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 /* code page info common to SBCS and DBCS */
  31 struct cp_info
  32 {
  33     unsigned int          codepage;          /* codepage id */
  34     unsigned int          char_size;         /* char size (1 or 2 bytes) */
  35     WCHAR                 def_char;          /* default char value (can be double-byte) */
  36     WCHAR                 def_unicode_char;  /* default Unicode char value */
  37     const char           *name;              /* code page name */
  38 };
  39
  40 struct sbcs_table
  41 {
  42     struct cp_info        info;
  43     const WCHAR          *cp2uni;            /* code page -> Unicode map */
  44     const unsigned char  *uni2cp_low;        /* Unicode -> code page map */
  45     const unsigned short *uni2cp_high;
  46 };
  47
  48 struct dbcs_table
  49 {
  50     struct cp_info        info;
  51     const WCHAR          *cp2uni;            /* code page -> Unicode map */
  52     const unsigned char  *cp2uni_leadbytes;
  53     const unsigned short *uni2cp_low;        /* Unicode -> code page map */
  54     const unsigned short *uni2cp_high;
  55     unsigned char         lead_bytes[12];    /* lead bytes ranges */
  56 };
  57
  58 union cptable
  59 {
  60     struct cp_info    info;
  61     struct sbcs_table sbcs;
  62     struct dbcs_table dbcs;
  63 };
  64
  65 extern const union cptable *wine_cp_get_table( unsigned int codepage );
  66 extern const union cptable *wine_cp_enum_table( unsigned int index );
  67
  68 extern int wine_cp_mbstowcs( const union cptable *table, int flags,
  69                              const char *src, int srclen,
  70                              WCHAR *dst, int dstlen );
  71 extern int wine_cp_wcstombs( const union cptable *table, int flags,
  72                              const WCHAR *src, int srclen,
  73                              char *dst, int dstlen, const char *defchar, int *used );
  74 extern int wine_utf8_wcstombs( const WCHAR *src, int srclen, char *dst, int dstlen );
  75 extern int wine_utf8_mbstowcs( int flags, const char *src, int srclen, WCHAR *dst, int dstlen );
  76
  77 extern int wine_get_sortkey( int flags, const WCHAR *src, int srclen, char *dst, int dstlen );
  78
  79 extern int strcmpiW( const WCHAR *str1, const WCHAR *str2 );
  80 extern int strncmpiW( const WCHAR *str1, const WCHAR *str2, int n );
  81 extern WCHAR *strstrW( const WCHAR *str, const WCHAR *sub );
  82 extern long int strtolW( const WCHAR *nptr, WCHAR **endptr, int base );
  83 extern unsigned long int strtoulW( const WCHAR *nptr, WCHAR **endptr, int base );
  84 extern int sprintfW( WCHAR *str, const WCHAR *format, ... );
  85 extern int snprintfW( WCHAR *str, size_t len, const WCHAR *format, ... );
  86 extern int vsprintfW( WCHAR *str, const WCHAR *format, va_list valist );
  87 extern int vsnprintfW( WCHAR *str, size_t len, const WCHAR *format, va_list valist );
  88
  89 static inline int is_dbcs_leadbyte( const union cptable *table, unsigned char ch )
  90 {
  91     return (table->info.char_size == 2) && (table->dbcs.cp2uni_leadbytes[ch]);
  92 }
  93
  94 static inline WCHAR tolowerW( WCHAR ch )
  95 {
  96     extern const WCHAR wine_casemap_lower[];
  97     return ch + wine_casemap_lower[wine_casemap_lower[ch >> 8] + (ch & 0xff)];
  98 }
  99
 100 static inline WCHAR toupperW( WCHAR ch )
 101 {
 102     extern const WCHAR wine_casemap_upper[];
 103     return ch + wine_casemap_upper[wine_casemap_upper[ch >> 8] + (ch & 0xff)];
 104 }
 105
 106 /* the character type contains the C1_* flags in the low 12 bits */
 107 /* and the C2_* type in the high 4 bits */
 108 static inline unsigned short get_char_typeW( WCHAR ch )
 109 {
 110     extern const unsigned short wine_wctype_table[];
 111     return wine_wctype_table[wine_wctype_table[ch >> 8] + (ch & 0xff)];
 112 }
 113
 114 inline static int iscntrlW( WCHAR wc )
 115 {
 116     return get_char_typeW(wc) & C1_CNTRL;
 117 }
 118
 119 inline static int ispunctW( WCHAR wc )
 120 {
 121     return get_char_typeW(wc) & C1_PUNCT;
 122 }
 123
 124 inline static int isspaceW( WCHAR wc )
 125 {
 126     return get_char_typeW(wc) & C1_SPACE;
 127 }
 128
 129 inline static int isdigitW( WCHAR wc )
 130 {
 131     return get_char_typeW(wc) & C1_DIGIT;
 132 }
 133
 134 inline static int isxdigitW( WCHAR wc )
 135 {
 136     return get_char_typeW(wc) & C1_XDIGIT;
 137 }
 138
 139 inline static int islowerW( WCHAR wc )
 140 {
 141     return get_char_typeW(wc) & C1_LOWER;
 142 }
 143
 144 inline static int isupperW( WCHAR wc )
 145 {
 146     return get_char_typeW(wc) & C1_UPPER;
 147 }
 148
 149 inline static int isalnumW( WCHAR wc )
 150 {
 151     return get_char_typeW(wc) & (C1_ALPHA|C1_DIGIT|C1_LOWER|C1_UPPER);
 152 }
 153
 154 inline static int isalphaW( WCHAR wc )
 155 {
 156     return get_char_typeW(wc) & (C1_ALPHA|C1_LOWER|C1_UPPER);
 157 }
 158
 159 inline static int isgraphW( WCHAR wc )
 160 {
 161     return get_char_typeW(wc) & (C1_ALPHA|C1_PUNCT|C1_DIGIT|C1_LOWER|C1_UPPER);
 162 }
 163
 164 inline static int isprintW( WCHAR wc )
 165 {
 166     return get_char_typeW(wc) & (C1_ALPHA|C1_BLANK|C1_PUNCT|C1_DIGIT|C1_LOWER|C1_UPPER);
 167 }
 168
 169 /* some useful string manipulation routines */
 170
 171 static inline unsigned int strlenW( const WCHAR *str )
 172 {
 173 #if defined(__i386__) && defined(__GNUC__)
 174     int dummy, res;
 175     __asm__ __volatile__( "cld\n\t"
 176                           "repnz\n\t"
 177                           "scasw\n\t"
 178                           "notl %0"
 179                           : "=c" (res), "=&D" (dummy)
 180                           : "0" (0xffffffff), "1" (str), "a" (0) );
 181     return res - 1;
 182 #else
 183     const WCHAR *s = str;
 184     while (*s) s++;
 185     return s - str;
 186 #endif
 187 }
 188
 189 static inline WCHAR *strcpyW( WCHAR *dst, const WCHAR *src )
 190 {
 191 #if defined(__i386__) && defined(__GNUC__)
 192     int dummy1, dummy2, dummy3;
 193     __asm__ __volatile__( "cld\n"
 194                           "1:\tlodsw\n\t"
 195                           "stosw\n\t"
 196                           "testw %%ax,%%ax\n\t"
 197                           "jne 1b"
 198                           : "=&S" (dummy1), "=&D" (dummy2), "=&a" (dummy3)
 199                           : "0" (src), "1" (dst)
 200                           : "memory" );
 201 #else
 202     WCHAR *p = dst;
 203     while ((*p++ = *src++));
 204 #endif
 205     return dst;
 206 }
 207
 208 static inline int strcmpW( const WCHAR *str1, const WCHAR *str2 )
 209 {
 210     while (*str1 && (*str1 == *str2)) { str1++; str2++; }
 211     return *str1 - *str2;
 212 }
 213
 214 static inline int strncmpW( const WCHAR *str1, const WCHAR *str2, int n )
 215 {
 216     if (n <= 0) return 0;
 217     while ((--n > 0) && *str1 && (*str1 == *str2)) { str1++; str2++; }
 218     return *str1 - *str2;
 219 }
 220
 221 static inline WCHAR *strncpyW( WCHAR *str1, const WCHAR *str2, int n )
 222 {
 223     WCHAR *ret = str1;
 224     while (n-- > 0) if (!(*str1++ = *str2++)) break;
 225     while (n-- > 0) *str1++ = 0;
 226     return ret;
 227 }
 228
 229 static inline WCHAR *strcatW( WCHAR *dst, const WCHAR *src )
 230 {
 231     strcpyW( dst + strlenW(dst), src );
 232     return dst;
 233 }
 234
 235 static inline WCHAR *strchrW( const WCHAR *str, WCHAR ch )
 236 {
 237     for ( ; *str; str++) if (*str == ch) return (WCHAR *)str;
 238     return NULL;
 239 }
 240
 241 static inline WCHAR *strrchrW( const WCHAR *str, WCHAR ch )
 242 {
 243     WCHAR *ret = NULL;
 244     for ( ; *str; str++) if (*str == ch) ret = (WCHAR *)str;
 245     return ret;
 246 }
 247
 248 static inline WCHAR *strpbrkW( const WCHAR *str, const WCHAR *accept )
 249 {
 250     for ( ; *str; str++) if (strchrW( accept, *str )) return (WCHAR *)str;
 251     return NULL;
 252 }
 253
 254 static inline WCHAR *strlwrW( WCHAR *str )
 255 {
 256     WCHAR *ret = str;
 257     while ((*str = tolowerW(*str))) str++;
 258     return ret;
 259 }
 260
 261 static inline WCHAR *struprW( WCHAR *str )
 262 {
 263     WCHAR *ret = str;
 264     while ((*str = toupperW(*str))) str++;
 265     return ret;
 266 }
 267
 268 static inline long int atolW( const WCHAR *str )
 269 {
 270     return strtolW( str, (WCHAR **)0, 10 );
 271 }
 272
 273 static inline int atoiW( const WCHAR *str )
 274 {
 275     return (int)atolW( str );
 276 }
 277
 278 #endif  /* __WINE_UNICODE_H */