dlls/msvcrt/ctype.c

   1 /*
   2  * msvcrt.dll ctype functions
   3  *
   4  * Copyright 2000 Jon Griffiths
   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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA
  19  */
  20
  21 #include "msvcrt.h"
  22 #include "winnls.h"
  23 #include "wine/unicode.h"
  24
  25 /* Some abbreviations to make the following table readable */
  26 #define _C_ MSVCRT__CONTROL
  27 #define _S_ MSVCRT__SPACE
  28 #define _P_ MSVCRT__PUNCT
  29 #define _D_ MSVCRT__DIGIT
  30 #define _H_ MSVCRT__HEX
  31 #define _U_ MSVCRT__UPPER
  32 #define _L_ MSVCRT__LOWER
  33
  34 WORD MSVCRT__ctype [257] = {
  35   0, _C_, _C_, _C_, _C_, _C_, _C_, _C_, _C_, _C_, _S_|_C_, _S_|_C_,
  36   _S_|_C_, _S_|_C_, _S_|_C_, _C_, _C_, _C_, _C_, _C_, _C_, _C_, _C_,
  37   _C_, _C_, _C_, _C_, _C_, _C_, _C_, _C_, _C_, _C_, _S_|MSVCRT__BLANK,
  38   _P_, _P_, _P_, _P_, _P_, _P_, _P_, _P_, _P_, _P_, _P_, _P_, _P_, _P_,
  39   _P_, _D_|_H_, _D_|_H_, _D_|_H_, _D_|_H_, _D_|_H_, _D_|_H_, _D_|_H_,
  40   _D_|_H_, _D_|_H_, _D_|_H_, _P_, _P_, _P_, _P_, _P_, _P_, _P_, _U_|_H_,
  41   _U_|_H_, _U_|_H_, _U_|_H_, _U_|_H_, _U_|_H_, _U_, _U_, _U_, _U_, _U_,
  42   _U_, _U_, _U_, _U_, _U_, _U_, _U_, _U_, _U_, _U_, _U_, _U_, _U_, _U_,
  43   _U_, _P_, _P_, _P_, _P_, _P_, _P_, _L_|_H_, _L_|_H_, _L_|_H_, _L_|_H_,
  44   _L_|_H_, _L_|_H_, _L_, _L_, _L_, _L_, _L_, _L_, _L_, _L_, _L_, _L_,
  45   _L_, _L_, _L_, _L_, _L_, _L_, _L_, _L_, _L_, _L_, _P_, _P_, _P_, _P_,
  46   _C_, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  47   0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  48   0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  49   0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  50   0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  51   0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
  52 };
  53
  54 /* pctype is used by macros in the Win32 headers. It must point
  55  * To a table of flags exactly like ctype. To allow locale
  56  * changes to affect ctypes (i.e. isleadbyte), we use a second table
  57  * and update its flags whenever the current locale changes.
  58  */
  59 unsigned short *MSVCRT__pctype = NULL;
  60
  61 /*********************************************************************
  62  *              __pctype_func (MSVCRT.@)
  63  */
  64 const unsigned short* CDECL MSVCRT___pctype_func(void)
  65 {
  66     return get_locinfo()->pctype;
  67 }
  68
  69 /*********************************************************************
  70  *              _isctype_l (MSVCRT.@)
  71  */
  72 int CDECL MSVCRT__isctype_l(int c, int type, MSVCRT__locale_t locale)
  73 {
  74   MSVCRT_pthreadlocinfo locinfo;
  75
  76   if(!locale)
  77     locinfo = get_locinfo();
  78   else
  79     locinfo = locale->locinfo;
  80
  81   if (c >= -1 && c <= 255)
  82     return locinfo->pctype[c] & type;
  83
  84   if (locinfo->mb_cur_max != 1 && c > 0)
  85   {
  86     /* FIXME: Is there a faster way to do this? */
  87     WORD typeInfo;
  88     char convert[3], *pconv = convert;
  89
  90     if (locinfo->pctype[(UINT)c >> 8] & MSVCRT__LEADBYTE)
  91       *pconv++ = (UINT)c >> 8;
  92     *pconv++ = c & 0xff;
  93     *pconv = 0;
  94
  95     if (GetStringTypeExA(locinfo->lc_handle[MSVCRT_LC_CTYPE],
  96                 CT_CTYPE1, convert, convert[1] ? 2 : 1, &typeInfo))
  97       return typeInfo & type;
  98   }
  99   return 0;
 100 }
 101
 102 /*********************************************************************
 103  *              _isctype (MSVCRT.@)
 104  */
 105 int CDECL MSVCRT__isctype(int c, int type)
 106 {
 107     return MSVCRT__isctype_l(c, type, NULL);
 108 }
 109
 110 /*********************************************************************
 111  *              _isalnum_l (MSVCRT.@)
 112  */
 113 int CDECL MSVCRT__isalnum_l(int c, MSVCRT__locale_t locale)
 114 {
 115   return MSVCRT__isctype_l( c, MSVCRT__ALPHA | MSVCRT__DIGIT, locale );
 116 }
 117
 118 /*********************************************************************
 119  *              isalnum (MSVCRT.@)
 120  */
 121 int CDECL MSVCRT_isalnum(int c)
 122 {
 123   return MSVCRT__isctype( c, MSVCRT__ALPHA | MSVCRT__DIGIT );
 124 }
 125
 126 /*********************************************************************
 127  *              _isalpha_l (MSVCRT.@)
 128  */
 129 int CDECL MSVCRT__isalpha_l(int c, MSVCRT__locale_t locale)
 130 {
 131   return MSVCRT__isctype_l( c, MSVCRT__ALPHA, locale );
 132 }
 133
 134 /*********************************************************************
 135  *              isalpha (MSVCRT.@)
 136  */
 137 int CDECL MSVCRT_isalpha(int c)
 138 {
 139   return MSVCRT__isctype( c, MSVCRT__ALPHA );
 140 }
 141
 142 /*********************************************************************
 143  *              _iscntrl_l (MSVCRT.@)
 144  */
 145 int CDECL MSVCRT__iscntrl_l(int c, MSVCRT__locale_t locale)
 146 {
 147   return MSVCRT__isctype_l( c, MSVCRT__CONTROL, locale );
 148 }
 149
 150 /*********************************************************************
 151  *              iscntrl (MSVCRT.@)
 152  */
 153 int CDECL MSVCRT_iscntrl(int c)
 154 {
 155   return MSVCRT__isctype( c, MSVCRT__CONTROL );
 156 }
 157
 158 /*********************************************************************
 159  *              _isdigit_l (MSVCRT.@)
 160  */
 161 int CDECL MSVCRT__isdigit_l(int c, MSVCRT__locale_t locale)
 162 {
 163   return MSVCRT__isctype_l( c, MSVCRT__DIGIT, locale );
 164 }
 165
 166 /*********************************************************************
 167  *              isdigit (MSVCRT.@)
 168  */
 169 int CDECL MSVCRT_isdigit(int c)
 170 {
 171   return MSVCRT__isctype( c, MSVCRT__DIGIT );
 172 }
 173
 174 /*********************************************************************
 175  *              _isgraph_l (MSVCRT.@)
 176  */
 177 int CDECL MSVCRT__isgraph_l(int c, MSVCRT__locale_t locale)
 178 {
 179   return MSVCRT__isctype_l( c, MSVCRT__ALPHA | MSVCRT__DIGIT | MSVCRT__PUNCT, locale );
 180 }
 181
 182 /*********************************************************************
 183  *              isgraph (MSVCRT.@)
 184  */
 185 int CDECL MSVCRT_isgraph(int c)
 186 {
 187   return MSVCRT__isctype( c, MSVCRT__ALPHA | MSVCRT__DIGIT | MSVCRT__PUNCT );
 188 }
 189
 190 /*********************************************************************
 191  *              _isleadbyte_l (MSVCRT.@)
 192  */
 193 int CDECL MSVCRT__isleadbyte_l(int c, MSVCRT__locale_t locale)
 194 {
 195   return MSVCRT__isctype_l( c, MSVCRT__LEADBYTE, locale );
 196 }
 197
 198 /*********************************************************************
 199  *              isleadbyte (MSVCRT.@)
 200  */
 201 int CDECL MSVCRT_isleadbyte(int c)
 202 {
 203   return MSVCRT__isctype( c, MSVCRT__LEADBYTE );
 204 }
 205
 206 /*********************************************************************
 207  *              _islower_l (MSVCRT.@)
 208  */
 209 int CDECL MSVCRT__islower_l(int c, MSVCRT__locale_t locale)
 210 {
 211   return MSVCRT__isctype_l( c, MSVCRT__LOWER, locale );
 212 }
 213
 214 /*********************************************************************
 215  *              islower (MSVCRT.@)
 216  */
 217 int CDECL MSVCRT_islower(int c)
 218 {
 219   return MSVCRT__isctype( c, MSVCRT__LOWER );
 220 }
 221
 222 /*********************************************************************
 223  *              _isprint_l (MSVCRT.@)
 224  */
 225 int CDECL MSVCRT__isprint_l(int c, MSVCRT__locale_t locale)
 226 {
 227   return MSVCRT__isctype_l( c, MSVCRT__ALPHA | MSVCRT__DIGIT | MSVCRT__BLANK | MSVCRT__PUNCT, locale );
 228 }
 229
 230 /*********************************************************************
 231  *              isprint (MSVCRT.@)
 232  */
 233 int CDECL MSVCRT_isprint(int c)
 234 {
 235   return MSVCRT__isctype( c, MSVCRT__ALPHA | MSVCRT__DIGIT | MSVCRT__BLANK | MSVCRT__PUNCT );
 236 }
 237
 238 /*********************************************************************
 239  *              ispunct (MSVCRT.@)
 240  */
 241 int CDECL MSVCRT_ispunct(int c)
 242 {
 243   return MSVCRT__isctype( c, MSVCRT__PUNCT );
 244 }
 245
 246 /*********************************************************************
 247  *              _isspace_l (MSVCRT.@)
 248  */
 249 int CDECL MSVCRT__isspace_l(int c, MSVCRT__locale_t locale)
 250 {
 251   return MSVCRT__isctype_l( c, MSVCRT__SPACE, locale );
 252 }
 253
 254 /*********************************************************************
 255  *              isspace (MSVCRT.@)
 256  */
 257 int CDECL MSVCRT_isspace(int c)
 258 {
 259   return MSVCRT__isctype( c, MSVCRT__SPACE );
 260 }
 261
 262 /*********************************************************************
 263  *              _isupper_l (MSVCRT.@)
 264  */
 265 int CDECL MSVCRT__isupper_l(int c, MSVCRT__locale_t locale)
 266 {
 267   return MSVCRT__isctype_l( c, MSVCRT__UPPER, locale );
 268 }
 269
 270 /*********************************************************************
 271  *              isupper (MSVCRT.@)
 272  */
 273 int CDECL MSVCRT_isupper(int c)
 274 {
 275   return MSVCRT__isctype( c, MSVCRT__UPPER );
 276 }
 277
 278 /*********************************************************************
 279  *              _isxdigit_l (MSVCRT.@)
 280  */
 281 int CDECL MSVCRT__isxdigit_l(int c, MSVCRT__locale_t locale)
 282 {
 283   return MSVCRT__isctype_l( c, MSVCRT__HEX, locale );
 284 }
 285
 286 /*********************************************************************
 287  *              isxdigit (MSVCRT.@)
 288  */
 289 int CDECL MSVCRT_isxdigit(int c)
 290 {
 291   return MSVCRT__isctype( c, MSVCRT__HEX );
 292 }
 293
 294 /*********************************************************************
 295  *              __isascii (MSVCRT.@)
 296  */
 297 int CDECL MSVCRT___isascii(int c)
 298 {
 299   return isascii((unsigned)c);
 300 }
 301
 302 /*********************************************************************
 303  *              __toascii (MSVCRT.@)
 304  */
 305 int CDECL MSVCRT___toascii(int c)
 306 {
 307   return (unsigned)c & 0x7f;
 308 }
 309
 310 /*********************************************************************
 311  *              iswascii (MSVCRT.@)
 312  *
 313  */
 314 int CDECL MSVCRT_iswascii(MSVCRT_wchar_t c)
 315 {
 316   return ((unsigned)c < 0x80);
 317 }
 318
 319 /*********************************************************************
 320  *              __iscsym (MSVCRT.@)
 321  */
 322 int CDECL MSVCRT___iscsym(int c)
 323 {
 324   return (c < 127 && (isalnum(c) || c == '_'));
 325 }
 326
 327 /*********************************************************************
 328  *              __iscsymf (MSVCRT.@)
 329  */
 330 int CDECL MSVCRT___iscsymf(int c)
 331 {
 332   return (c < 127 && (isalpha(c) || c == '_'));
 333 }
 334
 335 /*********************************************************************
 336  *              _toupper_l (MSVCRT.@)
 337  */
 338 int CDECL MSVCRT__toupper_l(int c, MSVCRT__locale_t locale)
 339 {
 340     MSVCRT_pthreadlocinfo locinfo;
 341
 342     if(!locale)
 343         locinfo = get_locinfo();
 344     else
 345         locinfo = locale->locinfo;
 346
 347     if(c < 256)
 348         return locinfo->pcumap[(unsigned char)c];
 349
 350     if(locinfo->pctype[(c>>8)&255] & MSVCRT__LEADBYTE)
 351     {
 352         WCHAR wide, upper;
 353         char str[2], *p = str;
 354         *p++ = (c>>8) & 255;
 355         *p++ = c & 255;
 356
 357         if(!MultiByteToWideChar(locinfo->lc_codepage,
 358                     MB_ERR_INVALID_CHARS, str, 2, &wide, 1))
 359             return c;
 360
 361         upper = toupperW(wide);
 362         if(upper == wide)
 363             return c;
 364
 365         WideCharToMultiByte(locinfo->lc_codepage, 0,
 366                 &upper, 1, str, 2, NULL, NULL);
 367
 368         return str[0] + (str[1]<<8);
 369     }
 370
 371     return c;
 372 }
 373
 374 /*********************************************************************
 375  *              toupper (MSVCRT.@)
 376  */
 377 int CDECL MSVCRT_toupper(int c)
 378 {
 379     return MSVCRT__toupper_l(c, NULL);
 380 }
 381
 382 /*********************************************************************
 383  *              _toupper (MSVCRT.@)
 384  */
 385 int CDECL MSVCRT__toupper(int c)
 386 {
 387     return c - 0x20;  /* sic */
 388 }
 389
 390 /*********************************************************************
 391  *              _tolower_l (MSVCRT.@)
 392  */
 393 int CDECL MSVCRT__tolower_l(int c, MSVCRT__locale_t locale)
 394 {
 395     MSVCRT_pthreadlocinfo locinfo;
 396
 397     if(!locale)
 398         locinfo = get_locinfo();
 399     else
 400         locinfo = locale->locinfo;
 401
 402     if(c < 256)
 403         return locinfo->pclmap[(unsigned char)c];
 404
 405     if(locinfo->pctype[(c>>8)&255] & MSVCRT__LEADBYTE)
 406     {
 407         WCHAR wide, upper;
 408         char str[2], *p = str;
 409         *p++ = (c>>8) & 255;
 410         *p++ = c & 255;
 411
 412         if(!MultiByteToWideChar(locinfo->lc_codepage,
 413                     MB_ERR_INVALID_CHARS, str, 2, &wide, 1))
 414             return c;
 415
 416         upper = tolowerW(wide);
 417         if(upper == wide)
 418             return c;
 419
 420         WideCharToMultiByte(locinfo->lc_codepage, 0,
 421                 &upper, 1, str, 2, NULL, NULL);
 422
 423         return str[0] + (str[1]<<8);
 424     }
 425
 426     return c;
 427 }
 428
 429 /*********************************************************************
 430  *              tolower (MSVCRT.@)
 431  */
 432 int CDECL MSVCRT_tolower(int c)
 433 {
 434         return MSVCRT__tolower_l(c, NULL);
 435 }
 436
 437 /*********************************************************************
 438  *              _tolower (MSVCRT.@)
 439  */
 440 int CDECL MSVCRT__tolower(int c)
 441 {
 442     return c + 0x20;  /* sic */
 443 }