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 "windef.h"
  25 #include "winnls.h"
  26
  27 /* code page info common to SBCS and DBCS */
  28 struct cp_info
  29 {
  30     unsigned int          codepage;          /* codepage id */
  31     unsigned int          char_size;         /* char size (1 or 2 bytes) */
  32     WCHAR                 def_char;          /* default char value (can be double-byte) */
  33     WCHAR                 def_unicode_char;  /* default Unicode char value */
  34     const char           *name;              /* code page name */
  35 };
  36
  37 struct sbcs_table
  38 {
  39     struct cp_info        info;
  40     const WCHAR          *cp2uni;            /* code page -> Unicode map */
  41     const unsigned char  *uni2cp_low;        /* Unicode -> code page map */
  42     const unsigned short *uni2cp_high;
  43 };
  44
  45 struct dbcs_table
  46 {
  47     struct cp_info        info;
  48     const WCHAR          *cp2uni;            /* code page -> Unicode map */
  49     const unsigned char  *cp2uni_leadbytes;
  50     const unsigned short *uni2cp_low;        /* Unicode -> code page map */
  51     const unsigned short *uni2cp_high;
  52     unsigned char         lead_bytes[12];    /* lead bytes ranges */
  53 };
  54
  55 union cptable
  56 {
  57     struct cp_info    info;
  58     struct sbcs_table sbcs;
  59     struct dbcs_table dbcs;
  60 };
  61
  62 extern const union cptable *cp_get_table( unsigned int codepage );
  63 extern const union cptable *cp_enum_table( unsigned int index );
  64
  65 extern int cp_mbstowcs( const union cptable *table, int flags,
  66                         const char *src, int srclen,
  67                         WCHAR *dst, int dstlen );
  68 extern int cp_wcstombs( const union cptable *table, int flags,
  69                         const WCHAR *src, int srclen,
  70                         char *dst, int dstlen, const char *defchar, int *used );
  71 extern int utf8_wcstombs( const WCHAR *src, int srclen, char *dst, int dstlen );
  72 extern int utf8_mbstowcs( int flags, const char *src, int srclen, WCHAR *dst, int dstlen );
  73
  74 extern int strcmpiW( const WCHAR *str1, const WCHAR *str2 );
  75 extern int strncmpiW( const WCHAR *str1, const WCHAR *str2, int n );
  76 extern WCHAR *strstrW( const WCHAR *str, const WCHAR *sub );
  77 extern long int strtolW( const WCHAR *nptr, WCHAR **endptr, int base );
  78 extern unsigned long int strtoulW( const WCHAR *nptr, WCHAR **endptr, int base );
  79 extern int sprintfW( WCHAR *str, const WCHAR *format, ... );
  80 extern int snprintfW( WCHAR *str, size_t len, const WCHAR *format, ... );
  81 extern int vsprintfW( WCHAR *str, const WCHAR *format, va_list valist );
  82 extern int vsnprintfW( WCHAR *str, size_t len, const WCHAR *format, va_list valist );
  83
  84 static inline int is_dbcs_leadbyte( const union cptable *table, unsigned char ch )
  85 {
  86     return (table->info.char_size == 2) && (table->dbcs.cp2uni_leadbytes[ch]);
  87 }
  88
  89 static inline WCHAR tolowerW( WCHAR ch )
  90 {
  91     extern const WCHAR casemap_lower[];
  92     return ch + casemap_lower[casemap_lower[ch >> 8] + (ch & 0xff)];
  93 }
  94
  95 static inline WCHAR toupperW( WCHAR ch )
  96 {
  97     extern const WCHAR casemap_upper[];
  98     return ch + casemap_upper[casemap_upper[ch >> 8] + (ch & 0xff)];
  99 }
 100
 101 /* the character type contains the C1_* flags in the low 12 bits */
 102 /* and the C2_* type in the high 4 bits */
 103 static inline unsigned short get_char_typeW( WCHAR ch )
 104 {
 105     extern const unsigned short wctype_table[];
 106     return wctype_table[wctype_table[ch >> 8] + (ch & 0xff)];
 107 }
 108
 109 inline static int iscntrlW( WCHAR wc )
 110 {
 111     return get_char_typeW(wc) & C1_CNTRL;
 112 }
 113
 114 inline static int ispunctW( WCHAR wc )
 115 {
 116     return get_char_typeW(wc) & C1_PUNCT;
 117 }
 118
 119 inline static int isspaceW( WCHAR wc )
 120 {
 121     return get_char_typeW(wc) & C1_SPACE;
 122 }
 123
 124 inline static int isdigitW( WCHAR wc )
 125 {
 126     return get_char_typeW(wc) & C1_DIGIT;
 127 }
 128
 129 inline static int isxdigitW( WCHAR wc )
 130 {
 131     return get_char_typeW(wc) & C1_XDIGIT;
 132 }
 133
 134 inline static int islowerW( WCHAR wc )
 135 {
 136     return get_char_typeW(wc) & C1_LOWER;
 137 }
 138
 139 inline static int isupperW( WCHAR wc )
 140 {
 141     return get_char_typeW(wc) & C1_UPPER;
 142 }
 143
 144 inline static int isalnumW( WCHAR wc )
 145 {
 146     return get_char_typeW(wc) & (C1_ALPHA|C1_DIGIT|C1_LOWER|C1_UPPER);
 147 }
 148
 149 inline static int isalphaW( WCHAR wc )
 150 {
 151     return get_char_typeW(wc) & (C1_ALPHA|C1_LOWER|C1_UPPER);
 152 }
 153
 154 inline static int isgraphW( WCHAR wc )
 155 {
 156     return get_char_typeW(wc) & (C1_ALPHA|C1_PUNCT|C1_DIGIT|C1_LOWER|C1_UPPER);
 157 }
 158
 159 inline static int isprintW( WCHAR wc )
 160 {
 161     return get_char_typeW(wc) & (C1_ALPHA|C1_BLANK|C1_PUNCT|C1_DIGIT|C1_LOWER|C1_UPPER);
 162 }
 163
 164 /* some useful string manipulation routines */
 165
 166 static inline unsigned int strlenW( const WCHAR *str )
 167 {
 168 #if defined(__i386__) && defined(__GNUC__)
 169     int dummy, res;
 170     __asm__ __volatile__( "cld\n\t"
 171                           "repnz\n\t"
 172                           "scasw\n\t"
 173                           "notl %0"
 174                           : "=c" (res), "=&D" (dummy)
 175                           : "0" (0xffffffff), "1" (str), "a" (0) );
 176     return res - 1;
 177 #else
 178     const WCHAR *s = str;
 179     while (*s) s++;
 180     return s - str;
 181 #endif
 182 }
 183
 184 static inline WCHAR *strcpyW( WCHAR *dst, const WCHAR *src )
 185 {
 186 #if defined(__i386__) && defined(__GNUC__)
 187     int dummy1, dummy2, dummy3;
 188     __asm__ __volatile__( "cld\n"
 189                           "1:\tlodsw\n\t"
 190                           "stosw\n\t"
 191                           "testw %%ax,%%ax\n\t"
 192                           "jne 1b"
 193                           : "=&S" (dummy1), "=&D" (dummy2), "=&a" (dummy3)
 194                           : "0" (src), "1" (dst)
 195                           : "memory" );
 196 #else
 197     WCHAR *p = dst;
 198     while ((*p++ = *src++));
 199 #endif
 200     return dst;
 201 }
 202
 203 static inline int strcmpW( const WCHAR *str1, const WCHAR *str2 )
 204 {
 205     while (*str1 && (*str1 == *str2)) { str1++; str2++; }
 206     return *str1 - *str2;
 207 }
 208
 209 static inline int strncmpW( const WCHAR *str1, const WCHAR *str2, int n )
 210 {
 211     if (n <= 0) return 0;
 212     while ((--n > 0) && *str1 && (*str1 == *str2)) { str1++; str2++; }
 213     return *str1 - *str2;
 214 }
 215
 216 static inline WCHAR *strncpyW( WCHAR *str1, const WCHAR *str2, int n )
 217 {
 218     WCHAR *ret = str1;
 219     while (n-- > 0) if (!(*str1++ = *str2++)) break;
 220     while (n-- > 0) *str1++ = 0;
 221     return ret;
 222 }
 223
 224 static inline WCHAR *strcatW( WCHAR *dst, const WCHAR *src )
 225 {
 226     strcpyW( dst + strlenW(dst), src );
 227     return dst;
 228 }
 229
 230 static inline WCHAR *strchrW( const WCHAR *str, WCHAR ch )
 231 {
 232     for ( ; *str; str++) if (*str == ch) return (WCHAR *)str;
 233     return NULL;
 234 }
 235
 236 static inline WCHAR *strrchrW( const WCHAR *str, WCHAR ch )
 237 {
 238     WCHAR *ret = NULL;
 239     for ( ; *str; str++) if (*str == ch) ret = (WCHAR *)str;
 240     return ret;
 241 }
 242
 243 static inline WCHAR *strpbrkW( const WCHAR *str, const WCHAR *accept )
 244 {
 245     for ( ; *str; str++) if (strchrW( accept, *str )) return (WCHAR *)str;
 246     return NULL;
 247 }
 248
 249 static inline WCHAR *strlwrW( WCHAR *str )
 250 {
 251     WCHAR *ret = str;
 252     while ((*str = tolowerW(*str))) str++;
 253     return ret;
 254 }
 255
 256 static inline WCHAR *struprW( WCHAR *str )
 257 {
 258     WCHAR *ret = str;
 259     while ((*str = toupperW(*str))) str++;
 260     return ret;
 261 }
 262
 263 static inline long int atolW( const WCHAR *str )
 264 {
 265     return strtolW( str, (WCHAR **)0, 10 );
 266 }
 267
 268 static inline int atoiW( const WCHAR *str )
 269 {
 270     return (int)atolW( str );
 271 }
 272
 273 #endif  /* __WINE_UNICODE_H */