icuSources/common/unicode/utf_old.h

   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
  24 /**
  25  *
  26  * The macros in utf_old.h are all deprecated and their use discouraged.
  27  * Some of the design principles behind the set of UTF macros
  28  * have changed or proved impractical.
  29  * Almost all of the old "UTF macros" are at least renamed.
  30  * If you are looking for a new equivalent to an old macro, please see the
  31  * comment at the old one.
  32  *
  33  * Brief summary of reasons for deprecation:
  34  * - Switch on UTF_SIZE (selection of UTF-8/16/32 default string processing)
  35  *   was impractical.
  36  * - Switch on UTF_SAFE etc. (selection of unsafe/safe/strict default string processing)
  37  *   was of little use and impractical.
  38  * - Whole classes of macros became obsolete outside of the UTF_SIZE/UTF_SAFE
  39  *   selection framework: UTF32_ macros (all trivial)
  40  *   and UTF_ default and intermediate macros (all aliases).
  41  * - The selection framework also caused many macro aliases.
  42  * - Change in Unicode standard: "irregular" sequences (3.0) became illegal (3.2).
  43  * - Change of language in Unicode standard:
  44  *   Growing distinction between internal x-bit Unicode strings and external UTF-x
  45  *   forms, with the former more lenient.
  46  *   Suggests renaming of UTF16_ macros to U16_.
  47  * - The prefix "UTF_" without a width number confused some users.
  48  * - "Safe" append macros needed the addition of an error indicator output.
  49  * - "Safe" UTF-8 macros used legitimate (if rarely used) code point values
  50  *   to indicate error conditions.
  51  * - The use of the "_CHAR" infix for code point operations confused some users.
  52  *
  53  * More details:
  54  *
  55  * Until ICU 2.2, utf.h theoretically allowed to choose among UTF-8/16/32
  56  * for string processing, and among unsafe/safe/strict default macros for that.
  57  *
  58  * It proved nearly impossible to write non-trivial, high-performance code
  59  * that is UTF-generic.
  60  * Unsafe default macros would be dangerous for default string processing,
  61  * and the main reason for the "strict" versions disappeared:
  62  * Between Unicode 3.0 and 3.2 all "irregular" UTF-8 sequences became illegal.
  63  * The only other conditions that "strict" checked for were non-characters,
  64  * which are valid during processing. Only during text input/output should they
  65  * be checked, and at that time other well-formedness checks may be
  66  * necessary or useful as well.
  67  * This can still be done by using U16_NEXT and U_IS_UNICODE_NONCHAR
  68  * or U_IS_UNICODE_CHAR.
  69  *
  70  * The old UTF8_..._SAFE macros also used some normal Unicode code points
  71  * to indicate malformed sequences.
  72  * The new UTF8_ macros without suffix use negative values instead.
  73  *
  74  * The entire contents of utf32.h was moved here without replacement
  75  * because all those macros were trivial and
  76  * were meaningful only in the framework of choosing the UTF size.
  77  *
  78  * See Jitterbug 2150 and its discussion on the ICU mailing list
  79  * in September 2002.
  80  *
  81  * <hr>
  82  *
  83  * <em>Obsolete part</em> of pre-ICU 2.4 utf.h file documentation:
  84  *
  85  * <p>The original concept for these files was for ICU to allow
  86  * in principle to set which UTF (UTF-8/16/32) is used internally
  87  * by defining UTF_SIZE to either 8, 16, or 32. utf.h would then define the UChar type
  88  * accordingly. UTF-16 was the default.</p>
  89  *
  90  * <p>This concept has been abandoned.
  91  * A lot of the ICU source code assumes UChar strings are in UTF-16.
  92  * This is especially true for low-level code like
  93  * conversion, normalization, and collation.
  94  * The utf.h header enforces the default of UTF-16.
  95  * The UTF-8 and UTF-32 macros remain for now for completeness and backward compatibility.</p>
  96  *
  97  * <p>Accordingly, utf.h defines UChar to be an unsigned 16-bit integer. If this matches wchar_t, then
  98  * UChar is defined to be exactly wchar_t, otherwise uint16_t.</p>
  99  *
 100  * <p>UChar32 is defined to be a signed 32-bit integer (int32_t), large enough for a 21-bit
 101  * Unicode code point (Unicode scalar value, 0..0x10ffff).
 102  * Before ICU 2.4, the definition of UChar32 was similarly platform-dependent as
 103  * the definition of UChar. For details see the documentation for UChar32 itself.</p>
 104  *
 105  * <p>utf.h also defines a number of C macros for handling single Unicode code points and
 106  * for using UTF Unicode strings. It includes utf8.h, utf16.h, and utf32.h for the actual
 107  * implementations of those macros and then aliases one set of them (for UTF-16) for general use.
 108  * The UTF-specific macros have the UTF size in the macro name prefixes (UTF16_...), while
 109  * the general alias macros always begin with UTF_...</p>
 110  *
 111  * <p>Many string operations can be done with or without error checking.
 112  * Where such a distinction is useful, there are two versions of the macros, "unsafe" and "safe"
 113  * ones with ..._UNSAFE and ..._SAFE suffixes. The unsafe macros are fast but may cause
 114  * program failures if the strings are not well-formed. The safe macros have an additional, boolean
 115  * parameter "strict". If strict is FALSE, then only illegal sequences are detected.
 116  * Otherwise, irregular sequences and non-characters are detected as well (like single surrogates).
 117  * Safe macros return special error code points for illegal/irregular sequences:
 118  * Typically, U+ffff, or values that would result in a code unit sequence of the same length
 119  * as the erroneous input sequence.<br>
 120  * Note that _UNSAFE macros have fewer parameters: They do not have the strictness parameter, and
 121  * they do not have start/length parameters for boundary checking.</p>
 122  *
 123  * <p>Here, the macros are aliased in two steps:
 124  * In the first step, the UTF-specific macros with UTF16_ prefix and _UNSAFE and _SAFE suffixes are
 125  * aliased according to the UTF_SIZE to macros with UTF_ prefix and the same suffixes and signatures.
 126  * Then, in a second step, the default, general alias macros are set to use either the unsafe or
 127  * the safe/not strict (default) or the safe/strict macro;
 128  * these general macros do not have a strictness parameter.</p>
 129  *
 130  * <p>It is possible to change the default choice for the general alias macros to be unsafe, safe/not strict or safe/strict.
 131  * The default is safe/not strict. It is not recommended to select the unsafe macros as the basis for
 132  * Unicode string handling in ICU! To select this, define UTF_SAFE, UTF_STRICT, or UTF_UNSAFE.</p>
 133  *
 134  * <p>For general use, one should use the default, general macros with UTF_ prefix and no _SAFE/_UNSAFE suffix.
 135  * Only in some cases it may be necessary to control the choice of macro directly and use a less generic alias.
 136  * For example, if it can be assumed that a string is well-formed and the index will stay within the bounds,
 137  * then the _UNSAFE version may be used.
 138  * If a UTF-8 string is to be processed, then the macros with UTF8_ prefixes need to be used.</p>
 139  *
 140  * <hr>
 141  *
 142  * @deprecated ICU 2.4. Use the macros in utf.h, utf16.h, utf8.h instead.
 143  */
 144
 145 #ifndef __UTF_OLD_H__
 146 #define __UTF_OLD_H__
 147
 148 #ifndef U_HIDE_DEPRECATED_API
 149
 150 #include "unicode/utf.h"
 151 #include "unicode/utf8.h"
 152 #include "unicode/utf16.h"
 153
 154 /* Formerly utf.h, part 1 --------------------------------------------------- */
 155
 156 #ifdef U_USE_UTF_DEPRECATES
 157 /**
 158  * Unicode string and array offset and index type.
 159  * ICU always counts Unicode code units (UChars) for
 160  * string offsets, indexes, and lengths, not Unicode code points.
 161  *
 162  * @obsolete ICU 2.6. Use int32_t directly instead since this API will be removed in that release.
 163  */
 164 typedef int32_t UTextOffset;
 165 #endif
 166
 167 /** Number of bits in a Unicode string code unit - ICU uses 16-bit Unicode. @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 168 #define UTF_SIZE 16
 169
 170 /**
 171  * The default choice for general Unicode string macros is to use the ..._SAFE macro implementations
 172  * with strict=FALSE.
 173  *
 174  * @deprecated ICU 2.4. Obsolete, see utf_old.h.
 175  */
 176 #define UTF_SAFE
 177 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 178 #undef UTF_UNSAFE
 179 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 180 #undef UTF_STRICT
 181
 182 /**
 183  * UTF8_ERROR_VALUE_1 and UTF8_ERROR_VALUE_2 are special error values for UTF-8,
 184  * which need 1 or 2 bytes in UTF-8:
 185  * \code
 186  * U+0015 = NAK = Negative Acknowledge, C0 control character
 187  * U+009f = highest C1 control character
 188  * \endcode
 189  *
 190  * 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().
 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 * \var utf8_countTrailBytes
 272 * Internal array with numbers of trail bytes for any given byte used in
 273 * lead byte position.
 274 *
 275 * This is internal since it is not meant to be called directly by external clients;
 276 * however it is called by public macros in this file and thus must remain stable,
 277 * and should not be hidden when other internal functions are hidden (otherwise
 278 * public macros would fail to compile).
 279 * @internal
 280 */
 281 #ifdef U_UTF8_IMPL
 282 // No forward declaration if compiling utf_impl.cpp, which defines utf8_countTrailBytes.
 283 #elif defined(U_STATIC_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION)
 284 U_CFUNC const uint8_t utf8_countTrailBytes[];
 285 #else
 286 U_CFUNC U_IMPORT const uint8_t utf8_countTrailBytes[];    /* U_IMPORT2? */ /*U_IMPORT*/
 287 #endif
 288
 289 /**
 290  * Count the trail bytes for a UTF-8 lead byte.
 291  * @deprecated ICU 2.4. Renamed to U8_COUNT_TRAIL_BYTES, see utf_old.h.
 292  */
 293 #define UTF8_COUNT_TRAIL_BYTES(leadByte) (utf8_countTrailBytes[(uint8_t)leadByte])
 294
 295 /**
 296  * Mask a UTF-8 lead byte, leave only the lower bits that form part of the code point value.
 297  * @deprecated ICU 2.4. Renamed to U8_MASK_LEAD_BYTE, see utf_old.h.
 298  */
 299 #define UTF8_MASK_LEAD_BYTE(leadByte, countTrailBytes) ((leadByte)&=(1<<(6-(countTrailBytes)))-1)
 300
 301 /** Is this this code point a single code unit (byte)? @deprecated ICU 2.4. Renamed to U8_IS_SINGLE, see utf_old.h. */
 302 #define UTF8_IS_SINGLE(uchar) (((uchar)&0x80)==0)
 303 /** 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. */
 304 #define UTF8_IS_LEAD(uchar) ((uint8_t)((uchar)-0xc0)<0x3e)
 305 /** 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. */
 306 #define UTF8_IS_TRAIL(uchar) (((uchar)&0xc0)==0x80)
 307
 308 /** 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. */
 309 #define UTF8_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0x7f)
 310
 311 /**
 312  * Given the lead character, how many bytes are taken by this code point.
 313  * ICU does not deal with code points >0x10ffff
 314  * unless necessary for advancing in the byte stream.
 315  *
 316  * These length macros take into account that for values >0x10ffff
 317  * the UTF8_APPEND_CHAR_SAFE macros would write the error code point 0xffff
 318  * with 3 bytes.
 319  * Code point comparisons need to be in uint32_t because UChar32
 320  * may be a signed type, and negative values must be recognized.
 321  *
 322  * @deprecated ICU 2.4. Use U8_LENGTH instead, see utf.h.
 323  */
 324 #if 1
 325 #   define UTF8_CHAR_LENGTH(c) \
 326         ((uint32_t)(c)<=0x7f ? 1 : \
 327             ((uint32_t)(c)<=0x7ff ? 2 : \
 328                 ((uint32_t)((c)-0x10000)>0xfffff ? 3 : 4) \
 329             ) \
 330         )
 331 #else
 332 #   define UTF8_CHAR_LENGTH(c) \
 333         ((uint32_t)(c)<=0x7f ? 1 : \
 334             ((uint32_t)(c)<=0x7ff ? 2 : \
 335                 ((uint32_t)(c)<=0xffff ? 3 : \
 336                     ((uint32_t)(c)<=0x10ffff ? 4 : \
 337                         ((uint32_t)(c)<=0x3ffffff ? 5 : \
 338                             ((uint32_t)(c)<=0x7fffffff ? 6 : 3) \
 339                         ) \
 340                     ) \
 341                 ) \
 342             ) \
 343         )
 344 #endif
 345
 346 /** The maximum number of bytes per code point. @deprecated ICU 2.4. Renamed to U8_MAX_LENGTH, see utf_old.h. */
 347 #define UTF8_MAX_CHAR_LENGTH 4
 348
 349 /** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 350 #define UTF8_ARRAY_SIZE(size) ((5*(size))/2)
 351
 352 /** @deprecated ICU 2.4. Renamed to U8_GET_UNSAFE, see utf_old.h. */
 353 #define UTF8_GET_CHAR_UNSAFE(s, i, c) { \
 354     int32_t _utf8_get_char_unsafe_index=(int32_t)(i); \
 355     UTF8_SET_CHAR_START_UNSAFE(s, _utf8_get_char_unsafe_index); \
 356     UTF8_NEXT_CHAR_UNSAFE(s, _utf8_get_char_unsafe_index, c); \
 357 }
 358
 359 /** @deprecated ICU 2.4. Use U8_GET instead, see utf_old.h. */
 360 #define UTF8_GET_CHAR_SAFE(s, start, i, length, c, strict) { \
 361     int32_t _utf8_get_char_safe_index=(int32_t)(i); \
 362     UTF8_SET_CHAR_START_SAFE(s, start, _utf8_get_char_safe_index); \
 363     UTF8_NEXT_CHAR_SAFE(s, _utf8_get_char_safe_index, length, c, strict); \
 364 }
 365
 366 /** @deprecated ICU 2.4. Renamed to U8_NEXT_UNSAFE, see utf_old.h. */
 367 #define UTF8_NEXT_CHAR_UNSAFE(s, i, c) { \
 368     (c)=(s)[(i)++]; \
 369     if((uint8_t)((c)-0xc0)<0x35) { \
 370         uint8_t __count=UTF8_COUNT_TRAIL_BYTES(c); \
 371         UTF8_MASK_LEAD_BYTE(c, __count); \
 372         switch(__count) { \
 373         /* each following branch falls through to the next one */ \
 374         case 3: \
 375             (c)=((c)<<6)|((s)[(i)++]&0x3f); \
 376         case 2: \
 377             (c)=((c)<<6)|((s)[(i)++]&0x3f); \
 378         case 1: \
 379             (c)=((c)<<6)|((s)[(i)++]&0x3f); \
 380         /* no other branches to optimize switch() */ \
 381             break; \
 382         } \
 383     } \
 384 }
 385
 386 /** @deprecated ICU 2.4. Renamed to U8_APPEND_UNSAFE, see utf_old.h. */
 387 #define UTF8_APPEND_CHAR_UNSAFE(s, i, c) { \
 388     if((uint32_t)(c)<=0x7f) { \
 389         (s)[(i)++]=(uint8_t)(c); \
 390     } else { \
 391         if((uint32_t)(c)<=0x7ff) { \
 392             (s)[(i)++]=(uint8_t)(((c)>>6)|0xc0); \
 393         } else { \
 394             if((uint32_t)(c)<=0xffff) { \
 395                 (s)[(i)++]=(uint8_t)(((c)>>12)|0xe0); \
 396             } else { \
 397                 (s)[(i)++]=(uint8_t)(((c)>>18)|0xf0); \
 398                 (s)[(i)++]=(uint8_t)((((c)>>12)&0x3f)|0x80); \
 399             } \
 400             (s)[(i)++]=(uint8_t)((((c)>>6)&0x3f)|0x80); \
 401         } \
 402         (s)[(i)++]=(uint8_t)(((c)&0x3f)|0x80); \
 403     } \
 404 }
 405
 406 /** @deprecated ICU 2.4. Renamed to U8_FWD_1_UNSAFE, see utf_old.h. */
 407 #define UTF8_FWD_1_UNSAFE(s, i) { \
 408     (i)+=1+UTF8_COUNT_TRAIL_BYTES((s)[i]); \
 409 }
 410
 411 /** @deprecated ICU 2.4. Renamed to U8_FWD_N_UNSAFE, see utf_old.h. */
 412 #define UTF8_FWD_N_UNSAFE(s, i, n) { \
 413     int32_t __N=(n); \
 414     while(__N>0) { \
 415         UTF8_FWD_1_UNSAFE(s, i); \
 416         --__N; \
 417     } \
 418 }
 419
 420 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_START_UNSAFE, see utf_old.h. */
 421 #define UTF8_SET_CHAR_START_UNSAFE(s, i) { \
 422     while(UTF8_IS_TRAIL((s)[i])) { --(i); } \
 423 }
 424
 425 /** @deprecated ICU 2.4. Use U8_NEXT instead, see utf_old.h. */
 426 #define UTF8_NEXT_CHAR_SAFE(s, i, length, c, strict) { \
 427     (c)=(s)[(i)++]; \
 428     if((c)>=0x80) { \
 429         if(UTF8_IS_LEAD(c)) { \
 430             (c)=utf8_nextCharSafeBody(s, &(i), (int32_t)(length), c, strict); \
 431         } else { \
 432             (c)=UTF8_ERROR_VALUE_1; \
 433         } \
 434     } \
 435 }
 436
 437 /** @deprecated ICU 2.4. Use U8_APPEND instead, see utf_old.h. */
 438 #define UTF8_APPEND_CHAR_SAFE(s, i, length, c)  { \
 439     if((uint32_t)(c)<=0x7f) { \
 440         (s)[(i)++]=(uint8_t)(c); \
 441     } else { \
 442         (i)=utf8_appendCharSafeBody(s, (int32_t)(i), (int32_t)(length), c, NULL); \
 443     } \
 444 }
 445
 446 /** @deprecated ICU 2.4. Renamed to U8_FWD_1, see utf_old.h. */
 447 #define UTF8_FWD_1_SAFE(s, i, length) U8_FWD_1(s, i, length)
 448
 449 /** @deprecated ICU 2.4. Renamed to U8_FWD_N, see utf_old.h. */
 450 #define UTF8_FWD_N_SAFE(s, i, length, n) U8_FWD_N(s, i, length, n)
 451
 452 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_START, see utf_old.h. */
 453 #define UTF8_SET_CHAR_START_SAFE(s, start, i) U8_SET_CP_START(s, start, i)
 454
 455 /** @deprecated ICU 2.4. Renamed to U8_PREV_UNSAFE, see utf_old.h. */
 456 #define UTF8_PREV_CHAR_UNSAFE(s, i, c) { \
 457     (c)=(s)[--(i)]; \
 458     if(UTF8_IS_TRAIL(c)) { \
 459         uint8_t __b, __count=1, __shift=6; \
 460 \
 461         /* c is a trail byte */ \
 462         (c)&=0x3f; \
 463         for(;;) { \
 464             __b=(s)[--(i)]; \
 465             if(__b>=0xc0) { \
 466                 UTF8_MASK_LEAD_BYTE(__b, __count); \
 467                 (c)|=(UChar32)__b<<__shift; \
 468                 break; \
 469             } else { \
 470                 (c)|=(UChar32)(__b&0x3f)<<__shift; \
 471                 ++__count; \
 472                 __shift+=6; \
 473             } \
 474         } \
 475     } \
 476 }
 477
 478 /** @deprecated ICU 2.4. Renamed to U8_BACK_1_UNSAFE, see utf_old.h. */
 479 #define UTF8_BACK_1_UNSAFE(s, i) { \
 480     while(UTF8_IS_TRAIL((s)[--(i)])) {} \
 481 }
 482
 483 /** @deprecated ICU 2.4. Renamed to U8_BACK_N_UNSAFE, see utf_old.h. */
 484 #define UTF8_BACK_N_UNSAFE(s, i, n) { \
 485     int32_t __N=(n); \
 486     while(__N>0) { \
 487         UTF8_BACK_1_UNSAFE(s, i); \
 488         --__N; \
 489     } \
 490 }
 491
 492 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT_UNSAFE, see utf_old.h. */
 493 #define UTF8_SET_CHAR_LIMIT_UNSAFE(s, i) { \
 494     UTF8_BACK_1_UNSAFE(s, i); \
 495     UTF8_FWD_1_UNSAFE(s, i); \
 496 }
 497
 498 /** @deprecated ICU 2.4. Use U8_PREV instead, see utf_old.h. */
 499 #define UTF8_PREV_CHAR_SAFE(s, start, i, c, strict) { \
 500     (c)=(s)[--(i)]; \
 501     if((c)>=0x80) { \
 502         if((c)<=0xbf) { \
 503             (c)=utf8_prevCharSafeBody(s, start, &(i), c, strict); \
 504         } else { \
 505             (c)=UTF8_ERROR_VALUE_1; \
 506         } \
 507     } \
 508 }
 509
 510 /** @deprecated ICU 2.4. Renamed to U8_BACK_1, see utf_old.h. */
 511 #define UTF8_BACK_1_SAFE(s, start, i) U8_BACK_1(s, start, i)
 512
 513 /** @deprecated ICU 2.4. Renamed to U8_BACK_N, see utf_old.h. */
 514 #define UTF8_BACK_N_SAFE(s, start, i, n) U8_BACK_N(s, start, i, n)
 515
 516 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT, see utf_old.h. */
 517 #define UTF8_SET_CHAR_LIMIT_SAFE(s, start, i, length) U8_SET_CP_LIMIT(s, start, i, length)
 518
 519 /* Formerly utf16.h --------------------------------------------------------- */
 520
 521 /** Is uchar a first/lead surrogate? @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h. */
 522 #define UTF_IS_FIRST_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xd800)
 523
 524 /** Is uchar a second/trail surrogate? @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h. */
 525 #define UTF_IS_SECOND_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xdc00)
 526
 527 /** 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. */
 528 #define UTF_IS_SURROGATE_FIRST(c) (((c)&0x400)==0)
 529
 530 /** Helper constant for UTF16_GET_PAIR_VALUE. @deprecated ICU 2.4. Renamed to U16_SURROGATE_OFFSET, see utf_old.h. */
 531 #define UTF_SURROGATE_OFFSET ((0xd800<<10UL)+0xdc00-0x10000)
 532
 533 /** Get the UTF-32 value from the surrogate code units. @deprecated ICU 2.4. Renamed to U16_GET_SUPPLEMENTARY, see utf_old.h. */
 534 #define UTF16_GET_PAIR_VALUE(first, second) \
 535     (((first)<<10UL)+(second)-UTF_SURROGATE_OFFSET)
 536
 537 /** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */
 538 #define UTF_FIRST_SURROGATE(supplementary) (UChar)(((supplementary)>>10)+0xd7c0)
 539
 540 /** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */
 541 #define UTF_SECOND_SURROGATE(supplementary) (UChar)(((supplementary)&0x3ff)|0xdc00)
 542
 543 /** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */
 544 #define UTF16_LEAD(supplementary) UTF_FIRST_SURROGATE(supplementary)
 545
 546 /** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */
 547 #define UTF16_TRAIL(supplementary) UTF_SECOND_SURROGATE(supplementary)
 548
 549 /** @deprecated ICU 2.4. Renamed to U16_IS_SINGLE, see utf_old.h. */
 550 #define UTF16_IS_SINGLE(uchar) !UTF_IS_SURROGATE(uchar)
 551
 552 /** @deprecated ICU 2.4. Renamed to U16_IS_LEAD, see utf_old.h. */
 553 #define UTF16_IS_LEAD(uchar) UTF_IS_FIRST_SURROGATE(uchar)
 554
 555 /** @deprecated ICU 2.4. Renamed to U16_IS_TRAIL, see utf_old.h. */
 556 #define UTF16_IS_TRAIL(uchar) UTF_IS_SECOND_SURROGATE(uchar)
 557
 558 /** 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. */
 559 #define UTF16_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0xffff)
 560
 561 /** @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h. */
 562 #define UTF16_CHAR_LENGTH(c) ((uint32_t)(c)<=0xffff ? 1 : 2)
 563
 564 /** @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h. */
 565 #define UTF16_MAX_CHAR_LENGTH 2
 566
 567 /** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 568 #define UTF16_ARRAY_SIZE(size) (size)
 569
 570 /**
 571  * Get a single code point from an offset that points to any
 572  * of the code units that belong to that code point.
 573  * Assume 0<=i<length.
 574  *
 575  * This could be used for iteration together with
 576  * UTF16_CHAR_LENGTH() and UTF_IS_ERROR(),
 577  * but the use of UTF16_NEXT_CHAR[_UNSAFE]() and
 578  * UTF16_PREV_CHAR[_UNSAFE]() is more efficient for that.
 579  * @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h.
 580  */
 581 #define UTF16_GET_CHAR_UNSAFE(s, i, c) { \
 582     (c)=(s)[i]; \
 583     if(UTF_IS_SURROGATE(c)) { \
 584         if(UTF_IS_SURROGATE_FIRST(c)) { \
 585             (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)+1]); \
 586         } else { \
 587             (c)=UTF16_GET_PAIR_VALUE((s)[(i)-1], (c)); \
 588         } \
 589     } \
 590 }
 591
 592 /** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */
 593 #define UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict) { \
 594     (c)=(s)[i]; \
 595     if(UTF_IS_SURROGATE(c)) { \
 596         uint16_t __c2; \
 597         if(UTF_IS_SURROGATE_FIRST(c)) { \
 598             if((i)+1<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)+1])) { \
 599                 (c)=UTF16_GET_PAIR_VALUE((c), __c2); \
 600                 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \
 601             } else if(strict) {\
 602                 /* unmatched first surrogate */ \
 603                 (c)=UTF_ERROR_VALUE; \
 604             } \
 605         } else { \
 606             if((i)-1>=(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \
 607                 (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \
 608                 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \
 609             } else if(strict) {\
 610                 /* unmatched second surrogate */ \
 611                 (c)=UTF_ERROR_VALUE; \
 612             } \
 613         } \
 614     } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \
 615         (c)=UTF_ERROR_VALUE; \
 616     } \
 617 }
 618
 619 /** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */
 620 #define UTF16_NEXT_CHAR_UNSAFE(s, i, c) { \
 621     (c)=(s)[(i)++]; \
 622     if(UTF_IS_FIRST_SURROGATE(c)) { \
 623         (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)++]); \
 624     } \
 625 }
 626
 627 /** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */
 628 #define UTF16_APPEND_CHAR_UNSAFE(s, i, c) { \
 629     if((uint32_t)(c)<=0xffff) { \
 630         (s)[(i)++]=(uint16_t)(c); \
 631     } else { \
 632         (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \
 633         (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \
 634     } \
 635 }
 636
 637 /** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */
 638 #define UTF16_FWD_1_UNSAFE(s, i) { \
 639     if(UTF_IS_FIRST_SURROGATE((s)[(i)++])) { \
 640         ++(i); \
 641     } \
 642 }
 643
 644 /** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */
 645 #define UTF16_FWD_N_UNSAFE(s, i, n) { \
 646     int32_t __N=(n); \
 647     while(__N>0) { \
 648         UTF16_FWD_1_UNSAFE(s, i); \
 649         --__N; \
 650     } \
 651 }
 652
 653 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */
 654 #define UTF16_SET_CHAR_START_UNSAFE(s, i) { \
 655     if(UTF_IS_SECOND_SURROGATE((s)[i])) { \
 656         --(i); \
 657     } \
 658 }
 659
 660 /** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */
 661 #define UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict) { \
 662     (c)=(s)[(i)++]; \
 663     if(UTF_IS_FIRST_SURROGATE(c)) { \
 664         uint16_t __c2; \
 665         if((i)<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)])) { \
 666             ++(i); \
 667             (c)=UTF16_GET_PAIR_VALUE((c), __c2); \
 668             /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \
 669         } else if(strict) {\
 670             /* unmatched first surrogate */ \
 671             (c)=UTF_ERROR_VALUE; \
 672         } \
 673     } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \
 674         /* unmatched second surrogate or other non-character */ \
 675         (c)=UTF_ERROR_VALUE; \
 676     } \
 677 }
 678
 679 /** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */
 680 #define UTF16_APPEND_CHAR_SAFE(s, i, length, c) { \
 681     if((uint32_t)(c)<=0xffff) { \
 682         (s)[(i)++]=(uint16_t)(c); \
 683     } else if((uint32_t)(c)<=0x10ffff) { \
 684         if((i)+1<(length)) { \
 685             (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \
 686             (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \
 687         } else /* not enough space */ { \
 688             (s)[(i)++]=UTF_ERROR_VALUE; \
 689         } \
 690     } else /* c>0x10ffff, write error value */ { \
 691         (s)[(i)++]=UTF_ERROR_VALUE; \
 692     } \
 693 }
 694
 695 /** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */
 696 #define UTF16_FWD_1_SAFE(s, i, length) U16_FWD_1(s, i, length)
 697
 698 /** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */
 699 #define UTF16_FWD_N_SAFE(s, i, length, n) U16_FWD_N(s, i, length, n)
 700
 701 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */
 702 #define UTF16_SET_CHAR_START_SAFE(s, start, i) U16_SET_CP_START(s, start, i)
 703
 704 /** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */
 705 #define UTF16_PREV_CHAR_UNSAFE(s, i, c) { \
 706     (c)=(s)[--(i)]; \
 707     if(UTF_IS_SECOND_SURROGATE(c)) { \
 708         (c)=UTF16_GET_PAIR_VALUE((s)[--(i)], (c)); \
 709     } \
 710 }
 711
 712 /** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */
 713 #define UTF16_BACK_1_UNSAFE(s, i) { \
 714     if(UTF_IS_SECOND_SURROGATE((s)[--(i)])) { \
 715         --(i); \
 716     } \
 717 }
 718
 719 /** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */
 720 #define UTF16_BACK_N_UNSAFE(s, i, n) { \
 721     int32_t __N=(n); \
 722     while(__N>0) { \
 723         UTF16_BACK_1_UNSAFE(s, i); \
 724         --__N; \
 725     } \
 726 }
 727
 728 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */
 729 #define UTF16_SET_CHAR_LIMIT_UNSAFE(s, i) { \
 730     if(UTF_IS_FIRST_SURROGATE((s)[(i)-1])) { \
 731         ++(i); \
 732     } \
 733 }
 734
 735 /** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */
 736 #define UTF16_PREV_CHAR_SAFE(s, start, i, c, strict) { \
 737     (c)=(s)[--(i)]; \
 738     if(UTF_IS_SECOND_SURROGATE(c)) { \
 739         uint16_t __c2; \
 740         if((i)>(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \
 741             --(i); \
 742             (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \
 743             /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \
 744         } else if(strict) {\
 745             /* unmatched second surrogate */ \
 746             (c)=UTF_ERROR_VALUE; \
 747         } \
 748     } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \
 749         /* unmatched first surrogate or other non-character */ \
 750         (c)=UTF_ERROR_VALUE; \
 751     } \
 752 }
 753
 754 /** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */
 755 #define UTF16_BACK_1_SAFE(s, start, i) U16_BACK_1(s, start, i)
 756
 757 /** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */
 758 #define UTF16_BACK_N_SAFE(s, start, i, n) U16_BACK_N(s, start, i, n)
 759
 760 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */
 761 #define UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, length)
 762
 763 /* Formerly utf32.h --------------------------------------------------------- */
 764
 765 /*
 766 * Old documentation:
 767 *
 768 *   This file defines macros to deal with UTF-32 code units and code points.
 769 *   Signatures and semantics are the same as for the similarly named macros
 770 *   in utf16.h.
 771 *   utf32.h is included by utf.h after unicode/umachine.h</p>
 772 *   and some common definitions.
 773 *   <p><b>Usage:</b>  ICU coding guidelines for if() statements should be followed when using these macros.
 774 *                  Compound statements (curly braces {}) must be used  for if-else-while...
 775 *                  bodies and all macro statements should be terminated with semicolon.</p>
 776 */
 777
 778 /* internal definitions ----------------------------------------------------- */
 779
 780 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 781 #define UTF32_IS_SAFE(c, strict) \
 782     (!(strict) ? \
 783         (uint32_t)(c)<=0x10ffff : \
 784         UTF_IS_UNICODE_CHAR(c))
 785
 786 /*
 787  * For the semantics of all of these macros, see utf16.h.
 788  * The UTF-32 versions are trivial because any code point is
 789  * encoded using exactly one code unit.
 790  */
 791
 792 /* single-code point definitions -------------------------------------------- */
 793
 794 /* classes of code unit values */
 795
 796 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 797 #define UTF32_IS_SINGLE(uchar) 1
 798 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 799 #define UTF32_IS_LEAD(uchar) 0
 800 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 801 #define UTF32_IS_TRAIL(uchar) 0
 802
 803 /* number of code units per code point */
 804
 805 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 806 #define UTF32_NEED_MULTIPLE_UCHAR(c) 0
 807 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 808 #define UTF32_CHAR_LENGTH(c) 1
 809 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 810 #define UTF32_MAX_CHAR_LENGTH 1
 811
 812 /* average number of code units compared to UTF-16 */
 813
 814 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 815 #define UTF32_ARRAY_SIZE(size) (size)
 816
 817 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 818 #define UTF32_GET_CHAR_UNSAFE(s, i, c) { \
 819     (c)=(s)[i]; \
 820 }
 821
 822 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 823 #define UTF32_GET_CHAR_SAFE(s, start, i, length, c, strict) { \
 824     (c)=(s)[i]; \
 825     if(!UTF32_IS_SAFE(c, strict)) { \
 826         (c)=UTF_ERROR_VALUE; \
 827     } \
 828 }
 829
 830 /* definitions with forward iteration --------------------------------------- */
 831
 832 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 833 #define UTF32_NEXT_CHAR_UNSAFE(s, i, c) { \
 834     (c)=(s)[(i)++]; \
 835 }
 836
 837 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 838 #define UTF32_APPEND_CHAR_UNSAFE(s, i, c) { \
 839     (s)[(i)++]=(c); \
 840 }
 841
 842 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 843 #define UTF32_FWD_1_UNSAFE(s, i) { \
 844     ++(i); \
 845 }
 846
 847 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 848 #define UTF32_FWD_N_UNSAFE(s, i, n) { \
 849     (i)+=(n); \
 850 }
 851
 852 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 853 #define UTF32_SET_CHAR_START_UNSAFE(s, i) { \
 854 }
 855
 856 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 857 #define UTF32_NEXT_CHAR_SAFE(s, i, length, c, strict) { \
 858     (c)=(s)[(i)++]; \
 859     if(!UTF32_IS_SAFE(c, strict)) { \
 860         (c)=UTF_ERROR_VALUE; \
 861     } \
 862 }
 863
 864 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 865 #define UTF32_APPEND_CHAR_SAFE(s, i, length, c) { \
 866     if((uint32_t)(c)<=0x10ffff) { \
 867         (s)[(i)++]=(c); \
 868     } else /* c>0x10ffff, write 0xfffd */ { \
 869         (s)[(i)++]=0xfffd; \
 870     } \
 871 }
 872
 873 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 874 #define UTF32_FWD_1_SAFE(s, i, length) { \
 875     ++(i); \
 876 }
 877
 878 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 879 #define UTF32_FWD_N_SAFE(s, i, length, n) { \
 880     if(((i)+=(n))>(length)) { \
 881         (i)=(length); \
 882     } \
 883 }
 884
 885 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 886 #define UTF32_SET_CHAR_START_SAFE(s, start, i) { \
 887 }
 888
 889 /* definitions with backward iteration -------------------------------------- */
 890
 891 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 892 #define UTF32_PREV_CHAR_UNSAFE(s, i, c) { \
 893     (c)=(s)[--(i)]; \
 894 }
 895
 896 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 897 #define UTF32_BACK_1_UNSAFE(s, i) { \
 898     --(i); \
 899 }
 900
 901 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 902 #define UTF32_BACK_N_UNSAFE(s, i, n) { \
 903     (i)-=(n); \
 904 }
 905
 906 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 907 #define UTF32_SET_CHAR_LIMIT_UNSAFE(s, i) { \
 908 }
 909
 910 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 911 #define UTF32_PREV_CHAR_SAFE(s, start, i, c, strict) { \
 912     (c)=(s)[--(i)]; \
 913     if(!UTF32_IS_SAFE(c, strict)) { \
 914         (c)=UTF_ERROR_VALUE; \
 915     } \
 916 }
 917
 918 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 919 #define UTF32_BACK_1_SAFE(s, start, i) { \
 920     --(i); \
 921 }
 922
 923 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 924 #define UTF32_BACK_N_SAFE(s, start, i, n) { \
 925     (i)-=(n); \
 926     if((i)<(start)) { \
 927         (i)=(start); \
 928     } \
 929 }
 930
 931 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
 932 #define UTF32_SET_CHAR_LIMIT_SAFE(s, i, length) { \
 933 }
 934
 935 /* Formerly utf.h, part 2 --------------------------------------------------- */
 936
 937 /**
 938  * Estimate the number of code units for a string based on the number of UTF-16 code units.
 939  *
 940  * @deprecated ICU 2.4. Obsolete, see utf_old.h.
 941  */
 942 #define UTF_ARRAY_SIZE(size) UTF16_ARRAY_SIZE(size)
 943
 944 /** @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h. */
 945 #define UTF_GET_CHAR_UNSAFE(s, i, c)                 UTF16_GET_CHAR_UNSAFE(s, i, c)
 946
 947 /** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */
 948 #define UTF_GET_CHAR_SAFE(s, start, i, length, c, strict) UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict)
 949
 950
 951 /** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */
 952 #define UTF_NEXT_CHAR_UNSAFE(s, i, c)                UTF16_NEXT_CHAR_UNSAFE(s, i, c)
 953
 954 /** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */
 955 #define UTF_NEXT_CHAR_SAFE(s, i, length, c, strict)  UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict)
 956
 957
 958 /** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */
 959 #define UTF_APPEND_CHAR_UNSAFE(s, i, c)              UTF16_APPEND_CHAR_UNSAFE(s, i, c)
 960
 961 /** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */
 962 #define UTF_APPEND_CHAR_SAFE(s, i, length, c)        UTF16_APPEND_CHAR_SAFE(s, i, length, c)
 963
 964
 965 /** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */
 966 #define UTF_FWD_1_UNSAFE(s, i)                       UTF16_FWD_1_UNSAFE(s, i)
 967
 968 /** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */
 969 #define UTF_FWD_1_SAFE(s, i, length)                 UTF16_FWD_1_SAFE(s, i, length)
 970
 971
 972 /** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */
 973 #define UTF_FWD_N_UNSAFE(s, i, n)                    UTF16_FWD_N_UNSAFE(s, i, n)
 974
 975 /** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */
 976 #define UTF_FWD_N_SAFE(s, i, length, n)              UTF16_FWD_N_SAFE(s, i, length, n)
 977
 978
 979 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */
 980 #define UTF_SET_CHAR_START_UNSAFE(s, i)              UTF16_SET_CHAR_START_UNSAFE(s, i)
 981
 982 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */
 983 #define UTF_SET_CHAR_START_SAFE(s, start, i)         UTF16_SET_CHAR_START_SAFE(s, start, i)
 984
 985
 986 /** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */
 987 #define UTF_PREV_CHAR_UNSAFE(s, i, c)                UTF16_PREV_CHAR_UNSAFE(s, i, c)
 988
 989 /** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */
 990 #define UTF_PREV_CHAR_SAFE(s, start, i, c, strict)   UTF16_PREV_CHAR_SAFE(s, start, i, c, strict)
 991
 992
 993 /** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */
 994 #define UTF_BACK_1_UNSAFE(s, i)                      UTF16_BACK_1_UNSAFE(s, i)
 995
 996 /** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */
 997 #define UTF_BACK_1_SAFE(s, start, i)                 UTF16_BACK_1_SAFE(s, start, i)
 998
 999
1000 /** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */
1001 #define UTF_BACK_N_UNSAFE(s, i, n)                   UTF16_BACK_N_UNSAFE(s, i, n)
1002
1003 /** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */
1004 #define UTF_BACK_N_SAFE(s, start, i, n)              UTF16_BACK_N_SAFE(s, start, i, n)
1005
1006
1007 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */
1008 #define UTF_SET_CHAR_LIMIT_UNSAFE(s, i)              UTF16_SET_CHAR_LIMIT_UNSAFE(s, i)
1009
1010 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */
1011 #define UTF_SET_CHAR_LIMIT_SAFE(s, start, i, length) UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length)
1012
1013 /* Define default macros (UTF-16 "safe") ------------------------------------ */
1014
1015 /**
1016  * Does this code unit alone encode a code point (BMP, not a surrogate)?
1017  * Same as UTF16_IS_SINGLE.
1018  * @deprecated ICU 2.4. Renamed to U_IS_SINGLE and U16_IS_SINGLE, see utf_old.h.
1019  */
1020 #define UTF_IS_SINGLE(uchar) U16_IS_SINGLE(uchar)
1021
1022 /**
1023  * Is this code unit the first one of several (a lead surrogate)?
1024  * Same as UTF16_IS_LEAD.
1025  * @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h.
1026  */
1027 #define UTF_IS_LEAD(uchar) U16_IS_LEAD(uchar)
1028
1029 /**
1030  * Is this code unit one of several but not the first one (a trail surrogate)?
1031  * Same as UTF16_IS_TRAIL.
1032  * @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h.
1033  */
1034 #define UTF_IS_TRAIL(uchar) U16_IS_TRAIL(uchar)
1035
1036 /**
1037  * Does this code point require multiple code units (is it a supplementary code point)?
1038  * Same as UTF16_NEED_MULTIPLE_UCHAR.
1039  * @deprecated ICU 2.4. Use U16_LENGTH or test ((uint32_t)(c)>0xffff) instead.
1040  */
1041 #define UTF_NEED_MULTIPLE_UCHAR(c) UTF16_NEED_MULTIPLE_UCHAR(c)
1042
1043 /**
1044  * How many code units are used to encode this code point (1 or 2)?
1045  * Same as UTF16_CHAR_LENGTH.
1046  * @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h.
1047  */
1048 #define UTF_CHAR_LENGTH(c) U16_LENGTH(c)
1049
1050 /**
1051  * How many code units are used at most for any Unicode code point (2)?
1052  * Same as UTF16_MAX_CHAR_LENGTH.
1053  * @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h.
1054  */
1055 #define UTF_MAX_CHAR_LENGTH U16_MAX_LENGTH
1056
1057 /**
1058  * Set c to the code point that contains the code unit i.
1059  * i could point to the lead or the trail surrogate for the code point.
1060  * i is not modified.
1061  * Same as UTF16_GET_CHAR.
1062  * \pre 0<=i<length
1063  *
1064  * @deprecated ICU 2.4. Renamed to U16_GET, see utf_old.h.
1065  */
1066 #define UTF_GET_CHAR(s, start, i, length, c) U16_GET(s, start, i, length, c)
1067
1068 /**
1069  * Set c to the code point that starts at code unit i
1070  * and advance i to beyond the code units of this code point (post-increment).
1071  * i must point to the first code unit of a code point.
1072  * Otherwise c is set to the trail unit (surrogate) itself.
1073  * Same as UTF16_NEXT_CHAR.
1074  * \pre 0<=i<length
1075  * \post 0<i<=length
1076  *
1077  * @deprecated ICU 2.4. Renamed to U16_NEXT, see utf_old.h.
1078  */
1079 #define UTF_NEXT_CHAR(s, i, length, c) U16_NEXT(s, i, length, c)
1080
1081 /**
1082  * Append the code units of code point c to the string at index i
1083  * and advance i to beyond the new code units (post-increment).
1084  * The code units beginning at index i will be overwritten.
1085  * Same as UTF16_APPEND_CHAR.
1086  * \pre 0<=c<=0x10ffff
1087  * \pre 0<=i<length
1088  * \post 0<i<=length
1089  *
1090  * @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h.
1091  */
1092 #define UTF_APPEND_CHAR(s, i, length, c) UTF16_APPEND_CHAR_SAFE(s, i, length, c)
1093
1094 /**
1095  * Advance i to beyond the code units of the code point that begins at i.
1096  * I.e., advance i by one code point.
1097  * Same as UTF16_FWD_1.
1098  * \pre 0<=i<length
1099  * \post 0<i<=length
1100  *
1101  * @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h.
1102  */
1103 #define UTF_FWD_1(s, i, length) U16_FWD_1(s, i, length)
1104
1105 /**
1106  * Advance i to beyond the code units of the n code points where the first one begins at i.
1107  * I.e., advance i by n code points.
1108  * Same as UT16_FWD_N.
1109  * \pre 0<=i<length
1110  * \post 0<i<=length
1111  *
1112  * @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h.
1113  */
1114 #define UTF_FWD_N(s, i, length, n) U16_FWD_N(s, i, length, n)
1115
1116 /**
1117  * Take the random-access index i and adjust it so that it points to the beginning
1118  * of a code point.
1119  * The input index points to any code unit of a code point and is moved to point to
1120  * the first code unit of the same code point. i is never incremented.
1121  * In other words, if i points to a trail surrogate that is preceded by a matching
1122  * lead surrogate, then i is decremented. Otherwise it is not modified.
1123  * This can be used to start an iteration with UTF_NEXT_CHAR() from a random index.
1124  * Same as UTF16_SET_CHAR_START.
1125  * \pre start<=i<length
1126  * \post start<=i<length
1127  *
1128  * @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h.
1129  */
1130 #define UTF_SET_CHAR_START(s, start, i) U16_SET_CP_START(s, start, i)
1131
1132 /**
1133  * Set c to the code point that has code units before i
1134  * and move i backward (towards the beginning of the string)
1135  * to the first code unit of this code point (pre-increment).
1136  * i must point to the first code unit after the last unit of a code point (i==length is allowed).
1137  * Same as UTF16_PREV_CHAR.
1138  * \pre start<i<=length
1139  * \post start<=i<length
1140  *
1141  * @deprecated ICU 2.4. Renamed to U16_PREV, see utf_old.h.
1142  */
1143 #define UTF_PREV_CHAR(s, start, i, c) U16_PREV(s, start, i, c)
1144
1145 /**
1146  * Move i backward (towards the beginning of the string)
1147  * to the first code unit of the code point that has code units before i.
1148  * I.e., move i backward by one code point.
1149  * i must point to the first code unit after the last unit of a code point (i==length is allowed).
1150  * Same as UTF16_BACK_1.
1151  * \pre start<i<=length
1152  * \post start<=i<length
1153  *
1154  * @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h.
1155  */
1156 #define UTF_BACK_1(s, start, i) U16_BACK_1(s, start, i)
1157
1158 /**
1159  * Move i backward (towards the beginning of the string)
1160  * to the first code unit of the n code points that have code units before i.
1161  * I.e., move i backward by n code points.
1162  * i must point to the first code unit after the last unit of a code point (i==length is allowed).
1163  * Same as UTF16_BACK_N.
1164  * \pre start<i<=length
1165  * \post start<=i<length
1166  *
1167  * @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h.
1168  */
1169 #define UTF_BACK_N(s, start, i, n) U16_BACK_N(s, start, i, n)
1170
1171 /**
1172  * Take the random-access index i and adjust it so that it points beyond
1173  * a code point. The input index points beyond any code unit
1174  * of a code point and is moved to point beyond the last code unit of the same
1175  * code point. i is never decremented.
1176  * In other words, if i points to a trail surrogate that is preceded by a matching
1177  * lead surrogate, then i is incremented. Otherwise it is not modified.
1178  * This can be used to start an iteration with UTF_PREV_CHAR() from a random index.
1179  * Same as UTF16_SET_CHAR_LIMIT.
1180  * \pre start<i<=length
1181  * \post start<i<=length
1182  *
1183  * @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h.
1184  */
1185 #define UTF_SET_CHAR_LIMIT(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, length)
1186
1187 #endif /* U_HIDE_DEPRECATED_API */
1188
1189 #endif
1190