utf_old.h (46954B)
1 // © 2016 and later: Unicode, Inc. and others. 2 // License & terms of use: http://www.unicode.org/copyright.html 3 /* 4 ******************************************************************************* 5 * 6 * Copyright (C) 2002-2012, International Business Machines 7 * Corporation and others. All Rights Reserved. 8 * 9 ******************************************************************************* 10 * file name: utf_old.h 11 * encoding: UTF-8 12 * tab size: 8 (not used) 13 * indentation:4 14 * 15 * created on: 2002sep21 16 * created by: Markus W. Scherer 17 */ 18 19 /** 20 * \file 21 * \brief C API: Deprecated macros for Unicode string handling 22 * 23 * The macros in utf_old.h are all deprecated and their use discouraged. 24 * Some of the design principles behind the set of UTF macros 25 * have changed or proved impractical. 26 * Almost all of the old "UTF macros" are at least renamed. 27 * If you are looking for a new equivalent to an old macro, please see the 28 * comment at the old one. 29 * 30 * Brief summary of reasons for deprecation: 31 * - Switch on UTF_SIZE (selection of UTF-8/16/32 default string processing) 32 * was impractical. 33 * - Switch on UTF_SAFE etc. (selection of unsafe/safe/strict default string processing) 34 * was of little use and impractical. 35 * - Whole classes of macros became obsolete outside of the UTF_SIZE/UTF_SAFE 36 * selection framework: UTF32_ macros (all trivial) 37 * and UTF_ default and intermediate macros (all aliases). 38 * - The selection framework also caused many macro aliases. 39 * - Change in Unicode standard: "irregular" sequences (3.0) became illegal (3.2). 40 * - Change of language in Unicode standard: 41 * Growing distinction between internal x-bit Unicode strings and external UTF-x 42 * forms, with the former more lenient. 43 * Suggests renaming of UTF16_ macros to U16_. 44 * - The prefix "UTF_" without a width number confused some users. 45 * - "Safe" append macros needed the addition of an error indicator output. 46 * - "Safe" UTF-8 macros used legitimate (if rarely used) code point values 47 * to indicate error conditions. 48 * - The use of the "_CHAR" infix for code point operations confused some users. 49 * 50 * More details: 51 * 52 * Until ICU 2.2, utf.h theoretically allowed to choose among UTF-8/16/32 53 * for string processing, and among unsafe/safe/strict default macros for that. 54 * 55 * It proved nearly impossible to write non-trivial, high-performance code 56 * that is UTF-generic. 57 * Unsafe default macros would be dangerous for default string processing, 58 * and the main reason for the "strict" versions disappeared: 59 * Between Unicode 3.0 and 3.2 all "irregular" UTF-8 sequences became illegal. 60 * The only other conditions that "strict" checked for were non-characters, 61 * which are valid during processing. Only during text input/output should they 62 * be checked, and at that time other well-formedness checks may be 63 * necessary or useful as well. 64 * This can still be done by using U16_NEXT and U_IS_UNICODE_NONCHAR 65 * or U_IS_UNICODE_CHAR. 66 * 67 * The old UTF8_..._SAFE macros also used some normal Unicode code points 68 * to indicate malformed sequences. 69 * The new UTF8_ macros without suffix use negative values instead. 70 * 71 * The entire contents of utf32.h was moved here without replacement 72 * because all those macros were trivial and 73 * were meaningful only in the framework of choosing the UTF size. 74 * 75 * See Jitterbug 2150 and its discussion on the ICU mailing list 76 * in September 2002. 77 * 78 * <hr> 79 * 80 * <em>Obsolete part</em> of pre-ICU 2.4 utf.h file documentation: 81 * 82 * <p>The original concept for these files was for ICU to allow 83 * in principle to set which UTF (UTF-8/16/32) is used internally 84 * by defining UTF_SIZE to either 8, 16, or 32. utf.h would then define the UChar type 85 * accordingly. UTF-16 was the default.</p> 86 * 87 * <p>This concept has been abandoned. 88 * A lot of the ICU source code assumes UChar strings are in UTF-16. 89 * This is especially true for low-level code like 90 * conversion, normalization, and collation. 91 * The utf.h header enforces the default of UTF-16. 92 * The UTF-8 and UTF-32 macros remain for now for completeness and backward compatibility.</p> 93 * 94 * <p>Accordingly, utf.h defines UChar to be an unsigned 16-bit integer. If this matches wchar_t, then 95 * UChar is defined to be exactly wchar_t, otherwise uint16_t.</p> 96 * 97 * <p>UChar32 is defined to be a signed 32-bit integer (int32_t), large enough for a 21-bit 98 * Unicode code point (Unicode scalar value, 0..0x10ffff). 99 * Before ICU 2.4, the definition of UChar32 was similarly platform-dependent as 100 * the definition of UChar. For details see the documentation for UChar32 itself.</p> 101 * 102 * <p>utf.h also defines a number of C macros for handling single Unicode code points and 103 * for using UTF Unicode strings. It includes utf8.h, utf16.h, and utf32.h for the actual 104 * implementations of those macros and then aliases one set of them (for UTF-16) for general use. 105 * The UTF-specific macros have the UTF size in the macro name prefixes (UTF16_...), while 106 * the general alias macros always begin with UTF_...</p> 107 * 108 * <p>Many string operations can be done with or without error checking. 109 * Where such a distinction is useful, there are two versions of the macros, "unsafe" and "safe" 110 * ones with ..._UNSAFE and ..._SAFE suffixes. The unsafe macros are fast but may cause 111 * program failures if the strings are not well-formed. The safe macros have an additional, boolean 112 * parameter "strict". If strict is false, then only illegal sequences are detected. 113 * Otherwise, irregular sequences and non-characters are detected as well (like single surrogates). 114 * Safe macros return special error code points for illegal/irregular sequences: 115 * Typically, U+ffff, or values that would result in a code unit sequence of the same length 116 * as the erroneous input sequence.<br> 117 * Note that _UNSAFE macros have fewer parameters: They do not have the strictness parameter, and 118 * they do not have start/length parameters for boundary checking.</p> 119 * 120 * <p>Here, the macros are aliased in two steps: 121 * In the first step, the UTF-specific macros with UTF16_ prefix and _UNSAFE and _SAFE suffixes are 122 * aliased according to the UTF_SIZE to macros with UTF_ prefix and the same suffixes and signatures. 123 * Then, in a second step, the default, general alias macros are set to use either the unsafe or 124 * the safe/not strict (default) or the safe/strict macro; 125 * these general macros do not have a strictness parameter.</p> 126 * 127 * <p>It is possible to change the default choice for the general alias macros to be unsafe, safe/not strict or safe/strict. 128 * The default is safe/not strict. It is not recommended to select the unsafe macros as the basis for 129 * Unicode string handling in ICU! To select this, define UTF_SAFE, UTF_STRICT, or UTF_UNSAFE.</p> 130 * 131 * <p>For general use, one should use the default, general macros with UTF_ prefix and no _SAFE/_UNSAFE suffix. 132 * Only in some cases it may be necessary to control the choice of macro directly and use a less generic alias. 133 * For example, if it can be assumed that a string is well-formed and the index will stay within the bounds, 134 * then the _UNSAFE version may be used. 135 * If a UTF-8 string is to be processed, then the macros with UTF8_ prefixes need to be used.</p> 136 * 137 * <hr> 138 * 139 * Deprecated ICU 2.4. Use the macros in utf.h, utf16.h, utf8.h instead. 140 */ 141 142 #ifndef __UTF_OLD_H__ 143 #define __UTF_OLD_H__ 144 145 #include "unicode/utf.h" 146 #include "unicode/utf8.h" 147 #include "unicode/utf16.h" 148 149 /** 150 * \def U_HIDE_OBSOLETE_UTF_OLD_H 151 * 152 * Hides the obsolete definitions in unicode/utf_old.h. 153 * Recommended to be set to 1 at compile time to make sure 154 * the long-deprecated macros are no longer used. 155 * 156 * For reasons for the deprecation see the utf_old.h file comments. 157 * 158 * @internal 159 */ 160 #ifndef U_HIDE_OBSOLETE_UTF_OLD_H 161 # define U_HIDE_OBSOLETE_UTF_OLD_H 0 162 #endif 163 164 #if !defined(U_HIDE_DEPRECATED_API) && !U_HIDE_OBSOLETE_UTF_OLD_H 165 166 /* Formerly utf.h, part 1 --------------------------------------------------- */ 167 168 #ifdef U_USE_UTF_DEPRECATES 169 /** 170 * Unicode string and array offset and index type. 171 * ICU always counts Unicode code units (UChars) for 172 * string offsets, indexes, and lengths, not Unicode code points. 173 * 174 * @obsolete ICU 2.6. Use int32_t directly instead since this API will be removed in that release. 175 */ 176 typedef int32_t UTextOffset; 177 #endif 178 179 /** Number of bits in a Unicode string code unit - ICU uses 16-bit Unicode. @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 180 #define UTF_SIZE 16 181 182 /** 183 * The default choice for general Unicode string macros is to use the ..._SAFE macro implementations 184 * with strict=false. 185 * 186 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 187 */ 188 #define UTF_SAFE 189 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 190 #undef UTF_UNSAFE 191 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 192 #undef UTF_STRICT 193 194 /** 195 * UTF8_ERROR_VALUE_1 and UTF8_ERROR_VALUE_2 are special error values for UTF-8, 196 * which need 1 or 2 bytes in UTF-8: 197 * \code 198 * U+0015 = NAK = Negative Acknowledge, C0 control character 199 * U+009f = highest C1 control character 200 * \endcode 201 * 202 * These are used by UTF8_..._SAFE macros so that they can return an error value 203 * that needs the same number of code units (bytes) as were seen by 204 * a macro. They should be tested with UTF_IS_ERROR() or UTF_IS_VALID(). 205 * 206 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 207 */ 208 #define UTF8_ERROR_VALUE_1 0x15 209 210 /** 211 * See documentation on UTF8_ERROR_VALUE_1 for details. 212 * 213 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 214 */ 215 #define UTF8_ERROR_VALUE_2 0x9f 216 217 /** 218 * Error value for all UTFs. This code point value will be set by macros with error 219 * checking if an error is detected. 220 * 221 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 222 */ 223 #define UTF_ERROR_VALUE 0xffff 224 225 /** 226 * Is a given 32-bit code an error value 227 * as returned by one of the macros for any UTF? 228 * 229 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 230 */ 231 #define UTF_IS_ERROR(c) \ 232 (((c)&0xfffe)==0xfffe || (c)==UTF8_ERROR_VALUE_1 || (c)==UTF8_ERROR_VALUE_2) 233 234 /** 235 * This is a combined macro: Is c a valid Unicode value _and_ not an error code? 236 * 237 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 238 */ 239 #define UTF_IS_VALID(c) \ 240 (UTF_IS_UNICODE_CHAR(c) && \ 241 (c)!=UTF8_ERROR_VALUE_1 && (c)!=UTF8_ERROR_VALUE_2) 242 243 /** 244 * Is this code unit or code point a surrogate (U+d800..U+dfff)? 245 * @deprecated ICU 2.4. Renamed to U_IS_SURROGATE and U16_IS_SURROGATE, see utf_old.h. 246 */ 247 #define UTF_IS_SURROGATE(uchar) (((uchar)&0xfffff800)==0xd800) 248 249 /** 250 * Is a given 32-bit code point a Unicode noncharacter? 251 * 252 * @deprecated ICU 2.4. Renamed to U_IS_UNICODE_NONCHAR, see utf_old.h. 253 */ 254 #define UTF_IS_UNICODE_NONCHAR(c) \ 255 ((c)>=0xfdd0 && \ 256 ((uint32_t)(c)<=0xfdef || ((c)&0xfffe)==0xfffe) && \ 257 (uint32_t)(c)<=0x10ffff) 258 259 /** 260 * Is a given 32-bit value a Unicode code point value (0..U+10ffff) 261 * that can be assigned a character? 262 * 263 * Code points that are not characters include: 264 * - single surrogate code points (U+d800..U+dfff, 2048 code points) 265 * - the last two code points on each plane (U+__fffe and U+__ffff, 34 code points) 266 * - U+fdd0..U+fdef (new with Unicode 3.1, 32 code points) 267 * - the highest Unicode code point value is U+10ffff 268 * 269 * This means that all code points below U+d800 are character code points, 270 * and that boundary is tested first for performance. 271 * 272 * @deprecated ICU 2.4. Renamed to U_IS_UNICODE_CHAR, see utf_old.h. 273 */ 274 #define UTF_IS_UNICODE_CHAR(c) \ 275 ((uint32_t)(c)<0xd800 || \ 276 ((uint32_t)(c)>0xdfff && \ 277 (uint32_t)(c)<=0x10ffff && \ 278 !UTF_IS_UNICODE_NONCHAR(c))) 279 280 /* Formerly utf8.h ---------------------------------------------------------- */ 281 282 /** 283 * \var utf8_countTrailBytes 284 * Internal array with numbers of trail bytes for any given byte used in 285 * lead byte position. 286 * 287 * This is internal since it is not meant to be called directly by external clients; 288 * however it is called by public macros in this file and thus must remain stable, 289 * and should not be hidden when other internal functions are hidden (otherwise 290 * public macros would fail to compile). 291 * @internal 292 */ 293 #ifdef U_UTF8_IMPL 294 // No forward declaration if compiling utf_impl.cpp, which defines utf8_countTrailBytes. 295 #elif defined(U_STATIC_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION) 296 U_CAPI const uint8_t utf8_countTrailBytes[]; 297 #else 298 U_CFUNC U_IMPORT const uint8_t utf8_countTrailBytes[]; 299 #endif 300 301 /** 302 * Count the trail bytes for a UTF-8 lead byte. 303 * @deprecated ICU 2.4. Renamed to U8_COUNT_TRAIL_BYTES, see utf_old.h. 304 */ 305 #define UTF8_COUNT_TRAIL_BYTES(leadByte) (utf8_countTrailBytes[(uint8_t)leadByte]) 306 307 /** 308 * Mask a UTF-8 lead byte, leave only the lower bits that form part of the code point value. 309 * @deprecated ICU 2.4. Renamed to U8_MASK_LEAD_BYTE, see utf_old.h. 310 */ 311 #define UTF8_MASK_LEAD_BYTE(leadByte, countTrailBytes) ((leadByte)&=(1<<(6-(countTrailBytes)))-1) 312 313 /** Is this this code point a single code unit (byte)? @deprecated ICU 2.4. Renamed to U8_IS_SINGLE, see utf_old.h. */ 314 #define UTF8_IS_SINGLE(uchar) (((uchar)&0x80)==0) 315 /** 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. */ 316 #define UTF8_IS_LEAD(uchar) ((uint8_t)((uchar)-0xc0)<0x3e) 317 /** 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. */ 318 #define UTF8_IS_TRAIL(uchar) (((uchar)&0xc0)==0x80) 319 320 /** 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. */ 321 #define UTF8_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0x7f) 322 323 /** 324 * Given the lead character, how many bytes are taken by this code point. 325 * ICU does not deal with code points >0x10ffff 326 * unless necessary for advancing in the byte stream. 327 * 328 * These length macros take into account that for values >0x10ffff 329 * the UTF8_APPEND_CHAR_SAFE macros would write the error code point 0xffff 330 * with 3 bytes. 331 * Code point comparisons need to be in uint32_t because UChar32 332 * may be a signed type, and negative values must be recognized. 333 * 334 * @deprecated ICU 2.4. Use U8_LENGTH instead, see utf.h. 335 */ 336 #if 1 337 # define UTF8_CHAR_LENGTH(c) \ 338 ((uint32_t)(c)<=0x7f ? 1 : \ 339 ((uint32_t)(c)<=0x7ff ? 2 : \ 340 ((uint32_t)((c)-0x10000)>0xfffff ? 3 : 4) \ 341 ) \ 342 ) 343 #else 344 # define UTF8_CHAR_LENGTH(c) \ 345 ((uint32_t)(c)<=0x7f ? 1 : \ 346 ((uint32_t)(c)<=0x7ff ? 2 : \ 347 ((uint32_t)(c)<=0xffff ? 3 : \ 348 ((uint32_t)(c)<=0x10ffff ? 4 : \ 349 ((uint32_t)(c)<=0x3ffffff ? 5 : \ 350 ((uint32_t)(c)<=0x7fffffff ? 6 : 3) \ 351 ) \ 352 ) \ 353 ) \ 354 ) \ 355 ) 356 #endif 357 358 /** The maximum number of bytes per code point. @deprecated ICU 2.4. Renamed to U8_MAX_LENGTH, see utf_old.h. */ 359 #define UTF8_MAX_CHAR_LENGTH 4 360 361 /** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 362 #define UTF8_ARRAY_SIZE(size) ((5*(size))/2) 363 364 /** @deprecated ICU 2.4. Renamed to U8_GET_UNSAFE, see utf_old.h. */ 365 #define UTF8_GET_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ 366 int32_t _utf8_get_char_unsafe_index=(int32_t)(i); \ 367 UTF8_SET_CHAR_START_UNSAFE(s, _utf8_get_char_unsafe_index); \ 368 UTF8_NEXT_CHAR_UNSAFE(s, _utf8_get_char_unsafe_index, c); \ 369 } UPRV_BLOCK_MACRO_END 370 371 /** @deprecated ICU 2.4. Use U8_GET instead, see utf_old.h. */ 372 #define UTF8_GET_CHAR_SAFE(s, start, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \ 373 int32_t _utf8_get_char_safe_index=(int32_t)(i); \ 374 UTF8_SET_CHAR_START_SAFE(s, start, _utf8_get_char_safe_index); \ 375 UTF8_NEXT_CHAR_SAFE(s, _utf8_get_char_safe_index, length, c, strict); \ 376 } UPRV_BLOCK_MACRO_END 377 378 /** @deprecated ICU 2.4. Renamed to U8_NEXT_UNSAFE, see utf_old.h. */ 379 #define UTF8_NEXT_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ 380 (c)=(s)[(i)++]; \ 381 if((uint8_t)((c)-0xc0)<0x35) { \ 382 uint8_t __count=UTF8_COUNT_TRAIL_BYTES(c); \ 383 UTF8_MASK_LEAD_BYTE(c, __count); \ 384 switch(__count) { \ 385 /* each following branch falls through to the next one */ \ 386 case 3: \ 387 (c)=((c)<<6)|((s)[(i)++]&0x3f); \ 388 U_FALLTHROUGH; \ 389 case 2: \ 390 (c)=((c)<<6)|((s)[(i)++]&0x3f); \ 391 U_FALLTHROUGH; \ 392 case 1: \ 393 (c)=((c)<<6)|((s)[(i)++]&0x3f); \ 394 /* no other branches to optimize switch() */ \ 395 break; \ 396 } \ 397 } \ 398 } UPRV_BLOCK_MACRO_END 399 400 /** @deprecated ICU 2.4. Renamed to U8_APPEND_UNSAFE, see utf_old.h. */ 401 #define UTF8_APPEND_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ 402 if((uint32_t)(c)<=0x7f) { \ 403 (s)[(i)++]=(uint8_t)(c); \ 404 } else { \ 405 if((uint32_t)(c)<=0x7ff) { \ 406 (s)[(i)++]=(uint8_t)(((c)>>6)|0xc0); \ 407 } else { \ 408 if((uint32_t)(c)<=0xffff) { \ 409 (s)[(i)++]=(uint8_t)(((c)>>12)|0xe0); \ 410 } else { \ 411 (s)[(i)++]=(uint8_t)(((c)>>18)|0xf0); \ 412 (s)[(i)++]=(uint8_t)((((c)>>12)&0x3f)|0x80); \ 413 } \ 414 (s)[(i)++]=(uint8_t)((((c)>>6)&0x3f)|0x80); \ 415 } \ 416 (s)[(i)++]=(uint8_t)(((c)&0x3f)|0x80); \ 417 } \ 418 } UPRV_BLOCK_MACRO_END 419 420 /** @deprecated ICU 2.4. Renamed to U8_FWD_1_UNSAFE, see utf_old.h. */ 421 #define UTF8_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ 422 (i)+=1+UTF8_COUNT_TRAIL_BYTES((s)[i]); \ 423 } UPRV_BLOCK_MACRO_END 424 425 /** @deprecated ICU 2.4. Renamed to U8_FWD_N_UNSAFE, see utf_old.h. */ 426 #define UTF8_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \ 427 int32_t __N=(n); \ 428 while(__N>0) { \ 429 UTF8_FWD_1_UNSAFE(s, i); \ 430 --__N; \ 431 } \ 432 } UPRV_BLOCK_MACRO_END 433 434 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_START_UNSAFE, see utf_old.h. */ 435 #define UTF8_SET_CHAR_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ 436 while(UTF8_IS_TRAIL((s)[i])) { --(i); } \ 437 } UPRV_BLOCK_MACRO_END 438 439 /** @deprecated ICU 2.4. Use U8_NEXT instead, see utf_old.h. */ 440 #define UTF8_NEXT_CHAR_SAFE(s, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \ 441 (c)=(s)[(i)++]; \ 442 if((c)>=0x80) { \ 443 if(UTF8_IS_LEAD(c)) { \ 444 (c)=utf8_nextCharSafeBody(s, &(i), (int32_t)(length), c, strict); \ 445 } else { \ 446 (c)=UTF8_ERROR_VALUE_1; \ 447 } \ 448 } \ 449 } UPRV_BLOCK_MACRO_END 450 451 /** @deprecated ICU 2.4. Use U8_APPEND instead, see utf_old.h. */ 452 #define UTF8_APPEND_CHAR_SAFE(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \ 453 if((uint32_t)(c)<=0x7f) { \ 454 (s)[(i)++]=(uint8_t)(c); \ 455 } else { \ 456 (i)=utf8_appendCharSafeBody(s, (int32_t)(i), (int32_t)(length), c, NULL); \ 457 } \ 458 } UPRV_BLOCK_MACRO_END 459 460 /** @deprecated ICU 2.4. Renamed to U8_FWD_1, see utf_old.h. */ 461 #define UTF8_FWD_1_SAFE(s, i, length) U8_FWD_1(s, i, length) 462 463 /** @deprecated ICU 2.4. Renamed to U8_FWD_N, see utf_old.h. */ 464 #define UTF8_FWD_N_SAFE(s, i, length, n) U8_FWD_N(s, i, length, n) 465 466 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_START, see utf_old.h. */ 467 #define UTF8_SET_CHAR_START_SAFE(s, start, i) U8_SET_CP_START(s, start, i) 468 469 /** @deprecated ICU 2.4. Renamed to U8_PREV_UNSAFE, see utf_old.h. */ 470 #define UTF8_PREV_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ 471 (c)=(s)[--(i)]; \ 472 if(UTF8_IS_TRAIL(c)) { \ 473 uint8_t __b, __count=1, __shift=6; \ 474 \ 475 /* c is a trail byte */ \ 476 (c)&=0x3f; \ 477 for(;;) { \ 478 __b=(s)[--(i)]; \ 479 if(__b>=0xc0) { \ 480 UTF8_MASK_LEAD_BYTE(__b, __count); \ 481 (c)|=(UChar32)__b<<__shift; \ 482 break; \ 483 } else { \ 484 (c)|=(UChar32)(__b&0x3f)<<__shift; \ 485 ++__count; \ 486 __shift+=6; \ 487 } \ 488 } \ 489 } \ 490 } UPRV_BLOCK_MACRO_END 491 492 /** @deprecated ICU 2.4. Renamed to U8_BACK_1_UNSAFE, see utf_old.h. */ 493 #define UTF8_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ 494 while(UTF8_IS_TRAIL((s)[--(i)])) {} \ 495 } UPRV_BLOCK_MACRO_END 496 497 /** @deprecated ICU 2.4. Renamed to U8_BACK_N_UNSAFE, see utf_old.h. */ 498 #define UTF8_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \ 499 int32_t __N=(n); \ 500 while(__N>0) { \ 501 UTF8_BACK_1_UNSAFE(s, i); \ 502 --__N; \ 503 } \ 504 } UPRV_BLOCK_MACRO_END 505 506 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT_UNSAFE, see utf_old.h. */ 507 #define UTF8_SET_CHAR_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ 508 UTF8_BACK_1_UNSAFE(s, i); \ 509 UTF8_FWD_1_UNSAFE(s, i); \ 510 } UPRV_BLOCK_MACRO_END 511 512 /** @deprecated ICU 2.4. Use U8_PREV instead, see utf_old.h. */ 513 #define UTF8_PREV_CHAR_SAFE(s, start, i, c, strict) UPRV_BLOCK_MACRO_BEGIN { \ 514 (c)=(s)[--(i)]; \ 515 if((c)>=0x80) { \ 516 if((c)<=0xbf) { \ 517 (c)=utf8_prevCharSafeBody(s, start, &(i), c, strict); \ 518 } else { \ 519 (c)=UTF8_ERROR_VALUE_1; \ 520 } \ 521 } \ 522 } UPRV_BLOCK_MACRO_END 523 524 /** @deprecated ICU 2.4. Renamed to U8_BACK_1, see utf_old.h. */ 525 #define UTF8_BACK_1_SAFE(s, start, i) U8_BACK_1(s, start, i) 526 527 /** @deprecated ICU 2.4. Renamed to U8_BACK_N, see utf_old.h. */ 528 #define UTF8_BACK_N_SAFE(s, start, i, n) U8_BACK_N(s, start, i, n) 529 530 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT, see utf_old.h. */ 531 #define UTF8_SET_CHAR_LIMIT_SAFE(s, start, i, length) U8_SET_CP_LIMIT(s, start, i, length) 532 533 /* Formerly utf16.h --------------------------------------------------------- */ 534 535 /** Is uchar a first/lead surrogate? @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h. */ 536 #define UTF_IS_FIRST_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xd800) 537 538 /** Is uchar a second/trail surrogate? @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h. */ 539 #define UTF_IS_SECOND_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xdc00) 540 541 /** 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. */ 542 #define UTF_IS_SURROGATE_FIRST(c) (((c)&0x400)==0) 543 544 /** Helper constant for UTF16_GET_PAIR_VALUE. @deprecated ICU 2.4. Renamed to U16_SURROGATE_OFFSET, see utf_old.h. */ 545 #define UTF_SURROGATE_OFFSET ((0xd800<<10UL)+0xdc00-0x10000) 546 547 /** Get the UTF-32 value from the surrogate code units. @deprecated ICU 2.4. Renamed to U16_GET_SUPPLEMENTARY, see utf_old.h. */ 548 #define UTF16_GET_PAIR_VALUE(first, second) \ 549 (((first)<<10UL)+(second)-UTF_SURROGATE_OFFSET) 550 551 /** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */ 552 #define UTF_FIRST_SURROGATE(supplementary) (UChar)(((supplementary)>>10)+0xd7c0) 553 554 /** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */ 555 #define UTF_SECOND_SURROGATE(supplementary) (UChar)(((supplementary)&0x3ff)|0xdc00) 556 557 /** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */ 558 #define UTF16_LEAD(supplementary) UTF_FIRST_SURROGATE(supplementary) 559 560 /** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */ 561 #define UTF16_TRAIL(supplementary) UTF_SECOND_SURROGATE(supplementary) 562 563 /** @deprecated ICU 2.4. Renamed to U16_IS_SINGLE, see utf_old.h. */ 564 #define UTF16_IS_SINGLE(uchar) !UTF_IS_SURROGATE(uchar) 565 566 /** @deprecated ICU 2.4. Renamed to U16_IS_LEAD, see utf_old.h. */ 567 #define UTF16_IS_LEAD(uchar) UTF_IS_FIRST_SURROGATE(uchar) 568 569 /** @deprecated ICU 2.4. Renamed to U16_IS_TRAIL, see utf_old.h. */ 570 #define UTF16_IS_TRAIL(uchar) UTF_IS_SECOND_SURROGATE(uchar) 571 572 /** 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. */ 573 #define UTF16_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0xffff) 574 575 /** @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h. */ 576 #define UTF16_CHAR_LENGTH(c) ((uint32_t)(c)<=0xffff ? 1 : 2) 577 578 /** @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h. */ 579 #define UTF16_MAX_CHAR_LENGTH 2 580 581 /** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 582 #define UTF16_ARRAY_SIZE(size) (size) 583 584 /** 585 * Get a single code point from an offset that points to any 586 * of the code units that belong to that code point. 587 * Assume 0<=i<length. 588 * 589 * This could be used for iteration together with 590 * UTF16_CHAR_LENGTH() and UTF_IS_ERROR(), 591 * but the use of UTF16_NEXT_CHAR[_UNSAFE]() and 592 * UTF16_PREV_CHAR[_UNSAFE]() is more efficient for that. 593 * @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h. 594 */ 595 #define UTF16_GET_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ 596 (c)=(s)[i]; \ 597 if(UTF_IS_SURROGATE(c)) { \ 598 if(UTF_IS_SURROGATE_FIRST(c)) { \ 599 (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)+1]); \ 600 } else { \ 601 (c)=UTF16_GET_PAIR_VALUE((s)[(i)-1], (c)); \ 602 } \ 603 } \ 604 } UPRV_BLOCK_MACRO_END 605 606 /** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */ 607 #define UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \ 608 (c)=(s)[i]; \ 609 if(UTF_IS_SURROGATE(c)) { \ 610 uint16_t __c2; \ 611 if(UTF_IS_SURROGATE_FIRST(c)) { \ 612 if((i)+1<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)+1])) { \ 613 (c)=UTF16_GET_PAIR_VALUE((c), __c2); \ 614 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \ 615 } else if(strict) {\ 616 /* unmatched first surrogate */ \ 617 (c)=UTF_ERROR_VALUE; \ 618 } \ 619 } else { \ 620 if((i)-1>=(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \ 621 (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \ 622 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \ 623 } else if(strict) {\ 624 /* unmatched second surrogate */ \ 625 (c)=UTF_ERROR_VALUE; \ 626 } \ 627 } \ 628 } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \ 629 (c)=UTF_ERROR_VALUE; \ 630 } \ 631 } UPRV_BLOCK_MACRO_END 632 633 /** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */ 634 #define UTF16_NEXT_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ 635 (c)=(s)[(i)++]; \ 636 if(UTF_IS_FIRST_SURROGATE(c)) { \ 637 (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)++]); \ 638 } \ 639 } UPRV_BLOCK_MACRO_END 640 641 /** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */ 642 #define UTF16_APPEND_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ 643 if((uint32_t)(c)<=0xffff) { \ 644 (s)[(i)++]=(uint16_t)(c); \ 645 } else { \ 646 (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ 647 (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ 648 } \ 649 } UPRV_BLOCK_MACRO_END 650 651 /** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */ 652 #define UTF16_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ 653 if(UTF_IS_FIRST_SURROGATE((s)[(i)++])) { \ 654 ++(i); \ 655 } \ 656 } UPRV_BLOCK_MACRO_END 657 658 /** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */ 659 #define UTF16_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \ 660 int32_t __N=(n); \ 661 while(__N>0) { \ 662 UTF16_FWD_1_UNSAFE(s, i); \ 663 --__N; \ 664 } \ 665 } UPRV_BLOCK_MACRO_END 666 667 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */ 668 #define UTF16_SET_CHAR_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ 669 if(UTF_IS_SECOND_SURROGATE((s)[i])) { \ 670 --(i); \ 671 } \ 672 } UPRV_BLOCK_MACRO_END 673 674 /** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */ 675 #define UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \ 676 (c)=(s)[(i)++]; \ 677 if(UTF_IS_FIRST_SURROGATE(c)) { \ 678 uint16_t __c2; \ 679 if((i)<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)])) { \ 680 ++(i); \ 681 (c)=UTF16_GET_PAIR_VALUE((c), __c2); \ 682 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \ 683 } else if(strict) {\ 684 /* unmatched first surrogate */ \ 685 (c)=UTF_ERROR_VALUE; \ 686 } \ 687 } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \ 688 /* unmatched second surrogate or other non-character */ \ 689 (c)=UTF_ERROR_VALUE; \ 690 } \ 691 } UPRV_BLOCK_MACRO_END 692 693 /** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */ 694 #define UTF16_APPEND_CHAR_SAFE(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \ 695 if((uint32_t)(c)<=0xffff) { \ 696 (s)[(i)++]=(uint16_t)(c); \ 697 } else if((uint32_t)(c)<=0x10ffff) { \ 698 if((i)+1<(length)) { \ 699 (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ 700 (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ 701 } else /* not enough space */ { \ 702 (s)[(i)++]=UTF_ERROR_VALUE; \ 703 } \ 704 } else /* c>0x10ffff, write error value */ { \ 705 (s)[(i)++]=UTF_ERROR_VALUE; \ 706 } \ 707 } UPRV_BLOCK_MACRO_END 708 709 /** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */ 710 #define UTF16_FWD_1_SAFE(s, i, length) U16_FWD_1(s, i, length) 711 712 /** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */ 713 #define UTF16_FWD_N_SAFE(s, i, length, n) U16_FWD_N(s, i, length, n) 714 715 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */ 716 #define UTF16_SET_CHAR_START_SAFE(s, start, i) U16_SET_CP_START(s, start, i) 717 718 /** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */ 719 #define UTF16_PREV_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ 720 (c)=(s)[--(i)]; \ 721 if(UTF_IS_SECOND_SURROGATE(c)) { \ 722 (c)=UTF16_GET_PAIR_VALUE((s)[--(i)], (c)); \ 723 } \ 724 } UPRV_BLOCK_MACRO_END 725 726 /** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */ 727 #define UTF16_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ 728 if(UTF_IS_SECOND_SURROGATE((s)[--(i)])) { \ 729 --(i); \ 730 } \ 731 } UPRV_BLOCK_MACRO_END 732 733 /** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */ 734 #define UTF16_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \ 735 int32_t __N=(n); \ 736 while(__N>0) { \ 737 UTF16_BACK_1_UNSAFE(s, i); \ 738 --__N; \ 739 } \ 740 } UPRV_BLOCK_MACRO_END 741 742 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */ 743 #define UTF16_SET_CHAR_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ 744 if(UTF_IS_FIRST_SURROGATE((s)[(i)-1])) { \ 745 ++(i); \ 746 } \ 747 } UPRV_BLOCK_MACRO_END 748 749 /** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */ 750 #define UTF16_PREV_CHAR_SAFE(s, start, i, c, strict) UPRV_BLOCK_MACRO_BEGIN { \ 751 (c)=(s)[--(i)]; \ 752 if(UTF_IS_SECOND_SURROGATE(c)) { \ 753 uint16_t __c2; \ 754 if((i)>(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \ 755 --(i); \ 756 (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \ 757 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \ 758 } else if(strict) {\ 759 /* unmatched second surrogate */ \ 760 (c)=UTF_ERROR_VALUE; \ 761 } \ 762 } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \ 763 /* unmatched first surrogate or other non-character */ \ 764 (c)=UTF_ERROR_VALUE; \ 765 } \ 766 } UPRV_BLOCK_MACRO_END 767 768 /** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */ 769 #define UTF16_BACK_1_SAFE(s, start, i) U16_BACK_1(s, start, i) 770 771 /** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */ 772 #define UTF16_BACK_N_SAFE(s, start, i, n) U16_BACK_N(s, start, i, n) 773 774 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */ 775 #define UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, length) 776 777 /* Formerly utf32.h --------------------------------------------------------- */ 778 779 /* 780 * Old documentation: 781 * 782 * This file defines macros to deal with UTF-32 code units and code points. 783 * Signatures and semantics are the same as for the similarly named macros 784 * in utf16.h. 785 * utf32.h is included by utf.h after unicode/umachine.h</p> 786 * and some common definitions. 787 * <p><b>Usage:</b> ICU coding guidelines for if() statements should be followed when using these macros. 788 * Compound statements (curly braces {}) must be used for if-else-while... 789 * bodies and all macro statements should be terminated with semicolon.</p> 790 */ 791 792 /* internal definitions ----------------------------------------------------- */ 793 794 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 795 #define UTF32_IS_SAFE(c, strict) \ 796 (!(strict) ? \ 797 (uint32_t)(c)<=0x10ffff : \ 798 UTF_IS_UNICODE_CHAR(c)) 799 800 /* 801 * For the semantics of all of these macros, see utf16.h. 802 * The UTF-32 versions are trivial because any code point is 803 * encoded using exactly one code unit. 804 */ 805 806 /* single-code point definitions -------------------------------------------- */ 807 808 /* classes of code unit values */ 809 810 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 811 #define UTF32_IS_SINGLE(uchar) 1 812 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 813 #define UTF32_IS_LEAD(uchar) 0 814 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 815 #define UTF32_IS_TRAIL(uchar) 0 816 817 /* number of code units per code point */ 818 819 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 820 #define UTF32_NEED_MULTIPLE_UCHAR(c) 0 821 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 822 #define UTF32_CHAR_LENGTH(c) 1 823 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 824 #define UTF32_MAX_CHAR_LENGTH 1 825 826 /* average number of code units compared to UTF-16 */ 827 828 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 829 #define UTF32_ARRAY_SIZE(size) (size) 830 831 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 832 #define UTF32_GET_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ 833 (c)=(s)[i]; \ 834 } UPRV_BLOCK_MACRO_END 835 836 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 837 #define UTF32_GET_CHAR_SAFE(s, start, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \ 838 (c)=(s)[i]; \ 839 if(!UTF32_IS_SAFE(c, strict)) { \ 840 (c)=UTF_ERROR_VALUE; \ 841 } \ 842 } UPRV_BLOCK_MACRO_END 843 844 /* definitions with forward iteration --------------------------------------- */ 845 846 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 847 #define UTF32_NEXT_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ 848 (c)=(s)[(i)++]; \ 849 } UPRV_BLOCK_MACRO_END 850 851 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 852 #define UTF32_APPEND_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ 853 (s)[(i)++]=(c); \ 854 } UPRV_BLOCK_MACRO_END 855 856 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 857 #define UTF32_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ 858 ++(i); \ 859 } UPRV_BLOCK_MACRO_END 860 861 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 862 #define UTF32_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \ 863 (i)+=(n); \ 864 } UPRV_BLOCK_MACRO_END 865 866 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 867 #define UTF32_SET_CHAR_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ 868 } UPRV_BLOCK_MACRO_END 869 870 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 871 #define UTF32_NEXT_CHAR_SAFE(s, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \ 872 (c)=(s)[(i)++]; \ 873 if(!UTF32_IS_SAFE(c, strict)) { \ 874 (c)=UTF_ERROR_VALUE; \ 875 } \ 876 } UPRV_BLOCK_MACRO_END 877 878 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 879 #define UTF32_APPEND_CHAR_SAFE(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \ 880 if((uint32_t)(c)<=0x10ffff) { \ 881 (s)[(i)++]=(c); \ 882 } else /* c>0x10ffff, write 0xfffd */ { \ 883 (s)[(i)++]=0xfffd; \ 884 } \ 885 } UPRV_BLOCK_MACRO_END 886 887 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 888 #define UTF32_FWD_1_SAFE(s, i, length) UPRV_BLOCK_MACRO_BEGIN { \ 889 ++(i); \ 890 } UPRV_BLOCK_MACRO_END 891 892 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 893 #define UTF32_FWD_N_SAFE(s, i, length, n) UPRV_BLOCK_MACRO_BEGIN { \ 894 if(((i)+=(n))>(length)) { \ 895 (i)=(length); \ 896 } \ 897 } UPRV_BLOCK_MACRO_END 898 899 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 900 #define UTF32_SET_CHAR_START_SAFE(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \ 901 } UPRV_BLOCK_MACRO_END 902 903 /* definitions with backward iteration -------------------------------------- */ 904 905 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 906 #define UTF32_PREV_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ 907 (c)=(s)[--(i)]; \ 908 } UPRV_BLOCK_MACRO_END 909 910 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 911 #define UTF32_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ 912 --(i); \ 913 } UPRV_BLOCK_MACRO_END 914 915 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 916 #define UTF32_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \ 917 (i)-=(n); \ 918 } UPRV_BLOCK_MACRO_END 919 920 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 921 #define UTF32_SET_CHAR_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ 922 } UPRV_BLOCK_MACRO_END 923 924 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 925 #define UTF32_PREV_CHAR_SAFE(s, start, i, c, strict) UPRV_BLOCK_MACRO_BEGIN { \ 926 (c)=(s)[--(i)]; \ 927 if(!UTF32_IS_SAFE(c, strict)) { \ 928 (c)=UTF_ERROR_VALUE; \ 929 } \ 930 } UPRV_BLOCK_MACRO_END 931 932 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 933 #define UTF32_BACK_1_SAFE(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \ 934 --(i); \ 935 } UPRV_BLOCK_MACRO_END 936 937 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 938 #define UTF32_BACK_N_SAFE(s, start, i, n) UPRV_BLOCK_MACRO_BEGIN { \ 939 (i)-=(n); \ 940 if((i)<(start)) { \ 941 (i)=(start); \ 942 } \ 943 } UPRV_BLOCK_MACRO_END 944 945 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */ 946 #define UTF32_SET_CHAR_LIMIT_SAFE(s, i, length) UPRV_BLOCK_MACRO_BEGIN { \ 947 } UPRV_BLOCK_MACRO_END 948 949 /* Formerly utf.h, part 2 --------------------------------------------------- */ 950 951 /** 952 * Estimate the number of code units for a string based on the number of UTF-16 code units. 953 * 954 * @deprecated ICU 2.4. Obsolete, see utf_old.h. 955 */ 956 #define UTF_ARRAY_SIZE(size) UTF16_ARRAY_SIZE(size) 957 958 /** @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h. */ 959 #define UTF_GET_CHAR_UNSAFE(s, i, c) UTF16_GET_CHAR_UNSAFE(s, i, c) 960 961 /** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */ 962 #define UTF_GET_CHAR_SAFE(s, start, i, length, c, strict) UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict) 963 964 965 /** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */ 966 #define UTF_NEXT_CHAR_UNSAFE(s, i, c) UTF16_NEXT_CHAR_UNSAFE(s, i, c) 967 968 /** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */ 969 #define UTF_NEXT_CHAR_SAFE(s, i, length, c, strict) UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict) 970 971 972 /** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */ 973 #define UTF_APPEND_CHAR_UNSAFE(s, i, c) UTF16_APPEND_CHAR_UNSAFE(s, i, c) 974 975 /** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */ 976 #define UTF_APPEND_CHAR_SAFE(s, i, length, c) UTF16_APPEND_CHAR_SAFE(s, i, length, c) 977 978 979 /** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */ 980 #define UTF_FWD_1_UNSAFE(s, i) UTF16_FWD_1_UNSAFE(s, i) 981 982 /** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */ 983 #define UTF_FWD_1_SAFE(s, i, length) UTF16_FWD_1_SAFE(s, i, length) 984 985 986 /** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */ 987 #define UTF_FWD_N_UNSAFE(s, i, n) UTF16_FWD_N_UNSAFE(s, i, n) 988 989 /** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */ 990 #define UTF_FWD_N_SAFE(s, i, length, n) UTF16_FWD_N_SAFE(s, i, length, n) 991 992 993 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */ 994 #define UTF_SET_CHAR_START_UNSAFE(s, i) UTF16_SET_CHAR_START_UNSAFE(s, i) 995 996 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */ 997 #define UTF_SET_CHAR_START_SAFE(s, start, i) UTF16_SET_CHAR_START_SAFE(s, start, i) 998 999 1000 /** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */ 1001 #define UTF_PREV_CHAR_UNSAFE(s, i, c) UTF16_PREV_CHAR_UNSAFE(s, i, c) 1002 1003 /** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */ 1004 #define UTF_PREV_CHAR_SAFE(s, start, i, c, strict) UTF16_PREV_CHAR_SAFE(s, start, i, c, strict) 1005 1006 1007 /** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */ 1008 #define UTF_BACK_1_UNSAFE(s, i) UTF16_BACK_1_UNSAFE(s, i) 1009 1010 /** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */ 1011 #define UTF_BACK_1_SAFE(s, start, i) UTF16_BACK_1_SAFE(s, start, i) 1012 1013 1014 /** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */ 1015 #define UTF_BACK_N_UNSAFE(s, i, n) UTF16_BACK_N_UNSAFE(s, i, n) 1016 1017 /** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */ 1018 #define UTF_BACK_N_SAFE(s, start, i, n) UTF16_BACK_N_SAFE(s, start, i, n) 1019 1020 1021 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */ 1022 #define UTF_SET_CHAR_LIMIT_UNSAFE(s, i) UTF16_SET_CHAR_LIMIT_UNSAFE(s, i) 1023 1024 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */ 1025 #define UTF_SET_CHAR_LIMIT_SAFE(s, start, i, length) UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length) 1026 1027 /* Define default macros (UTF-16 "safe") ------------------------------------ */ 1028 1029 /** 1030 * Does this code unit alone encode a code point (BMP, not a surrogate)? 1031 * Same as UTF16_IS_SINGLE. 1032 * @deprecated ICU 2.4. Renamed to U_IS_SINGLE and U16_IS_SINGLE, see utf_old.h. 1033 */ 1034 #define UTF_IS_SINGLE(uchar) U16_IS_SINGLE(uchar) 1035 1036 /** 1037 * Is this code unit the first one of several (a lead surrogate)? 1038 * Same as UTF16_IS_LEAD. 1039 * @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h. 1040 */ 1041 #define UTF_IS_LEAD(uchar) U16_IS_LEAD(uchar) 1042 1043 /** 1044 * Is this code unit one of several but not the first one (a trail surrogate)? 1045 * Same as UTF16_IS_TRAIL. 1046 * @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h. 1047 */ 1048 #define UTF_IS_TRAIL(uchar) U16_IS_TRAIL(uchar) 1049 1050 /** 1051 * Does this code point require multiple code units (is it a supplementary code point)? 1052 * Same as UTF16_NEED_MULTIPLE_UCHAR. 1053 * @deprecated ICU 2.4. Use U16_LENGTH or test ((uint32_t)(c)>0xffff) instead. 1054 */ 1055 #define UTF_NEED_MULTIPLE_UCHAR(c) UTF16_NEED_MULTIPLE_UCHAR(c) 1056 1057 /** 1058 * How many code units are used to encode this code point (1 or 2)? 1059 * Same as UTF16_CHAR_LENGTH. 1060 * @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h. 1061 */ 1062 #define UTF_CHAR_LENGTH(c) U16_LENGTH(c) 1063 1064 /** 1065 * How many code units are used at most for any Unicode code point (2)? 1066 * Same as UTF16_MAX_CHAR_LENGTH. 1067 * @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h. 1068 */ 1069 #define UTF_MAX_CHAR_LENGTH U16_MAX_LENGTH 1070 1071 /** 1072 * Set c to the code point that contains the code unit i. 1073 * i could point to the lead or the trail surrogate for the code point. 1074 * i is not modified. 1075 * Same as UTF16_GET_CHAR. 1076 * \pre 0<=i<length 1077 * 1078 * @deprecated ICU 2.4. Renamed to U16_GET, see utf_old.h. 1079 */ 1080 #define UTF_GET_CHAR(s, start, i, length, c) U16_GET(s, start, i, length, c) 1081 1082 /** 1083 * Set c to the code point that starts at code unit i 1084 * and advance i to beyond the code units of this code point (post-increment). 1085 * i must point to the first code unit of a code point. 1086 * Otherwise c is set to the trail unit (surrogate) itself. 1087 * Same as UTF16_NEXT_CHAR. 1088 * \pre 0<=i<length 1089 * \post 0<i<=length 1090 * 1091 * @deprecated ICU 2.4. Renamed to U16_NEXT, see utf_old.h. 1092 */ 1093 #define UTF_NEXT_CHAR(s, i, length, c) U16_NEXT(s, i, length, c) 1094 1095 /** 1096 * Append the code units of code point c to the string at index i 1097 * and advance i to beyond the new code units (post-increment). 1098 * The code units beginning at index i will be overwritten. 1099 * Same as UTF16_APPEND_CHAR. 1100 * \pre 0<=c<=0x10ffff 1101 * \pre 0<=i<length 1102 * \post 0<i<=length 1103 * 1104 * @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. 1105 */ 1106 #define UTF_APPEND_CHAR(s, i, length, c) UTF16_APPEND_CHAR_SAFE(s, i, length, c) 1107 1108 /** 1109 * Advance i to beyond the code units of the code point that begins at i. 1110 * I.e., advance i by one code point. 1111 * Same as UTF16_FWD_1. 1112 * \pre 0<=i<length 1113 * \post 0<i<=length 1114 * 1115 * @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. 1116 */ 1117 #define UTF_FWD_1(s, i, length) U16_FWD_1(s, i, length) 1118 1119 /** 1120 * Advance i to beyond the code units of the n code points where the first one begins at i. 1121 * I.e., advance i by n code points. 1122 * Same as UT16_FWD_N. 1123 * \pre 0<=i<length 1124 * \post 0<i<=length 1125 * 1126 * @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. 1127 */ 1128 #define UTF_FWD_N(s, i, length, n) U16_FWD_N(s, i, length, n) 1129 1130 /** 1131 * Take the random-access index i and adjust it so that it points to the beginning 1132 * of a code point. 1133 * The input index points to any code unit of a code point and is moved to point to 1134 * the first code unit of the same code point. i is never incremented. 1135 * In other words, if i points to a trail surrogate that is preceded by a matching 1136 * lead surrogate, then i is decremented. Otherwise it is not modified. 1137 * This can be used to start an iteration with UTF_NEXT_CHAR() from a random index. 1138 * Same as UTF16_SET_CHAR_START. 1139 * \pre start<=i<length 1140 * \post start<=i<length 1141 * 1142 * @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. 1143 */ 1144 #define UTF_SET_CHAR_START(s, start, i) U16_SET_CP_START(s, start, i) 1145 1146 /** 1147 * Set c to the code point that has code units before i 1148 * and move i backward (towards the beginning of the string) 1149 * to the first code unit of this code point (pre-increment). 1150 * i must point to the first code unit after the last unit of a code point (i==length is allowed). 1151 * Same as UTF16_PREV_CHAR. 1152 * \pre start<i<=length 1153 * \post start<=i<length 1154 * 1155 * @deprecated ICU 2.4. Renamed to U16_PREV, see utf_old.h. 1156 */ 1157 #define UTF_PREV_CHAR(s, start, i, c) U16_PREV(s, start, i, c) 1158 1159 /** 1160 * Move i backward (towards the beginning of the string) 1161 * to the first code unit of the code point that has code units before i. 1162 * I.e., move i backward by one code point. 1163 * i must point to the first code unit after the last unit of a code point (i==length is allowed). 1164 * Same as UTF16_BACK_1. 1165 * \pre start<i<=length 1166 * \post start<=i<length 1167 * 1168 * @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. 1169 */ 1170 #define UTF_BACK_1(s, start, i) U16_BACK_1(s, start, i) 1171 1172 /** 1173 * Move i backward (towards the beginning of the string) 1174 * to the first code unit of the n code points that have code units before i. 1175 * I.e., move i backward by n code points. 1176 * i must point to the first code unit after the last unit of a code point (i==length is allowed). 1177 * Same as UTF16_BACK_N. 1178 * \pre start<i<=length 1179 * \post start<=i<length 1180 * 1181 * @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. 1182 */ 1183 #define UTF_BACK_N(s, start, i, n) U16_BACK_N(s, start, i, n) 1184 1185 /** 1186 * Take the random-access index i and adjust it so that it points beyond 1187 * a code point. The input index points beyond any code unit 1188 * of a code point and is moved to point beyond the last code unit of the same 1189 * code point. i is never decremented. 1190 * In other words, if i points to a trail surrogate that is preceded by a matching 1191 * lead surrogate, then i is incremented. Otherwise it is not modified. 1192 * This can be used to start an iteration with UTF_PREV_CHAR() from a random index. 1193 * Same as UTF16_SET_CHAR_LIMIT. 1194 * \pre start<i<=length 1195 * \post start<i<=length 1196 * 1197 * @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. 1198 */ 1199 #define UTF_SET_CHAR_LIMIT(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, length) 1200 1201 #endif // !U_HIDE_DEPRECATED_API && !U_HIDE_OBSOLETE_UTF_OLD_H 1202 1203 #endif