1 /* 2 ******************************************************************************* 3 * 4 * Copyright (C) 2002-2005, 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 processing) 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 UChar 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 compatibility.</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 for 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 points and 107 * for using UTF Unicode strings. It includes utf8.h, utf16.h, and utf32.h for the 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, "unsafe" and "safe" 114 * ones with ..._UNSAFE and ..._SAFE suffixes. The unsafe macros are fast but may 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 detected. 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 strictness 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 suffixes and signatures. 127 * Then, in a second step, the default, general alias macros are set to use either 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 to be unsafe, safe/not strict or safe/strict. 132 * The default is safe/not strict. It is not recommended to select the unsafe macros 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_ prefix and no _SAFE/_UNSAFE suffix. 136 * Only in some cases it may be necessary to control the choice of macro directly 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 need 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 removed 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. @deprecated 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 macro 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 * <p>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:<br> 187 * U+0015 = NAK = Negative Acknowledge, C0 control character<br> 188 * U+009f = highest C1 control character</p> 189 * 190 * <p>These are used by UTF8_..._SAFE macros so that they can return an error value 191 * that needs the same number of code units (bytes) as were seen by 192 * a macro. They should be tested with UTF_IS_ERROR() or UTF_IS_VALID().</p> 193 * 194 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 195 */ 196 #define UTF8_ERROR_VALUE_1 0x15 197 198 /** 199 * See documentation on UTF8_ERROR_VALUE_1 for details. 200 * 201 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 202 */ 203 #define UTF8_ERROR_VALUE_2 0x9f 204 205 /** 206 * Error value for all UTFs. This code point value will be set by macros with error 207 * checking if an error is detected. 208 * 209 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 210 */ 211 #define UTF_ERROR_VALUE 0xffff 212 213 /** 214 * Is a given 32-bit code an error value 215 * as returned by one of the macros for any UTF? 216 * 217 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 218 */ 219 #define UTF_IS_ERROR(c) \ 220 (((c)&0xfffe)==0xfffe || (c)==UTF8_ERROR_VALUE_1 || (c)==UTF8_ERROR_VALUE_2) 221 222 /** 223 * This is a combined macro: Is c a valid Unicode value _and_ not an error code? 224 * 225 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 226 */ 227 #define UTF_IS_VALID(c) \ 228 (UTF_IS_UNICODE_CHAR(c) && \ 229 (c)!=UTF8_ERROR_VALUE_1 && (c)!=UTF8_ERROR_VALUE_2) 230 231 /** 232 * Is this code unit or code point a surrogate (U+d800..U+dfff)? 233 * @deprecated ICU 2.4. Renamed to U_IS_SURROGATE and U16_IS_SURROGATE, see utf_old.h. 234 */ 235 #define UTF_IS_SURROGATE(uchar) (((uchar)&0xfffff800)==0xd800) 236 237 /** 238 * Is a given 32-bit code point a Unicode noncharacter? 239 * 240 * @deprecated ICU 2.4. Renamed to U_IS_UNICODE_NONCHAR, see utf_old.h. 241 */ 242 #define UTF_IS_UNICODE_NONCHAR(c) \ 243 ((c)>=0xfdd0 && \ 244 ((uint32_t)(c)<=0xfdef || ((c)&0xfffe)==0xfffe) && \ 245 (uint32_t)(c)<=0x10ffff) 246 247 /** 248 * Is a given 32-bit value a Unicode code point value (0..U+10ffff) 249 * that can be assigned a character? 250 * 251 * Code points that are not characters include: 252 * - single surrogate code points (U+d800..U+dfff, 2048 code points) 253 * - the last two code points on each plane (U+__fffe and U+__ffff, 34 code points) 254 * - U+fdd0..U+fdef (new with Unicode 3.1, 32 code points) 255 * - the highest Unicode code point value is U+10ffff 256 * 257 * This means that all code points below U+d800 are character code points, 258 * and that boundary is tested first for performance. 259 * 260 * @deprecated ICU 2.4. Renamed to U_IS_UNICODE_CHAR, see utf_old.h. 261 */ 262 #define UTF_IS_UNICODE_CHAR(c) \ 263 ((uint32_t)(c)<0xd800 || \ 264 ((uint32_t)(c)>0xdfff && \ 265 (uint32_t)(c)<=0x10ffff && \ 266 !UTF_IS_UNICODE_NONCHAR(c))) 267 268 /* Formerly utf8.h ---------------------------------------------------------- */ 269 270 /** 271 * Count the trail bytes for a UTF-8 lead byte. 272 * @deprecated ICU 2.4. Renamed to U8_COUNT_TRAIL_BYTES, see utf_old.h. 273 */ 274 #define UTF8_COUNT_TRAIL_BYTES(leadByte) (utf8_countTrailBytes[(uint8_t)leadByte]) 275 276 /** 277 * Mask a UTF-8 lead byte, leave only the lower bits that form part of the code point value. 278 * @deprecated ICU 2.4. Renamed to U8_MASK_LEAD_BYTE, see utf_old.h. 279 */ 280 #define UTF8_MASK_LEAD_BYTE(leadByte, countTrailBytes) ((leadByte)&=(1<<(6-(countTrailBytes)))-1) 281 282 /** Is this this code point a single code unit (byte)? @deprecated ICU 2.4. Renamed to U8_IS_SINGLE, see utf_old.h. */ 283 #define UTF8_IS_SINGLE(uchar) (((uchar)&0x80)==0) 284 /** 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. */ 285 #define UTF8_IS_LEAD(uchar) ((uint8_t)((uchar)-0xc0)<0x3e) 286 /** 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. */ 287 #define UTF8_IS_TRAIL(uchar) (((uchar)&0xc0)==0x80) 288 289 /** 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. */ 290 #define UTF8_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0x7f) 291 292 /** 293 * Given the lead character, how many bytes are taken by this code point. 294 * ICU does not deal with code points >0x10ffff 295 * unless necessary for advancing in the byte stream. 296 * 297 * These length macros take into account that for values >0x10ffff 298 * the UTF8_APPEND_CHAR_SAFE macros would write the error code point 0xffff 299 * with 3 bytes. 300 * Code point comparisons need to be in uint32_t because UChar32 301 * may be a signed type, and negative values must be recognized. 302 * 303 * @deprecated ICU 2.4. Use U8_LENGTH instead, see utf_old.h. 304 */ 305 #if 1 306 # define UTF8_CHAR_LENGTH(c) \ 307 ((uint32_t)(c)<=0x7f ? 1 : \ 308 ((uint32_t)(c)<=0x7ff ? 2 : \ 309 ((uint32_t)((c)-0x10000)>0xfffff ? 3 : 4) \ 310 ) \ 311 ) 312 #else 313 # define UTF8_CHAR_LENGTH(c) \ 314 ((uint32_t)(c)<=0x7f ? 1 : \ 315 ((uint32_t)(c)<=0x7ff ? 2 : \ 316 ((uint32_t)(c)<=0xffff ? 3 : \ 317 ((uint32_t)(c)<=0x10ffff ? 4 : \ 318 ((uint32_t)(c)<=0x3ffffff ? 5 : \ 319 ((uint32_t)(c)<=0x7fffffff ? 6 : 3) \ 320 ) \ 321 ) \ 322 ) \ 323 ) \ 324 ) 325 #endif 326 327 /** The maximum number of bytes per code point. @deprecated ICU 2.4. Renamed to U8_MAX_LENGTH, see utf_old.h. */ 328 #define UTF8_MAX_CHAR_LENGTH 4 329 330 /** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 331 #define UTF8_ARRAY_SIZE(size) ((5*(size))/2) 332 333 /** @deprecated ICU 2.4. Renamed to U8_GET_UNSAFE, see utf_old.h. */ 334 #define UTF8_GET_CHAR_UNSAFE(s, i, c) { \ 335 int32_t _utf8_get_char_unsafe_index=(int32_t)(i); \ 336 UTF8_SET_CHAR_START_UNSAFE(s, _utf8_get_char_unsafe_index); \ 337 UTF8_NEXT_CHAR_UNSAFE(s, _utf8_get_char_unsafe_index, c); \ 338 } 339 340 /** @deprecated ICU 2.4. Use U8_GET instead, see utf_old.h. */ 341 #define UTF8_GET_CHAR_SAFE(s, start, i, length, c, strict) { \ 342 int32_t _utf8_get_char_safe_index=(int32_t)(i); \ 343 UTF8_SET_CHAR_START_SAFE(s, start, _utf8_get_char_safe_index); \ 344 UTF8_NEXT_CHAR_SAFE(s, _utf8_get_char_safe_index, length, c, strict); \ 345 } 346 347 /** @deprecated ICU 2.4. Renamed to U8_NEXT_UNSAFE, see utf_old.h. */ 348 #define UTF8_NEXT_CHAR_UNSAFE(s, i, c) { \ 349 (c)=(s)[(i)++]; \ 350 if((uint8_t)((c)-0xc0)<0x35) { \ 351 uint8_t __count=UTF8_COUNT_TRAIL_BYTES(c); \ 352 UTF8_MASK_LEAD_BYTE(c, __count); \ 353 switch(__count) { \ 354 /* each following branch falls through to the next one */ \ 355 case 3: \ 356 (c)=((c)<<6)|((s)[(i)++]&0x3f); \ 357 case 2: \ 358 (c)=((c)<<6)|((s)[(i)++]&0x3f); \ 359 case 1: \ 360 (c)=((c)<<6)|((s)[(i)++]&0x3f); \ 361 /* no other branches to optimize switch() */ \ 362 break; \ 363 } \ 364 } \ 365 } 366 367 /** @deprecated ICU 2.4. Renamed to U8_APPEND_UNSAFE, see utf_old.h. */ 368 #define UTF8_APPEND_CHAR_UNSAFE(s, i, c) { \ 369 if((uint32_t)(c)<=0x7f) { \ 370 (s)[(i)++]=(uint8_t)(c); \ 371 } else { \ 372 if((uint32_t)(c)<=0x7ff) { \ 373 (s)[(i)++]=(uint8_t)(((c)>>6)|0xc0); \ 374 } else { \ 375 if((uint32_t)(c)<=0xffff) { \ 376 (s)[(i)++]=(uint8_t)(((c)>>12)|0xe0); \ 377 } else { \ 378 (s)[(i)++]=(uint8_t)(((c)>>18)|0xf0); \ 379 (s)[(i)++]=(uint8_t)((((c)>>12)&0x3f)|0x80); \ 380 } \ 381 (s)[(i)++]=(uint8_t)((((c)>>6)&0x3f)|0x80); \ 382 } \ 383 (s)[(i)++]=(uint8_t)(((c)&0x3f)|0x80); \ 384 } \ 385 } 386 387 /** @deprecated ICU 2.4. Renamed to U8_FWD_1_UNSAFE, see utf_old.h. */ 388 #define UTF8_FWD_1_UNSAFE(s, i) { \ 389 (i)+=1+UTF8_COUNT_TRAIL_BYTES((s)[i]); \ 390 } 391 392 /** @deprecated ICU 2.4. Renamed to U8_FWD_N_UNSAFE, see utf_old.h. */ 393 #define UTF8_FWD_N_UNSAFE(s, i, n) { \ 394 int32_t __N=(n); \ 395 while(__N>0) { \ 396 UTF8_FWD_1_UNSAFE(s, i); \ 397 --__N; \ 398 } \ 399 } 400 401 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_START_UNSAFE, see utf_old.h. */ 402 #define UTF8_SET_CHAR_START_UNSAFE(s, i) { \ 403 while(UTF8_IS_TRAIL((s)[i])) { --(i); } \ 404 } 405 406 /** @deprecated ICU 2.4. Use U8_NEXT instead, see utf_old.h. */ 407 #define UTF8_NEXT_CHAR_SAFE(s, i, length, c, strict) { \ 408 (c)=(s)[(i)++]; \ 409 if((c)>=0x80) { \ 410 if(UTF8_IS_LEAD(c)) { \ 411 (c)=utf8_nextCharSafeBody(s, &(i), (int32_t)(length), c, strict); \ 412 } else { \ 413 (c)=UTF8_ERROR_VALUE_1; \ 414 } \ 415 } \ 416 } 417 418 /** @deprecated ICU 2.4. Use U8_APPEND instead, see utf_old.h. */ 419 #define UTF8_APPEND_CHAR_SAFE(s, i, length, c) { \ 420 if((uint32_t)(c)<=0x7f) { \ 421 (s)[(i)++]=(uint8_t)(c); \ 422 } else { \ 423 (i)=utf8_appendCharSafeBody(s, (int32_t)(i), (int32_t)(length), c, NULL); \ 424 } \ 425 } 426 427 /** @deprecated ICU 2.4. Renamed to U8_FWD_1, see utf_old.h. */ 428 #define UTF8_FWD_1_SAFE(s, i, length) U8_FWD_1(s, i, length) 429 430 /** @deprecated ICU 2.4. Renamed to U8_FWD_N, see utf_old.h. */ 431 #define UTF8_FWD_N_SAFE(s, i, length, n) U8_FWD_N(s, i, length, n) 432 433 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_START, see utf_old.h. */ 434 #define UTF8_SET_CHAR_START_SAFE(s, start, i) U8_SET_CP_START(s, start, i) 435 436 /** @deprecated ICU 2.4. Renamed to U8_PREV_UNSAFE, see utf_old.h. */ 437 #define UTF8_PREV_CHAR_UNSAFE(s, i, c) { \ 438 (c)=(s)[--(i)]; \ 439 if(UTF8_IS_TRAIL(c)) { \ 440 uint8_t __b, __count=1, __shift=6; \ 441 \ 442 /* c is a trail byte */ \ 443 (c)&=0x3f; \ 444 for(;;) { \ 445 __b=(s)[--(i)]; \ 446 if(__b>=0xc0) { \ 447 UTF8_MASK_LEAD_BYTE(__b, __count); \ 448 (c)|=(UChar32)__b<<__shift; \ 449 break; \ 450 } else { \ 451 (c)|=(UChar32)(__b&0x3f)<<__shift; \ 452 ++__count; \ 453 __shift+=6; \ 454 } \ 455 } \ 456 } \ 457 } 458 459 /** @deprecated ICU 2.4. Renamed to U8_BACK_1_UNSAFE, see utf_old.h. */ 460 #define UTF8_BACK_1_UNSAFE(s, i) { \ 461 while(UTF8_IS_TRAIL((s)[--(i)])) {} \ 462 } 463 464 /** @deprecated ICU 2.4. Renamed to U8_BACK_N_UNSAFE, see utf_old.h. */ 465 #define UTF8_BACK_N_UNSAFE(s, i, n) { \ 466 int32_t __N=(n); \ 467 while(__N>0) { \ 468 UTF8_BACK_1_UNSAFE(s, i); \ 469 --__N; \ 470 } \ 471 } 472 473 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT_UNSAFE, see utf_old.h. */ 474 #define UTF8_SET_CHAR_LIMIT_UNSAFE(s, i) { \ 475 UTF8_BACK_1_UNSAFE(s, i); \ 476 UTF8_FWD_1_UNSAFE(s, i); \ 477 } 478 479 /** @deprecated ICU 2.4. Use U8_PREV instead, see utf_old.h. */ 480 #define UTF8_PREV_CHAR_SAFE(s, start, i, c, strict) { \ 481 (c)=(s)[--(i)]; \ 482 if((c)>=0x80) { \ 483 if((c)<=0xbf) { \ 484 (c)=utf8_prevCharSafeBody(s, start, &(i), c, strict); \ 485 } else { \ 486 (c)=UTF8_ERROR_VALUE_1; \ 487 } \ 488 } \ 489 } 490 491 /** @deprecated ICU 2.4. Renamed to U8_BACK_1, see utf_old.h. */ 492 #define UTF8_BACK_1_SAFE(s, start, i) U8_BACK_1(s, start, i) 493 494 /** @deprecated ICU 2.4. Renamed to U8_BACK_N, see utf_old.h. */ 495 #define UTF8_BACK_N_SAFE(s, start, i, n) U8_BACK_N(s, start, i, n) 496 497 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT, see utf_old.h. */ 498 #define UTF8_SET_CHAR_LIMIT_SAFE(s, start, i, length) U8_SET_CP_LIMIT(s, start, i, length) 499 500 /* Formerly utf16.h --------------------------------------------------------- */ 501 502 /** Is uchar a first/lead surrogate? @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h. */ 503 #define UTF_IS_FIRST_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xd800) 504 505 /** Is uchar a second/trail surrogate? @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h. */ 506 #define UTF_IS_SECOND_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xdc00) 507 508 /** 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. */ 509 #define UTF_IS_SURROGATE_FIRST(c) (((c)&0x400)==0) 510 511 /** Helper constant for UTF16_GET_PAIR_VALUE. @deprecated ICU 2.4. Renamed to U16_SURROGATE_OFFSET, see utf_old.h. */ 512 #define UTF_SURROGATE_OFFSET ((0xd800<<10UL)+0xdc00-0x10000) 513 514 /** Get the UTF-32 value from the surrogate code units. @deprecated ICU 2.4. Renamed to U16_GET_SUPPLEMENTARY, see utf_old.h. */ 515 #define UTF16_GET_PAIR_VALUE(first, second) \ 516 (((first)<<10UL)+(second)-UTF_SURROGATE_OFFSET) 517 518 /** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */ 519 #define UTF_FIRST_SURROGATE(supplementary) (UChar)(((supplementary)>>10)+0xd7c0) 520 521 /** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */ 522 #define UTF_SECOND_SURROGATE(supplementary) (UChar)(((supplementary)&0x3ff)|0xdc00) 523 524 /** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */ 525 #define UTF16_LEAD(supplementary) UTF_FIRST_SURROGATE(supplementary) 526 527 /** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */ 528 #define UTF16_TRAIL(supplementary) UTF_SECOND_SURROGATE(supplementary) 529 530 /** @deprecated ICU 2.4. Renamed to U16_IS_SINGLE, see utf_old.h. */ 531 #define UTF16_IS_SINGLE(uchar) !UTF_IS_SURROGATE(uchar) 532 533 /** @deprecated ICU 2.4. Renamed to U16_IS_LEAD, see utf_old.h. */ 534 #define UTF16_IS_LEAD(uchar) UTF_IS_FIRST_SURROGATE(uchar) 535 536 /** @deprecated ICU 2.4. Renamed to U16_IS_TRAIL, see utf_old.h. */ 537 #define UTF16_IS_TRAIL(uchar) UTF_IS_SECOND_SURROGATE(uchar) 538 539 /** 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. */ 540 #define UTF16_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0xffff) 541 542 /** @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h. */ 543 #define UTF16_CHAR_LENGTH(c) ((uint32_t)(c)<=0xffff ? 1 : 2) 544 545 /** @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h. */ 546 #define UTF16_MAX_CHAR_LENGTH 2 547 548 /** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 549 #define UTF16_ARRAY_SIZE(size) (size) 550 551 /** 552 * Get a single code point from an offset that points to any 553 * of the code units that belong to that code point. 554 * Assume 0<=i<length. 555 * 556 * This could be used for iteration together with 557 * UTF16_CHAR_LENGTH() and UTF_IS_ERROR(), 558 * but the use of UTF16_NEXT_CHAR[_UNSAFE]() and 559 * UTF16_PREV_CHAR[_UNSAFE]() is more efficient for that. 560 * @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h. 561 */ 562 #define UTF16_GET_CHAR_UNSAFE(s, i, c) { \ 563 (c)=(s)[i]; \ 564 if(UTF_IS_SURROGATE(c)) { \ 565 if(UTF_IS_SURROGATE_FIRST(c)) { \ 566 (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)+1]); \ 567 } else { \ 568 (c)=UTF16_GET_PAIR_VALUE((s)[(i)-1], (c)); \ 569 } \ 570 } \ 571 } 572 573 /** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */ 574 #define UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict) { \ 575 (c)=(s)[i]; \ 576 if(UTF_IS_SURROGATE(c)) { \ 577 uint16_t __c2; \ 578 if(UTF_IS_SURROGATE_FIRST(c)) { \ 579 if((i)+1<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)+1])) { \ 580 (c)=UTF16_GET_PAIR_VALUE((c), __c2); \ 581 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \ 582 } else if(strict) {\ 583 /* unmatched first surrogate */ \ 584 (c)=UTF_ERROR_VALUE; \ 585 } \ 586 } else { \ 587 if((i)-1>=(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \ 588 (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \ 589 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \ 590 } else if(strict) {\ 591 /* unmatched second surrogate */ \ 592 (c)=UTF_ERROR_VALUE; \ 593 } \ 594 } \ 595 } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \ 596 (c)=UTF_ERROR_VALUE; \ 597 } \ 598 } 599 600 /** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */ 601 #define UTF16_NEXT_CHAR_UNSAFE(s, i, c) { \ 602 (c)=(s)[(i)++]; \ 603 if(UTF_IS_FIRST_SURROGATE(c)) { \ 604 (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)++]); \ 605 } \ 606 } 607 608 /** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */ 609 #define UTF16_APPEND_CHAR_UNSAFE(s, i, c) { \ 610 if((uint32_t)(c)<=0xffff) { \ 611 (s)[(i)++]=(uint16_t)(c); \ 612 } else { \ 613 (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ 614 (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ 615 } \ 616 } 617 618 /** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */ 619 #define UTF16_FWD_1_UNSAFE(s, i) { \ 620 if(UTF_IS_FIRST_SURROGATE((s)[(i)++])) { \ 621 ++(i); \ 622 } \ 623 } 624 625 /** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */ 626 #define UTF16_FWD_N_UNSAFE(s, i, n) { \ 627 int32_t __N=(n); \ 628 while(__N>0) { \ 629 UTF16_FWD_1_UNSAFE(s, i); \ 630 --__N; \ 631 } \ 632 } 633 634 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */ 635 #define UTF16_SET_CHAR_START_UNSAFE(s, i) { \ 636 if(UTF_IS_SECOND_SURROGATE((s)[i])) { \ 637 --(i); \ 638 } \ 639 } 640 641 /** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */ 642 #define UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict) { \ 643 (c)=(s)[(i)++]; \ 644 if(UTF_IS_FIRST_SURROGATE(c)) { \ 645 uint16_t __c2; \ 646 if((i)<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)])) { \ 647 ++(i); \ 648 (c)=UTF16_GET_PAIR_VALUE((c), __c2); \ 649 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \ 650 } else if(strict) {\ 651 /* unmatched first surrogate */ \ 652 (c)=UTF_ERROR_VALUE; \ 653 } \ 654 } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \ 655 /* unmatched second surrogate or other non-character */ \ 656 (c)=UTF_ERROR_VALUE; \ 657 } \ 658 } 659 660 /** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */ 661 #define UTF16_APPEND_CHAR_SAFE(s, i, length, c) { \ 662 if((uint32_t)(c)<=0xffff) { \ 663 (s)[(i)++]=(uint16_t)(c); \ 664 } else if((uint32_t)(c)<=0x10ffff) { \ 665 if((i)+1<(length)) { \ 666 (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ 667 (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ 668 } else /* not enough space */ { \ 669 (s)[(i)++]=UTF_ERROR_VALUE; \ 670 } \ 671 } else /* c>0x10ffff, write error value */ { \ 672 (s)[(i)++]=UTF_ERROR_VALUE; \ 673 } \ 674 } 675 676 /** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */ 677 #define UTF16_FWD_1_SAFE(s, i, length) U16_FWD_1(s, i, length) 678 679 /** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */ 680 #define UTF16_FWD_N_SAFE(s, i, length, n) U16_FWD_N(s, i, length, n) 681 682 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */ 683 #define UTF16_SET_CHAR_START_SAFE(s, start, i) U16_SET_CP_START(s, start, i) 684 685 /** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */ 686 #define UTF16_PREV_CHAR_UNSAFE(s, i, c) { \ 687 (c)=(s)[--(i)]; \ 688 if(UTF_IS_SECOND_SURROGATE(c)) { \ 689 (c)=UTF16_GET_PAIR_VALUE((s)[--(i)], (c)); \ 690 } \ 691 } 692 693 /** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */ 694 #define UTF16_BACK_1_UNSAFE(s, i) { \ 695 if(UTF_IS_SECOND_SURROGATE((s)[--(i)])) { \ 696 --(i); \ 697 } \ 698 } 699 700 /** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */ 701 #define UTF16_BACK_N_UNSAFE(s, i, n) { \ 702 int32_t __N=(n); \ 703 while(__N>0) { \ 704 UTF16_BACK_1_UNSAFE(s, i); \ 705 --__N; \ 706 } \ 707 } 708 709 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */ 710 #define UTF16_SET_CHAR_LIMIT_UNSAFE(s, i) { \ 711 if(UTF_IS_FIRST_SURROGATE((s)[(i)-1])) { \ 712 ++(i); \ 713 } \ 714 } 715 716 /** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */ 717 #define UTF16_PREV_CHAR_SAFE(s, start, i, c, strict) { \ 718 (c)=(s)[--(i)]; \ 719 if(UTF_IS_SECOND_SURROGATE(c)) { \ 720 uint16_t __c2; \ 721 if((i)>(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \ 722 --(i); \ 723 (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \ 724 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \ 725 } else if(strict) {\ 726 /* unmatched second surrogate */ \ 727 (c)=UTF_ERROR_VALUE; \ 728 } \ 729 } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \ 730 /* unmatched first surrogate or other non-character */ \ 731 (c)=UTF_ERROR_VALUE; \ 732 } \ 733 } 734 735 /** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */ 736 #define UTF16_BACK_1_SAFE(s, start, i) U16_BACK_1(s, start, i) 737 738 /** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */ 739 #define UTF16_BACK_N_SAFE(s, start, i, n) U16_BACK_N(s, start, i, n) 740 741 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */ 742 #define UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, length) 743 744 /* Formerly utf32.h --------------------------------------------------------- */ 745 746 /* 747 * Old documentation: 748 * 749 * This file defines macros to deal with UTF-32 code units and code points. 750 * Signatures and semantics are the same as for the similarly named macros 751 * in utf16.h. 752 * utf32.h is included by utf.h after unicode/umachine.h</p> 753 * and some common definitions. 754 * <p><b>Usage:</b> ICU coding guidelines for if() statements should be followed when using these macros. 755 * Compound statements (curly braces {}) must be used for if-else-while... 756 * bodies and all macro statements should be terminated with semicolon.</p> 757 */ 758 759 /* internal definitions ----------------------------------------------------- */ 760 761 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 762 #define UTF32_IS_SAFE(c, strict) \ 763 (!(strict) ? \ 764 (uint32_t)(c)<=0x10ffff : \ 765 UTF_IS_UNICODE_CHAR(c)) 766 767 /* 768 * For the semantics of all of these macros, see utf16.h. 769 * The UTF-32 versions are trivial because any code point is 770 * encoded using exactly one code unit. 771 */ 772 773 /* single-code point definitions -------------------------------------------- */ 774 775 /* classes of code unit values */ 776 777 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 778 #define UTF32_IS_SINGLE(uchar) 1 779 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 780 #define UTF32_IS_LEAD(uchar) 0 781 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 782 #define UTF32_IS_TRAIL(uchar) 0 783 784 /* number of code units per code point */ 785 786 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 787 #define UTF32_NEED_MULTIPLE_UCHAR(c) 0 788 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 789 #define UTF32_CHAR_LENGTH(c) 1 790 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 791 #define UTF32_MAX_CHAR_LENGTH 1 792 793 /* average number of code units compared to UTF-16 */ 794 795 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 796 #define UTF32_ARRAY_SIZE(size) (size) 797 798 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 799 #define UTF32_GET_CHAR_UNSAFE(s, i, c) { \ 800 (c)=(s)[i]; \ 801 } 802 803 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 804 #define UTF32_GET_CHAR_SAFE(s, start, i, length, c, strict) { \ 805 (c)=(s)[i]; \ 806 if(!UTF32_IS_SAFE(c, strict)) { \ 807 (c)=UTF_ERROR_VALUE; \ 808 } \ 809 } 810 811 /* definitions with forward iteration --------------------------------------- */ 812 813 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 814 #define UTF32_NEXT_CHAR_UNSAFE(s, i, c) { \ 815 (c)=(s)[(i)++]; \ 816 } 817 818 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 819 #define UTF32_APPEND_CHAR_UNSAFE(s, i, c) { \ 820 (s)[(i)++]=(c); \ 821 } 822 823 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 824 #define UTF32_FWD_1_UNSAFE(s, i) { \ 825 ++(i); \ 826 } 827 828 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 829 #define UTF32_FWD_N_UNSAFE(s, i, n) { \ 830 (i)+=(n); \ 831 } 832 833 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 834 #define UTF32_SET_CHAR_START_UNSAFE(s, i) { \ 835 } 836 837 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 838 #define UTF32_NEXT_CHAR_SAFE(s, i, length, c, strict) { \ 839 (c)=(s)[(i)++]; \ 840 if(!UTF32_IS_SAFE(c, strict)) { \ 841 (c)=UTF_ERROR_VALUE; \ 842 } \ 843 } 844 845 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 846 #define UTF32_APPEND_CHAR_SAFE(s, i, length, c) { \ 847 if((uint32_t)(c)<=0x10ffff) { \ 848 (s)[(i)++]=(c); \ 849 } else /* c>0x10ffff, write 0xfffd */ { \ 850 (s)[(i)++]=0xfffd; \ 851 } \ 852 } 853 854 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 855 #define UTF32_FWD_1_SAFE(s, i, length) { \ 856 ++(i); \ 857 } 858 859 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 860 #define UTF32_FWD_N_SAFE(s, i, length, n) { \ 861 if(((i)+=(n))>(length)) { \ 862 (i)=(length); \ 863 } \ 864 } 865 866 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 867 #define UTF32_SET_CHAR_START_SAFE(s, start, i) { \ 868 } 869 870 /* definitions with backward iteration -------------------------------------- */ 871 872 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 873 #define UTF32_PREV_CHAR_UNSAFE(s, i, c) { \ 874 (c)=(s)[--(i)]; \ 875 } 876 877 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 878 #define UTF32_BACK_1_UNSAFE(s, i) { \ 879 --(i); \ 880 } 881 882 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 883 #define UTF32_BACK_N_UNSAFE(s, i, n) { \ 884 (i)-=(n); \ 885 } 886 887 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 888 #define UTF32_SET_CHAR_LIMIT_UNSAFE(s, i) { \ 889 } 890 891 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 892 #define UTF32_PREV_CHAR_SAFE(s, start, i, c, strict) { \ 893 (c)=(s)[--(i)]; \ 894 if(!UTF32_IS_SAFE(c, strict)) { \ 895 (c)=UTF_ERROR_VALUE; \ 896 } \ 897 } 898 899 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 900 #define UTF32_BACK_1_SAFE(s, start, i) { \ 901 --(i); \ 902 } 903 904 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 905 #define UTF32_BACK_N_SAFE(s, start, i, n) { \ 906 (i)-=(n); \ 907 if((i)<(start)) { \ 908 (i)=(start); \ 909 } \ 910 } 911 912 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 913 #define UTF32_SET_CHAR_LIMIT_SAFE(s, i, length) { \ 914 } 915 916 /* Formerly utf.h, part 2 --------------------------------------------------- */ 917 918 /** 919 * Estimate the number of code units for a string based on the number of UTF-16 code units. 920 * 921 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 922 */ 923 #define UTF_ARRAY_SIZE(size) UTF16_ARRAY_SIZE(size) 924 925 /** @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h. */ 926 #define UTF_GET_CHAR_UNSAFE(s, i, c) UTF16_GET_CHAR_UNSAFE(s, i, c) 927 928 /** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */ 929 #define UTF_GET_CHAR_SAFE(s, start, i, length, c, strict) UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict) 930 931 932 /** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */ 933 #define UTF_NEXT_CHAR_UNSAFE(s, i, c) UTF16_NEXT_CHAR_UNSAFE(s, i, c) 934 935 /** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */ 936 #define UTF_NEXT_CHAR_SAFE(s, i, length, c, strict) UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict) 937 938 939 /** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */ 940 #define UTF_APPEND_CHAR_UNSAFE(s, i, c) UTF16_APPEND_CHAR_UNSAFE(s, i, c) 941 942 /** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */ 943 #define UTF_APPEND_CHAR_SAFE(s, i, length, c) UTF16_APPEND_CHAR_SAFE(s, i, length, c) 944 945 946 /** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */ 947 #define UTF_FWD_1_UNSAFE(s, i) UTF16_FWD_1_UNSAFE(s, i) 948 949 /** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */ 950 #define UTF_FWD_1_SAFE(s, i, length) UTF16_FWD_1_SAFE(s, i, length) 951 952 953 /** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */ 954 #define UTF_FWD_N_UNSAFE(s, i, n) UTF16_FWD_N_UNSAFE(s, i, n) 955 956 /** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */ 957 #define UTF_FWD_N_SAFE(s, i, length, n) UTF16_FWD_N_SAFE(s, i, length, n) 958 959 960 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */ 961 #define UTF_SET_CHAR_START_UNSAFE(s, i) UTF16_SET_CHAR_START_UNSAFE(s, i) 962 963 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */ 964 #define UTF_SET_CHAR_START_SAFE(s, start, i) UTF16_SET_CHAR_START_SAFE(s, start, i) 965 966 967 /** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */ 968 #define UTF_PREV_CHAR_UNSAFE(s, i, c) UTF16_PREV_CHAR_UNSAFE(s, i, c) 969 970 /** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */ 971 #define UTF_PREV_CHAR_SAFE(s, start, i, c, strict) UTF16_PREV_CHAR_SAFE(s, start, i, c, strict) 972 973 974 /** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */ 975 #define UTF_BACK_1_UNSAFE(s, i) UTF16_BACK_1_UNSAFE(s, i) 976 977 /** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */ 978 #define UTF_BACK_1_SAFE(s, start, i) UTF16_BACK_1_SAFE(s, start, i) 979 980 981 /** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */ 982 #define UTF_BACK_N_UNSAFE(s, i, n) UTF16_BACK_N_UNSAFE(s, i, n) 983 984 /** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */ 985 #define UTF_BACK_N_SAFE(s, start, i, n) UTF16_BACK_N_SAFE(s, start, i, n) 986 987 988 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */ 989 #define UTF_SET_CHAR_LIMIT_UNSAFE(s, i) UTF16_SET_CHAR_LIMIT_UNSAFE(s, i) 990 991 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */ 992 #define UTF_SET_CHAR_LIMIT_SAFE(s, start, i, length) UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length) 993 994 /* Define default macros (UTF-16 "safe") ------------------------------------ */ 995 996 /** 997 * Does this code unit alone encode a code point (BMP, not a surrogate)? 998 * Same as UTF16_IS_SINGLE. 999 * @deprecated ICU 2.4. Renamed to U_IS_SINGLE and U16_IS_SINGLE, see utf_old.h. 1000 */ 1001 #define UTF_IS_SINGLE(uchar) U16_IS_SINGLE(uchar) 1002 1003 /** 1004 * Is this code unit the first one of several (a lead surrogate)? 1005 * Same as UTF16_IS_LEAD. 1006 * @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h. 1007 */ 1008 #define UTF_IS_LEAD(uchar) U16_IS_LEAD(uchar) 1009 1010 /** 1011 * Is this code unit one of several but not the first one (a trail surrogate)? 1012 * Same as UTF16_IS_TRAIL. 1013 * @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h. 1014 */ 1015 #define UTF_IS_TRAIL(uchar) U16_IS_TRAIL(uchar) 1016 1017 /** 1018 * Does this code point require multiple code units (is it a supplementary code point)? 1019 * Same as UTF16_NEED_MULTIPLE_UCHAR. 1020 * @deprecated ICU 2.4. Use U16_LENGTH or test ((uint32_t)(c)>0xffff) instead. 1021 */ 1022 #define UTF_NEED_MULTIPLE_UCHAR(c) UTF16_NEED_MULTIPLE_UCHAR(c) 1023 1024 /** 1025 * How many code units are used to encode this code point (1 or 2)? 1026 * Same as UTF16_CHAR_LENGTH. 1027 * @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h. 1028 */ 1029 #define UTF_CHAR_LENGTH(c) U16_LENGTH(c) 1030 1031 /** 1032 * How many code units are used at most for any Unicode code point (2)? 1033 * Same as UTF16_MAX_CHAR_LENGTH. 1034 * @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h. 1035 */ 1036 #define UTF_MAX_CHAR_LENGTH U16_MAX_LENGTH 1037 1038 /** 1039 * Set c to the code point that contains the code unit i. 1040 * i could point to the lead or the trail surrogate for the code point. 1041 * i is not modified. 1042 * Same as UTF16_GET_CHAR. 1043 * \pre 0<=i<length 1044 * 1045 * @deprecated ICU 2.4. Renamed to U16_GET, see utf_old.h. 1046 */ 1047 #define UTF_GET_CHAR(s, start, i, length, c) U16_GET(s, start, i, length, c) 1048 1049 /** 1050 * Set c to the code point that starts at code unit i 1051 * and advance i to beyond the code units of this code point (post-increment). 1052 * i must point to the first code unit of a code point. 1053 * Otherwise c is set to the trail unit (surrogate) itself. 1054 * Same as UTF16_NEXT_CHAR. 1055 * \pre 0<=i<length 1056 * \post 0<i<=length 1057 * 1058 * @deprecated ICU 2.4. Renamed to U16_NEXT, see utf_old.h. 1059 */ 1060 #define UTF_NEXT_CHAR(s, i, length, c) U16_NEXT(s, i, length, c) 1061 1062 /** 1063 * Append the code units of code point c to the string at index i 1064 * and advance i to beyond the new code units (post-increment). 1065 * The code units beginning at index i will be overwritten. 1066 * Same as UTF16_APPEND_CHAR. 1067 * \pre 0<=c<=0x10ffff 1068 * \pre 0<=i<length 1069 * \post 0<i<=length 1070 * 1071 * @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. 1072 */ 1073 #define UTF_APPEND_CHAR(s, i, length, c) UTF16_APPEND_CHAR_SAFE(s, i, length, c) 1074 1075 /** 1076 * Advance i to beyond the code units of the code point that begins at i. 1077 * I.e., advance i by one code point. 1078 * Same as UTF16_FWD_1. 1079 * \pre 0<=i<length 1080 * \post 0<i<=length 1081 * 1082 * @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. 1083 */ 1084 #define UTF_FWD_1(s, i, length) U16_FWD_1(s, i, length) 1085 1086 /** 1087 * Advance i to beyond the code units of the n code points where the first one begins at i. 1088 * I.e., advance i by n code points. 1089 * Same as UT16_FWD_N. 1090 * \pre 0<=i<length 1091 * \post 0<i<=length 1092 * 1093 * @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. 1094 */ 1095 #define UTF_FWD_N(s, i, length, n) U16_FWD_N(s, i, length, n) 1096 1097 /** 1098 * Take the random-access index i and adjust it so that it points to the beginning 1099 * of a code point. 1100 * The input index points to any code unit of a code point and is moved to point to 1101 * the first code unit of the same code point. i is never incremented. 1102 * In other words, if i points to a trail surrogate that is preceded by a matching 1103 * lead surrogate, then i is decremented. Otherwise it is not modified. 1104 * This can be used to start an iteration with UTF_NEXT_CHAR() from a random index. 1105 * Same as UTF16_SET_CHAR_START. 1106 * \pre start<=i<length 1107 * \post start<=i<length 1108 * 1109 * @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. 1110 */ 1111 #define UTF_SET_CHAR_START(s, start, i) U16_SET_CP_START(s, start, i) 1112 1113 /** 1114 * Set c to the code point that has code units before i 1115 * and move i backward (towards the beginning of the string) 1116 * to the first code unit of this code point (pre-increment). 1117 * i must point to the first code unit after the last unit of a code point (i==length is allowed). 1118 * Same as UTF16_PREV_CHAR. 1119 * \pre start<i<=length 1120 * \post start<=i<length 1121 * 1122 * @deprecated ICU 2.4. Renamed to U16_PREV, see utf_old.h. 1123 */ 1124 #define UTF_PREV_CHAR(s, start, i, c) U16_PREV(s, start, i, c) 1125 1126 /** 1127 * Move i backward (towards the beginning of the string) 1128 * to the first code unit of the code point that has code units before i. 1129 * I.e., move i backward by one code point. 1130 * i must point to the first code unit after the last unit of a code point (i==length is allowed). 1131 * Same as UTF16_BACK_1. 1132 * \pre start<i<=length 1133 * \post start<=i<length 1134 * 1135 * @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. 1136 */ 1137 #define UTF_BACK_1(s, start, i) U16_BACK_1(s, start, i) 1138 1139 /** 1140 * Move i backward (towards the beginning of the string) 1141 * to the first code unit of the n code points that have code units before i. 1142 * I.e., move i backward by n code points. 1143 * i must point to the first code unit after the last unit of a code point (i==length is allowed). 1144 * Same as UTF16_BACK_N. 1145 * \pre start<i<=length 1146 * \post start<=i<length 1147 * 1148 * @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. 1149 */ 1150 #define UTF_BACK_N(s, start, i, n) U16_BACK_N(s, start, i, n) 1151 1152 /** 1153 * Take the random-access index i and adjust it so that it points beyond 1154 * a code point. The input index points beyond any code unit 1155 * of a code point and is moved to point beyond the last code unit of the same 1156 * code point. i is never decremented. 1157 * In other words, if i points to a trail surrogate that is preceded by a matching 1158 * lead surrogate, then i is incremented. Otherwise it is not modified. 1159 * This can be used to start an iteration with UTF_PREV_CHAR() from a random index. 1160 * Same as UTF16_SET_CHAR_LIMIT. 1161 * \pre start<i<=length 1162 * \post start<i<=length 1163 * 1164 * @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. 1165 */ 1166 #define UTF_SET_CHAR_LIMIT(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, length) 1167 1168 #endif /* U_HIDE_DEPRECATED_API */ 1169 1170 #endif 1171 1172