprecated ICU 2.4. Renamed to U_IS_SURROGATE and U16_IS_SURROGATE, see utf_old.h. |
232 | */ |
233 | #define UTF_IS_SURROGATE(uchar) (((uchar)&0xfffff800)==0xd800) |
234 | |
235 | /** |
236 | * Is a given 32-bit code point a Unicode noncharacter? |
237 | * |
238 | * @deprecated ICU 2.4. Renamed to U_IS_UNICODE_NONCHAR, see utf_old.h. |
239 | */ |
240 | #define UTF_IS_UNICODE_NONCHAR(c) \ |
241 | ((c)>=0xfdd0 && \ |
242 | ((uint32_t)(c)<=0xfdef || ((c)&0xfffe)==0xfffe) && \ |
243 | (uint32_t)(c)<=0x10ffff) |
244 | |
245 | /** |
246 | * Is a given 32-bit value a Unicode code point value (0..U+10ffff) |
247 | * that can be assigned a character? |
248 | * |
249 | * Code points that are not characters include: |
250 | * - single surrogate code points (U+d800..U+dfff, 2048 code points) |
251 | * - the last two code points on each plane (U+__fffe and U+__ffff, 34 code points) |
252 | * - U+fdd0..U+fdef (new with Unicode 3.1, 32 code points) |
253 | * - the highest Unicode code point value is U+10ffff |
254 | * |
255 | * This means that all code points below U+d800 are character code points, |
256 | * and that boundary is tested first for performance. |
257 | * |
258 | * @deprecated ICU 2.4. Renamed to U_IS_UNICODE_CHAR, see utf_old.h. |
259 | */ |
260 | #define UTF_IS_UNICODE_CHAR(c) \ |
261 | ((uint32_t)(c)<0xd800 || \ |
262 | ((uint32_t)(c)>0xdfff && \ |
263 | (uint32_t)(c)<=0x10ffff && \ |
264 | !UTF_IS_UNICODE_NONCHAR(c))) |
265 | |
266 | /* Formerly utf8.h ---------------------------------------------------------- */ |
267 | |
268 | /** |
269 | * Count the trail bytes for a UTF-8 lead byte. |
270 | * @deprecated ICU 2.4. Renamed to U8_COUNT_TRAIL_BYTES, see utf_old.h. |
271 | */ |
272 | #define UTF8_COUNT_TRAIL_BYTES(leadByte) (utf8_countTrailBytes[(uint8_t)leadByte]) |
273 | |
274 | /** |
275 | * Mask a UTF-8 lead byte, leave only the lower bits that form part of the code point value. |
276 | * @deprecated ICU 2.4. Renamed to U8_MASK_LEAD_BYTE, see utf_old.h. |
277 | */ |
278 | #define UTF8_MASK_LEAD_BYTE(leadByte, countTrailBytes) ((leadByte)&=(1<<(6-(countTrailBytes)))-1) |
279 | |
280 | /** Is this this code point a single code unit (byte)? @deprecated ICU 2.4. Renamed to U8_IS_SINGLE, see utf_old.h. */ |
281 | #define UTF8_IS_SINGLE(uchar) (((uchar)&0x80)==0) |
282 | /** Is this this code unit the lead code unit (byte) of a code point? @deprecated ICU 2.4. Renamed to U8_IS_LEAD, see utf_old.h. */ |
283 | #define UTF8_IS_LEAD(uchar) ((uint8_t)((uchar)-0xc0)<0x3e) |
284 | /** Is this this code unit a trailing code unit (byte) of a code point? @deprecated ICU 2.4. Renamed to U8_IS_TRAIL, see utf_old.h. */ |
285 | #define UTF8_IS_TRAIL(uchar) (((uchar)&0xc0)==0x80) |
286 | |
287 | /** Does this scalar Unicode value need multiple code units for storage? @deprecated ICU 2.4. Use U8_LENGTH or test ((uint32_t)(c)>0x7f) instead, see utf_old.h. */ |
288 | #define UTF8_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0x7f) |
289 | |
290 | /** |
291 | * Given the lead character, how many bytes are taken by this code point. |
292 | * ICU does not deal with code points >0x10ffff |
293 | * unless necessary for advancing in the byte stream. |
294 | * |
295 | * These length macros take into account that for values >0x10ffff |
296 | * the UTF8_APPEND_CHAR_SAFE macros would write the error code point 0xffff |
297 | * with 3 bytes. |
298 | * Code point comparisons need to be in uint32_t because UChar32 |
299 | * may be a signed type, and negative values must be recognized. |
300 | * |
301 | * @deprecated ICU 2.4. Use U8_LENGTH instead, see utf.h. |
302 | */ |
303 | #if 1 |
304 | # define UTF8_CHAR_LENGTH(c) \ |
305 | ((uint32_t)(c)<=0x7f ? 1 : \ |
306 | ((uint32_t)(c)<=0x7ff ? 2 : \ |
307 | ((uint32_t)((c)-0x10000)>0xfffff ? 3 : 4) \ |
308 | ) \ |
309 | ) |
310 | #else |
311 | # define UTF8_CHAR_LENGTH(c) \ |
312 | ((uint32_t)(c)<=0x7f ? 1 : \ |
313 | ((uint32_t)(c)<=0x7ff ? 2 : \ |
314 | ((uint32_t)(c)<=0xffff ? 3 : \ |
315 | ((uint32_t)(c)<=0x10ffff ? 4 : \ |
316 | ((uint32_t)(c)<=0x3ffffff ? 5 : \ |
317 | ((uint32_t)(c)<=0x7fffffff ? 6 : 3) \ |
318 | ) \ |
319 | ) \ |
320 | ) \ |
321 | ) \ |
322 | ) |
323 | #endif |
324 | |
325 | /** The maximum number of bytes per code point. @deprecated ICU 2.4. Renamed to U8_MAX_LENGTH, see utf_old.h. */ |
326 | #define UTF8_MAX_CHAR_LENGTH 4 |
327 | |
328 | /** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
329 | #define UTF8_ARRAY_SIZE(size) ((5*(size))/2) |
330 | |
331 | /** @deprecated ICU 2.4. Renamed to U8_GET_UNSAFE, see utf_old.h. */ |
332 | #define UTF8_GET_CHAR_UNSAFE(s, i, c) { \ |
333 | int32_t _utf8_get_char_unsafe_index=(int32_t)(i); \ |
334 | UTF8_SET_CHAR_START_UNSAFE(s, _utf8_get_char_unsafe_index); \ |
335 | UTF8_NEXT_CHAR_UNSAFE(s, _utf8_get_char_unsafe_index, c); \ |
336 | } |
337 | |
338 | /** @deprecated ICU 2.4. Use U8_GET instead, see utf_old.h. */ |
339 | #define UTF8_GET_CHAR_SAFE(s, start, i, length, c, strict) { \ |
340 | int32_t _utf8_get_char_safe_index=(int32_t)(i); \ |
341 | UTF8_SET_CHAR_START_SAFE(s, start, _utf8_get_char_safe_index); \ |
342 | UTF8_NEXT_CHAR_SAFE(s, _utf8_get_char_safe_index, length, c, strict); \ |
343 | } |
344 | |
345 | /** @deprecated ICU 2.4. Renamed to U8_NEXT_UNSAFE, see utf_old.h. */ |
346 | #define UTF8_NEXT_CHAR_UNSAFE(s, i, c) { \ |
347 | (c)=(s)[(i)++]; \ |
348 | if((uint8_t)((c)-0xc0)<0x35) { \ |
349 | uint8_t __count=UTF8_COUNT_TRAIL_BYTES(c); \ |
350 | UTF8_MASK_LEAD_BYTE(c, __count); \ |
351 | switch(__count) { \ |
352 | /* each following branch falls through to the next one */ \ |
353 | case 3: \ |
354 | (c)=((c)<<6)|((s)[(i)++]&0x3f); \ |
355 | case 2: \ |
356 | (c)=((c)<<6)|((s)[(i)++]&0x3f); \ |
357 | case 1: \ |
358 | (c)=((c)<<6)|((s)[(i)++]&0x3f); \ |
359 | /* no other branches to optimize switch() */ \ |
360 | break; \ |
361 | } \ |
362 | } \ |
363 | } |
364 | |
365 | /** @deprecated ICU 2.4. Renamed to U8_APPEND_UNSAFE, see utf_old.h. */ |
366 | #define UTF8_APPEND_CHAR_UNSAFE(s, i, c) { \ |
367 | if((uint32_t)(c)<=0x7f) { \ |
368 | (s)[(i)++]=(uint8_t)(c); \ |
369 | } else { \ |
370 | if((uint32_t)(c)<=0x7ff) { \ |
371 | (s)[(i)++]=(uint8_t)(((c)>>6)|0xc0); \ |
372 | } else { \ |
373 | if((uint32_t)(c)<=0xffff) { \ |
374 | (s)[(i)++]=(uint8_t)(((c)>>12)|0xe0); \ |
375 | } else { \ |
376 | (s)[(i)++]=(uint8_t)(((c)>>18)|0xf0); \ |
377 | (s)[(i)++]=(uint8_t)((((c)>>12)&0x3f)|0x80); \ |
378 | } \ |
379 | (s)[(i)++]=(uint8_t)((((c)>>6)&0x3f)|0x80); \ |
380 | } \ |
381 | (s)[(i)++]=(uint8_t)(((c)&0x3f)|0x80); \ |
382 | } \ |
383 | } |
384 | |
385 | /** @deprecated ICU 2.4. Renamed to U8_FWD_1_UNSAFE, see utf_old.h. */ |
386 | #define UTF8_FWD_1_UNSAFE(s, i) { \ |
387 | (i)+=1+UTF8_COUNT_TRAIL_BYTES((s)[i]); \ |
388 | } |
389 | |
390 | /** @deprecated ICU 2.4. Renamed to U8_FWD_N_UNSAFE, see utf_old.h. */ |
391 | #define UTF8_FWD_N_UNSAFE(s, i, n) { \ |
392 | int32_t __N=(n); \ |
393 | while(__N>0) { \ |
394 | UTF8_FWD_1_UNSAFE(s, i); \ |
395 | --__N; \ |
396 | } \ |
397 | } |
398 | |
399 | /** @deprecated ICU 2.4. Renamed to U8_SET_CP_START_UNSAFE, see utf_old.h. */ |
400 | #define UTF8_SET_CHAR_START_UNSAFE(s, i) { \ |
401 | while(UTF8_IS_TRAIL((s)[i])) { --(i); } \ |
402 | } |
403 | |
404 | /** @deprecated ICU 2.4. Use U8_NEXT instead, see utf_old.h. */ |
405 | #define UTF8_NEXT_CHAR_SAFE(s, i, length, c, strict) { \ |
406 | (c)=(s)[(i)++]; \ |
407 | if((c)>=0x80) { \ |
408 | if(UTF8_IS_LEAD(c)) { \ |
409 | (c)=utf8_nextCharSafeBody(s, &(i), (int32_t)(length), c, strict); \ |
410 | } else { \ |
411 | (c)=UTF8_ERROR_VALUE_1; \ |
412 | } \ |
413 | } \ |
414 | } |
415 | |
416 | /** @deprecated ICU 2.4. Use U8_APPEND instead, see utf_old.h. */ |
417 | #define UTF8_APPEND_CHAR_SAFE(s, i, length, c) { \ |
418 | if((uint32_t)(c)<=0x7f) { \ |
419 | (s)[(i)++]=(uint8_t)(c); \ |
420 | } else { \ |
421 | (i)=utf8_appendCharSafeBody(s, (int32_t)(i), (int32_t)(length), c, NULL); \ |
422 | } \ |
423 | } |
424 | |
425 | /** @deprecated ICU 2.4. Renamed to U8_FWD_1, see utf_old.h. */ |
426 | #define UTF8_FWD_1_SAFE(s, i, length) U8_FWD_1(s, i, length) |
427 | |
428 | /** @deprecated ICU 2.4. Renamed to U8_FWD_N, see utf_old.h. */ |
429 | #define UTF8_FWD_N_SAFE(s, i, length, n) U8_FWD_N(s, i, length, n) |
430 | |
431 | /** @deprecated ICU 2.4. Renamed to U8_SET_CP_START, see utf_old.h. */ |
432 | #define UTF8_SET_CHAR_START_SAFE(s, start, i) U8_SET_CP_START(s, start, i) |
433 | |
434 | /** @deprecated ICU 2.4. Renamed to U8_PREV_UNSAFE, see utf_old.h. */ |
435 | #define UTF8_PREV_CHAR_UNSAFE(s, i, c) { \ |
436 | (c)=(s)[--(i)]; \ |
437 | if(UTF8_IS_TRAIL(c)) { \ |
438 | uint8_t __b, __count=1, __shift=6; \ |
439 | \ |
440 | /* c is a trail byte */ \ |
441 | (c)&=0x3f; \ |
442 | for(;;) { \ |
443 | __b=(s)[--(i)]; \ |
444 | if(__b>=0xc0) { \ |
445 | UTF8_MASK_LEAD_BYTE(__b, __count); \ |
446 | (c)|=(UChar32)__b<<__shift; \ |
447 | break; \ |
448 | } else { \ |
449 | (c)|=(UChar32)(__b&0x3f)<<__shift; \ |
450 | ++__count; \ |
451 | __shift+=6; \ |
452 | } \ |
453 | } \ |
454 | } \ |
455 | } |
456 | |
457 | /** @deprecated ICU 2.4. Renamed to U8_BACK_1_UNSAFE, see utf_old.h. */ |
458 | #define UTF8_BACK_1_UNSAFE(s, i) { \ |
459 | while(UTF8_IS_TRAIL((s)[--(i)])) {} \ |
460 | } |
461 | |
462 | /** @deprecated ICU 2.4. Renamed to U8_BACK_N_UNSAFE, see utf_old.h. */ |
463 | #define UTF8_BACK_N_UNSAFE(s, i, n) { \ |
464 | int32_t __N=(n); \ |
465 | while(__N>0) { \ |
466 | UTF8_BACK_1_UNSAFE(s, i); \ |
467 | --__N; \ |
468 | } \ |
469 | } |
470 | |
471 | /** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT_UNSAFE, see utf_old.h. */ |
472 | #define UTF8_SET_CHAR_LIMIT_UNSAFE(s, i) { \ |
473 | UTF8_BACK_1_UNSAFE(s, i); \ |
474 | UTF8_FWD_1_UNSAFE(s, i); \ |
475 | } |
476 | |
477 | /** @deprecated ICU 2.4. Use U8_PREV instead, see utf_old.h. */ |
478 | #define UTF8_PREV_CHAR_SAFE(s, start, i, c, strict) { \ |
479 | (c)=(s)[--(i)]; \ |
480 | if((c)>=0x80) { \ |
481 | if((c)<=0xbf) { \ |
482 | (c)=utf8_prevCharSafeBody(s, start, &(i), c, strict); \ |
483 | } else { \ |
484 | (c)=UTF8_ERROR_VALUE_1; \ |
485 | } \ |
486 | } \ |
487 | } |
488 | |
489 | /** @deprecated ICU 2.4. Renamed to U8_BACK_1, see utf_old.h. */ |
490 | #define UTF8_BACK_1_SAFE(s, start, i) U8_BACK_1(s, start, i) |
491 | |
492 | /** @deprecated ICU 2.4. Renamed to U8_BACK_N, see utf_old.h. */ |
493 | #define UTF8_BACK_N_SAFE(s, start, i, n) U8_BACK_N(s, start, i, n) |
494 | |
495 | /** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT, see utf_old.h. */ |
496 | #define UTF8_SET_CHAR_LIMIT_SAFE(s, start, i, length) U8_SET_CP_LIMIT(s, start, i, length) |
497 | |
498 | /* Formerly utf16.h --------------------------------------------------------- */ |
499 | |
500 | /** Is uchar a first/lead surrogate? @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h. */ |
501 | #define UTF_IS_FIRST_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xd800) |
502 | |
503 | /** Is uchar a second/trail surrogate? @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h. */ |
504 | #define UTF_IS_SECOND_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xdc00) |
505 | |
506 | /** Assuming c is a surrogate, is it a first/lead surrogate? @deprecated ICU 2.4. Renamed to U_IS_SURROGATE_LEAD and U16_IS_SURROGATE_LEAD, see utf_old.h. */ |
507 | #define UTF_IS_SURROGATE_FIRST(c) (((c)&0x400)==0) |
508 | |
509 | /** Helper constant for UTF16_GET_PAIR_VALUE. @deprecated ICU 2.4. Renamed to U16_SURROGATE_OFFSET, see utf_old.h. */ |
510 | #define UTF_SURROGATE_OFFSET ((0xd800<<10UL)+0xdc00-0x10000) |
511 | |
512 | /** Get the UTF-32 value from the surrogate code units. @deprecated ICU 2.4. Renamed to U16_GET_SUPPLEMENTARY, see utf_old.h. */ |
513 | #define UTF16_GET_PAIR_VALUE(first, second) \ |
514 | (((first)<<10UL)+(second)-UTF_SURROGATE_OFFSET) |
515 | |
516 | /** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */ |
517 | #define UTF_FIRST_SURROGATE(supplementary) (UChar)(((supplementary)>>10)+0xd7c0) |
518 | |
519 | /** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */ |
520 | #define UTF_SECOND_SURROGATE(supplementary) (UChar)(((supplementary)&0x3ff)|0xdc00) |
521 | |
522 | /** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */ |
523 | #define UTF16_LEAD(supplementary) UTF_FIRST_SURROGATE(supplementary) |
524 | |
525 | /** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */ |
526 | #define UTF16_TRAIL(supplementary) UTF_SECOND_SURROGATE(supplementary) |
527 | |
528 | /** @deprecated ICU 2.4. Renamed to U16_IS_SINGLE, see utf_old.h. */ |
529 | #define UTF16_IS_SINGLE(uchar) !UTF_IS_SURROGATE(uchar) |
530 | |
531 | /** @deprecated ICU 2.4. Renamed to U16_IS_LEAD, see utf_old.h. */ |
532 | #define UTF16_IS_LEAD(uchar) UTF_IS_FIRST_SURROGATE(uchar) |
533 | |
534 | /** @deprecated ICU 2.4. Renamed to U16_IS_TRAIL, see utf_old.h. */ |
535 | #define UTF16_IS_TRAIL(uchar) UTF_IS_SECOND_SURROGATE(uchar) |
536 | |
537 | /** Does this scalar Unicode value need multiple code units for storage? @deprecated ICU 2.4. Use U16_LENGTH or test ((uint32_t)(c)>0xffff) instead, see utf_old.h. */ |
538 | #define UTF16_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0xffff) |
539 | |
540 | /** @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h. */ |
541 | #define UTF16_CHAR_LENGTH(c) ((uint32_t)(c)<=0xffff ? 1 : 2) |
542 | |
543 | /** @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h. */ |
544 | #define UTF16_MAX_CHAR_LENGTH 2 |
545 | |
546 | /** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
547 | #define UTF16_ARRAY_SIZE(size) (size) |
548 | |
549 | /** |
550 | * Get a single code point from an offset that points to any |
551 | * of the code units that belong to that code point. |
552 | * Assume 0<=i<length. |
553 | * |
554 | * This could be used for iteration together with |
555 | * UTF16_CHAR_LENGTH() and UTF_IS_ERROR(), |
556 | * but the use of UTF16_NEXT_CHAR[_UNSAFE]() and |
557 | * UTF16_PREV_CHAR[_UNSAFE]() is more efficient for that. |
558 | * @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h. |
559 | */ |
560 | #define UTF16_GET_CHAR_UNSAFE(s, i, c) { \ |
561 | (c)=(s)[i]; \ |
562 | if(UTF_IS_SURROGATE(c)) { \ |
563 | if(UTF_IS_SURROGATE_FIRST(c)) { \ |
564 | (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)+1]); \ |
565 | } else { \ |
566 | (c)=UTF16_GET_PAIR_VALUE((s)[(i)-1], (c)); \ |
567 | } \ |
568 | } \ |
569 | } |
570 | |
571 | /** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */ |
572 | #define UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict) { \ |
573 | (c)=(s)[i]; \ |
574 | if(UTF_IS_SURROGATE(c)) { \ |
575 | uint16_t __c2; \ |
576 | if(UTF_IS_SURROGATE_FIRST(c)) { \ |
577 | if((i)+1<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)+1])) { \ |
578 | (c)=UTF16_GET_PAIR_VALUE((c), __c2); \ |
579 | /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \ |
580 | } else if(strict) {\ |
581 | /* unmatched first surrogate */ \ |
582 | (c)=UTF_ERROR_VALUE; \ |
583 | } \ |
584 | } else { \ |
585 | if((i)-1>=(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \ |
586 | (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \ |
587 | /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \ |
588 | } else if(strict) {\ |
589 | /* unmatched second surrogate */ \ |
590 | (c)=UTF_ERROR_VALUE; \ |
591 | } \ |
592 | } \ |
593 | } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \ |
594 | (c)=UTF_ERROR_VALUE; \ |
595 | } \ |
596 | } |
597 | |
598 | /** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */ |
599 | #define UTF16_NEXT_CHAR_UNSAFE(s, i, c) { \ |
600 | (c)=(s)[(i)++]; \ |
601 | if(UTF_IS_FIRST_SURROGATE(c)) { \ |
602 | (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)++]); \ |
603 | } \ |
604 | } |
605 | |
606 | /** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */ |
607 | #define UTF16_APPEND_CHAR_UNSAFE(s, i, c) { \ |
608 | if((uint32_t)(c)<=0xffff) { \ |
609 | (s)[(i)++]=(uint16_t)(c); \ |
610 | } else { \ |
611 | (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ |
612 | (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ |
613 | } \ |
614 | } |
615 | |
616 | /** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */ |
617 | #define UTF16_FWD_1_UNSAFE(s, i) { \ |
618 | if(UTF_IS_FIRST_SURROGATE((s)[(i)++])) { \ |
619 | ++(i); \ |
620 | } \ |
621 | } |
622 | |
623 | /** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */ |
624 | #define UTF16_FWD_N_UNSAFE(s, i, n) { \ |
625 | int32_t __N=(n); \ |
626 | while(__N>0) { \ |
627 | UTF16_FWD_1_UNSAFE(s, i); \ |
628 | --__N; \ |
629 | } \ |
630 | } |
631 | |
632 | /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */ |
633 | #define UTF16_SET_CHAR_START_UNSAFE(s, i) { \ |
634 | if(UTF_IS_SECOND_SURROGATE((s)[i])) { \ |
635 | --(i); \ |
636 | } \ |
637 | } |
638 | |
639 | /** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */ |
640 | #define UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict) { \ |
641 | (c)=(s)[(i)++]; \ |
642 | if(UTF_IS_FIRST_SURROGATE(c)) { \ |
643 | uint16_t __c2; \ |
644 | if((i)<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)])) { \ |
645 | ++(i); \ |
646 | (c)=UTF16_GET_PAIR_VALUE((c), __c2); \ |
647 | /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \ |
648 | } else if(strict) {\ |
649 | /* unmatched first surrogate */ \ |
650 | (c)=UTF_ERROR_VALUE; \ |
651 | } \ |
652 | } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \ |
653 | /* unmatched second surrogate or other non-character */ \ |
654 | (c)=UTF_ERROR_VALUE; \ |
655 | } \ |
656 | } |
657 | |
658 | /** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */ |
659 | #define UTF16_APPEND_CHAR_SAFE(s, i, length, c) { \ |
660 | if((uint32_t)(c)<=0xffff) { \ |
661 | (s)[(i)++]=(uint16_t)(c); \ |
662 | } else if((uint32_t)(c)<=0x10ffff) { \ |
663 | if((i)+1<(length)) { \ |
664 | (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ |
665 | (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ |
666 | } else /* not enough space */ { \ |
667 | (s)[(i)++]=UTF_ERROR_VALUE; \ |
668 | } \ |
669 | } else /* c>0x10ffff, write error value */ { \ |
670 | (s)[(i)++]=UTF_ERROR_VALUE; \ |
671 | } \ |
672 | } |
673 | |
674 | /** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */ |
675 | #define UTF16_FWD_1_SAFE(s, i, length) U16_FWD_1(s, i, length) |
676 | |
677 | /** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */ |
678 | #define UTF16_FWD_N_SAFE(s, i, length, n) U16_FWD_N(s, i, length, n) |
679 | |
680 | /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */ |
681 | #define UTF16_SET_CHAR_START_SAFE(s, start, i) U16_SET_CP_START(s, start, i) |
682 | |
683 | /** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */ |
684 | #define UTF16_PREV_CHAR_UNSAFE(s, i, c) { \ |
685 | (c)=(s)[--(i)]; \ |
686 | if(UTF_IS_SECOND_SURROGATE(c)) { \ |
687 | (c)=UTF16_GET_PAIR_VALUE((s)[--(i)], (c)); \ |
688 | } \ |
689 | } |
690 | |
691 | /** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */ |
692 | #define UTF16_BACK_1_UNSAFE(s, i) { \ |
693 | if(UTF_IS_SECOND_SURROGATE((s)[--(i)])) { \ |
694 | --(i); \ |
695 | } \ |
696 | } |
697 | |
698 | /** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */ |
699 | #define UTF16_BACK_N_UNSAFE(s, i, n) { \ |
700 | int32_t __N=(n); \ |
701 | while(__N>0) { \ |
702 | UTF16_BACK_1_UNSAFE(s, i); \ |
703 | --__N; \ |
704 | } \ |
705 | } |
706 | |
707 | /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */ |
708 | #define UTF16_SET_CHAR_LIMIT_UNSAFE(s, i) { \ |
709 | if(UTF_IS_FIRST_SURROGATE((s)[(i)-1])) { \ |
710 | ++(i); \ |
711 | } \ |
712 | } |
713 | |
714 | /** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */ |
715 | #define UTF16_PREV_CHAR_SAFE(s, start, i, c, strict) { \ |
716 | (c)=(s)[--(i)]; \ |
717 | if(UTF_IS_SECOND_SURROGATE(c)) { \ |
718 | uint16_t __c2; \ |
719 | if((i)>(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \ |
720 | --(i); \ |
721 | (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \ |
722 | /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \ |
723 | } else if(strict) {\ |
724 | /* unmatched second surrogate */ \ |
725 | (c)=UTF_ERROR_VALUE; \ |
726 | } \ |
727 | } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \ |
728 | /* unmatched first surrogate or other non-character */ \ |
729 | (c)=UTF_ERROR_VALUE; \ |
730 | } \ |
731 | } |
732 | |
733 | /** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */ |
734 | #define UTF16_BACK_1_SAFE(s, start, i) U16_BACK_1(s, start, i) |
735 | |
736 | /** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */ |
737 | #define UTF16_BACK_N_SAFE(s, start, i, n) U16_BACK_N(s, start, i, n) |
738 | |
739 | /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */ |
740 | #define UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, length) |
741 | |
742 | /* Formerly utf32.h --------------------------------------------------------- */ |
743 | |
744 | /* |
745 | * Old documentation: |
746 | * |
747 | * This file defines macros to deal with UTF-32 code units and code points. |
748 | * Signatures and semantics are the same as for the similarly named macros |
749 | * in utf16.h. |
750 | * utf32.h is included by utf.h after unicode/umachine.h</p> |
751 | * and some common definitions. |
752 | * <p><b>Usage:</b> ICU coding guidelines for if() statements should be followed when using these macros. |
753 | * Compound statements (curly braces {}) must be used for if-else-while... |
754 | * bodies and all macro statements should be terminated with semicolon.</p> |
755 | */ |
756 | |
757 | /* internal definitions ----------------------------------------------------- */ |
758 | |
759 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
760 | #define UTF32_IS_SAFE(c, strict) \ |
761 | (!(strict) ? \ |
762 | (uint32_t)(c)<=0x10ffff : \ |
763 | UTF_IS_UNICODE_CHAR(c)) |
764 | |
765 | /* |
766 | * For the semantics of all of these macros, see utf16.h. |
767 | * The UTF-32 versions are trivial because any code point is |
768 | * encoded using exactly one code unit. |
769 | */ |
770 | |
771 | /* single-code point definitions -------------------------------------------- */ |
772 | |
773 | /* classes of code unit values */ |
774 | |
775 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
776 | #define UTF32_IS_SINGLE(uchar) 1 |
777 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
778 | #define UTF32_IS_LEAD(uchar) 0 |
779 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
780 | #define UTF32_IS_TRAIL(uchar) 0 |
781 | |
782 | /* number of code units per code point */ |
783 | |
784 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
785 | #define UTF32_NEED_MULTIPLE_UCHAR(c) 0 |
786 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
787 | #define UTF32_CHAR_LENGTH(c) 1 |
788 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
789 | #define UTF32_MAX_CHAR_LENGTH 1 |
790 | |
791 | /* average number of code units compared to UTF-16 */ |
792 | |
793 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
794 | #define UTF32_ARRAY_SIZE(size) (size) |
795 | |
796 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
797 | #define UTF32_GET_CHAR_UNSAFE(s, i, c) { \ |
798 | (c)=(s)[i]; \ |
799 | } |
800 | |
801 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
802 | #define UTF32_GET_CHAR_SAFE(s, start, i, length, c, strict) { \ |
803 | (c)=(s)[i]; \ |
804 | if(!UTF32_IS_SAFE(c, strict)) { \ |
805 | (c)=UTF_ERROR_VALUE; \ |
806 | } \ |
807 | } |
808 | |
809 | /* definitions with forward iteration --------------------------------------- */ |
810 | |
811 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
812 | #define UTF32_NEXT_CHAR_UNSAFE(s, i, c) { \ |
813 | (c)=(s)[(i)++]; \ |
814 | } |
815 | |
816 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
817 | #define UTF32_APPEND_CHAR_UNSAFE(s, i, c) { \ |
818 | (s)[(i)++]=(c); \ |
819 | } |
820 | |
821 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
822 | #define UTF32_FWD_1_UNSAFE(s, i) { \ |
823 | ++(i); \ |
824 | } |
825 | |
826 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
827 | #define UTF32_FWD_N_UNSAFE(s, i, n) { \ |
828 | (i)+=(n); \ |
829 | } |
830 | |
831 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
832 | #define UTF32_SET_CHAR_START_UNSAFE(s, i) { \ |
833 | } |
834 | |
835 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
836 | #define UTF32_NEXT_CHAR_SAFE(s, i, length, c, strict) { \ |
837 | (c)=(s)[(i)++]; \ |
838 | if(!UTF32_IS_SAFE(c, strict)) { \ |
839 | (c)=UTF_ERROR_VALUE; \ |
840 | } \ |
841 | } |
842 | |
843 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
844 | #define UTF32_APPEND_CHAR_SAFE(s, i, length, c) { \ |
845 | if((uint32_t)(c)<=0x10ffff) { \ |
846 | (s)[(i)++]=(c); \ |
847 | } else /* c>0x10ffff, write 0xfffd */ { \ |
848 | (s)[(i)++]=0xfffd; \ |
849 | } \ |
850 | } |
851 | |
852 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
853 | #define UTF32_FWD_1_SAFE(s, i, length) { \ |
854 | ++(i); \ |
855 | } |
856 | |
857 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
858 | #define UTF32_FWD_N_SAFE(s, i, length, n) { \ |
859 | if(((i)+=(n))>(length)) { \ |
860 | (i)=(length); \ |
861 | } \ |
862 | } |
863 | |
864 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
865 | #define UTF32_SET_CHAR_START_SAFE(s, start, i) { \ |
866 | } |
867 | |
868 | /* definitions with backward iteration -------------------------------------- */ |
869 | |
870 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
871 | #define UTF32_PREV_CHAR_UNSAFE(s, i, c) { \ |
872 | (c)=(s)[--(i)]; \ |
873 | } |
874 | |
875 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
876 | #define UTF32_BACK_1_UNSAFE(s, i) { \ |
877 | --(i); \ |
878 | } |
879 | |
880 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
881 | #define UTF32_BACK_N_UNSAFE(s, i, n) { \ |
882 | (i)-=(n); \ |
883 | } |
884 | |
885 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
886 | #define UTF32_SET_CHAR_LIMIT_UNSAFE(s, i) { \ |
887 | } |
888 | |
889 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
890 | #define UTF32_PREV_CHAR_SAFE(s, start, i, c, strict) { \ |
891 | (c)=(s)[--(i)]; \ |
892 | if(!UTF32_IS_SAFE(c, strict)) { \ |
893 | (c)=UTF_ERROR_VALUE; \ |
894 | } \ |
895 | } |
896 | |
897 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
898 | #define UTF32_BACK_1_SAFE(s, start, i) { \ |
899 | --(i); \ |
900 | } |
901 | |
902 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
903 | #define UTF32_BACK_N_SAFE(s, start, i, n) { \ |
904 | (i)-=(n); \ |
905 | if((i)<(start)) { \ |
906 | (i)=(start); \ |
907 | } \ |
908 | } |
909 | |
910 | /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
911 | #define UTF32_SET_CHAR_LIMIT_SAFE(s, i, length) { \ |
912 | } |
913 | |
914 | /* Formerly utf.h, part 2 --------------------------------------------------- */ |
915 | |
916 | /** |
917 | * Estimate the number of code units for a string based on the number of UTF-16 code units. |
918 | * |
919 | * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
920 | */ |
921 | #define UTF_ARRAY_SIZE(size) UTF16_ARRAY_SIZE(size) |
922 | |
923 | /** @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h. */ |
924 | #define UTF_GET_CHAR_UNSAFE(s, i, c) UTF16_GET_CHAR_UNSAFE(s, i, c) |
925 | |
926 | /** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */ |
927 | #define UTF_GET_CHAR_SAFE(s, start, i, length, c, strict) UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict) |
928 | |
929 | |
930 | /** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */ |
931 | #define UTF_NEXT_CHAR_UNSAFE(s, i, c) UTF16_NEXT_CHAR_UNSAFE(s, i, c) |
932 | |
933 | /** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */ |
934 | #define UTF_NEXT_CHAR_SAFE(s, i, length, c, strict) UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict) |
935 | |
936 | |
937 | /** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */ |
938 | #define UTF_APPEND_CHAR_UNSAFE(s, i, c) UTF16_APPEND_CHAR_UNSAFE(s, i, c) |
939 | |
940 | /** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */ |
941 | #define UTF_APPEND_CHAR_SAFE(s, i, length, c) UTF16_APPEND_CHAR_SAFE(s, i, length, c) |
942 | |
943 | |
944 | /** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */ |
945 | #define UTF_FWD_1_UNSAFE(s, i) UTF16_FWD_1_UNSAFE(s, i) |
946 | |
947 | /** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */ |
948 | #define UTF_FWD_1_SAFE(s, i, length) UTF16_FWD_1_SAFE(s, i, length) |
949 | |
950 | |
951 | /** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */ |
952 | #define UTF_FWD_N_UNSAFE(s, i, n) UTF16_FWD_N_UNSAFE(s, i, n) |
953 | |
954 | /** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */ |
955 | #define UTF_FWD_N_SAFE(s, i, length, n) UTF16_FWD_N_SAFE(s, i, length, n) |
956 | |
957 | |
958 | /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */ |
959 | #define UTF_SET_CHAR_START_UNSAFE(s, i) UTF16_SET_CHAR_START_UNSAFE(s, i) |
960 | |
961 | /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */ |
962 | #define UTF_SET_CHAR_START_SAFE(s, start, i) UTF16_SET_CHAR_START_SAFE(s, start, i) |
963 | |
964 | |
965 | /** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */ |
966 | #define UTF_PREV_CHAR_UNSAFE(s, i, c) UTF16_PREV_CHAR_UNSAFE(s, i, c) |
967 | |
968 | /** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */ |
969 | #define UTF_PREV_CHAR_SAFE(s, start, i, c, strict) UTF16_PREV_CHAR_SAFE(s, start, i, c, strict) |
970 | |
971 | |
972 | /** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */ |
973 | #define UTF_BACK_1_UNSAFE(s, i) UTF16_BACK_1_UNSAFE(s, i) |
974 | |
975 | /** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */ |
976 | #define UTF_BACK_1_SAFE(s, start, i) UTF16_BACK_1_SAFE(s, start, i) |
977 | |
978 | |
979 | /** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */ |
980 | #define UTF_BACK_N_UNSAFE(s, i, n) UTF16_BACK_N_UNSAFE(s, i, n) |
981 | |
982 | /** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */ |
983 | #define UTF_BACK_N_SAFE(s, start, i, n) UTF16_BACK_N_SAFE(s, start, i, n) |
984 | |
985 | |
986 | /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */ |
987 | #define UTF_SET_CHAR_LIMIT_UNSAFE(s, i) UTF16_SET_CHAR_LIMIT_UNSAFE(s, i) |
988 | |
989 | /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */ |
990 | #define UTF_SET_CHAR_LIMIT_SAFE(s, start, i, length) UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length) |
991 | |
992 | /* Define default macros (UTF-16 "safe") ------------------------------------ */ |
993 | |
994 | /** |
995 | * Does this code unit alone encode a code point (BMP, not a surrogate)? |
996 | * Same as UTF16_IS_SINGLE. |
997 | * @deprecated ICU 2.4. Renamed to U_IS_SINGLE and U16_IS_SINGLE, see utf_old.h. |
998 | */ |
999 | #define UTF_IS_SINGLE(uchar) U16_IS_SINGLE(uchar) |
1000 | |
1001 | /** |
1002 | * Is this code unit the first one of several (a lead surrogate)? |
1003 | * Same as UTF16_IS_LEAD. |
1004 | * @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h. |
1005 | */ |
1006 | #define UTF_IS_LEAD(uchar) U16_IS_LEAD(uchar) |
1007 | |
1008 | /** |
1009 | * Is this code unit one of several but not the first one (a trail surrogate)? |
1010 | * Same as UTF16_IS_TRAIL. |
1011 | * @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h. |
1012 | */ |
1013 | #define UTF_IS_TRAIL(uchar) U16_IS_TRAIL(uchar) |
1014 | |
1015 | /** |
1016 | * Does this code point require multiple code units (is it a supplementary code point)? |
1017 | * Same as UTF16_NEED_MULTIPLE_UCHAR. |
1018 | * @deprecated ICU 2.4. Use U16_LENGTH or test ((uint32_t)(c)>0xffff) instead. |
1019 | */ |
1020 | #define UTF_NEED_MULTIPLE_UCHAR(c) UTF16_NEED_MULTIPLE_UCHAR(c) |
1021 | |
1022 | /** |
1023 | * How many code units are used to encode this code point (1 or 2)? |
1024 | * Same as UTF16_CHAR_LENGTH. |
1025 | * @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h. |
1026 | */ |
1027 | #define UTF_CHAR_LENGTH(c) U16_LENGTH(c) |
1028 | |
1029 | /** |
1030 | * How many code units are used at most for any Unicode code point (2)? |
1031 | * Same as UTF16_MAX_CHAR_LENGTH. |
1032 | * @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h. |
1033 | */ |
1034 | #define UTF_MAX_CHAR_LENGTH U16_MAX_LENGTH |
1035 | |
1036 | /** |
1037 | * Set c to the code point that contains the code unit i. |
1038 | * i could point to the lead or the trail surrogate for the code point. |
1039 | * i is not modified. |
1040 | * Same as UTF16_GET_CHAR. |
1041 | * \pre 0<=i<length |
1042 | * |
1043 | * @deprecated ICU 2.4. Renamed to U16_GET, see utf_old.h. |
1044 | */ |
1045 | #define UTF_GET_CHAR(s, start, i, length, c) U16_GET(s, start, i, length, c) |
1046 | |
1047 | /** |
1048 | * Set c to the code point that starts at code unit i |
1049 | * and advance i to beyond the code units of this code point (post-increment). |
1050 | * i must point to the first code unit of a code point. |
1051 | * Otherwise c is set to the trail unit (surrogate) itself. |
1052 | * Same as UTF16_NEXT_CHAR. |
1053 | * \pre 0<=i<length |
1054 | * \post 0<i<=length |
1055 | * |
1056 | * @deprecated ICU 2.4. Renamed to U16_NEXT, see utf_old.h. |
1057 | */ |
1058 | #define UTF_NEXT_CHAR(s, i, length, c) U16_NEXT(s, i, length, c) |
1059 | |
1060 | /** |
1061 | * Append the code units of code point c to the string at index i |
1062 | * and advance i to beyond the new code units (post-increment). |
1063 | * The code units beginning at index i will be overwritten. |
1064 | * Same as UTF16_APPEND_CHAR. |
1065 | * \pre 0<=c<=0x10ffff |
1066 | * \pre 0<=i<length |
1067 | * \post 0<i<=length |
1068 | * |
1069 | * @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. |
1070 | */ |
1071 | #define UTF_APPEND_CHAR(s, i, length, c) UTF16_APPEND_CHAR_SAFE(s, i, length, c) |
1072 | |
1073 | /** |
1074 | * Advance i to beyond the code units of the code point that begins at i. |
1075 | * I.e., advance i by one code point. |
1076 | * Same as UTF16_FWD_1. |
1077 | * \pre 0<=i<length |
1078 | * \post 0<i<=length |
1079 | * |
1080 | * @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. |
1081 | */ |
1082 | #define UTF_FWD_1(s, i, length) U16_FWD_1(s, i, length) |
1083 | |
1084 | /** |
1085 | * Advance i to beyond the code units of the n code points where the first one begins at i. |
1086 | * I.e., advance i by n code points. |
1087 | * Same as UT16_FWD_N. |
1088 | * \pre 0<=i<length |
1089 | * \post 0<i<=length |
1090 | * |
1091 | * @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. |
1092 | */ |
1093 | #define UTF_FWD_N(s, i, length, n) U16_FWD_N(s, i, length, n) |
1094 | |
1095 | /** |
1096 | * Take the random-access index i and adjust it so that it points to the beginning |
1097 | * of a code point. |
1098 | * The input index points to any code unit of a code point and is moved to point to |
1099 | * the first code unit of the same code point. i is never incremented. |
1100 | * In other words, if i points to a trail surrogate that is preceded by a matching |
1101 | * lead surrogate, then i is decremented. Otherwise it is not modified. |
1102 | * This can be used to start an iteration with UTF_NEXT_CHAR() from a random index. |
1103 | * Same as UTF16_SET_CHAR_START. |
1104 | * \pre start<=i<length |
1105 | * \post start<=i<length |
1106 | * |
1107 | * @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. |
1108 | */ |
1109 | #define UTF_SET_CHAR_START(s, start, i) U16_SET_CP_START(s, start, i) |
1110 | |
1111 | /** |
1112 | * Set c to the code point that has code units before i |
1113 | * and move i backward (towards the beginning of the string) |
1114 | * to the first code unit of this code point (pre-increment). |
1115 | * i must point to the first code unit after the last unit of a code point (i==length is allowed). |
1116 | * Same as UTF16_PREV_CHAR. |
1117 | * \pre start<i<=length |
1118 | * \post start<=i<length |
1119 | * |
1120 | * @deprecated ICU 2.4. Renamed to U16_PREV, see utf_old.h. |
1121 | */ |
1122 | #define UTF_PREV_CHAR(s, start, i, c) U16_PREV(s, start, i, c) |
1123 | |
1124 | /** |
1125 | * Move i backward (towards the beginning of the string) |
1126 | * to the first code unit of the code point that has code units before i. |
1127 | * I.e., move i backward by one code point. |
1128 | * i must point to the first code unit after the last unit of a code point (i==length is allowed). |
1129 | * Same as UTF16_BACK_1. |
1130 | * \pre start<i<=length |
1131 | * \post start<=i<length |
1132 | * |
1133 | * @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. |
1134 | */ |
1135 | #define UTF_BACK_1(s, start, i) U16_BACK_1(s, start, i) |
1136 | |
1137 | /** |
1138 | * Move i backward (towards the beginning of the string) |
1139 | * to the first code unit of the n code points that have code units before i. |
1140 | * I.e., move i backward by n code points. |
1141 | * i must point to the first code unit after the last unit of a code point (i==length is allowed). |
1142 | * Same as UTF16_BACK_N. |
1143 | * \pre start<i<=length |
1144 | * \post start<=i<length |
1145 | * |
1146 | * @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. |
1147 | */ |
1148 | #define UTF_BACK_N(s, start, i, n) U16_BACK_N(s, start, i, n) |
1149 | |
1150 | /** |
1151 | * Take the random-access index i and adjust it so that it points beyond |
1152 | * a code point. The input index points beyond any code unit |
1153 | * of a code point and is moved to point beyond the last code unit of the same |
1154 | * code point. i is never decremented. |
1155 | * In other words, if i points to a trail surrogate that is preceded by a matching |
1156 | * lead surrogate, then i is incremented. Otherwise it is not modified. |
1157 | * This can be used to start an iteration with UTF_PREV_CHAR() from a random index. |
1158 | * Same as UTF16_SET_CHAR_LIMIT. |
1159 | * \pre start<i<=length |
1160 | * \post start<i<=length |
1161 | * |
1162 | * @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. |
1163 | */ |
1164 | #define UTF_SET_CHAR_LIMIT(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, length) |
1165 | |
1166 | #endif /* U_HIDE_DEPRECATED_API */ |
1167 | |
1168 | #endif |
1169 | |
1170 | |