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
  80 static inline int is_dbcs_leadbyte( const union cptable *table, unsigned char ch )
  81 {
  82     return (table->info.char_size == 2) && (table->dbcs.cp2uni_leadbytes[ch]);
  83 }
  84
  85 static inline WCHAR tolowerW( WCHAR ch )
  86 {
  87     extern const WCHAR casemap_lower[];
  88     return ch + casemap_lower[casemap_lower[ch >> 8] + (ch & 0xff)];
  89 }
  90
  91 static inline WCHAR toupperW( WCHAR ch )
  92 {
  93     extern const WCHAR casemap_upper[];
  94     return ch + casemap_upper[casemap_upper[ch >> 8] + (ch & 0xff)];
  95 }
  96
  97 /* the character type contains the C1_* flags in the low 12 bits */
  98 /* and the C2_* type in the high 4 bits */
  99 static inline unsigned short get_char_typeW( WCHAR ch )
 100 {
 101     extern const unsigned short wctype_table[];
 102     return wctype_table[wctype_table[ch >> 8] + (ch & 0xff)];
 103 }
 104
 105 inline static int iscntrlW( WCHAR wc )
 106 {
 107     return get_char_typeW(wc) & C1_CNTRL;
 108 }
 109
 110 inline static int ispunctW( WCHAR wc )
 111 {
 112     return get_char_typeW(wc) & C1_PUNCT;
 113 }
 114
 115 inline static int isspaceW( WCHAR wc )
 116 {
 117     return get_char_typeW(wc) & C1_SPACE;
 118 }
 119
 120 inline static int isdigitW( WCHAR wc )
 121 {
 122     return get_char_typeW(wc) & C1_DIGIT;
 123 }
 124
 125 inline static int isxdigitW( WCHAR wc )
 126 {
 127     return get_char_typeW(wc) & C1_XDIGIT;
 128 }
 129
 130 inline static int islowerW( WCHAR wc )
 131 {
 132     return get_char_typeW(wc) & C1_LOWER;
 133 }
 134
 135 inline static int isupperW( WCHAR wc )
 136 {
 137     return get_char_typeW(wc) & C1_UPPER;
 138 }
 139
 140 inline static int isalnumW( WCHAR wc )
 141 {
 142     return get_char_typeW(wc) & (C1_ALPHA|C1_DIGIT|C1_LOWER|C1_UPPER);
 143 }
 144
 145 inline static int isalphaW( WCHAR wc )
 146 {
 147     return get_char_typeW(wc) & (C1_ALPHA|C1_LOWER|C1_UPPER);
 148 }
 149
 150 inline static int isgraphW( WCHAR wc )
 151 {
 152     return get_char_typeW(wc) & (C1_ALPHA|C1_PUNCT|C1_DIGIT|C1_LOWER|C1_UPPER);
 153 }
 154
 155 inline static int isprintW( WCHAR wc )
 156 {
 157     return get_char_typeW(wc) & (C1_ALPHA|C1_BLANK|C1_PUNCT|C1_DIGIT|C1_LOWER|C1_UPPER);
 158 }
 159
 160 /* some useful string manipulation routines */
 161
 162 static inline unsigned int strlenW( const WCHAR *str )
 163 {
 164 #if defined(__i386__) && defined(__GNUC__)
 165     int dummy, res;
 166     __asm__ __volatile__( "cld\n\t"
 167                           "repnz\n\t"
 168                           "scasw\n\t"
 169                           "notl %0"
 170                           : "=c" (res), "=&D" (dummy)
 171                           : "0" (0xffffffff), "1" (str), "a" (0) );
 172     return res - 1;
 173 #else
 174     const WCHAR *s = str;
 175     while (*s) s++;
 176     return s - str;
 177 #endif
 178 }
 179
 180 static inline WCHAR *strcpyW( WCHAR *dst, const WCHAR *src )
 181 {
 182 #if defined(__i386__) && defined(__GNUC__)
 183     int dummy1, dummy2, dummy3;
 184     __asm__ __volatile__( "cld\n"
 185                           "1:\tlodsw\n\t"
 186                           "stosw\n\t"
 187                           "testw %%ax,%%ax\n\t"
 188                           "jne 1b"
 189                           : "=&S" (dummy1), "=&D" (dummy2), "=&a" (dummy3)
 190                           : "0" (src), "1" (dst)
 191                           : "memory" );
 192 #else
 193     WCHAR *p = dst;
 194     while ((*p++ = *src++));
 195 #endif
 196     return dst;
 197 }
 198
 199 static inline int strcmpW( const WCHAR *str1, const WCHAR *str2 )
 200 {
 201     while (*str1 && (*str1 == *str2)) { str1++; str2++; }
 202     return *str1 - *str2;
 203 }
 204
 205 static inline int strncmpW( const WCHAR *str1, const WCHAR *str2, int n )
 206 {
 207     if (n <= 0) return 0;
 208     while ((--n > 0) && *str1 && (*str1 == *str2)) { str1++; str2++; }
 209     return *str1 - *str2;
 210 }
 211
 212 static inline WCHAR *strncpyW( WCHAR *str1, const WCHAR *str2, int n )
 213 {
 214     WCHAR *ret = str1;
 215     while (n-- > 0) if (!(*str1++ = *str2++)) break;
 216     while (n-- > 0) *str1++ = 0;
 217     return ret;
 218 }
 219
 220 static inline WCHAR *strcatW( WCHAR *dst, const WCHAR *src )
 221 {
 222     strcpyW( dst + strlenW(dst), src );
 223     return dst;
 224 }
 225
 226 static inline WCHAR *strchrW( const WCHAR *str, WCHAR ch )
 227 {
 228     for ( ; *str; str++) if (*str == ch) return (WCHAR *)str;
 229     return NULL;
 230 }
 231
 232 static inline WCHAR *strrchrW( const WCHAR *str, WCHAR ch )
 233 {
 234     WCHAR *ret = NULL;
 235     for ( ; *str; str++) if (*str == ch) ret = (WCHAR *)str;
 236     return ret;
 237 }
 238
 239 static inline WCHAR *strlwrW( WCHAR *str )
 240 {
 241     WCHAR *ret = str;
 242     while ((*str = tolowerW(*str))) str++;
 243     return ret;
 244 }
 245
 246 static inline WCHAR *struprW( WCHAR *str )
 247 {
 248     WCHAR *ret = str;
 249     while ((*str = toupperW(*str))) str++;
 250     return ret;
 251 }
 252
 253 static inline long int atolW( const WCHAR *str )
 254 {
 255     return strtolW( str, (WCHAR **)0, 10 );
 256 }
 257
 258 static inline int atoiW( const WCHAR *str )
 259 {
 260     return (int)atolW( str );
 261 }
 262
 263 #endif  /* __WINE_UNICODE_H */