include/unicode/utf_old.h

0001 // © 2016 and later: Unicode, Inc. and others.
0002 // License & terms of use: http://www.unicode.org/copyright.html
0003 /*
0004 *******************************************************************************
0005 *
0006 *   Copyright (C) 2002-2012, International Business Machines
0007 *   Corporation and others.  All Rights Reserved.
0008 *
0009 *******************************************************************************
0010 *   file name:  utf_old.h
0011 *   encoding:   UTF-8
0012 *   tab size:   8 (not used)
0013 *   indentation:4
0014 *
0015 *   created on: 2002sep21
0016 *   created by: Markus W. Scherer
0017 */
0018
0019 /**
0020  * \file
0021  * \brief C API: Deprecated macros for Unicode string handling
0022  *
0023  * The macros in utf_old.h are all deprecated and their use discouraged.
0024  * Some of the design principles behind the set of UTF macros
0025  * have changed or proved impractical.
0026  * Almost all of the old "UTF macros" are at least renamed.
0027  * If you are looking for a new equivalent to an old macro, please see the
0028  * comment at the old one.
0029  *
0030  * Brief summary of reasons for deprecation:
0031  * - Switch on UTF_SIZE (selection of UTF-8/16/32 default string processing)
0032  *   was impractical.
0033  * - Switch on UTF_SAFE etc. (selection of unsafe/safe/strict default string processing)
0034  *   was of little use and impractical.
0035  * - Whole classes of macros became obsolete outside of the UTF_SIZE/UTF_SAFE
0036  *   selection framework: UTF32_ macros (all trivial)
0037  *   and UTF_ default and intermediate macros (all aliases).
0038  * - The selection framework also caused many macro aliases.
0039  * - Change in Unicode standard: "irregular" sequences (3.0) became illegal (3.2).
0040  * - Change of language in Unicode standard:
0041  *   Growing distinction between internal x-bit Unicode strings and external UTF-x
0042  *   forms, with the former more lenient.
0043  *   Suggests renaming of UTF16_ macros to U16_.
0044  * - The prefix "UTF_" without a width number confused some users.
0045  * - "Safe" append macros needed the addition of an error indicator output.
0046  * - "Safe" UTF-8 macros used legitimate (if rarely used) code point values
0047  *   to indicate error conditions.
0048  * - The use of the "_CHAR" infix for code point operations confused some users.
0049  *
0050  * More details:
0051  *
0052  * Until ICU 2.2, utf.h theoretically allowed to choose among UTF-8/16/32
0053  * for string processing, and among unsafe/safe/strict default macros for that.
0054  *
0055  * It proved nearly impossible to write non-trivial, high-performance code
0056  * that is UTF-generic.
0057  * Unsafe default macros would be dangerous for default string processing,
0058  * and the main reason for the "strict" versions disappeared:
0059  * Between Unicode 3.0 and 3.2 all "irregular" UTF-8 sequences became illegal.
0060  * The only other conditions that "strict" checked for were non-characters,
0061  * which are valid during processing. Only during text input/output should they
0062  * be checked, and at that time other well-formedness checks may be
0063  * necessary or useful as well.
0064  * This can still be done by using U16_NEXT and U_IS_UNICODE_NONCHAR
0065  * or U_IS_UNICODE_CHAR.
0066  *
0067  * The old UTF8_..._SAFE macros also used some normal Unicode code points
0068  * to indicate malformed sequences.
0069  * The new UTF8_ macros without suffix use negative values instead.
0070  *
0071  * The entire contents of utf32.h was moved here without replacement
0072  * because all those macros were trivial and
0073  * were meaningful only in the framework of choosing the UTF size.
0074  *
0075  * See Jitterbug 2150 and its discussion on the ICU mailing list
0076  * in September 2002.
0077  *
0078  * <hr>
0079  *
0080  * <em>Obsolete part</em> of pre-ICU 2.4 utf.h file documentation:
0081  *
0082  * <p>The original concept for these files was for ICU to allow
0083  * in principle to set which UTF (UTF-8/16/32) is used internally
0084  * by defining UTF_SIZE to either 8, 16, or 32. utf.h would then define the UChar type
0085  * accordingly. UTF-16 was the default.</p>
0086  *
0087  * <p>This concept has been abandoned.
0088  * A lot of the ICU source code assumes UChar strings are in UTF-16.
0089  * This is especially true for low-level code like
0090  * conversion, normalization, and collation.
0091  * The utf.h header enforces the default of UTF-16.
0092  * The UTF-8 and UTF-32 macros remain for now for completeness and backward compatibility.</p>
0093  *
0094  * <p>Accordingly, utf.h defines UChar to be an unsigned 16-bit integer. If this matches wchar_t, then
0095  * UChar is defined to be exactly wchar_t, otherwise uint16_t.</p>
0096  *
0097  * <p>UChar32 is defined to be a signed 32-bit integer (int32_t), large enough for a 21-bit
0098  * Unicode code point (Unicode scalar value, 0..0x10ffff).
0099  * Before ICU 2.4, the definition of UChar32 was similarly platform-dependent as
0100  * the definition of UChar. For details see the documentation for UChar32 itself.</p>
0101  *
0102  * <p>utf.h also defines a number of C macros for handling single Unicode code points and
0103  * for using UTF Unicode strings. It includes utf8.h, utf16.h, and utf32.h for the actual
0104  * implementations of those macros and then aliases one set of them (for UTF-16) for general use.
0105  * The UTF-specific macros have the UTF size in the macro name prefixes (UTF16_...), while
0106  * the general alias macros always begin with UTF_...</p>
0107  *
0108  * <p>Many string operations can be done with or without error checking.
0109  * Where such a distinction is useful, there are two versions of the macros, "unsafe" and "safe"
0110  * ones with ..._UNSAFE and ..._SAFE suffixes. The unsafe macros are fast but may cause
0111  * program failures if the strings are not well-formed. The safe macros have an additional, boolean
0112  * parameter "strict". If strict is false, then only illegal sequences are detected.
0113  * Otherwise, irregular sequences and non-characters are detected as well (like single surrogates).
0114  * Safe macros return special error code points for illegal/irregular sequences:
0115  * Typically, U+ffff, or values that would result in a code unit sequence of the same length
0116  * as the erroneous input sequence.<br>
0117  * Note that _UNSAFE macros have fewer parameters: They do not have the strictness parameter, and
0118  * they do not have start/length parameters for boundary checking.</p>
0119  *
0120  * <p>Here, the macros are aliased in two steps:
0121  * In the first step, the UTF-specific macros with UTF16_ prefix and _UNSAFE and _SAFE suffixes are
0122  * aliased according to the UTF_SIZE to macros with UTF_ prefix and the same suffixes and signatures.
0123  * Then, in a second step, the default, general alias macros are set to use either the unsafe or
0124  * the safe/not strict (default) or the safe/strict macro;
0125  * these general macros do not have a strictness parameter.</p>
0126  *
0127  * <p>It is possible to change the default choice for the general alias macros to be unsafe, safe/not strict or safe/strict.
0128  * The default is safe/not strict. It is not recommended to select the unsafe macros as the basis for
0129  * Unicode string handling in ICU! To select this, define UTF_SAFE, UTF_STRICT, or UTF_UNSAFE.</p>
0130  *
0131  * <p>For general use, one should use the default, general macros with UTF_ prefix and no _SAFE/_UNSAFE suffix.
0132  * Only in some cases it may be necessary to control the choice of macro directly and use a less generic alias.
0133  * For example, if it can be assumed that a string is well-formed and the index will stay within the bounds,
0134  * then the _UNSAFE version may be used.
0135  * If a UTF-8 string is to be processed, then the macros with UTF8_ prefixes need to be used.</p>
0136  *
0137  * <hr>
0138  *
0139  * Deprecated ICU 2.4. Use the macros in utf.h, utf16.h, utf8.h instead.
0140  */
0141
0142 #ifndef __UTF_OLD_H__
0143 #define __UTF_OLD_H__
0144
0145 #include "unicode/utf.h"
0146 #include "unicode/utf8.h"
0147 #include "unicode/utf16.h"
0148
0149 /**
0150  * \def U_HIDE_OBSOLETE_UTF_OLD_H
0151  *
0152  * Hides the obsolete definitions in unicode/utf_old.h.
0153  * Recommended to be set to 1 at compile time to make sure
0154  * the long-deprecated macros are no longer used.
0155  *
0156  * For reasons for the deprecation see the utf_old.h file comments.
0157  *
0158  * @internal
0159  */
0160 #ifndef U_HIDE_OBSOLETE_UTF_OLD_H
0161 #   define U_HIDE_OBSOLETE_UTF_OLD_H 0
0162 #endif
0163
0164 #if !defined(U_HIDE_DEPRECATED_API) && !U_HIDE_OBSOLETE_UTF_OLD_H
0165
0166 /* Formerly utf.h, part 1 --------------------------------------------------- */
0167
0168 #ifdef U_USE_UTF_DEPRECATES
0169 /**
0170  * Unicode string and array offset and index type.
0171  * ICU always counts Unicode code units (UChars) for
0172  * string offsets, indexes, and lengths, not Unicode code points.
0173  *
0174  * @obsolete ICU 2.6. Use int32_t directly instead since this API will be removed in that release.
0175  */
0176 typedef int32_t UTextOffset;
0177 #endif
0178
0179 /** Number of bits in a Unicode string code unit - ICU uses 16-bit Unicode. @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0180 #define UTF_SIZE 16
0181
0182 /**
0183  * The default choice for general Unicode string macros is to use the ..._SAFE macro implementations
0184  * with strict=false.
0185  *
0186  * @deprecated ICU 2.4. Obsolete, see utf_old.h.
0187  */
0188 #define UTF_SAFE
0189 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0190 #undef UTF_UNSAFE
0191 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0192 #undef UTF_STRICT
0193
0194 /**
0195  * UTF8_ERROR_VALUE_1 and UTF8_ERROR_VALUE_2 are special error values for UTF-8,
0196  * which need 1 or 2 bytes in UTF-8:
0197  * \code
0198  * U+0015 = NAK = Negative Acknowledge, C0 control character
0199  * U+009f = highest C1 control character
0200  * \endcode
0201  *
0202  * These are used by UTF8_..._SAFE macros so that they can return an error value
0203  * that needs the same number of code units (bytes) as were seen by
0204  * a macro. They should be tested with UTF_IS_ERROR() or UTF_IS_VALID().
0205  *
0206  * @deprecated ICU 2.4. Obsolete, see utf_old.h.
0207  */
0208 #define UTF8_ERROR_VALUE_1 0x15
0209
0210 /**
0211  * See documentation on UTF8_ERROR_VALUE_1 for details.
0212  *
0213  * @deprecated ICU 2.4. Obsolete, see utf_old.h.
0214  */
0215 #define UTF8_ERROR_VALUE_2 0x9f
0216
0217 /**
0218  * Error value for all UTFs. This code point value will be set by macros with error
0219  * checking if an error is detected.
0220  *
0221  * @deprecated ICU 2.4. Obsolete, see utf_old.h.
0222  */
0223 #define UTF_ERROR_VALUE 0xffff
0224
0225 /**
0226  * Is a given 32-bit code an error value
0227  * as returned by one of the macros for any UTF?
0228  *
0229  * @deprecated ICU 2.4. Obsolete, see utf_old.h.
0230  */
0231 #define UTF_IS_ERROR(c) \
0232     (((c)&0xfffe)==0xfffe || (c)==UTF8_ERROR_VALUE_1 || (c)==UTF8_ERROR_VALUE_2)
0233
0234 /**
0235  * This is a combined macro: Is c a valid Unicode value _and_ not an error code?
0236  *
0237  * @deprecated ICU 2.4. Obsolete, see utf_old.h.
0238  */
0239 #define UTF_IS_VALID(c) \
0240     (UTF_IS_UNICODE_CHAR(c) && \
0241      (c)!=UTF8_ERROR_VALUE_1 && (c)!=UTF8_ERROR_VALUE_2)
0242
0243 /**
0244  * Is this code unit or code point a surrogate (U+d800..U+dfff)?
0245  * @deprecated ICU 2.4. Renamed to U_IS_SURROGATE and U16_IS_SURROGATE, see utf_old.h.
0246  */
0247 #define UTF_IS_SURROGATE(uchar) (((uchar)&0xfffff800)==0xd800)
0248
0249 /**
0250  * Is a given 32-bit code point a Unicode noncharacter?
0251  *
0252  * @deprecated ICU 2.4. Renamed to U_IS_UNICODE_NONCHAR, see utf_old.h.
0253  */
0254 #define UTF_IS_UNICODE_NONCHAR(c) \
0255     ((c)>=0xfdd0 && \
0256      ((uint32_t)(c)<=0xfdef || ((c)&0xfffe)==0xfffe) && \
0257      (uint32_t)(c)<=0x10ffff)
0258
0259 /**
0260  * Is a given 32-bit value a Unicode code point value (0..U+10ffff)
0261  * that can be assigned a character?
0262  *
0263  * Code points that are not characters include:
0264  * - single surrogate code points (U+d800..U+dfff, 2048 code points)
0265  * - the last two code points on each plane (U+__fffe and U+__ffff, 34 code points)
0266  * - U+fdd0..U+fdef (new with Unicode 3.1, 32 code points)
0267  * - the highest Unicode code point value is U+10ffff
0268  *
0269  * This means that all code points below U+d800 are character code points,
0270  * and that boundary is tested first for performance.
0271  *
0272  * @deprecated ICU 2.4. Renamed to U_IS_UNICODE_CHAR, see utf_old.h.
0273  */
0274 #define UTF_IS_UNICODE_CHAR(c) \
0275     ((uint32_t)(c)<0xd800 || \
0276         ((uint32_t)(c)>0xdfff && \
0277          (uint32_t)(c)<=0x10ffff && \
0278          !UTF_IS_UNICODE_NONCHAR(c)))
0279
0280 /* Formerly utf8.h ---------------------------------------------------------- */
0281
0282 /**
0283 * \var utf8_countTrailBytes
0284 * Internal array with numbers of trail bytes for any given byte used in
0285 * lead byte position.
0286 *
0287 * This is internal since it is not meant to be called directly by external clients;
0288 * however it is called by public macros in this file and thus must remain stable,
0289 * and should not be hidden when other internal functions are hidden (otherwise
0290 * public macros would fail to compile).
0291 * @internal
0292 */
0293 #ifdef U_UTF8_IMPL
0294 // No forward declaration if compiling utf_impl.cpp, which defines utf8_countTrailBytes.
0295 #elif defined(U_STATIC_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION)
0296 U_CAPI const uint8_t utf8_countTrailBytes[];
0297 #else
0298 U_CFUNC U_IMPORT const uint8_t utf8_countTrailBytes[];
0299 #endif
0300
0301 /**
0302  * Count the trail bytes for a UTF-8 lead byte.
0303  * @deprecated ICU 2.4. Renamed to U8_COUNT_TRAIL_BYTES, see utf_old.h.
0304  */
0305 #define UTF8_COUNT_TRAIL_BYTES(leadByte) (utf8_countTrailBytes[(uint8_t)leadByte])
0306
0307 /**
0308  * Mask a UTF-8 lead byte, leave only the lower bits that form part of the code point value.
0309  * @deprecated ICU 2.4. Renamed to U8_MASK_LEAD_BYTE, see utf_old.h.
0310  */
0311 #define UTF8_MASK_LEAD_BYTE(leadByte, countTrailBytes) ((leadByte)&=(1<<(6-(countTrailBytes)))-1)
0312
0313 /** Is this this code point a single code unit (byte)? @deprecated ICU 2.4. Renamed to U8_IS_SINGLE, see utf_old.h. */
0314 #define UTF8_IS_SINGLE(uchar) (((uchar)&0x80)==0)
0315 /** 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. */
0316 #define UTF8_IS_LEAD(uchar) ((uint8_t)((uchar)-0xc0)<0x3e)
0317 /** 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. */
0318 #define UTF8_IS_TRAIL(uchar) (((uchar)&0xc0)==0x80)
0319
0320 /** 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. */
0321 #define UTF8_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0x7f)
0322
0323 /**
0324  * Given the lead character, how many bytes are taken by this code point.
0325  * ICU does not deal with code points >0x10ffff
0326  * unless necessary for advancing in the byte stream.
0327  *
0328  * These length macros take into account that for values >0x10ffff
0329  * the UTF8_APPEND_CHAR_SAFE macros would write the error code point 0xffff
0330  * with 3 bytes.
0331  * Code point comparisons need to be in uint32_t because UChar32
0332  * may be a signed type, and negative values must be recognized.
0333  *
0334  * @deprecated ICU 2.4. Use U8_LENGTH instead, see utf.h.
0335  */
0336 #if 1
0337 #   define UTF8_CHAR_LENGTH(c) \
0338         ((uint32_t)(c)<=0x7f ? 1 : \
0339             ((uint32_t)(c)<=0x7ff ? 2 : \
0340                 ((uint32_t)((c)-0x10000)>0xfffff ? 3 : 4) \
0341             ) \
0342         )
0343 #else
0344 #   define UTF8_CHAR_LENGTH(c) \
0345         ((uint32_t)(c)<=0x7f ? 1 : \
0346             ((uint32_t)(c)<=0x7ff ? 2 : \
0347                 ((uint32_t)(c)<=0xffff ? 3 : \
0348                     ((uint32_t)(c)<=0x10ffff ? 4 : \
0349                         ((uint32_t)(c)<=0x3ffffff ? 5 : \
0350                             ((uint32_t)(c)<=0x7fffffff ? 6 : 3) \
0351                         ) \
0352                     ) \
0353                 ) \
0354             ) \
0355         )
0356 #endif
0357
0358 /** The maximum number of bytes per code point. @deprecated ICU 2.4. Renamed to U8_MAX_LENGTH, see utf_old.h. */
0359 #define UTF8_MAX_CHAR_LENGTH 4
0360
0361 /** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0362 #define UTF8_ARRAY_SIZE(size) ((5*(size))/2)
0363
0364 /** @deprecated ICU 2.4. Renamed to U8_GET_UNSAFE, see utf_old.h. */
0365 #define UTF8_GET_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
0366     int32_t _utf8_get_char_unsafe_index=(int32_t)(i); \
0367     UTF8_SET_CHAR_START_UNSAFE(s, _utf8_get_char_unsafe_index); \
0368     UTF8_NEXT_CHAR_UNSAFE(s, _utf8_get_char_unsafe_index, c); \
0369 } UPRV_BLOCK_MACRO_END
0370
0371 /** @deprecated ICU 2.4. Use U8_GET instead, see utf_old.h. */
0372 #define UTF8_GET_CHAR_SAFE(s, start, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
0373     int32_t _utf8_get_char_safe_index=(int32_t)(i); \
0374     UTF8_SET_CHAR_START_SAFE(s, start, _utf8_get_char_safe_index); \
0375     UTF8_NEXT_CHAR_SAFE(s, _utf8_get_char_safe_index, length, c, strict); \
0376 } UPRV_BLOCK_MACRO_END
0377
0378 /** @deprecated ICU 2.4. Renamed to U8_NEXT_UNSAFE, see utf_old.h. */
0379 #define UTF8_NEXT_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
0380     (c)=(s)[(i)++]; \
0381     if((uint8_t)((c)-0xc0)<0x35) { \
0382         uint8_t __count=UTF8_COUNT_TRAIL_BYTES(c); \
0383         UTF8_MASK_LEAD_BYTE(c, __count); \
0384         switch(__count) { \
0385         /* each following branch falls through to the next one */ \
0386         case 3: \
0387             (c)=((c)<<6)|((s)[(i)++]&0x3f); \
0388         case 2: \
0389             (c)=((c)<<6)|((s)[(i)++]&0x3f); \
0390         case 1: \
0391             (c)=((c)<<6)|((s)[(i)++]&0x3f); \
0392         /* no other branches to optimize switch() */ \
0393             break; \
0394         } \
0395     } \
0396 } UPRV_BLOCK_MACRO_END
0397
0398 /** @deprecated ICU 2.4. Renamed to U8_APPEND_UNSAFE, see utf_old.h. */
0399 #define UTF8_APPEND_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
0400     if((uint32_t)(c)<=0x7f) { \
0401         (s)[(i)++]=(uint8_t)(c); \
0402     } else { \
0403         if((uint32_t)(c)<=0x7ff) { \
0404             (s)[(i)++]=(uint8_t)(((c)>>6)|0xc0); \
0405         } else { \
0406             if((uint32_t)(c)<=0xffff) { \
0407                 (s)[(i)++]=(uint8_t)(((c)>>12)|0xe0); \
0408             } else { \
0409                 (s)[(i)++]=(uint8_t)(((c)>>18)|0xf0); \
0410                 (s)[(i)++]=(uint8_t)((((c)>>12)&0x3f)|0x80); \
0411             } \
0412             (s)[(i)++]=(uint8_t)((((c)>>6)&0x3f)|0x80); \
0413         } \
0414         (s)[(i)++]=(uint8_t)(((c)&0x3f)|0x80); \
0415     } \
0416 } UPRV_BLOCK_MACRO_END
0417
0418 /** @deprecated ICU 2.4. Renamed to U8_FWD_1_UNSAFE, see utf_old.h. */
0419 #define UTF8_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
0420     (i)+=1+UTF8_COUNT_TRAIL_BYTES((s)[i]); \
0421 } UPRV_BLOCK_MACRO_END
0422
0423 /** @deprecated ICU 2.4. Renamed to U8_FWD_N_UNSAFE, see utf_old.h. */
0424 #define UTF8_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
0425     int32_t __N=(n); \
0426     while(__N>0) { \
0427         UTF8_FWD_1_UNSAFE(s, i); \
0428         --__N; \
0429     } \
0430 } UPRV_BLOCK_MACRO_END
0431
0432 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_START_UNSAFE, see utf_old.h. */
0433 #define UTF8_SET_CHAR_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
0434     while(UTF8_IS_TRAIL((s)[i])) { --(i); } \
0435 } UPRV_BLOCK_MACRO_END
0436
0437 /** @deprecated ICU 2.4. Use U8_NEXT instead, see utf_old.h. */
0438 #define UTF8_NEXT_CHAR_SAFE(s, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
0439     (c)=(s)[(i)++]; \
0440     if((c)>=0x80) { \
0441         if(UTF8_IS_LEAD(c)) { \
0442             (c)=utf8_nextCharSafeBody(s, &(i), (int32_t)(length), c, strict); \
0443         } else { \
0444             (c)=UTF8_ERROR_VALUE_1; \
0445         } \
0446     } \
0447 } UPRV_BLOCK_MACRO_END
0448
0449 /** @deprecated ICU 2.4. Use U8_APPEND instead, see utf_old.h. */
0450 #define UTF8_APPEND_CHAR_SAFE(s, i, length, c)  UPRV_BLOCK_MACRO_BEGIN { \
0451     if((uint32_t)(c)<=0x7f) { \
0452         (s)[(i)++]=(uint8_t)(c); \
0453     } else { \
0454         (i)=utf8_appendCharSafeBody(s, (int32_t)(i), (int32_t)(length), c, NULL); \
0455     } \
0456 } UPRV_BLOCK_MACRO_END
0457
0458 /** @deprecated ICU 2.4. Renamed to U8_FWD_1, see utf_old.h. */
0459 #define UTF8_FWD_1_SAFE(s, i, length) U8_FWD_1(s, i, length)
0460
0461 /** @deprecated ICU 2.4. Renamed to U8_FWD_N, see utf_old.h. */
0462 #define UTF8_FWD_N_SAFE(s, i, length, n) U8_FWD_N(s, i, length, n)
0463
0464 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_START, see utf_old.h. */
0465 #define UTF8_SET_CHAR_START_SAFE(s, start, i) U8_SET_CP_START(s, start, i)
0466
0467 /** @deprecated ICU 2.4. Renamed to U8_PREV_UNSAFE, see utf_old.h. */
0468 #define UTF8_PREV_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
0469     (c)=(s)[--(i)]; \
0470     if(UTF8_IS_TRAIL(c)) { \
0471         uint8_t __b, __count=1, __shift=6; \
0472 \
0473         /* c is a trail byte */ \
0474         (c)&=0x3f; \
0475         for(;;) { \
0476             __b=(s)[--(i)]; \
0477             if(__b>=0xc0) { \
0478                 UTF8_MASK_LEAD_BYTE(__b, __count); \
0479                 (c)|=(UChar32)__b<<__shift; \
0480                 break; \
0481             } else { \
0482                 (c)|=(UChar32)(__b&0x3f)<<__shift; \
0483                 ++__count; \
0484                 __shift+=6; \
0485             } \
0486         } \
0487     } \
0488 } UPRV_BLOCK_MACRO_END
0489
0490 /** @deprecated ICU 2.4. Renamed to U8_BACK_1_UNSAFE, see utf_old.h. */
0491 #define UTF8_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
0492     while(UTF8_IS_TRAIL((s)[--(i)])) {} \
0493 } UPRV_BLOCK_MACRO_END
0494
0495 /** @deprecated ICU 2.4. Renamed to U8_BACK_N_UNSAFE, see utf_old.h. */
0496 #define UTF8_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
0497     int32_t __N=(n); \
0498     while(__N>0) { \
0499         UTF8_BACK_1_UNSAFE(s, i); \
0500         --__N; \
0501     } \
0502 } UPRV_BLOCK_MACRO_END
0503
0504 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT_UNSAFE, see utf_old.h. */
0505 #define UTF8_SET_CHAR_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
0506     UTF8_BACK_1_UNSAFE(s, i); \
0507     UTF8_FWD_1_UNSAFE(s, i); \
0508 } UPRV_BLOCK_MACRO_END
0509
0510 /** @deprecated ICU 2.4. Use U8_PREV instead, see utf_old.h. */
0511 #define UTF8_PREV_CHAR_SAFE(s, start, i, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
0512     (c)=(s)[--(i)]; \
0513     if((c)>=0x80) { \
0514         if((c)<=0xbf) { \
0515             (c)=utf8_prevCharSafeBody(s, start, &(i), c, strict); \
0516         } else { \
0517             (c)=UTF8_ERROR_VALUE_1; \
0518         } \
0519     } \
0520 } UPRV_BLOCK_MACRO_END
0521
0522 /** @deprecated ICU 2.4. Renamed to U8_BACK_1, see utf_old.h. */
0523 #define UTF8_BACK_1_SAFE(s, start, i) U8_BACK_1(s, start, i)
0524
0525 /** @deprecated ICU 2.4. Renamed to U8_BACK_N, see utf_old.h. */
0526 #define UTF8_BACK_N_SAFE(s, start, i, n) U8_BACK_N(s, start, i, n)
0527
0528 /** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT, see utf_old.h. */
0529 #define UTF8_SET_CHAR_LIMIT_SAFE(s, start, i, length) U8_SET_CP_LIMIT(s, start, i, length)
0530
0531 /* Formerly utf16.h --------------------------------------------------------- */
0532
0533 /** Is uchar a first/lead surrogate? @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h. */
0534 #define UTF_IS_FIRST_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xd800)
0535
0536 /** Is uchar a second/trail surrogate? @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h. */
0537 #define UTF_IS_SECOND_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xdc00)
0538
0539 /** 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. */
0540 #define UTF_IS_SURROGATE_FIRST(c) (((c)&0x400)==0)
0541
0542 /** Helper constant for UTF16_GET_PAIR_VALUE. @deprecated ICU 2.4. Renamed to U16_SURROGATE_OFFSET, see utf_old.h. */
0543 #define UTF_SURROGATE_OFFSET ((0xd800<<10UL)+0xdc00-0x10000)
0544
0545 /** Get the UTF-32 value from the surrogate code units. @deprecated ICU 2.4. Renamed to U16_GET_SUPPLEMENTARY, see utf_old.h. */
0546 #define UTF16_GET_PAIR_VALUE(first, second) \
0547     (((first)<<10UL)+(second)-UTF_SURROGATE_OFFSET)
0548
0549 /** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */
0550 #define UTF_FIRST_SURROGATE(supplementary) (UChar)(((supplementary)>>10)+0xd7c0)
0551
0552 /** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */
0553 #define UTF_SECOND_SURROGATE(supplementary) (UChar)(((supplementary)&0x3ff)|0xdc00)
0554
0555 /** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */
0556 #define UTF16_LEAD(supplementary) UTF_FIRST_SURROGATE(supplementary)
0557
0558 /** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */
0559 #define UTF16_TRAIL(supplementary) UTF_SECOND_SURROGATE(supplementary)
0560
0561 /** @deprecated ICU 2.4. Renamed to U16_IS_SINGLE, see utf_old.h. */
0562 #define UTF16_IS_SINGLE(uchar) !UTF_IS_SURROGATE(uchar)
0563
0564 /** @deprecated ICU 2.4. Renamed to U16_IS_LEAD, see utf_old.h. */
0565 #define UTF16_IS_LEAD(uchar) UTF_IS_FIRST_SURROGATE(uchar)
0566
0567 /** @deprecated ICU 2.4. Renamed to U16_IS_TRAIL, see utf_old.h. */
0568 #define UTF16_IS_TRAIL(uchar) UTF_IS_SECOND_SURROGATE(uchar)
0569
0570 /** 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. */
0571 #define UTF16_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0xffff)
0572
0573 /** @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h. */
0574 #define UTF16_CHAR_LENGTH(c) ((uint32_t)(c)<=0xffff ? 1 : 2)
0575
0576 /** @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h. */
0577 #define UTF16_MAX_CHAR_LENGTH 2
0578
0579 /** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0580 #define UTF16_ARRAY_SIZE(size) (size)
0581
0582 /**
0583  * Get a single code point from an offset that points to any
0584  * of the code units that belong to that code point.
0585  * Assume 0<=i<length.
0586  *
0587  * This could be used for iteration together with
0588  * UTF16_CHAR_LENGTH() and UTF_IS_ERROR(),
0589  * but the use of UTF16_NEXT_CHAR[_UNSAFE]() and
0590  * UTF16_PREV_CHAR[_UNSAFE]() is more efficient for that.
0591  * @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h.
0592  */
0593 #define UTF16_GET_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
0594     (c)=(s)[i]; \
0595     if(UTF_IS_SURROGATE(c)) { \
0596         if(UTF_IS_SURROGATE_FIRST(c)) { \
0597             (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)+1]); \
0598         } else { \
0599             (c)=UTF16_GET_PAIR_VALUE((s)[(i)-1], (c)); \
0600         } \
0601     } \
0602 } UPRV_BLOCK_MACRO_END
0603
0604 /** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */
0605 #define UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
0606     (c)=(s)[i]; \
0607     if(UTF_IS_SURROGATE(c)) { \
0608         uint16_t __c2; \
0609         if(UTF_IS_SURROGATE_FIRST(c)) { \
0610             if((i)+1<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)+1])) { \
0611                 (c)=UTF16_GET_PAIR_VALUE((c), __c2); \
0612                 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \
0613             } else if(strict) {\
0614                 /* unmatched first surrogate */ \
0615                 (c)=UTF_ERROR_VALUE; \
0616             } \
0617         } else { \
0618             if((i)-1>=(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \
0619                 (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \
0620                 /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \
0621             } else if(strict) {\
0622                 /* unmatched second surrogate */ \
0623                 (c)=UTF_ERROR_VALUE; \
0624             } \
0625         } \
0626     } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \
0627         (c)=UTF_ERROR_VALUE; \
0628     } \
0629 } UPRV_BLOCK_MACRO_END
0630
0631 /** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */
0632 #define UTF16_NEXT_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
0633     (c)=(s)[(i)++]; \
0634     if(UTF_IS_FIRST_SURROGATE(c)) { \
0635         (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)++]); \
0636     } \
0637 } UPRV_BLOCK_MACRO_END
0638
0639 /** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */
0640 #define UTF16_APPEND_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
0641     if((uint32_t)(c)<=0xffff) { \
0642         (s)[(i)++]=(uint16_t)(c); \
0643     } else { \
0644         (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \
0645         (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \
0646     } \
0647 } UPRV_BLOCK_MACRO_END
0648
0649 /** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */
0650 #define UTF16_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
0651     if(UTF_IS_FIRST_SURROGATE((s)[(i)++])) { \
0652         ++(i); \
0653     } \
0654 } UPRV_BLOCK_MACRO_END
0655
0656 /** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */
0657 #define UTF16_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
0658     int32_t __N=(n); \
0659     while(__N>0) { \
0660         UTF16_FWD_1_UNSAFE(s, i); \
0661         --__N; \
0662     } \
0663 } UPRV_BLOCK_MACRO_END
0664
0665 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */
0666 #define UTF16_SET_CHAR_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
0667     if(UTF_IS_SECOND_SURROGATE((s)[i])) { \
0668         --(i); \
0669     } \
0670 } UPRV_BLOCK_MACRO_END
0671
0672 /** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */
0673 #define UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
0674     (c)=(s)[(i)++]; \
0675     if(UTF_IS_FIRST_SURROGATE(c)) { \
0676         uint16_t __c2; \
0677         if((i)<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)])) { \
0678             ++(i); \
0679             (c)=UTF16_GET_PAIR_VALUE((c), __c2); \
0680             /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \
0681         } else if(strict) {\
0682             /* unmatched first surrogate */ \
0683             (c)=UTF_ERROR_VALUE; \
0684         } \
0685     } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \
0686         /* unmatched second surrogate or other non-character */ \
0687         (c)=UTF_ERROR_VALUE; \
0688     } \
0689 } UPRV_BLOCK_MACRO_END
0690
0691 /** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */
0692 #define UTF16_APPEND_CHAR_SAFE(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
0693     if((uint32_t)(c)<=0xffff) { \
0694         (s)[(i)++]=(uint16_t)(c); \
0695     } else if((uint32_t)(c)<=0x10ffff) { \
0696         if((i)+1<(length)) { \
0697             (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \
0698             (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \
0699         } else /* not enough space */ { \
0700             (s)[(i)++]=UTF_ERROR_VALUE; \
0701         } \
0702     } else /* c>0x10ffff, write error value */ { \
0703         (s)[(i)++]=UTF_ERROR_VALUE; \
0704     } \
0705 } UPRV_BLOCK_MACRO_END
0706
0707 /** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */
0708 #define UTF16_FWD_1_SAFE(s, i, length) U16_FWD_1(s, i, length)
0709
0710 /** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */
0711 #define UTF16_FWD_N_SAFE(s, i, length, n) U16_FWD_N(s, i, length, n)
0712
0713 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */
0714 #define UTF16_SET_CHAR_START_SAFE(s, start, i) U16_SET_CP_START(s, start, i)
0715
0716 /** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */
0717 #define UTF16_PREV_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
0718     (c)=(s)[--(i)]; \
0719     if(UTF_IS_SECOND_SURROGATE(c)) { \
0720         (c)=UTF16_GET_PAIR_VALUE((s)[--(i)], (c)); \
0721     } \
0722 } UPRV_BLOCK_MACRO_END
0723
0724 /** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */
0725 #define UTF16_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
0726     if(UTF_IS_SECOND_SURROGATE((s)[--(i)])) { \
0727         --(i); \
0728     } \
0729 } UPRV_BLOCK_MACRO_END
0730
0731 /** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */
0732 #define UTF16_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
0733     int32_t __N=(n); \
0734     while(__N>0) { \
0735         UTF16_BACK_1_UNSAFE(s, i); \
0736         --__N; \
0737     } \
0738 } UPRV_BLOCK_MACRO_END
0739
0740 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */
0741 #define UTF16_SET_CHAR_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
0742     if(UTF_IS_FIRST_SURROGATE((s)[(i)-1])) { \
0743         ++(i); \
0744     } \
0745 } UPRV_BLOCK_MACRO_END
0746
0747 /** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */
0748 #define UTF16_PREV_CHAR_SAFE(s, start, i, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
0749     (c)=(s)[--(i)]; \
0750     if(UTF_IS_SECOND_SURROGATE(c)) { \
0751         uint16_t __c2; \
0752         if((i)>(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \
0753             --(i); \
0754             (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \
0755             /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \
0756         } else if(strict) {\
0757             /* unmatched second surrogate */ \
0758             (c)=UTF_ERROR_VALUE; \
0759         } \
0760     } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \
0761         /* unmatched first surrogate or other non-character */ \
0762         (c)=UTF_ERROR_VALUE; \
0763     } \
0764 } UPRV_BLOCK_MACRO_END
0765
0766 /** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */
0767 #define UTF16_BACK_1_SAFE(s, start, i) U16_BACK_1(s, start, i)
0768
0769 /** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */
0770 #define UTF16_BACK_N_SAFE(s, start, i, n) U16_BACK_N(s, start, i, n)
0771
0772 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */
0773 #define UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, length)
0774
0775 /* Formerly utf32.h --------------------------------------------------------- */
0776
0777 /*
0778 * Old documentation:
0779 *
0780 *   This file defines macros to deal with UTF-32 code units and code points.
0781 *   Signatures and semantics are the same as for the similarly named macros
0782 *   in utf16.h.
0783 *   utf32.h is included by utf.h after unicode/umachine.h</p>
0784 *   and some common definitions.
0785 *   <p><b>Usage:</b>  ICU coding guidelines for if() statements should be followed when using these macros.
0786 *                  Compound statements (curly braces {}) must be used  for if-else-while...
0787 *                  bodies and all macro statements should be terminated with semicolon.</p>
0788 */
0789
0790 /* internal definitions ----------------------------------------------------- */
0791
0792 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0793 #define UTF32_IS_SAFE(c, strict) \
0794     (!(strict) ? \
0795         (uint32_t)(c)<=0x10ffff : \
0796         UTF_IS_UNICODE_CHAR(c))
0797
0798 /*
0799  * For the semantics of all of these macros, see utf16.h.
0800  * The UTF-32 versions are trivial because any code point is
0801  * encoded using exactly one code unit.
0802  */
0803
0804 /* single-code point definitions -------------------------------------------- */
0805
0806 /* classes of code unit values */
0807
0808 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0809 #define UTF32_IS_SINGLE(uchar) 1
0810 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0811 #define UTF32_IS_LEAD(uchar) 0
0812 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0813 #define UTF32_IS_TRAIL(uchar) 0
0814
0815 /* number of code units per code point */
0816
0817 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0818 #define UTF32_NEED_MULTIPLE_UCHAR(c) 0
0819 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0820 #define UTF32_CHAR_LENGTH(c) 1
0821 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0822 #define UTF32_MAX_CHAR_LENGTH 1
0823
0824 /* average number of code units compared to UTF-16 */
0825
0826 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0827 #define UTF32_ARRAY_SIZE(size) (size)
0828
0829 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0830 #define UTF32_GET_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
0831     (c)=(s)[i]; \
0832 } UPRV_BLOCK_MACRO_END
0833
0834 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0835 #define UTF32_GET_CHAR_SAFE(s, start, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
0836     (c)=(s)[i]; \
0837     if(!UTF32_IS_SAFE(c, strict)) { \
0838         (c)=UTF_ERROR_VALUE; \
0839     } \
0840 } UPRV_BLOCK_MACRO_END
0841
0842 /* definitions with forward iteration --------------------------------------- */
0843
0844 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0845 #define UTF32_NEXT_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
0846     (c)=(s)[(i)++]; \
0847 } UPRV_BLOCK_MACRO_END
0848
0849 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0850 #define UTF32_APPEND_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
0851     (s)[(i)++]=(c); \
0852 } UPRV_BLOCK_MACRO_END
0853
0854 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0855 #define UTF32_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
0856     ++(i); \
0857 } UPRV_BLOCK_MACRO_END
0858
0859 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0860 #define UTF32_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
0861     (i)+=(n); \
0862 } UPRV_BLOCK_MACRO_END
0863
0864 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0865 #define UTF32_SET_CHAR_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
0866 } UPRV_BLOCK_MACRO_END
0867
0868 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0869 #define UTF32_NEXT_CHAR_SAFE(s, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
0870     (c)=(s)[(i)++]; \
0871     if(!UTF32_IS_SAFE(c, strict)) { \
0872         (c)=UTF_ERROR_VALUE; \
0873     } \
0874 } UPRV_BLOCK_MACRO_END
0875
0876 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0877 #define UTF32_APPEND_CHAR_SAFE(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
0878     if((uint32_t)(c)<=0x10ffff) { \
0879         (s)[(i)++]=(c); \
0880     } else /* c>0x10ffff, write 0xfffd */ { \
0881         (s)[(i)++]=0xfffd; \
0882     } \
0883 } UPRV_BLOCK_MACRO_END
0884
0885 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0886 #define UTF32_FWD_1_SAFE(s, i, length) UPRV_BLOCK_MACRO_BEGIN { \
0887     ++(i); \
0888 } UPRV_BLOCK_MACRO_END
0889
0890 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0891 #define UTF32_FWD_N_SAFE(s, i, length, n) UPRV_BLOCK_MACRO_BEGIN { \
0892     if(((i)+=(n))>(length)) { \
0893         (i)=(length); \
0894     } \
0895 } UPRV_BLOCK_MACRO_END
0896
0897 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0898 #define UTF32_SET_CHAR_START_SAFE(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \
0899 } UPRV_BLOCK_MACRO_END
0900
0901 /* definitions with backward iteration -------------------------------------- */
0902
0903 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0904 #define UTF32_PREV_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
0905     (c)=(s)[--(i)]; \
0906 } UPRV_BLOCK_MACRO_END
0907
0908 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0909 #define UTF32_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
0910     --(i); \
0911 } UPRV_BLOCK_MACRO_END
0912
0913 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0914 #define UTF32_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
0915     (i)-=(n); \
0916 } UPRV_BLOCK_MACRO_END
0917
0918 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0919 #define UTF32_SET_CHAR_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
0920 } UPRV_BLOCK_MACRO_END
0921
0922 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0923 #define UTF32_PREV_CHAR_SAFE(s, start, i, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
0924     (c)=(s)[--(i)]; \
0925     if(!UTF32_IS_SAFE(c, strict)) { \
0926         (c)=UTF_ERROR_VALUE; \
0927     } \
0928 } UPRV_BLOCK_MACRO_END
0929
0930 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0931 #define UTF32_BACK_1_SAFE(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \
0932     --(i); \
0933 } UPRV_BLOCK_MACRO_END
0934
0935 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0936 #define UTF32_BACK_N_SAFE(s, start, i, n) UPRV_BLOCK_MACRO_BEGIN { \
0937     (i)-=(n); \
0938     if((i)<(start)) { \
0939         (i)=(start); \
0940     } \
0941 } UPRV_BLOCK_MACRO_END
0942
0943 /** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
0944 #define UTF32_SET_CHAR_LIMIT_SAFE(s, i, length) UPRV_BLOCK_MACRO_BEGIN { \
0945 } UPRV_BLOCK_MACRO_END
0946
0947 /* Formerly utf.h, part 2 --------------------------------------------------- */
0948
0949 /**
0950  * Estimate the number of code units for a string based on the number of UTF-16 code units.
0951  *
0952  * @deprecated ICU 2.4. Obsolete, see utf_old.h.
0953  */
0954 #define UTF_ARRAY_SIZE(size) UTF16_ARRAY_SIZE(size)
0955
0956 /** @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h. */
0957 #define UTF_GET_CHAR_UNSAFE(s, i, c)                 UTF16_GET_CHAR_UNSAFE(s, i, c)
0958
0959 /** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */
0960 #define UTF_GET_CHAR_SAFE(s, start, i, length, c, strict) UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict)
0961
0962
0963 /** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */
0964 #define UTF_NEXT_CHAR_UNSAFE(s, i, c)                UTF16_NEXT_CHAR_UNSAFE(s, i, c)
0965
0966 /** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */
0967 #define UTF_NEXT_CHAR_SAFE(s, i, length, c, strict)  UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict)
0968
0969
0970 /** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */
0971 #define UTF_APPEND_CHAR_UNSAFE(s, i, c)              UTF16_APPEND_CHAR_UNSAFE(s, i, c)
0972
0973 /** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */
0974 #define UTF_APPEND_CHAR_SAFE(s, i, length, c)        UTF16_APPEND_CHAR_SAFE(s, i, length, c)
0975
0976
0977 /** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */
0978 #define UTF_FWD_1_UNSAFE(s, i)                       UTF16_FWD_1_UNSAFE(s, i)
0979
0980 /** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */
0981 #define UTF_FWD_1_SAFE(s, i, length)                 UTF16_FWD_1_SAFE(s, i, length)
0982
0983
0984 /** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */
0985 #define UTF_FWD_N_UNSAFE(s, i, n)                    UTF16_FWD_N_UNSAFE(s, i, n)
0986
0987 /** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */
0988 #define UTF_FWD_N_SAFE(s, i, length, n)              UTF16_FWD_N_SAFE(s, i, length, n)
0989
0990
0991 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */
0992 #define UTF_SET_CHAR_START_UNSAFE(s, i)              UTF16_SET_CHAR_START_UNSAFE(s, i)
0993
0994 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */
0995 #define UTF_SET_CHAR_START_SAFE(s, start, i)         UTF16_SET_CHAR_START_SAFE(s, start, i)
0996
0997
0998 /** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */
0999 #define UTF_PREV_CHAR_UNSAFE(s, i, c)                UTF16_PREV_CHAR_UNSAFE(s, i, c)
1000
1001 /** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */
1002 #define UTF_PREV_CHAR_SAFE(s, start, i, c, strict)   UTF16_PREV_CHAR_SAFE(s, start, i, c, strict)
1003
1004
1005 /** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */
1006 #define UTF_BACK_1_UNSAFE(s, i)                      UTF16_BACK_1_UNSAFE(s, i)
1007
1008 /** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */
1009 #define UTF_BACK_1_SAFE(s, start, i)                 UTF16_BACK_1_SAFE(s, start, i)
1010
1011
1012 /** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */
1013 #define UTF_BACK_N_UNSAFE(s, i, n)                   UTF16_BACK_N_UNSAFE(s, i, n)
1014
1015 /** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */
1016 #define UTF_BACK_N_SAFE(s, start, i, n)              UTF16_BACK_N_SAFE(s, start, i, n)
1017
1018
1019 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */
1020 #define UTF_SET_CHAR_LIMIT_UNSAFE(s, i)              UTF16_SET_CHAR_LIMIT_UNSAFE(s, i)
1021
1022 /** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */
1023 #define UTF_SET_CHAR_LIMIT_SAFE(s, start, i, length) UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length)
1024
1025 /* Define default macros (UTF-16 "safe") ------------------------------------ */
1026
1027 /**
1028  * Does this code unit alone encode a code point (BMP, not a surrogate)?
1029  * Same as UTF16_IS_SINGLE.
1030  * @deprecated ICU 2.4. Renamed to U_IS_SINGLE and U16_IS_SINGLE, see utf_old.h.
1031  */
1032 #define UTF_IS_SINGLE(uchar) U16_IS_SINGLE(uchar)
1033
1034 /**
1035  * Is this code unit the first one of several (a lead surrogate)?
1036  * Same as UTF16_IS_LEAD.
1037  * @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h.
1038  */
1039 #define UTF_IS_LEAD(uchar) U16_IS_LEAD(uchar)
1040
1041 /**
1042  * Is this code unit one of several but not the first one (a trail surrogate)?
1043  * Same as UTF16_IS_TRAIL.
1044  * @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h.
1045  */
1046 #define UTF_IS_TRAIL(uchar) U16_IS_TRAIL(uchar)
1047
1048 /**
1049  * Does this code point require multiple code units (is it a supplementary code point)?
1050  * Same as UTF16_NEED_MULTIPLE_UCHAR.
1051  * @deprecated ICU 2.4. Use U16_LENGTH or test ((uint32_t)(c)>0xffff) instead.
1052  */
1053 #define UTF_NEED_MULTIPLE_UCHAR(c) UTF16_NEED_MULTIPLE_UCHAR(c)
1054
1055 /**
1056  * How many code units are used to encode this code point (1 or 2)?
1057  * Same as UTF16_CHAR_LENGTH.
1058  * @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h.
1059  */
1060 #define UTF_CHAR_LENGTH(c) U16_LENGTH(c)
1061
1062 /**
1063  * How many code units are used at most for any Unicode code point (2)?
1064  * Same as UTF16_MAX_CHAR_LENGTH.
1065  * @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h.
1066  */
1067 #define UTF_MAX_CHAR_LENGTH U16_MAX_LENGTH
1068
1069 /**
1070  * Set c to the code point that contains the code unit i.
1071  * i could point to the lead or the trail surrogate for the code point.
1072  * i is not modified.
1073  * Same as UTF16_GET_CHAR.
1074  * \pre 0<=i<length
1075  *
1076  * @deprecated ICU 2.4. Renamed to U16_GET, see utf_old.h.
1077  */
1078 #define UTF_GET_CHAR(s, start, i, length, c) U16_GET(s, start, i, length, c)
1079
1080 /**
1081  * Set c to the code point that starts at code unit i
1082  * and advance i to beyond the code units of this code point (post-increment).
1083  * i must point to the first code unit of a code point.
1084  * Otherwise c is set to the trail unit (surrogate) itself.
1085  * Same as UTF16_NEXT_CHAR.
1086  * \pre 0<=i<length
1087  * \post 0<i<=length
1088  *
1089  * @deprecated ICU 2.4. Renamed to U16_NEXT, see utf_old.h.
1090  */
1091 #define UTF_NEXT_CHAR(s, i, length, c) U16_NEXT(s, i, length, c)
1092
1093 /**
1094  * Append the code units of code point c to the string at index i
1095  * and advance i to beyond the new code units (post-increment).
1096  * The code units beginning at index i will be overwritten.
1097  * Same as UTF16_APPEND_CHAR.
1098  * \pre 0<=c<=0x10ffff
1099  * \pre 0<=i<length
1100  * \post 0<i<=length
1101  *
1102  * @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h.
1103  */
1104 #define UTF_APPEND_CHAR(s, i, length, c) UTF16_APPEND_CHAR_SAFE(s, i, length, c)
1105
1106 /**
1107  * Advance i to beyond the code units of the code point that begins at i.
1108  * I.e., advance i by one code point.
1109  * Same as UTF16_FWD_1.
1110  * \pre 0<=i<length
1111  * \post 0<i<=length
1112  *
1113  * @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h.
1114  */
1115 #define UTF_FWD_1(s, i, length) U16_FWD_1(s, i, length)
1116
1117 /**
1118  * Advance i to beyond the code units of the n code points where the first one begins at i.
1119  * I.e., advance i by n code points.
1120  * Same as UT16_FWD_N.
1121  * \pre 0<=i<length
1122  * \post 0<i<=length
1123  *
1124  * @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h.
1125  */
1126 #define UTF_FWD_N(s, i, length, n) U16_FWD_N(s, i, length, n)
1127
1128 /**
1129  * Take the random-access index i and adjust it so that it points to the beginning
1130  * of a code point.
1131  * The input index points to any code unit of a code point and is moved to point to
1132  * the first code unit of the same code point. i is never incremented.
1133  * In other words, if i points to a trail surrogate that is preceded by a matching
1134  * lead surrogate, then i is decremented. Otherwise it is not modified.
1135  * This can be used to start an iteration with UTF_NEXT_CHAR() from a random index.
1136  * Same as UTF16_SET_CHAR_START.
1137  * \pre start<=i<length
1138  * \post start<=i<length
1139  *
1140  * @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h.
1141  */
1142 #define UTF_SET_CHAR_START(s, start, i) U16_SET_CP_START(s, start, i)
1143
1144 /**
1145  * Set c to the code point that has code units before i
1146  * and move i backward (towards the beginning of the string)
1147  * to the first code unit of this code point (pre-increment).
1148  * i must point to the first code unit after the last unit of a code point (i==length is allowed).
1149  * Same as UTF16_PREV_CHAR.
1150  * \pre start<i<=length
1151  * \post start<=i<length
1152  *
1153  * @deprecated ICU 2.4. Renamed to U16_PREV, see utf_old.h.
1154  */
1155 #define UTF_PREV_CHAR(s, start, i, c) U16_PREV(s, start, i, c)
1156
1157 /**
1158  * Move i backward (towards the beginning of the string)
1159  * to the first code unit of the code point that has code units before i.
1160  * I.e., move i backward by one code point.
1161  * i must point to the first code unit after the last unit of a code point (i==length is allowed).
1162  * Same as UTF16_BACK_1.
1163  * \pre start<i<=length
1164  * \post start<=i<length
1165  *
1166  * @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h.
1167  */
1168 #define UTF_BACK_1(s, start, i) U16_BACK_1(s, start, i)
1169
1170 /**
1171  * Move i backward (towards the beginning of the string)
1172  * to the first code unit of the n code points that have code units before i.
1173  * I.e., move i backward by n code points.
1174  * i must point to the first code unit after the last unit of a code point (i==length is allowed).
1175  * Same as UTF16_BACK_N.
1176  * \pre start<i<=length
1177  * \post start<=i<length
1178  *
1179  * @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h.
1180  */
1181 #define UTF_BACK_N(s, start, i, n) U16_BACK_N(s, start, i, n)
1182
1183 /**
1184  * Take the random-access index i and adjust it so that it points beyond
1185  * a code point. The input index points beyond any code unit
1186  * of a code point and is moved to point beyond the last code unit of the same
1187  * code point. i is never decremented.
1188  * In other words, if i points to a trail surrogate that is preceded by a matching
1189  * lead surrogate, then i is incremented. Otherwise it is not modified.
1190  * This can be used to start an iteration with UTF_PREV_CHAR() from a random index.
1191  * Same as UTF16_SET_CHAR_LIMIT.
1192  * \pre start<i<=length
1193  * \post start<i<=length
1194  *
1195  * @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h.
1196  */
1197 #define UTF_SET_CHAR_LIMIT(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, length)
1198
1199 #endif  // !U_HIDE_DEPRECATED_API && !U_HIDE_OBSOLETE_UTF_OLD_H
1200
1201 #endif