include/wine/unicode.h

   1 /*
   2  * Wine internal Unicode definitions
   3  *
   4  * Copyright 2000 Alexandre Julliard
   5  *
   6  * This library is free software; you can redistribute it and/or
   7  * modify it under the terms of the GNU Lesser General Public
   8  * License as published by the Free Software Foundation; either
   9  * version 2.1 of the License, or (at your option) any later version.
  10  *
  11  * This library is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14  * Lesser General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU Lesser General Public
  17  * License along with this library; if not, write to the Free Software
  18  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  19  */
  20
  21 #ifndef __WINE_UNICODE_H
  22 #define __WINE_UNICODE_H
  23
  24 #include <stdarg.h>
  25
  26 #include <windef.h>
  27 #include <winbase.h>
  28 #include <winnls.h>
  29
  30 #ifndef WINE_UNICODE_API
  31 #define WINE_UNICODE_API DECLSPEC_IMPORT
  32 #endif
  33
  34 /* code page info common to SBCS and DBCS */
  35 struct cp_info
  36 {
  37     unsigned int          codepage;          /* codepage id */
  38     unsigned int          char_size;         /* char size (1 or 2 bytes) */
  39     WCHAR                 def_char;          /* default char value (can be double-byte) */
  40     WCHAR                 def_unicode_char;  /* default Unicode char value */
  41     const char           *name;              /* code page name */
  42 };
  43
  44 struct sbcs_table
  45 {
  46     struct cp_info        info;
  47     const WCHAR          *cp2uni;            /* code page -> Unicode map */
  48     const unsigned char  *uni2cp_low;        /* Unicode -> code page map */
  49     const unsigned short *uni2cp_high;
  50 };
  51
  52 struct dbcs_table
  53 {
  54     struct cp_info        info;
  55     const WCHAR          *cp2uni;            /* code page -> Unicode map */
  56     const unsigned char  *cp2uni_leadbytes;
  57     const unsigned short *uni2cp_low;        /* Unicode -> code page map */
  58     const unsigned short *uni2cp_high;
  59     unsigned char         lead_bytes[12];    /* lead bytes ranges */
  60 };
  61
  62 union cptable
  63 {
  64     struct cp_info    info;
  65     struct sbcs_table sbcs;
  66     struct dbcs_table dbcs;
  67 };
  68
  69 extern const union cptable *wine_cp_get_table( unsigned int codepage );
  70 extern const union cptable *wine_cp_enum_table( unsigned int index );
  71
  72 extern int wine_cp_mbstowcs( const union cptable *table, int flags,
  73                              const char *src, int srclen,
  74                              WCHAR *dst, int dstlen );
  75 extern int wine_cp_wcstombs( const union cptable *table, int flags,
  76                              const WCHAR *src, int srclen,
  77                              char *dst, int dstlen, const char *defchar, int *used );
  78 extern int wine_cpsymbol_mbstowcs( const char *src, int srclen, WCHAR *dst, int dstlen );
  79 extern int wine_cpsymbol_wcstombs( const WCHAR *src, int srclen, char *dst, int dstlen );
  80 extern int wine_utf8_mbstowcs( int flags, const char *src, int srclen, WCHAR *dst, int dstlen );
  81 extern int wine_utf8_wcstombs( const WCHAR *src, int srclen, char *dst, int dstlen );
  82
  83 extern int wine_compare_string( int flags, const WCHAR *str1, int len1, const WCHAR *str2, int len2 );
  84 extern int wine_get_sortkey( int flags, const WCHAR *src, int srclen, char *dst, int dstlen );
  85 extern int wine_fold_string( int flags, const WCHAR *src, int srclen , WCHAR *dst, int dstlen );
  86
  87 extern int strcmpiW( const WCHAR *str1, const WCHAR *str2 );
  88 extern int strncmpiW( const WCHAR *str1, const WCHAR *str2, int n );
  89 extern int memicmpW( const WCHAR *str1, const WCHAR *str2, int n );
  90 extern WCHAR *strstrW( const WCHAR *str, const WCHAR *sub );
  91 extern long int strtolW( const WCHAR *nptr, WCHAR **endptr, int base );
  92 extern unsigned long int strtoulW( const WCHAR *nptr, WCHAR **endptr, int base );
  93 extern int sprintfW( WCHAR *str, const WCHAR *format, ... );
  94 extern int snprintfW( WCHAR *str, size_t len, const WCHAR *format, ... );
  95 extern int vsprintfW( WCHAR *str, const WCHAR *format, va_list valist );
  96 extern int vsnprintfW( WCHAR *str, size_t len, const WCHAR *format, va_list valist );
  97
  98 static inline int is_dbcs_leadbyte( const union cptable *table, unsigned char ch )
  99 {
 100     return (table->info.char_size == 2) && (table->dbcs.cp2uni_leadbytes[ch]);
 101 }
 102
 103 static inline WCHAR tolowerW( WCHAR ch )
 104 {
 105     extern WINE_UNICODE_API const WCHAR wine_casemap_lower[];
 106     return ch + wine_casemap_lower[wine_casemap_lower[ch >> 8] + (ch & 0xff)];
 107 }
 108
 109 static inline WCHAR toupperW( WCHAR ch )
 110 {
 111     extern WINE_UNICODE_API const WCHAR wine_casemap_upper[];
 112     return ch + wine_casemap_upper[wine_casemap_upper[ch >> 8] + (ch & 0xff)];
 113 }
 114
 115 /* the character type contains the C1_* flags in the low 12 bits */
 116 /* and the C2_* type in the high 4 bits */
 117 static inline unsigned short get_char_typeW( WCHAR ch )
 118 {
 119     extern WINE_UNICODE_API const unsigned short wine_wctype_table[];
 120     return wine_wctype_table[wine_wctype_table[ch >> 8] + (ch & 0xff)];
 121 }
 122
 123 inline static int iscntrlW( WCHAR wc )
 124 {
 125     return get_char_typeW(wc) & C1_CNTRL;
 126 }
 127
 128 inline static int ispunctW( WCHAR wc )
 129 {
 130     return get_char_typeW(wc) & C1_PUNCT;
 131 }
 132
 133 inline static int isspaceW( WCHAR wc )
 134 {
 135     return get_char_typeW(wc) & C1_SPACE;
 136 }
 137
 138 inline static int isdigitW( WCHAR wc )
 139 {
 140     return get_char_typeW(wc) & C1_DIGIT;
 141 }
 142
 143 inline static int isxdigitW( WCHAR wc )
 144 {
 145     return get_char_typeW(wc) & C1_XDIGIT;
 146 }
 147
 148 inline static int islowerW( WCHAR wc )
 149 {
 150     return get_char_typeW(wc) & C1_LOWER;
 151 }
 152
 153 inline static int isupperW( WCHAR wc )
 154 {
 155     return get_char_typeW(wc) & C1_UPPER;
 156 }
 157
 158 inline static int isalnumW( WCHAR wc )
 159 {
 160     return get_char_typeW(wc) & (C1_ALPHA|C1_DIGIT|C1_LOWER|C1_UPPER);
 161 }
 162
 163 inline static int isalphaW( WCHAR wc )
 164 {
 165     return get_char_typeW(wc) & (C1_ALPHA|C1_LOWER|C1_UPPER);
 166 }
 167
 168 inline static int isgraphW( WCHAR wc )
 169 {
 170     return get_char_typeW(wc) & (C1_ALPHA|C1_PUNCT|C1_DIGIT|C1_LOWER|C1_UPPER);
 171 }
 172
 173 inline static int isprintW( WCHAR wc )
 174 {
 175     return get_char_typeW(wc) & (C1_ALPHA|C1_BLANK|C1_PUNCT|C1_DIGIT|C1_LOWER|C1_UPPER);
 176 }
 177
 178 /* some useful string manipulation routines */
 179
 180 static inline unsigned int strlenW( const WCHAR *str )
 181 {
 182 #if defined(__i386__) && defined(__GNUC__)
 183     int dummy, res;
 184     __asm__ __volatile__( "cld\n\t"
 185                           "repnz\n\t"
 186                           "scasw\n\t"
 187                           "notl %0"
 188                           : "=c" (res), "=&D" (dummy)
 189                           : "0" (0xffffffff), "1" (str), "a" (0) );
 190     return res - 1;
 191 #else
 192     const WCHAR *s = str;
 193     while (*s) s++;
 194     return s - str;
 195 #endif
 196 }
 197
 198 static inline WCHAR *strcpyW( WCHAR *dst, const WCHAR *src )
 199 {
 200 #if defined(__i386__) && defined(__GNUC__)
 201     int dummy1, dummy2, dummy3;
 202     __asm__ __volatile__( "cld\n"
 203                           "1:\tlodsw\n\t"
 204                           "stosw\n\t"
 205                           "testw %%ax,%%ax\n\t"
 206                           "jne 1b"
 207                           : "=&S" (dummy1), "=&D" (dummy2), "=&a" (dummy3)
 208                           : "0" (src), "1" (dst)
 209                           : "memory" );
 210 #else
 211     WCHAR *p = dst;
 212     while ((*p++ = *src++));
 213 #endif
 214     return dst;
 215 }
 216
 217 static inline int strcmpW( const WCHAR *str1, const WCHAR *str2 )
 218 {
 219     while (*str1 && (*str1 == *str2)) { str1++; str2++; }
 220     return *str1 - *str2;
 221 }
 222
 223 static inline int strncmpW( const WCHAR *str1, const WCHAR *str2, int n )
 224 {
 225     if (n <= 0) return 0;
 226     while ((--n > 0) && *str1 && (*str1 == *str2)) { str1++; str2++; }
 227     return *str1 - *str2;
 228 }
 229
 230 static inline WCHAR *strncpyW( WCHAR *str1, const WCHAR *str2, int n )
 231 {
 232     WCHAR *ret = str1;
 233     while (n-- > 0) if (!(*str1++ = *str2++)) break;
 234     while (n-- > 0) *str1++ = 0;
 235     return ret;
 236 }
 237
 238 static inline WCHAR *strcatW( WCHAR *dst, const WCHAR *src )
 239 {
 240     strcpyW( dst + strlenW(dst), src );
 241     return dst;
 242 }
 243
 244 static inline WCHAR *strchrW( const WCHAR *str, WCHAR ch )
 245 {
 246     for ( ; *str; str++) if (*str == ch) return (WCHAR *)str;
 247     return NULL;
 248 }
 249
 250 static inline WCHAR *strrchrW( const WCHAR *str, WCHAR ch )
 251 {
 252     WCHAR *ret = NULL;
 253     for ( ; *str; str++) if (*str == ch) ret = (WCHAR *)str;
 254     return ret;
 255 }
 256
 257 static inline WCHAR *strpbrkW( const WCHAR *str, const WCHAR *accept )
 258 {
 259     for ( ; *str; str++) if (strchrW( accept, *str )) return (WCHAR *)str;
 260     return NULL;
 261 }
 262
 263 static inline size_t strspnW( const WCHAR *str, const WCHAR *accept )
 264 {
 265     const WCHAR *ptr;
 266     for (ptr = str; *ptr; ptr++) if (!strchrW( accept, *ptr )) break;
 267     return ptr - str;
 268 }
 269
 270 static inline size_t strcspnW( const WCHAR *str, const WCHAR *reject )
 271 {
 272     const WCHAR *ptr;
 273     for (ptr = str; *ptr; ptr++) if (strchrW( reject, *ptr )) break;
 274     return ptr - str;
 275 }
 276
 277 static inline WCHAR *strlwrW( WCHAR *str )
 278 {
 279     WCHAR *ret = str;
 280     while ((*str = tolowerW(*str))) str++;
 281     return ret;
 282 }
 283
 284 static inline WCHAR *struprW( WCHAR *str )
 285 {
 286     WCHAR *ret = str;
 287     while ((*str = toupperW(*str))) str++;
 288     return ret;
 289 }
 290
 291 static inline WCHAR *memchrW( const WCHAR *ptr, WCHAR ch, size_t n )
 292 {
 293     const WCHAR *end;
 294     for (end = ptr + n; ptr < end; ptr++) if (*ptr == ch) return (WCHAR *)ptr;
 295     return NULL;
 296 }
 297
 298 static inline WCHAR *memrchrW( const WCHAR *ptr, WCHAR ch, size_t n )
 299 {
 300     const WCHAR *end, *ret = NULL;
 301     for (end = ptr + n; ptr < end; ptr++) if (*ptr == ch) ret = ptr;
 302     return (WCHAR *)ret;
 303 }
 304
 305 static inline long int atolW( const WCHAR *str )
 306 {
 307     return strtolW( str, (WCHAR **)0, 10 );
 308 }
 309
 310 static inline int atoiW( const WCHAR *str )
 311 {
 312     return (int)atolW( str );
 313 }
 314
 315 #endif  /* __WINE_UNICODE_H */