Main Page   Class Hierarchy   Alphabetical List   Compound List   File List   Compound Members   File Members   Search  

utf8.h

Go to the documentation of this file.
00001 /*
00002 *******************************************************************************
00003 *
00004 *   Copyright (C) 1999-2001, International Business Machines
00005 *   Corporation and others.  All Rights Reserved.
00006 *
00007 *******************************************************************************
00008 *   file name:  utf8.h
00009 *   encoding:   US-ASCII
00010 *   tab size:   8 (not used)
00011 *   indentation:4
00012 *
00013 *   created on: 1999sep13
00014 *   created by: Markus W. Scherer
00015 */
00016 
00032 /* utf.h must be included first. */
00033 #ifndef __UTF_H__
00034 # include "unicode/utf.h"
00035 #endif
00036 
00037 #ifndef __UTF8_H__
00038 #define __UTF8_H__
00039 
00040 /* internal definitions ----------------------------------------------------- */
00041 
00042 #ifdef U_UTF8_IMPL
00043 U_CAPI const uint8_t 
00044 utf8_countTrailBytes[256];
00045 #else
00046 U_CFUNC U_IMPORT const uint8_t /* U_IMPORT2? */ /*U_IMPORT*/ 
00047 utf8_countTrailBytes[256];
00048 #endif
00049 
00055 #define UTF8_COUNT_TRAIL_BYTES(leadByte) (utf8_countTrailBytes[(uint8_t)leadByte])
00056 
00057 /* use a macro here, too - there may be a simpler way with some machines */
00058 #define UTF8_MASK_LEAD_BYTE(leadByte, countTrailBytes) ((leadByte)&=(1<<(6-(countTrailBytes)))-1)
00059 
00060 U_CAPI UChar32 U_EXPORT2
00061 utf8_nextCharSafeBody(const uint8_t *s, int32_t *pi, int32_t length, UChar32 c, UBool strict, UBool *pIsError);
00062 
00063 U_CAPI int32_t U_EXPORT2
00064 utf8_appendCharSafeBody(uint8_t *s, int32_t i, int32_t length, UChar32 c);
00065 
00066 U_CAPI UChar32 U_EXPORT2
00067 utf8_prevCharSafeBody(const uint8_t *s, int32_t start, int32_t *pi, UChar32 c, UBool strict);
00068 
00069 U_CAPI int32_t U_EXPORT2
00070 utf8_back1SafeBody(const uint8_t *s, int32_t start, int32_t i);
00071 
00072 /*
00073  * For the semantics of all of these macros, see utf16.h.
00074  * The UTF-8 macros favor sequences more the shorter they are.
00075  * Sometimes, only the single-byte case is covered by a macro,
00076  * while longer sequences are handled by a function call.
00077  */
00078 
00079 /* single-code point definitions -------------------------------------------- */
00080 
00082 #define UTF8_IS_SINGLE(uchar) (((uchar)&0x80)==0)
00083 
00084 #define UTF8_IS_LEAD(uchar) ((uint8_t)((uchar)-0xc0)<0x3e)
00085 
00086 #define UTF8_IS_TRAIL(uchar) (((uchar)&0xc0)==0x80)
00087 
00089 #define UTF8_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0x7f)
00090 
00102 #if 1
00103 #   define UTF8_CHAR_LENGTH(c) \
00104         ((uint32_t)(c)<=0x7f ? 1 : \
00105             ((uint32_t)(c)<=0x7ff ? 2 : \
00106                 ((uint32_t)((c)-0x10000)>0xfffff ? 3 : 4) \
00107             ) \
00108         )
00109 #else
00110 #   define UTF8_CHAR_LENGTH(c) \
00111         ((uint32_t)(c)<=0x7f ? 1 : \
00112             ((uint32_t)(c)<=0x7ff ? 2 : \
00113                 ((uint32_t)(c)<=0xffff ? 3 : \
00114                     ((uint32_t)(c)<=0x10ffff ? 4 : \
00115                         ((uint32_t)(c)<=0x3ffffff ? 5 : \
00116                             ((uint32_t)(c)<=0x7fffffff ? 6 : 3) \
00117                         ) \
00118                     ) \
00119                 ) \
00120             ) \
00121         )
00122 #endif
00123 
00125 #define UTF8_MAX_CHAR_LENGTH 4
00126 
00128 #define UTF8_ARRAY_SIZE(size) ((5*(size))/2)
00129 
00130 #define UTF8_GET_CHAR_UNSAFE(s, i, c) { \
00131     int32_t __I=(int32_t)(i); \
00132     UTF8_SET_CHAR_START_UNSAFE(s, __I); \
00133     UTF8_NEXT_CHAR_UNSAFE(s, __I, c); \
00134 }
00135 
00136 #define UTF8_GET_CHAR_SAFE(s, start, i, length, c, strict) { \
00137     int32_t __I=(int32_t)(i); \
00138     UTF8_SET_CHAR_START_SAFE(s, start, __I); \
00139     UTF8_NEXT_CHAR_SAFE(s, __I, length, c, strict); \
00140 }
00141 
00142 /* definitions with forward iteration --------------------------------------- */
00143 
00154 #define UTF8_NEXT_CHAR_UNSAFE(s, i, c) { \
00155     (c)=(s)[(i)++]; \
00156     if((uint8_t)((c)-0xc0)<0x35) { \
00157         uint8_t __count=UTF8_COUNT_TRAIL_BYTES(c); \
00158         UTF8_MASK_LEAD_BYTE(c, __count); \
00159         switch(__count) { \
00160         /* each following branch falls through to the next one */ \
00161         case 3: \
00162             (c)=((c)<<6)|((s)[(i)++]&0x3f); \
00163         case 2: \
00164             (c)=((c)<<6)|((s)[(i)++]&0x3f); \
00165         case 1: \
00166             (c)=((c)<<6)|((s)[(i)++]&0x3f); \
00167         /* no other branches to optimize switch() */ \
00168             break; \
00169         } \
00170     } \
00171 }
00172 
00173 #define UTF8_APPEND_CHAR_UNSAFE(s, i, c) { \
00174     if((uint32_t)(c)<=0x7f) { \
00175         (s)[(i)++]=(uint8_t)(c); \
00176     } else { \
00177         if((uint32_t)(c)<=0x7ff) { \
00178             (s)[(i)++]=(uint8_t)(((c)>>6)|0xc0); \
00179         } else { \
00180             if((uint32_t)(c)<=0xffff) { \
00181                 (s)[(i)++]=(uint8_t)(((c)>>12)|0xe0); \
00182             } else { \
00183                 (s)[(i)++]=(uint8_t)(((c)>>18)|0xf0); \
00184                 (s)[(i)++]=(uint8_t)((((c)>>12)&0x3f)|0x80); \
00185             } \
00186             (s)[(i)++]=(uint8_t)((((c)>>6)&0x3f)|0x80); \
00187         } \
00188         (s)[(i)++]=(uint8_t)(((c)&0x3f)|0x80); \
00189     } \
00190 }
00191 
00192 #define UTF8_FWD_1_UNSAFE(s, i) { \
00193     (i)+=1+UTF8_COUNT_TRAIL_BYTES((s)[i]); \
00194 }
00195 
00196 #define UTF8_FWD_N_UNSAFE(s, i, n) { \
00197     int32_t __N=(n); \
00198     while(__N>0) { \
00199         UTF8_FWD_1_UNSAFE(s, i); \
00200         --__N; \
00201     } \
00202 }
00203 
00204 #define UTF8_SET_CHAR_START_UNSAFE(s, i) { \
00205     while(UTF8_IS_TRAIL((s)[i])) { --(i); } \
00206 }
00207 
00208 #define UTF8_NEXT_CHAR_SAFE(s, i, length, c, strict) { \
00209     (c)=(s)[(i)++]; \
00210     if((c)>=0x80) { \
00211         if(UTF8_IS_LEAD(c)) { \
00212             (c)=utf8_nextCharSafeBody(s, &(i), (int32_t)(length), c, strict, NULL); \
00213         } else { \
00214             (c)=UTF8_ERROR_VALUE_1; \
00215         } \
00216     } \
00217 }
00218 
00219 #define UTF8_APPEND_CHAR_SAFE(s, i, length, c) { \
00220     if((uint32_t)(c)<=0x7f) { \
00221         (s)[(i)++]=(uint8_t)(c); \
00222     } else { \
00223         (i)=utf8_appendCharSafeBody(s, (int32_t)(i), (int32_t)(length), c); \
00224     } \
00225 }
00226 
00227 #define UTF8_FWD_1_SAFE(s, i, length) { \
00228     uint8_t __b=(s)[(i)++]; \
00229     if(UTF8_IS_LEAD(__b)) { \
00230         uint8_t __count=UTF8_COUNT_TRAIL_BYTES(__b); \
00231         if((i)+__count>(length)) { \
00232             __count=(uint8_t)((length)-(i)); \
00233         } \
00234         while(__count>0 && UTF8_IS_TRAIL((s)[i])) { \
00235             ++(i); \
00236             --__count; \
00237         } \
00238     } \
00239 }
00240 
00241 #define UTF8_FWD_N_SAFE(s, i, length, n) { \
00242     int32_t __N=(n); \
00243     while(__N>0 && (i)<(length)) { \
00244         UTF8_FWD_1_SAFE(s, i, length); \
00245         --__N; \
00246     } \
00247 }
00248 
00249 #define UTF8_SET_CHAR_START_SAFE(s, start, i) { \
00250     if(UTF8_IS_TRAIL((s)[(i)])) { \
00251         (i)=utf8_back1SafeBody(s, start, (int32_t)(i)); \
00252     } \
00253 }
00254 
00255 /* definitions with backward iteration -------------------------------------- */
00256 
00257 #define UTF8_PREV_CHAR_UNSAFE(s, i, c) { \
00258     (c)=(s)[--(i)]; \
00259     if(UTF8_IS_TRAIL(c)) { \
00260         uint8_t __b, __count=1, __shift=6; \
00261 \
00262         /* c is a trail byte */ \
00263         (c)&=0x3f; \
00264         for(;;) { \
00265             __b=(s)[--(i)]; \
00266             if(__b>=0xc0) { \
00267                 UTF8_MASK_LEAD_BYTE(__b, __count); \
00268                 (c)|=(UChar32)__b<<__shift; \
00269                 break; \
00270             } else { \
00271                 (c)|=(UChar32)(__b&0x3f)<<__shift; \
00272                 ++__count; \
00273                 __shift+=6; \
00274             } \
00275         } \
00276     } \
00277 }
00278 
00279 #define UTF8_BACK_1_UNSAFE(s, i) { \
00280     while(UTF8_IS_TRAIL((s)[--(i)])) {} \
00281 }
00282 
00283 #define UTF8_BACK_N_UNSAFE(s, i, n) { \
00284     int32_t __N=(n); \
00285     while(__N>0) { \
00286         UTF8_BACK_1_UNSAFE(s, i); \
00287         --__N; \
00288     } \
00289 }
00290 
00291 #define UTF8_SET_CHAR_LIMIT_UNSAFE(s, i) { \
00292     UTF8_BACK_1_UNSAFE(s, i); \
00293     UTF8_FWD_1_UNSAFE(s, i); \
00294 }
00295 
00296 #define UTF8_PREV_CHAR_SAFE(s, start, i, c, strict) { \
00297     (c)=(s)[--(i)]; \
00298     if((c)>=0x80) { \
00299         if((c)<=0xbf) { \
00300             (c)=utf8_prevCharSafeBody(s, start, &(i), c, strict); \
00301         } else { \
00302             (c)=UTF8_ERROR_VALUE_1; \
00303         } \
00304     } \
00305 }
00306 
00307 #define UTF8_BACK_1_SAFE(s, start, i) { \
00308     if(UTF8_IS_TRAIL((s)[--(i)])) { \
00309         (i)=utf8_back1SafeBody(s, start, (int32_t)(i)); \
00310     } \
00311 }
00312 
00313 #define UTF8_BACK_N_SAFE(s, start, i, n) { \
00314     int32_t __N=(n); \
00315     while(__N>0 && (i)>(start)) { \
00316         UTF8_BACK_1_SAFE(s, start, i); \
00317         --__N; \
00318     } \
00319 }
00320 
00321 /*
00322  * Need to use UTF8_FWD_1_SAFE() because UTF8_BACK_1_SAFE()
00323  * may have started from the middle of the sequence and not checked
00324  * all trail bytes.
00325  */
00326 #define UTF8_SET_CHAR_LIMIT_SAFE(s, start, i, length) { \
00327     if((start)<(i) && (i)<(length)) { \
00328         UTF8_BACK_1_SAFE(s, start, i); \
00329         UTF8_FWD_1_SAFE(s, i, length); \
00330     } \
00331 }
00332 
00333 #endif

Generated on Thu Aug 15 14:13:34 2002 for ICU 2.2 by doxygen1.2.11.1 written by Dimitri van Heesch, © 1997-2001