OLD | NEW |
(Empty) | |
| 1 /* |
| 2 ******************************************************************************* |
| 3 * |
| 4 * Copyright (C) 2002-2008, International Business Machines |
| 5 * Corporation and others. All Rights Reserved. |
| 6 * |
| 7 ******************************************************************************* |
| 8 * file name: utf.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 * utf_old.h is included by utf.h after unicode/umachine.h |
| 32 * and some common definitions, to not break old code. |
| 33 * |
| 34 * Brief summary of reasons for deprecation: |
| 35 * - Switch on UTF_SIZE (selection of UTF-8/16/32 default string processing) |
| 36 * was impractical. |
| 37 * - Switch on UTF_SAFE etc. (selection of unsafe/safe/strict default string pro
cessing) |
| 38 * was of little use and impractical. |
| 39 * - Whole classes of macros became obsolete outside of the UTF_SIZE/UTF_SAFE |
| 40 * selection framework: UTF32_ macros (all trivial) |
| 41 * and UTF_ default and intermediate macros (all aliases). |
| 42 * - The selection framework also caused many macro aliases. |
| 43 * - Change in Unicode standard: "irregular" sequences (3.0) became illegal (3.2
). |
| 44 * - Change of language in Unicode standard: |
| 45 * Growing distinction between internal x-bit Unicode strings and external UTF
-x |
| 46 * forms, with the former more lenient. |
| 47 * Suggests renaming of UTF16_ macros to U16_. |
| 48 * - The prefix "UTF_" without a width number confused some users. |
| 49 * - "Safe" append macros needed the addition of an error indicator output. |
| 50 * - "Safe" UTF-8 macros used legitimate (if rarely used) code point values |
| 51 * to indicate error conditions. |
| 52 * - The use of the "_CHAR" infix for code point operations confused some users. |
| 53 * |
| 54 * More details: |
| 55 * |
| 56 * Until ICU 2.2, utf.h theoretically allowed to choose among UTF-8/16/32 |
| 57 * for string processing, and among unsafe/safe/strict default macros for that. |
| 58 * |
| 59 * It proved nearly impossible to write non-trivial, high-performance code |
| 60 * that is UTF-generic. |
| 61 * Unsafe default macros would be dangerous for default string processing, |
| 62 * and the main reason for the "strict" versions disappeared: |
| 63 * Between Unicode 3.0 and 3.2 all "irregular" UTF-8 sequences became illegal. |
| 64 * The only other conditions that "strict" checked for were non-characters, |
| 65 * which are valid during processing. Only during text input/output should they |
| 66 * be checked, and at that time other well-formedness checks may be |
| 67 * necessary or useful as well. |
| 68 * This can still be done by using U16_NEXT and U_IS_UNICODE_NONCHAR |
| 69 * or U_IS_UNICODE_CHAR. |
| 70 * |
| 71 * The old UTF8_..._SAFE macros also used some normal Unicode code points |
| 72 * to indicate malformed sequences. |
| 73 * The new UTF8_ macros without suffix use negative values instead. |
| 74 * |
| 75 * The entire contents of utf32.h was moved here without replacement |
| 76 * because all those macros were trivial and |
| 77 * were meaningful only in the framework of choosing the UTF size. |
| 78 * |
| 79 * See Jitterbug 2150 and its discussion on the ICU mailing list |
| 80 * in September 2002. |
| 81 * |
| 82 * <hr> |
| 83 * |
| 84 * <em>Obsolete part</em> of pre-ICU 2.4 utf.h file documentation: |
| 85 * |
| 86 * <p>The original concept for these files was for ICU to allow |
| 87 * in principle to set which UTF (UTF-8/16/32) is used internally |
| 88 * by defining UTF_SIZE to either 8, 16, or 32. utf.h would then define the UCha
r type |
| 89 * accordingly. UTF-16 was the default.</p> |
| 90 * |
| 91 * <p>This concept has been abandoned. |
| 92 * A lot of the ICU source code assumes UChar strings are in UTF-16. |
| 93 * This is especially true for low-level code like |
| 94 * conversion, normalization, and collation. |
| 95 * The utf.h header enforces the default of UTF-16. |
| 96 * The UTF-8 and UTF-32 macros remain for now for completeness and backward comp
atibility.</p> |
| 97 * |
| 98 * <p>Accordingly, utf.h defines UChar to be an unsigned 16-bit integer. If this
matches wchar_t, then |
| 99 * UChar is defined to be exactly wchar_t, otherwise uint16_t.</p> |
| 100 * |
| 101 * <p>UChar32 is defined to be a signed 32-bit integer (int32_t), large enough f
or a 21-bit |
| 102 * Unicode code point (Unicode scalar value, 0..0x10ffff). |
| 103 * Before ICU 2.4, the definition of UChar32 was similarly platform-dependent as |
| 104 * the definition of UChar. For details see the documentation for UChar32 itself
.</p> |
| 105 * |
| 106 * <p>utf.h also defines a number of C macros for handling single Unicode code p
oints and |
| 107 * for using UTF Unicode strings. It includes utf8.h, utf16.h, and utf32.h for t
he actual |
| 108 * implementations of those macros and then aliases one set of them (for UTF-16)
for general use. |
| 109 * The UTF-specific macros have the UTF size in the macro name prefixes (UTF16_.
..), while |
| 110 * the general alias macros always begin with UTF_...</p> |
| 111 * |
| 112 * <p>Many string operations can be done with or without error checking. |
| 113 * Where such a distinction is useful, there are two versions of the macros, "un
safe" and "safe" |
| 114 * ones with ..._UNSAFE and ..._SAFE suffixes. The unsafe macros are fast but ma
y cause |
| 115 * program failures if the strings are not well-formed. The safe macros have an
additional, boolean |
| 116 * parameter "strict". If strict is FALSE, then only illegal sequences are detec
ted. |
| 117 * Otherwise, irregular sequences and non-characters are detected as well (like
single surrogates). |
| 118 * Safe macros return special error code points for illegal/irregular sequences: |
| 119 * Typically, U+ffff, or values that would result in a code unit sequence of the
same length |
| 120 * as the erroneous input sequence.<br> |
| 121 * Note that _UNSAFE macros have fewer parameters: They do not have the strictne
ss parameter, and |
| 122 * they do not have start/length parameters for boundary checking.</p> |
| 123 * |
| 124 * <p>Here, the macros are aliased in two steps: |
| 125 * In the first step, the UTF-specific macros with UTF16_ prefix and _UNSAFE and
_SAFE suffixes are |
| 126 * aliased according to the UTF_SIZE to macros with UTF_ prefix and the same suf
fixes and signatures. |
| 127 * Then, in a second step, the default, general alias macros are set to use eith
er the unsafe or |
| 128 * the safe/not strict (default) or the safe/strict macro; |
| 129 * these general macros do not have a strictness parameter.</p> |
| 130 * |
| 131 * <p>It is possible to change the default choice for the general alias macros t
o be unsafe, safe/not strict or safe/strict. |
| 132 * The default is safe/not strict. It is not recommended to select the unsafe ma
cros as the basis for |
| 133 * Unicode string handling in ICU! To select this, define UTF_SAFE, UTF_STRICT,
or UTF_UNSAFE.</p> |
| 134 * |
| 135 * <p>For general use, one should use the default, general macros with UTF_ pref
ix and no _SAFE/_UNSAFE suffix. |
| 136 * Only in some cases it may be necessary to control the choice of macro directl
y and use a less generic alias. |
| 137 * For example, if it can be assumed that a string is well-formed and the index
will stay within the bounds, |
| 138 * then the _UNSAFE version may be used. |
| 139 * If a UTF-8 string is to be processed, then the macros with UTF8_ prefixes nee
d to be used.</p> |
| 140 * |
| 141 * <hr> |
| 142 * |
| 143 * @deprecated ICU 2.4. Use the macros in utf.h, utf16.h, utf8.h instead. |
| 144 */ |
| 145 |
| 146 #ifndef __UTF_OLD_H__ |
| 147 #define __UTF_OLD_H__ |
| 148 |
| 149 #ifndef U_HIDE_DEPRECATED_API |
| 150 |
| 151 /* utf.h must be included first. */ |
| 152 #ifndef __UTF_H__ |
| 153 # include "unicode/utf.h" |
| 154 #endif |
| 155 |
| 156 /* Formerly utf.h, part 1 --------------------------------------------------- */ |
| 157 |
| 158 #ifdef U_USE_UTF_DEPRECATES |
| 159 /** |
| 160 * Unicode string and array offset and index type. |
| 161 * ICU always counts Unicode code units (UChars) for |
| 162 * string offsets, indexes, and lengths, not Unicode code points. |
| 163 * |
| 164 * @obsolete ICU 2.6. Use int32_t directly instead since this API will be remove
d in that release. |
| 165 */ |
| 166 typedef int32_t UTextOffset; |
| 167 #endif |
| 168 |
| 169 /** Number of bits in a Unicode string code unit - ICU uses 16-bit Unicode. @dep
recated ICU 2.4. Obsolete, see utf_old.h. */ |
| 170 #define UTF_SIZE 16 |
| 171 |
| 172 /** |
| 173 * The default choice for general Unicode string macros is to use the ..._SAFE m
acro implementations |
| 174 * with strict=FALSE. |
| 175 * |
| 176 * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
| 177 */ |
| 178 #define UTF_SAFE |
| 179 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 180 #undef UTF_UNSAFE |
| 181 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 182 #undef UTF_STRICT |
| 183 |
| 184 /** |
| 185 * UTF8_ERROR_VALUE_1 and UTF8_ERROR_VALUE_2 are special error values for UTF-8, |
| 186 * which need 1 or 2 bytes in UTF-8: |
| 187 * \code |
| 188 * U+0015 = NAK = Negative Acknowledge, C0 control character |
| 189 * U+009f = highest C1 control character |
| 190 * \endcode |
| 191 * |
| 192 * These are used by UTF8_..._SAFE macros so that they can return an error value |
| 193 * that needs the same number of code units (bytes) as were seen by |
| 194 * a macro. They should be tested with UTF_IS_ERROR() or UTF_IS_VALID(). |
| 195 * |
| 196 * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
| 197 */ |
| 198 #define UTF8_ERROR_VALUE_1 0x15 |
| 199 |
| 200 /** |
| 201 * See documentation on UTF8_ERROR_VALUE_1 for details. |
| 202 * |
| 203 * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
| 204 */ |
| 205 #define UTF8_ERROR_VALUE_2 0x9f |
| 206 |
| 207 /** |
| 208 * Error value for all UTFs. This code point value will be set by macros with er
ror |
| 209 * checking if an error is detected. |
| 210 * |
| 211 * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
| 212 */ |
| 213 #define UTF_ERROR_VALUE 0xffff |
| 214 |
| 215 /** |
| 216 * Is a given 32-bit code an error value |
| 217 * as returned by one of the macros for any UTF? |
| 218 * |
| 219 * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
| 220 */ |
| 221 #define UTF_IS_ERROR(c) \ |
| 222 (((c)&0xfffe)==0xfffe || (c)==UTF8_ERROR_VALUE_1 || (c)==UTF8_ERROR_VALUE_2) |
| 223 |
| 224 /** |
| 225 * This is a combined macro: Is c a valid Unicode value _and_ not an error code? |
| 226 * |
| 227 * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
| 228 */ |
| 229 #define UTF_IS_VALID(c) \ |
| 230 (UTF_IS_UNICODE_CHAR(c) && \ |
| 231 (c)!=UTF8_ERROR_VALUE_1 && (c)!=UTF8_ERROR_VALUE_2) |
| 232 |
| 233 /** |
| 234 * Is this code unit or code point a surrogate (U+d800..U+dfff)? |
| 235 * @deprecated ICU 2.4. Renamed to U_IS_SURROGATE and U16_IS_SURROGATE, see utf_
old.h. |
| 236 */ |
| 237 #define UTF_IS_SURROGATE(uchar) (((uchar)&0xfffff800)==0xd800) |
| 238 |
| 239 /** |
| 240 * Is a given 32-bit code point a Unicode noncharacter? |
| 241 * |
| 242 * @deprecated ICU 2.4. Renamed to U_IS_UNICODE_NONCHAR, see utf_old.h. |
| 243 */ |
| 244 #define UTF_IS_UNICODE_NONCHAR(c) \ |
| 245 ((c)>=0xfdd0 && \ |
| 246 ((uint32_t)(c)<=0xfdef || ((c)&0xfffe)==0xfffe) && \ |
| 247 (uint32_t)(c)<=0x10ffff) |
| 248 |
| 249 /** |
| 250 * Is a given 32-bit value a Unicode code point value (0..U+10ffff) |
| 251 * that can be assigned a character? |
| 252 * |
| 253 * Code points that are not characters include: |
| 254 * - single surrogate code points (U+d800..U+dfff, 2048 code points) |
| 255 * - the last two code points on each plane (U+__fffe and U+__ffff, 34 code poin
ts) |
| 256 * - U+fdd0..U+fdef (new with Unicode 3.1, 32 code points) |
| 257 * - the highest Unicode code point value is U+10ffff |
| 258 * |
| 259 * This means that all code points below U+d800 are character code points, |
| 260 * and that boundary is tested first for performance. |
| 261 * |
| 262 * @deprecated ICU 2.4. Renamed to U_IS_UNICODE_CHAR, see utf_old.h. |
| 263 */ |
| 264 #define UTF_IS_UNICODE_CHAR(c) \ |
| 265 ((uint32_t)(c)<0xd800 || \ |
| 266 ((uint32_t)(c)>0xdfff && \ |
| 267 (uint32_t)(c)<=0x10ffff && \ |
| 268 !UTF_IS_UNICODE_NONCHAR(c))) |
| 269 |
| 270 /* Formerly utf8.h ---------------------------------------------------------- */ |
| 271 |
| 272 /** |
| 273 * Count the trail bytes for a UTF-8 lead byte. |
| 274 * @deprecated ICU 2.4. Renamed to U8_COUNT_TRAIL_BYTES, see utf_old.h. |
| 275 */ |
| 276 #define UTF8_COUNT_TRAIL_BYTES(leadByte) (utf8_countTrailBytes[(uint8_t)leadByte
]) |
| 277 |
| 278 /** |
| 279 * Mask a UTF-8 lead byte, leave only the lower bits that form part of the code
point value. |
| 280 * @deprecated ICU 2.4. Renamed to U8_MASK_LEAD_BYTE, see utf_old.h. |
| 281 */ |
| 282 #define UTF8_MASK_LEAD_BYTE(leadByte, countTrailBytes) ((leadByte)&=(1<<(6-(coun
tTrailBytes)))-1) |
| 283 |
| 284 /** Is this this code point a single code unit (byte)? @deprecated ICU 2.4. Rena
med to U8_IS_SINGLE, see utf_old.h. */ |
| 285 #define UTF8_IS_SINGLE(uchar) (((uchar)&0x80)==0) |
| 286 /** Is this this code unit the lead code unit (byte) of a code point? @deprecate
d ICU 2.4. Renamed to U8_IS_LEAD, see utf_old.h. */ |
| 287 #define UTF8_IS_LEAD(uchar) ((uint8_t)((uchar)-0xc0)<0x3e) |
| 288 /** Is this this code unit a trailing code unit (byte) of a code point? @depreca
ted ICU 2.4. Renamed to U8_IS_TRAIL, see utf_old.h. */ |
| 289 #define UTF8_IS_TRAIL(uchar) (((uchar)&0xc0)==0x80) |
| 290 |
| 291 /** Does this scalar Unicode value need multiple code units for storage? @deprec
ated ICU 2.4. Use U8_LENGTH or test ((uint32_t)(c)>0x7f) instead, see utf_old.h.
*/ |
| 292 #define UTF8_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0x7f) |
| 293 |
| 294 /** |
| 295 * Given the lead character, how many bytes are taken by this code point. |
| 296 * ICU does not deal with code points >0x10ffff |
| 297 * unless necessary for advancing in the byte stream. |
| 298 * |
| 299 * These length macros take into account that for values >0x10ffff |
| 300 * the UTF8_APPEND_CHAR_SAFE macros would write the error code point 0xffff |
| 301 * with 3 bytes. |
| 302 * Code point comparisons need to be in uint32_t because UChar32 |
| 303 * may be a signed type, and negative values must be recognized. |
| 304 * |
| 305 * @deprecated ICU 2.4. Use U8_LENGTH instead, see utf_old.h. |
| 306 */ |
| 307 #if 1 |
| 308 # define UTF8_CHAR_LENGTH(c) \ |
| 309 ((uint32_t)(c)<=0x7f ? 1 : \ |
| 310 ((uint32_t)(c)<=0x7ff ? 2 : \ |
| 311 ((uint32_t)((c)-0x10000)>0xfffff ? 3 : 4) \ |
| 312 ) \ |
| 313 ) |
| 314 #else |
| 315 # define UTF8_CHAR_LENGTH(c) \ |
| 316 ((uint32_t)(c)<=0x7f ? 1 : \ |
| 317 ((uint32_t)(c)<=0x7ff ? 2 : \ |
| 318 ((uint32_t)(c)<=0xffff ? 3 : \ |
| 319 ((uint32_t)(c)<=0x10ffff ? 4 : \ |
| 320 ((uint32_t)(c)<=0x3ffffff ? 5 : \ |
| 321 ((uint32_t)(c)<=0x7fffffff ? 6 : 3) \ |
| 322 ) \ |
| 323 ) \ |
| 324 ) \ |
| 325 ) \ |
| 326 ) |
| 327 #endif |
| 328 |
| 329 /** The maximum number of bytes per code point. @deprecated ICU 2.4. Renamed to
U8_MAX_LENGTH, see utf_old.h. */ |
| 330 #define UTF8_MAX_CHAR_LENGTH 4 |
| 331 |
| 332 /** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsole
te, see utf_old.h. */ |
| 333 #define UTF8_ARRAY_SIZE(size) ((5*(size))/2) |
| 334 |
| 335 /** @deprecated ICU 2.4. Renamed to U8_GET_UNSAFE, see utf_old.h. */ |
| 336 #define UTF8_GET_CHAR_UNSAFE(s, i, c) { \ |
| 337 int32_t _utf8_get_char_unsafe_index=(int32_t)(i); \ |
| 338 UTF8_SET_CHAR_START_UNSAFE(s, _utf8_get_char_unsafe_index); \ |
| 339 UTF8_NEXT_CHAR_UNSAFE(s, _utf8_get_char_unsafe_index, c); \ |
| 340 } |
| 341 |
| 342 /** @deprecated ICU 2.4. Use U8_GET instead, see utf_old.h. */ |
| 343 #define UTF8_GET_CHAR_SAFE(s, start, i, length, c, strict) { \ |
| 344 int32_t _utf8_get_char_safe_index=(int32_t)(i); \ |
| 345 UTF8_SET_CHAR_START_SAFE(s, start, _utf8_get_char_safe_index); \ |
| 346 UTF8_NEXT_CHAR_SAFE(s, _utf8_get_char_safe_index, length, c, strict); \ |
| 347 } |
| 348 |
| 349 /** @deprecated ICU 2.4. Renamed to U8_NEXT_UNSAFE, see utf_old.h. */ |
| 350 #define UTF8_NEXT_CHAR_UNSAFE(s, i, c) { \ |
| 351 (c)=(s)[(i)++]; \ |
| 352 if((uint8_t)((c)-0xc0)<0x35) { \ |
| 353 uint8_t __count=UTF8_COUNT_TRAIL_BYTES(c); \ |
| 354 UTF8_MASK_LEAD_BYTE(c, __count); \ |
| 355 switch(__count) { \ |
| 356 /* each following branch falls through to the next one */ \ |
| 357 case 3: \ |
| 358 (c)=((c)<<6)|((s)[(i)++]&0x3f); \ |
| 359 case 2: \ |
| 360 (c)=((c)<<6)|((s)[(i)++]&0x3f); \ |
| 361 case 1: \ |
| 362 (c)=((c)<<6)|((s)[(i)++]&0x3f); \ |
| 363 /* no other branches to optimize switch() */ \ |
| 364 break; \ |
| 365 } \ |
| 366 } \ |
| 367 } |
| 368 |
| 369 /** @deprecated ICU 2.4. Renamed to U8_APPEND_UNSAFE, see utf_old.h. */ |
| 370 #define UTF8_APPEND_CHAR_UNSAFE(s, i, c) { \ |
| 371 if((uint32_t)(c)<=0x7f) { \ |
| 372 (s)[(i)++]=(uint8_t)(c); \ |
| 373 } else { \ |
| 374 if((uint32_t)(c)<=0x7ff) { \ |
| 375 (s)[(i)++]=(uint8_t)(((c)>>6)|0xc0); \ |
| 376 } else { \ |
| 377 if((uint32_t)(c)<=0xffff) { \ |
| 378 (s)[(i)++]=(uint8_t)(((c)>>12)|0xe0); \ |
| 379 } else { \ |
| 380 (s)[(i)++]=(uint8_t)(((c)>>18)|0xf0); \ |
| 381 (s)[(i)++]=(uint8_t)((((c)>>12)&0x3f)|0x80); \ |
| 382 } \ |
| 383 (s)[(i)++]=(uint8_t)((((c)>>6)&0x3f)|0x80); \ |
| 384 } \ |
| 385 (s)[(i)++]=(uint8_t)(((c)&0x3f)|0x80); \ |
| 386 } \ |
| 387 } |
| 388 |
| 389 /** @deprecated ICU 2.4. Renamed to U8_FWD_1_UNSAFE, see utf_old.h. */ |
| 390 #define UTF8_FWD_1_UNSAFE(s, i) { \ |
| 391 (i)+=1+UTF8_COUNT_TRAIL_BYTES((s)[i]); \ |
| 392 } |
| 393 |
| 394 /** @deprecated ICU 2.4. Renamed to U8_FWD_N_UNSAFE, see utf_old.h. */ |
| 395 #define UTF8_FWD_N_UNSAFE(s, i, n) { \ |
| 396 int32_t __N=(n); \ |
| 397 while(__N>0) { \ |
| 398 UTF8_FWD_1_UNSAFE(s, i); \ |
| 399 --__N; \ |
| 400 } \ |
| 401 } |
| 402 |
| 403 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_START_UNSAFE, see utf_old.h. */ |
| 404 #define UTF8_SET_CHAR_START_UNSAFE(s, i) { \ |
| 405 while(UTF8_IS_TRAIL((s)[i])) { --(i); } \ |
| 406 } |
| 407 |
| 408 /** @deprecated ICU 2.4. Use U8_NEXT instead, see utf_old.h. */ |
| 409 #define UTF8_NEXT_CHAR_SAFE(s, i, length, c, strict) { \ |
| 410 (c)=(s)[(i)++]; \ |
| 411 if((c)>=0x80) { \ |
| 412 if(UTF8_IS_LEAD(c)) { \ |
| 413 (c)=utf8_nextCharSafeBody(s, &(i), (int32_t)(length), c, strict); \ |
| 414 } else { \ |
| 415 (c)=UTF8_ERROR_VALUE_1; \ |
| 416 } \ |
| 417 } \ |
| 418 } |
| 419 |
| 420 /** @deprecated ICU 2.4. Use U8_APPEND instead, see utf_old.h. */ |
| 421 #define UTF8_APPEND_CHAR_SAFE(s, i, length, c) { \ |
| 422 if((uint32_t)(c)<=0x7f) { \ |
| 423 (s)[(i)++]=(uint8_t)(c); \ |
| 424 } else { \ |
| 425 (i)=utf8_appendCharSafeBody(s, (int32_t)(i), (int32_t)(length), c, NULL)
; \ |
| 426 } \ |
| 427 } |
| 428 |
| 429 /** @deprecated ICU 2.4. Renamed to U8_FWD_1, see utf_old.h. */ |
| 430 #define UTF8_FWD_1_SAFE(s, i, length) U8_FWD_1(s, i, length) |
| 431 |
| 432 /** @deprecated ICU 2.4. Renamed to U8_FWD_N, see utf_old.h. */ |
| 433 #define UTF8_FWD_N_SAFE(s, i, length, n) U8_FWD_N(s, i, length, n) |
| 434 |
| 435 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_START, see utf_old.h. */ |
| 436 #define UTF8_SET_CHAR_START_SAFE(s, start, i) U8_SET_CP_START(s, start, i) |
| 437 |
| 438 /** @deprecated ICU 2.4. Renamed to U8_PREV_UNSAFE, see utf_old.h. */ |
| 439 #define UTF8_PREV_CHAR_UNSAFE(s, i, c) { \ |
| 440 (c)=(s)[--(i)]; \ |
| 441 if(UTF8_IS_TRAIL(c)) { \ |
| 442 uint8_t __b, __count=1, __shift=6; \ |
| 443 \ |
| 444 /* c is a trail byte */ \ |
| 445 (c)&=0x3f; \ |
| 446 for(;;) { \ |
| 447 __b=(s)[--(i)]; \ |
| 448 if(__b>=0xc0) { \ |
| 449 UTF8_MASK_LEAD_BYTE(__b, __count); \ |
| 450 (c)|=(UChar32)__b<<__shift; \ |
| 451 break; \ |
| 452 } else { \ |
| 453 (c)|=(UChar32)(__b&0x3f)<<__shift; \ |
| 454 ++__count; \ |
| 455 __shift+=6; \ |
| 456 } \ |
| 457 } \ |
| 458 } \ |
| 459 } |
| 460 |
| 461 /** @deprecated ICU 2.4. Renamed to U8_BACK_1_UNSAFE, see utf_old.h. */ |
| 462 #define UTF8_BACK_1_UNSAFE(s, i) { \ |
| 463 while(UTF8_IS_TRAIL((s)[--(i)])) {} \ |
| 464 } |
| 465 |
| 466 /** @deprecated ICU 2.4. Renamed to U8_BACK_N_UNSAFE, see utf_old.h. */ |
| 467 #define UTF8_BACK_N_UNSAFE(s, i, n) { \ |
| 468 int32_t __N=(n); \ |
| 469 while(__N>0) { \ |
| 470 UTF8_BACK_1_UNSAFE(s, i); \ |
| 471 --__N; \ |
| 472 } \ |
| 473 } |
| 474 |
| 475 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT_UNSAFE, see utf_old.h. */ |
| 476 #define UTF8_SET_CHAR_LIMIT_UNSAFE(s, i) { \ |
| 477 UTF8_BACK_1_UNSAFE(s, i); \ |
| 478 UTF8_FWD_1_UNSAFE(s, i); \ |
| 479 } |
| 480 |
| 481 /** @deprecated ICU 2.4. Use U8_PREV instead, see utf_old.h. */ |
| 482 #define UTF8_PREV_CHAR_SAFE(s, start, i, c, strict) { \ |
| 483 (c)=(s)[--(i)]; \ |
| 484 if((c)>=0x80) { \ |
| 485 if((c)<=0xbf) { \ |
| 486 (c)=utf8_prevCharSafeBody(s, start, &(i), c, strict); \ |
| 487 } else { \ |
| 488 (c)=UTF8_ERROR_VALUE_1; \ |
| 489 } \ |
| 490 } \ |
| 491 } |
| 492 |
| 493 /** @deprecated ICU 2.4. Renamed to U8_BACK_1, see utf_old.h. */ |
| 494 #define UTF8_BACK_1_SAFE(s, start, i) U8_BACK_1(s, start, i) |
| 495 |
| 496 /** @deprecated ICU 2.4. Renamed to U8_BACK_N, see utf_old.h. */ |
| 497 #define UTF8_BACK_N_SAFE(s, start, i, n) U8_BACK_N(s, start, i, n) |
| 498 |
| 499 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT, see utf_old.h. */ |
| 500 #define UTF8_SET_CHAR_LIMIT_SAFE(s, start, i, length) U8_SET_CP_LIMIT(s, start,
i, length) |
| 501 |
| 502 /* Formerly utf16.h --------------------------------------------------------- */ |
| 503 |
| 504 /** Is uchar a first/lead surrogate? @deprecated ICU 2.4. Renamed to U_IS_LEAD a
nd U16_IS_LEAD, see utf_old.h. */ |
| 505 #define UTF_IS_FIRST_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xd800) |
| 506 |
| 507 /** Is uchar a second/trail surrogate? @deprecated ICU 2.4. Renamed to U_IS_TRAI
L and U16_IS_TRAIL, see utf_old.h. */ |
| 508 #define UTF_IS_SECOND_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xdc00) |
| 509 |
| 510 /** 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. */ |
| 511 #define UTF_IS_SURROGATE_FIRST(c) (((c)&0x400)==0) |
| 512 |
| 513 /** Helper constant for UTF16_GET_PAIR_VALUE. @deprecated ICU 2.4. Renamed to U1
6_SURROGATE_OFFSET, see utf_old.h. */ |
| 514 #define UTF_SURROGATE_OFFSET ((0xd800<<10UL)+0xdc00-0x10000) |
| 515 |
| 516 /** Get the UTF-32 value from the surrogate code units. @deprecated ICU 2.4. Ren
amed to U16_GET_SUPPLEMENTARY, see utf_old.h. */ |
| 517 #define UTF16_GET_PAIR_VALUE(first, second) \ |
| 518 (((first)<<10UL)+(second)-UTF_SURROGATE_OFFSET) |
| 519 |
| 520 /** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */ |
| 521 #define UTF_FIRST_SURROGATE(supplementary) (UChar)(((supplementary)>>10)+0xd7c0) |
| 522 |
| 523 /** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */ |
| 524 #define UTF_SECOND_SURROGATE(supplementary) (UChar)(((supplementary)&0x3ff)|0xdc
00) |
| 525 |
| 526 /** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */ |
| 527 #define UTF16_LEAD(supplementary) UTF_FIRST_SURROGATE(supplementary) |
| 528 |
| 529 /** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */ |
| 530 #define UTF16_TRAIL(supplementary) UTF_SECOND_SURROGATE(supplementary) |
| 531 |
| 532 /** @deprecated ICU 2.4. Renamed to U16_IS_SINGLE, see utf_old.h. */ |
| 533 #define UTF16_IS_SINGLE(uchar) !UTF_IS_SURROGATE(uchar) |
| 534 |
| 535 /** @deprecated ICU 2.4. Renamed to U16_IS_LEAD, see utf_old.h. */ |
| 536 #define UTF16_IS_LEAD(uchar) UTF_IS_FIRST_SURROGATE(uchar) |
| 537 |
| 538 /** @deprecated ICU 2.4. Renamed to U16_IS_TRAIL, see utf_old.h. */ |
| 539 #define UTF16_IS_TRAIL(uchar) UTF_IS_SECOND_SURROGATE(uchar) |
| 540 |
| 541 /** Does this scalar Unicode value need multiple code units for storage? @deprec
ated ICU 2.4. Use U16_LENGTH or test ((uint32_t)(c)>0xffff) instead, see utf_old
.h. */ |
| 542 #define UTF16_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0xffff) |
| 543 |
| 544 /** @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h. */ |
| 545 #define UTF16_CHAR_LENGTH(c) ((uint32_t)(c)<=0xffff ? 1 : 2) |
| 546 |
| 547 /** @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h. */ |
| 548 #define UTF16_MAX_CHAR_LENGTH 2 |
| 549 |
| 550 /** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsole
te, see utf_old.h. */ |
| 551 #define UTF16_ARRAY_SIZE(size) (size) |
| 552 |
| 553 /** |
| 554 * Get a single code point from an offset that points to any |
| 555 * of the code units that belong to that code point. |
| 556 * Assume 0<=i<length. |
| 557 * |
| 558 * This could be used for iteration together with |
| 559 * UTF16_CHAR_LENGTH() and UTF_IS_ERROR(), |
| 560 * but the use of UTF16_NEXT_CHAR[_UNSAFE]() and |
| 561 * UTF16_PREV_CHAR[_UNSAFE]() is more efficient for that. |
| 562 * @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h. |
| 563 */ |
| 564 #define UTF16_GET_CHAR_UNSAFE(s, i, c) { \ |
| 565 (c)=(s)[i]; \ |
| 566 if(UTF_IS_SURROGATE(c)) { \ |
| 567 if(UTF_IS_SURROGATE_FIRST(c)) { \ |
| 568 (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)+1]); \ |
| 569 } else { \ |
| 570 (c)=UTF16_GET_PAIR_VALUE((s)[(i)-1], (c)); \ |
| 571 } \ |
| 572 } \ |
| 573 } |
| 574 |
| 575 /** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */ |
| 576 #define UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict) { \ |
| 577 (c)=(s)[i]; \ |
| 578 if(UTF_IS_SURROGATE(c)) { \ |
| 579 uint16_t __c2; \ |
| 580 if(UTF_IS_SURROGATE_FIRST(c)) { \ |
| 581 if((i)+1<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)+1])) { \ |
| 582 (c)=UTF16_GET_PAIR_VALUE((c), __c2); \ |
| 583 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and
UTF_IS_UNICODE_CHAR() */ \ |
| 584 } else if(strict) {\ |
| 585 /* unmatched first surrogate */ \ |
| 586 (c)=UTF_ERROR_VALUE; \ |
| 587 } \ |
| 588 } else { \ |
| 589 if((i)-1>=(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \ |
| 590 (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \ |
| 591 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and
UTF_IS_UNICODE_CHAR() */ \ |
| 592 } else if(strict) {\ |
| 593 /* unmatched second surrogate */ \ |
| 594 (c)=UTF_ERROR_VALUE; \ |
| 595 } \ |
| 596 } \ |
| 597 } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \ |
| 598 (c)=UTF_ERROR_VALUE; \ |
| 599 } \ |
| 600 } |
| 601 |
| 602 /** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */ |
| 603 #define UTF16_NEXT_CHAR_UNSAFE(s, i, c) { \ |
| 604 (c)=(s)[(i)++]; \ |
| 605 if(UTF_IS_FIRST_SURROGATE(c)) { \ |
| 606 (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)++]); \ |
| 607 } \ |
| 608 } |
| 609 |
| 610 /** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */ |
| 611 #define UTF16_APPEND_CHAR_UNSAFE(s, i, c) { \ |
| 612 if((uint32_t)(c)<=0xffff) { \ |
| 613 (s)[(i)++]=(uint16_t)(c); \ |
| 614 } else { \ |
| 615 (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ |
| 616 (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ |
| 617 } \ |
| 618 } |
| 619 |
| 620 /** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */ |
| 621 #define UTF16_FWD_1_UNSAFE(s, i) { \ |
| 622 if(UTF_IS_FIRST_SURROGATE((s)[(i)++])) { \ |
| 623 ++(i); \ |
| 624 } \ |
| 625 } |
| 626 |
| 627 /** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */ |
| 628 #define UTF16_FWD_N_UNSAFE(s, i, n) { \ |
| 629 int32_t __N=(n); \ |
| 630 while(__N>0) { \ |
| 631 UTF16_FWD_1_UNSAFE(s, i); \ |
| 632 --__N; \ |
| 633 } \ |
| 634 } |
| 635 |
| 636 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */ |
| 637 #define UTF16_SET_CHAR_START_UNSAFE(s, i) { \ |
| 638 if(UTF_IS_SECOND_SURROGATE((s)[i])) { \ |
| 639 --(i); \ |
| 640 } \ |
| 641 } |
| 642 |
| 643 /** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */ |
| 644 #define UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict) { \ |
| 645 (c)=(s)[(i)++]; \ |
| 646 if(UTF_IS_FIRST_SURROGATE(c)) { \ |
| 647 uint16_t __c2; \ |
| 648 if((i)<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)])) { \ |
| 649 ++(i); \ |
| 650 (c)=UTF16_GET_PAIR_VALUE((c), __c2); \ |
| 651 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_
IS_UNICODE_CHAR() */ \ |
| 652 } else if(strict) {\ |
| 653 /* unmatched first surrogate */ \ |
| 654 (c)=UTF_ERROR_VALUE; \ |
| 655 } \ |
| 656 } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \ |
| 657 /* unmatched second surrogate or other non-character */ \ |
| 658 (c)=UTF_ERROR_VALUE; \ |
| 659 } \ |
| 660 } |
| 661 |
| 662 /** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */ |
| 663 #define UTF16_APPEND_CHAR_SAFE(s, i, length, c) { \ |
| 664 if((uint32_t)(c)<=0xffff) { \ |
| 665 (s)[(i)++]=(uint16_t)(c); \ |
| 666 } else if((uint32_t)(c)<=0x10ffff) { \ |
| 667 if((i)+1<(length)) { \ |
| 668 (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ |
| 669 (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ |
| 670 } else /* not enough space */ { \ |
| 671 (s)[(i)++]=UTF_ERROR_VALUE; \ |
| 672 } \ |
| 673 } else /* c>0x10ffff, write error value */ { \ |
| 674 (s)[(i)++]=UTF_ERROR_VALUE; \ |
| 675 } \ |
| 676 } |
| 677 |
| 678 /** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */ |
| 679 #define UTF16_FWD_1_SAFE(s, i, length) U16_FWD_1(s, i, length) |
| 680 |
| 681 /** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */ |
| 682 #define UTF16_FWD_N_SAFE(s, i, length, n) U16_FWD_N(s, i, length, n) |
| 683 |
| 684 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */ |
| 685 #define UTF16_SET_CHAR_START_SAFE(s, start, i) U16_SET_CP_START(s, start, i) |
| 686 |
| 687 /** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */ |
| 688 #define UTF16_PREV_CHAR_UNSAFE(s, i, c) { \ |
| 689 (c)=(s)[--(i)]; \ |
| 690 if(UTF_IS_SECOND_SURROGATE(c)) { \ |
| 691 (c)=UTF16_GET_PAIR_VALUE((s)[--(i)], (c)); \ |
| 692 } \ |
| 693 } |
| 694 |
| 695 /** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */ |
| 696 #define UTF16_BACK_1_UNSAFE(s, i) { \ |
| 697 if(UTF_IS_SECOND_SURROGATE((s)[--(i)])) { \ |
| 698 --(i); \ |
| 699 } \ |
| 700 } |
| 701 |
| 702 /** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */ |
| 703 #define UTF16_BACK_N_UNSAFE(s, i, n) { \ |
| 704 int32_t __N=(n); \ |
| 705 while(__N>0) { \ |
| 706 UTF16_BACK_1_UNSAFE(s, i); \ |
| 707 --__N; \ |
| 708 } \ |
| 709 } |
| 710 |
| 711 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */ |
| 712 #define UTF16_SET_CHAR_LIMIT_UNSAFE(s, i) { \ |
| 713 if(UTF_IS_FIRST_SURROGATE((s)[(i)-1])) { \ |
| 714 ++(i); \ |
| 715 } \ |
| 716 } |
| 717 |
| 718 /** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */ |
| 719 #define UTF16_PREV_CHAR_SAFE(s, start, i, c, strict) { \ |
| 720 (c)=(s)[--(i)]; \ |
| 721 if(UTF_IS_SECOND_SURROGATE(c)) { \ |
| 722 uint16_t __c2; \ |
| 723 if((i)>(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \ |
| 724 --(i); \ |
| 725 (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \ |
| 726 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_
IS_UNICODE_CHAR() */ \ |
| 727 } else if(strict) {\ |
| 728 /* unmatched second surrogate */ \ |
| 729 (c)=UTF_ERROR_VALUE; \ |
| 730 } \ |
| 731 } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \ |
| 732 /* unmatched first surrogate or other non-character */ \ |
| 733 (c)=UTF_ERROR_VALUE; \ |
| 734 } \ |
| 735 } |
| 736 |
| 737 /** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */ |
| 738 #define UTF16_BACK_1_SAFE(s, start, i) U16_BACK_1(s, start, i) |
| 739 |
| 740 /** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */ |
| 741 #define UTF16_BACK_N_SAFE(s, start, i, n) U16_BACK_N(s, start, i, n) |
| 742 |
| 743 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */ |
| 744 #define UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length) U16_SET_CP_LIMIT(s, start
, i, length) |
| 745 |
| 746 /* Formerly utf32.h --------------------------------------------------------- */ |
| 747 |
| 748 /* |
| 749 * Old documentation: |
| 750 * |
| 751 * This file defines macros to deal with UTF-32 code units and code points. |
| 752 * Signatures and semantics are the same as for the similarly named macros |
| 753 * in utf16.h. |
| 754 * utf32.h is included by utf.h after unicode/umachine.h</p> |
| 755 * and some common definitions. |
| 756 * <p><b>Usage:</b> ICU coding guidelines for if() statements should be follow
ed when using these macros. |
| 757 * Compound statements (curly braces {}) must be used for if-el
se-while... |
| 758 * bodies and all macro statements should be terminated with sem
icolon.</p> |
| 759 */ |
| 760 |
| 761 /* internal definitions ----------------------------------------------------- */ |
| 762 |
| 763 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 764 #define UTF32_IS_SAFE(c, strict) \ |
| 765 (!(strict) ? \ |
| 766 (uint32_t)(c)<=0x10ffff : \ |
| 767 UTF_IS_UNICODE_CHAR(c)) |
| 768 |
| 769 /* |
| 770 * For the semantics of all of these macros, see utf16.h. |
| 771 * The UTF-32 versions are trivial because any code point is |
| 772 * encoded using exactly one code unit. |
| 773 */ |
| 774 |
| 775 /* single-code point definitions -------------------------------------------- */ |
| 776 |
| 777 /* classes of code unit values */ |
| 778 |
| 779 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 780 #define UTF32_IS_SINGLE(uchar) 1 |
| 781 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 782 #define UTF32_IS_LEAD(uchar) 0 |
| 783 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 784 #define UTF32_IS_TRAIL(uchar) 0 |
| 785 |
| 786 /* number of code units per code point */ |
| 787 |
| 788 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 789 #define UTF32_NEED_MULTIPLE_UCHAR(c) 0 |
| 790 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 791 #define UTF32_CHAR_LENGTH(c) 1 |
| 792 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 793 #define UTF32_MAX_CHAR_LENGTH 1 |
| 794 |
| 795 /* average number of code units compared to UTF-16 */ |
| 796 |
| 797 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 798 #define UTF32_ARRAY_SIZE(size) (size) |
| 799 |
| 800 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 801 #define UTF32_GET_CHAR_UNSAFE(s, i, c) { \ |
| 802 (c)=(s)[i]; \ |
| 803 } |
| 804 |
| 805 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 806 #define UTF32_GET_CHAR_SAFE(s, start, i, length, c, strict) { \ |
| 807 (c)=(s)[i]; \ |
| 808 if(!UTF32_IS_SAFE(c, strict)) { \ |
| 809 (c)=UTF_ERROR_VALUE; \ |
| 810 } \ |
| 811 } |
| 812 |
| 813 /* definitions with forward iteration --------------------------------------- */ |
| 814 |
| 815 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 816 #define UTF32_NEXT_CHAR_UNSAFE(s, i, c) { \ |
| 817 (c)=(s)[(i)++]; \ |
| 818 } |
| 819 |
| 820 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 821 #define UTF32_APPEND_CHAR_UNSAFE(s, i, c) { \ |
| 822 (s)[(i)++]=(c); \ |
| 823 } |
| 824 |
| 825 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 826 #define UTF32_FWD_1_UNSAFE(s, i) { \ |
| 827 ++(i); \ |
| 828 } |
| 829 |
| 830 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 831 #define UTF32_FWD_N_UNSAFE(s, i, n) { \ |
| 832 (i)+=(n); \ |
| 833 } |
| 834 |
| 835 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 836 #define UTF32_SET_CHAR_START_UNSAFE(s, i) { \ |
| 837 } |
| 838 |
| 839 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 840 #define UTF32_NEXT_CHAR_SAFE(s, i, length, c, strict) { \ |
| 841 (c)=(s)[(i)++]; \ |
| 842 if(!UTF32_IS_SAFE(c, strict)) { \ |
| 843 (c)=UTF_ERROR_VALUE; \ |
| 844 } \ |
| 845 } |
| 846 |
| 847 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 848 #define UTF32_APPEND_CHAR_SAFE(s, i, length, c) { \ |
| 849 if((uint32_t)(c)<=0x10ffff) { \ |
| 850 (s)[(i)++]=(c); \ |
| 851 } else /* c>0x10ffff, write 0xfffd */ { \ |
| 852 (s)[(i)++]=0xfffd; \ |
| 853 } \ |
| 854 } |
| 855 |
| 856 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 857 #define UTF32_FWD_1_SAFE(s, i, length) { \ |
| 858 ++(i); \ |
| 859 } |
| 860 |
| 861 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 862 #define UTF32_FWD_N_SAFE(s, i, length, n) { \ |
| 863 if(((i)+=(n))>(length)) { \ |
| 864 (i)=(length); \ |
| 865 } \ |
| 866 } |
| 867 |
| 868 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 869 #define UTF32_SET_CHAR_START_SAFE(s, start, i) { \ |
| 870 } |
| 871 |
| 872 /* definitions with backward iteration -------------------------------------- */ |
| 873 |
| 874 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 875 #define UTF32_PREV_CHAR_UNSAFE(s, i, c) { \ |
| 876 (c)=(s)[--(i)]; \ |
| 877 } |
| 878 |
| 879 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 880 #define UTF32_BACK_1_UNSAFE(s, i) { \ |
| 881 --(i); \ |
| 882 } |
| 883 |
| 884 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 885 #define UTF32_BACK_N_UNSAFE(s, i, n) { \ |
| 886 (i)-=(n); \ |
| 887 } |
| 888 |
| 889 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 890 #define UTF32_SET_CHAR_LIMIT_UNSAFE(s, i) { \ |
| 891 } |
| 892 |
| 893 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 894 #define UTF32_PREV_CHAR_SAFE(s, start, i, c, strict) { \ |
| 895 (c)=(s)[--(i)]; \ |
| 896 if(!UTF32_IS_SAFE(c, strict)) { \ |
| 897 (c)=UTF_ERROR_VALUE; \ |
| 898 } \ |
| 899 } |
| 900 |
| 901 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 902 #define UTF32_BACK_1_SAFE(s, start, i) { \ |
| 903 --(i); \ |
| 904 } |
| 905 |
| 906 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 907 #define UTF32_BACK_N_SAFE(s, start, i, n) { \ |
| 908 (i)-=(n); \ |
| 909 if((i)<(start)) { \ |
| 910 (i)=(start); \ |
| 911 } \ |
| 912 } |
| 913 |
| 914 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ |
| 915 #define UTF32_SET_CHAR_LIMIT_SAFE(s, i, length) { \ |
| 916 } |
| 917 |
| 918 /* Formerly utf.h, part 2 --------------------------------------------------- */ |
| 919 |
| 920 /** |
| 921 * Estimate the number of code units for a string based on the number of UTF-16
code units. |
| 922 * |
| 923 * @deprecated ICU 2.4. Obsolete, see utf_old.h. |
| 924 */ |
| 925 #define UTF_ARRAY_SIZE(size) UTF16_ARRAY_SIZE(size) |
| 926 |
| 927 /** @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h. */ |
| 928 #define UTF_GET_CHAR_UNSAFE(s, i, c) UTF16_GET_CHAR_UNSAFE(s, i,
c) |
| 929 |
| 930 /** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */ |
| 931 #define UTF_GET_CHAR_SAFE(s, start, i, length, c, strict) UTF16_GET_CHAR_SAFE(s,
start, i, length, c, strict) |
| 932 |
| 933 |
| 934 /** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */ |
| 935 #define UTF_NEXT_CHAR_UNSAFE(s, i, c) UTF16_NEXT_CHAR_UNSAFE(s, i
, c) |
| 936 |
| 937 /** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */ |
| 938 #define UTF_NEXT_CHAR_SAFE(s, i, length, c, strict) UTF16_NEXT_CHAR_SAFE(s, i,
length, c, strict) |
| 939 |
| 940 |
| 941 /** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */ |
| 942 #define UTF_APPEND_CHAR_UNSAFE(s, i, c) UTF16_APPEND_CHAR_UNSAFE(s,
i, c) |
| 943 |
| 944 /** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */ |
| 945 #define UTF_APPEND_CHAR_SAFE(s, i, length, c) UTF16_APPEND_CHAR_SAFE(s, i
, length, c) |
| 946 |
| 947 |
| 948 /** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */ |
| 949 #define UTF_FWD_1_UNSAFE(s, i) UTF16_FWD_1_UNSAFE(s, i) |
| 950 |
| 951 /** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */ |
| 952 #define UTF_FWD_1_SAFE(s, i, length) UTF16_FWD_1_SAFE(s, i, leng
th) |
| 953 |
| 954 |
| 955 /** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */ |
| 956 #define UTF_FWD_N_UNSAFE(s, i, n) UTF16_FWD_N_UNSAFE(s, i, n) |
| 957 |
| 958 /** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */ |
| 959 #define UTF_FWD_N_SAFE(s, i, length, n) UTF16_FWD_N_SAFE(s, i, leng
th, n) |
| 960 |
| 961 |
| 962 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */ |
| 963 #define UTF_SET_CHAR_START_UNSAFE(s, i) UTF16_SET_CHAR_START_UNSAFE
(s, i) |
| 964 |
| 965 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */ |
| 966 #define UTF_SET_CHAR_START_SAFE(s, start, i) UTF16_SET_CHAR_START_SAFE(s
, start, i) |
| 967 |
| 968 |
| 969 /** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */ |
| 970 #define UTF_PREV_CHAR_UNSAFE(s, i, c) UTF16_PREV_CHAR_UNSAFE(s, i
, c) |
| 971 |
| 972 /** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */ |
| 973 #define UTF_PREV_CHAR_SAFE(s, start, i, c, strict) UTF16_PREV_CHAR_SAFE(s, sta
rt, i, c, strict) |
| 974 |
| 975 |
| 976 /** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */ |
| 977 #define UTF_BACK_1_UNSAFE(s, i) UTF16_BACK_1_UNSAFE(s, i) |
| 978 |
| 979 /** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */ |
| 980 #define UTF_BACK_1_SAFE(s, start, i) UTF16_BACK_1_SAFE(s, start,
i) |
| 981 |
| 982 |
| 983 /** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */ |
| 984 #define UTF_BACK_N_UNSAFE(s, i, n) UTF16_BACK_N_UNSAFE(s, i, n
) |
| 985 |
| 986 /** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */ |
| 987 #define UTF_BACK_N_SAFE(s, start, i, n) UTF16_BACK_N_SAFE(s, start,
i, n) |
| 988 |
| 989 |
| 990 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */ |
| 991 #define UTF_SET_CHAR_LIMIT_UNSAFE(s, i) UTF16_SET_CHAR_LIMIT_UNSAFE
(s, i) |
| 992 |
| 993 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */ |
| 994 #define UTF_SET_CHAR_LIMIT_SAFE(s, start, i, length) UTF16_SET_CHAR_LIMIT_SAFE(s
, start, i, length) |
| 995 |
| 996 /* Define default macros (UTF-16 "safe") ------------------------------------ */ |
| 997 |
| 998 /** |
| 999 * Does this code unit alone encode a code point (BMP, not a surrogate)? |
| 1000 * Same as UTF16_IS_SINGLE. |
| 1001 * @deprecated ICU 2.4. Renamed to U_IS_SINGLE and U16_IS_SINGLE, see utf_old.h. |
| 1002 */ |
| 1003 #define UTF_IS_SINGLE(uchar) U16_IS_SINGLE(uchar) |
| 1004 |
| 1005 /** |
| 1006 * Is this code unit the first one of several (a lead surrogate)? |
| 1007 * Same as UTF16_IS_LEAD. |
| 1008 * @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h. |
| 1009 */ |
| 1010 #define UTF_IS_LEAD(uchar) U16_IS_LEAD(uchar) |
| 1011 |
| 1012 /** |
| 1013 * Is this code unit one of several but not the first one (a trail surrogate)? |
| 1014 * Same as UTF16_IS_TRAIL. |
| 1015 * @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h. |
| 1016 */ |
| 1017 #define UTF_IS_TRAIL(uchar) U16_IS_TRAIL(uchar) |
| 1018 |
| 1019 /** |
| 1020 * Does this code point require multiple code units (is it a supplementary code
point)? |
| 1021 * Same as UTF16_NEED_MULTIPLE_UCHAR. |
| 1022 * @deprecated ICU 2.4. Use U16_LENGTH or test ((uint32_t)(c)>0xffff) instead. |
| 1023 */ |
| 1024 #define UTF_NEED_MULTIPLE_UCHAR(c) UTF16_NEED_MULTIPLE_UCHAR(c) |
| 1025 |
| 1026 /** |
| 1027 * How many code units are used to encode this code point (1 or 2)? |
| 1028 * Same as UTF16_CHAR_LENGTH. |
| 1029 * @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h. |
| 1030 */ |
| 1031 #define UTF_CHAR_LENGTH(c) U16_LENGTH(c) |
| 1032 |
| 1033 /** |
| 1034 * How many code units are used at most for any Unicode code point (2)? |
| 1035 * Same as UTF16_MAX_CHAR_LENGTH. |
| 1036 * @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h. |
| 1037 */ |
| 1038 #define UTF_MAX_CHAR_LENGTH U16_MAX_LENGTH |
| 1039 |
| 1040 /** |
| 1041 * Set c to the code point that contains the code unit i. |
| 1042 * i could point to the lead or the trail surrogate for the code point. |
| 1043 * i is not modified. |
| 1044 * Same as UTF16_GET_CHAR. |
| 1045 * \pre 0<=i<length |
| 1046 * |
| 1047 * @deprecated ICU 2.4. Renamed to U16_GET, see utf_old.h. |
| 1048 */ |
| 1049 #define UTF_GET_CHAR(s, start, i, length, c) U16_GET(s, start, i, length, c) |
| 1050 |
| 1051 /** |
| 1052 * Set c to the code point that starts at code unit i |
| 1053 * and advance i to beyond the code units of this code point (post-increment). |
| 1054 * i must point to the first code unit of a code point. |
| 1055 * Otherwise c is set to the trail unit (surrogate) itself. |
| 1056 * Same as UTF16_NEXT_CHAR. |
| 1057 * \pre 0<=i<length |
| 1058 * \post 0<i<=length |
| 1059 * |
| 1060 * @deprecated ICU 2.4. Renamed to U16_NEXT, see utf_old.h. |
| 1061 */ |
| 1062 #define UTF_NEXT_CHAR(s, i, length, c) U16_NEXT(s, i, length, c) |
| 1063 |
| 1064 /** |
| 1065 * Append the code units of code point c to the string at index i |
| 1066 * and advance i to beyond the new code units (post-increment). |
| 1067 * The code units beginning at index i will be overwritten. |
| 1068 * Same as UTF16_APPEND_CHAR. |
| 1069 * \pre 0<=c<=0x10ffff |
| 1070 * \pre 0<=i<length |
| 1071 * \post 0<i<=length |
| 1072 * |
| 1073 * @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. |
| 1074 */ |
| 1075 #define UTF_APPEND_CHAR(s, i, length, c) UTF16_APPEND_CHAR_SAFE(s, i, length, c) |
| 1076 |
| 1077 /** |
| 1078 * Advance i to beyond the code units of the code point that begins at i. |
| 1079 * I.e., advance i by one code point. |
| 1080 * Same as UTF16_FWD_1. |
| 1081 * \pre 0<=i<length |
| 1082 * \post 0<i<=length |
| 1083 * |
| 1084 * @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. |
| 1085 */ |
| 1086 #define UTF_FWD_1(s, i, length) U16_FWD_1(s, i, length) |
| 1087 |
| 1088 /** |
| 1089 * Advance i to beyond the code units of the n code points where the first one b
egins at i. |
| 1090 * I.e., advance i by n code points. |
| 1091 * Same as UT16_FWD_N. |
| 1092 * \pre 0<=i<length |
| 1093 * \post 0<i<=length |
| 1094 * |
| 1095 * @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. |
| 1096 */ |
| 1097 #define UTF_FWD_N(s, i, length, n) U16_FWD_N(s, i, length, n) |
| 1098 |
| 1099 /** |
| 1100 * Take the random-access index i and adjust it so that it points to the beginni
ng |
| 1101 * of a code point. |
| 1102 * The input index points to any code unit of a code point and is moved to point
to |
| 1103 * the first code unit of the same code point. i is never incremented. |
| 1104 * In other words, if i points to a trail surrogate that is preceded by a matchi
ng |
| 1105 * lead surrogate, then i is decremented. Otherwise it is not modified. |
| 1106 * This can be used to start an iteration with UTF_NEXT_CHAR() from a random ind
ex. |
| 1107 * Same as UTF16_SET_CHAR_START. |
| 1108 * \pre start<=i<length |
| 1109 * \post start<=i<length |
| 1110 * |
| 1111 * @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. |
| 1112 */ |
| 1113 #define UTF_SET_CHAR_START(s, start, i) U16_SET_CP_START(s, start, i) |
| 1114 |
| 1115 /** |
| 1116 * Set c to the code point that has code units before i |
| 1117 * and move i backward (towards the beginning of the string) |
| 1118 * to the first code unit of this code point (pre-increment). |
| 1119 * i must point to the first code unit after the last unit of a code point (i==l
ength is allowed). |
| 1120 * Same as UTF16_PREV_CHAR. |
| 1121 * \pre start<i<=length |
| 1122 * \post start<=i<length |
| 1123 * |
| 1124 * @deprecated ICU 2.4. Renamed to U16_PREV, see utf_old.h. |
| 1125 */ |
| 1126 #define UTF_PREV_CHAR(s, start, i, c) U16_PREV(s, start, i, c) |
| 1127 |
| 1128 /** |
| 1129 * Move i backward (towards the beginning of the string) |
| 1130 * to the first code unit of the code point that has code units before i. |
| 1131 * I.e., move i backward by one code point. |
| 1132 * i must point to the first code unit after the last unit of a code point (i==l
ength is allowed). |
| 1133 * Same as UTF16_BACK_1. |
| 1134 * \pre start<i<=length |
| 1135 * \post start<=i<length |
| 1136 * |
| 1137 * @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. |
| 1138 */ |
| 1139 #define UTF_BACK_1(s, start, i) U16_BACK_1(s, start, i) |
| 1140 |
| 1141 /** |
| 1142 * Move i backward (towards the beginning of the string) |
| 1143 * to the first code unit of the n code points that have code units before i. |
| 1144 * I.e., move i backward by n code points. |
| 1145 * i must point to the first code unit after the last unit of a code point (i==l
ength is allowed). |
| 1146 * Same as UTF16_BACK_N. |
| 1147 * \pre start<i<=length |
| 1148 * \post start<=i<length |
| 1149 * |
| 1150 * @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. |
| 1151 */ |
| 1152 #define UTF_BACK_N(s, start, i, n) U16_BACK_N(s, start, i, n) |
| 1153 |
| 1154 /** |
| 1155 * Take the random-access index i and adjust it so that it points beyond |
| 1156 * a code point. The input index points beyond any code unit |
| 1157 * of a code point and is moved to point beyond the last code unit of the same |
| 1158 * code point. i is never decremented. |
| 1159 * In other words, if i points to a trail surrogate that is preceded by a matchi
ng |
| 1160 * lead surrogate, then i is incremented. Otherwise it is not modified. |
| 1161 * This can be used to start an iteration with UTF_PREV_CHAR() from a random ind
ex. |
| 1162 * Same as UTF16_SET_CHAR_LIMIT. |
| 1163 * \pre start<i<=length |
| 1164 * \post start<i<=length |
| 1165 * |
| 1166 * @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. |
| 1167 */ |
| 1168 #define UTF_SET_CHAR_LIMIT(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, le
ngth) |
| 1169 |
| 1170 #endif /* U_HIDE_DEPRECATED_API */ |
| 1171 |
| 1172 #endif |
| 1173 |
OLD | NEW |