src/common/encconv.cpp

   1 /////////////////////////////////////////////////////////////////////////////
   2 // Name:        encconv.cpp
   3 // Purpose:     wxEncodingConverter class for converting between different
   4 //              font encodings
   5 // Author:      Vaclav Slavik
   6 // Copyright:   (c) 1999 Vaclav Slavik
   7 // Licence:     wxWindows Licence
   8 /////////////////////////////////////////////////////////////////////////////
   9
  10 #if defined(__GNUG__) && !defined(NO_GCC_PRAGMA)
  11 #pragma implementation "encconv.h"
  12 #endif
  13
  14 // For compilers that support precompilation, includes "wx.h".
  15 #include "wx/wxprec.h"
  16
  17 #ifdef __BORLANDC__
  18   #pragma hdrstop
  19 #endif
  20
  21 #if wxUSE_FONTMAP
  22
  23 #include "wx/encconv.h"
  24
  25 #include <stdlib.h>
  26
  27 // conversion tables, generated by scripts in $(WXWIN)/misc/unictabl:
  28 #ifdef __BORLANDC__
  29 #include "../common/unictabl.inc"
  30 #else
  31 #include "unictabl.inc"
  32 #endif
  33
  34 #if wxUSE_WCHAR_T
  35 typedef wchar_t tchar;
  36 #else
  37 typedef char tchar;
  38 #endif
  39
  40 #ifdef __WXWINCE__
  41 #undef LINKAGEMODE
  42 #define LINKAGEMODE __cdecl
  43 #endif
  44
  45 #ifdef __WXMAC__
  46
  47 #include "ATSUnicode.h"
  48 #include "TextCommon.h"
  49 #include "TextEncodingConverter.h"
  50
  51 #include  "wx/mac/private.h"  // includes mac headers
  52
  53 typedef struct {
  54     wxFontEncoding enc ;
  55     TextEncodingBase mac ;
  56 } MacCP ;
  57
  58 MacCP gMacCodePages[] =
  59 {
  60     wxFONTENCODING_MACROMAN, kTextEncodingMacRoman,
  61     wxFONTENCODING_MACCENTRALEUR, kTextEncodingMacCentralEurRoman,
  62     wxFONTENCODING_MACHEBREW, kTextEncodingMacHebrew,
  63     wxFONTENCODING_MACGREEK, kTextEncodingMacGreek,
  64     wxFONTENCODING_MACARABIC, kTextEncodingMacArabic,
  65     wxFONTENCODING_MACTURKISH, kTextEncodingMacTurkish,
  66     wxFONTENCODING_MACCYRILLIC, kTextEncodingMacCyrillic,
  67 } ;
  68
  69 wxUint16 gMacEncodings[WXSIZEOF(gMacCodePages)][128] ;
  70 bool gMacEncodingsInited[WXSIZEOF(gMacCodePages)] ;
  71
  72 #endif
  73
  74 static wxUint16* LINKAGEMODE GetEncTable(wxFontEncoding enc)
  75 {
  76 #ifdef __WXMAC__
  77     for (int i = 0 ; i < WXSIZEOF(gMacCodePages) ; ++i )
  78     {
  79         if ( gMacCodePages[i].enc == enc )
  80         {
  81             if ( gMacEncodingsInited[i] == false )
  82             {
  83                 TECObjectRef converter ;
  84                 TextEncodingBase code = gMacCodePages[i].mac ;
  85                         TextEncodingBase unicode = CreateTextEncoding(kTextEncodingUnicodeDefault,0,kUnicode16BitFormat) ;
  86                     OSStatus status = TECCreateConverter(&converter,code,unicode);
  87                     char s[2] ;
  88                     s[1] = 0 ;
  89                     ByteCount byteInLen, byteOutLen ;
  90                     for( char c = 255 ; c >= 128 ; --c )
  91                     {
  92                         s[0] = c ;
  93                         status = TECConvertText(converter, (ConstTextPtr) &s , 1, &byteInLen,
  94                         (TextPtr) &gMacEncodings[i][c-128] , 2, &byteOutLen);
  95                     }
  96                     status = TECDisposeConverter(converter);
  97                     gMacEncodingsInited[i]=true;
  98             }
  99             return gMacEncodings[i] ;
 100         }
 101     }
 102 #endif
 103
 104     for (int i = 0; encodings_list[i].table != NULL; i++)
 105     {
 106         if (encodings_list[i].encoding == enc)
 107             return encodings_list[i].table;
 108     }
 109     return NULL;
 110 }
 111
 112 typedef struct {
 113     wxUint16 u;
 114     wxUint8  c;
 115 } CharsetItem;
 116
 117 extern "C" int LINKAGEMODE CompareCharsetItems(const void *i1, const void *i2)
 118 {
 119     return ( ((CharsetItem*)i1) -> u - ((CharsetItem*)i2) -> u );
 120 }
 121
 122
 123 static CharsetItem* LINKAGEMODE BuildReverseTable(wxUint16 *tbl)
 124 {
 125     CharsetItem *rev = new CharsetItem[128];
 126
 127     for (int i = 0; i < 128; i++)
 128         rev[i].c = 128 + i, rev[i].u = tbl[i];
 129
 130     qsort(rev, 128, sizeof(CharsetItem), CompareCharsetItems);
 131
 132     return rev;
 133 }
 134
 135
 136
 137 wxEncodingConverter::wxEncodingConverter()
 138 {
 139     m_Table = NULL;
 140     m_UnicodeInput = m_UnicodeOutput = FALSE;
 141     m_JustCopy = FALSE;
 142 }
 143
 144
 145
 146 bool wxEncodingConverter::Init(wxFontEncoding input_enc, wxFontEncoding output_enc, int method)
 147 {
 148     unsigned i;
 149     wxUint16 *in_tbl, *out_tbl = NULL;
 150
 151     if (m_Table) {delete[] m_Table; m_Table = NULL;}
 152
 153 #if !wxUSE_WCHAR_T
 154     if (input_enc == wxFONTENCODING_UNICODE || output_enc == wxFONTENCODING_UNICODE) return FALSE;
 155 #endif
 156
 157     if (input_enc == output_enc) {m_JustCopy = TRUE; return TRUE;}
 158
 159     m_UnicodeOutput = (output_enc == wxFONTENCODING_UNICODE);
 160     m_JustCopy = FALSE;
 161
 162     if (input_enc == wxFONTENCODING_UNICODE)
 163     {
 164         if ((out_tbl = GetEncTable(output_enc)) == NULL) return FALSE;
 165
 166         m_Table = new tchar[65536];
 167         for (i = 0; i < 128; i++)  m_Table[i] = (tchar)i; // 7bit ASCII
 168         for (i = 128; i < 65536; i++)  m_Table[i] = (tchar)'?';
 169                 // FIXME - this should be character that means `unicode to charset' impossible, not '?'
 170
 171         if (method == wxCONVERT_SUBSTITUTE)
 172         {
 173             for (i = 0; i < encoding_unicode_fallback_count; i++)
 174                 m_Table[encoding_unicode_fallback[i].c] = (tchar) encoding_unicode_fallback[i].s;
 175         }
 176
 177         for (i = 0; i < 128; i++)
 178             m_Table[out_tbl[i]] = (tchar)(128 + i);
 179
 180         m_UnicodeInput = TRUE;
 181     }
 182     else // input !Unicode
 183     {
 184         if ((in_tbl = GetEncTable(input_enc)) == NULL) return FALSE;
 185         if (output_enc != wxFONTENCODING_UNICODE)
 186             if ((out_tbl = GetEncTable(output_enc)) == NULL) return FALSE;
 187
 188         m_UnicodeInput = FALSE;
 189
 190         m_Table = new tchar[256];
 191         for (i = 0; i < 128; i++)  m_Table[i] = (tchar)i; // 7bit ASCII
 192
 193         if (output_enc == wxFONTENCODING_UNICODE)
 194         {
 195             for (i = 0; i < 128; i++)  m_Table[128 + i] = (tchar)in_tbl[i];
 196             return TRUE;
 197         }
 198         // FIXME: write a substitute for bsearch
 199 #ifndef __WXWINCE__
 200         else // output !Unicode
 201         {
 202             CharsetItem *rev = BuildReverseTable(out_tbl);
 203             CharsetItem *item;
 204             CharsetItem key;
 205
 206             for (i = 0; i < 128; i++)
 207             {
 208                 key.u = in_tbl[i];
 209                 item = (CharsetItem*) bsearch(&key, rev, 128, sizeof(CharsetItem), CompareCharsetItems);
 210                 if (item == NULL && method == wxCONVERT_SUBSTITUTE)
 211                     item = (CharsetItem*) bsearch(&key, encoding_unicode_fallback,
 212                                 encoding_unicode_fallback_count, sizeof(CharsetItem), CompareCharsetItems);
 213                 if (item)
 214                     m_Table[128 + i] = (tchar)item -> c;
 215                 else
 216 #if wxUSE_WCHAR_T
 217                     m_Table[128 + i] = (wchar_t)(128 + i);
 218 #else
 219                     m_Table[128 + i] = (char)(128 + i);
 220 #endif
 221             }
 222
 223             delete[] rev;
 224         }
 225 #endif // !__WXWINCE__
 226     }
 227
 228     return TRUE;
 229 }
 230
 231
 232
 233 void wxEncodingConverter::Convert(const char* input, char* output) const
 234 {
 235     wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
 236     wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
 237
 238     const char *i;
 239     char *o;
 240
 241     if (m_JustCopy)
 242     {
 243         strcpy(output, input);
 244         return;
 245     }
 246
 247     wxCHECK_RET(m_Table != NULL, wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 248
 249     for (i = input, o = output; *i != 0;)
 250         *(o++) = (char)(m_Table[(wxUint8)*(i++)]);
 251     *o = 0;
 252 }
 253
 254
 255 #if wxUSE_WCHAR_T
 256
 257 void wxEncodingConverter::Convert(const char* input, wchar_t* output) const
 258 {
 259     wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
 260     wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
 261
 262     const char *i;
 263     wchar_t *o;
 264
 265     if (m_JustCopy)
 266     {
 267         for (i = input, o = output; *i != 0;)
 268             *(o++) = (wchar_t)(*(i++));
 269         *o = 0;
 270         return;
 271     }
 272
 273     wxCHECK_RET(m_Table != NULL, wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 274
 275     for (i = input, o = output; *i != 0;)
 276         *(o++) = (wchar_t)(m_Table[(wxUint8)*(i++)]);
 277     *o = 0;
 278 }
 279
 280
 281
 282 void wxEncodingConverter::Convert(const wchar_t* input, char* output) const
 283 {
 284     wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
 285     wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
 286
 287     const wchar_t *i;
 288     char *o;
 289
 290     if (m_JustCopy)
 291     {
 292         for (i = input, o = output; *i != 0;)
 293             *(o++) = (char)(*(i++));
 294         *o = 0;
 295         return;
 296     }
 297
 298     wxCHECK_RET(m_Table != NULL, wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 299
 300     for (i = input, o = output; *i != 0;)
 301         *(o++) = (char)(m_Table[(wxUint16)*(i++)]);
 302     *o = 0;
 303 }
 304
 305
 306
 307 void wxEncodingConverter::Convert(const wchar_t* input, wchar_t* output) const
 308 {
 309     wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
 310     wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
 311
 312     const wchar_t *i;
 313     wchar_t *o;
 314
 315     if (m_JustCopy)
 316     {
 317         // wcscpy() is not guaranteed to exist
 318         for (i = input, o = output; *i != 0;)
 319             *(o++) = (*(i++));
 320         *o = 0;
 321         return;
 322     }
 323
 324     wxCHECK_RET(m_Table != NULL, wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 325
 326     for (i = input, o = output; *i != 0;)
 327         *(o++) = (wchar_t)(m_Table[(wxUint8)*(i++)]);
 328     *o = 0;
 329 }
 330
 331 #endif // wxUSE_WCHAR_T
 332
 333
 334 wxString wxEncodingConverter::Convert(const wxString& input) const
 335 {
 336     if (m_JustCopy) return input;
 337
 338     wxString s;
 339     const wxChar *i;
 340
 341     wxCHECK_MSG(m_Table != NULL, s,
 342                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 343
 344     if (m_UnicodeInput)
 345     {
 346         for (i = input.c_str(); *i != 0; i++)
 347             s << (wxChar)(m_Table[(wxUint16)*i]);
 348     }
 349     else
 350     {
 351         for (i = input.c_str(); *i != 0; i++)
 352             s << (wxChar)(m_Table[(wxUint8)*i]);
 353     }
 354
 355     return s;
 356 }
 357
 358
 359
 360
 361
 362
 363
 364 // Following tables describe classes of encoding equivalence.
 365 //
 366
 367 #define STOP wxFONTENCODING_SYSTEM
 368
 369 #define NUM_OF_PLATFORMS  4 /*must conform to enum wxPLATFORM_XXXX !!!*/
 370 #define ENC_PER_PLATFORM  5
 371            // max no. of encodings for one language used on one platform
 372            // Anybody thinks 5 is not enough? ;-)
 373
 374 static wxFontEncoding
 375     EquivalentEncodings[][NUM_OF_PLATFORMS][ENC_PER_PLATFORM+1] = {
 376
 377     // *** Please put more common encodings as first! ***
 378
 379     // Western European
 380     {
 381         /* unix    */ {wxFONTENCODING_ISO8859_1, wxFONTENCODING_ISO8859_15, STOP},
 382         /* windows */ {wxFONTENCODING_CP1252, STOP},
 383         /* os2     */ {STOP},
 384         /* mac     */ {wxFONTENCODING_MACROMAN, STOP}
 385     },
 386
 387     // Central European
 388     {
 389         /* unix    */ {wxFONTENCODING_ISO8859_2, STOP},
 390         /* windows */ {wxFONTENCODING_CP1250, STOP},
 391         /* os2     */ {STOP},
 392         /* mac     */ {wxFONTENCODING_MACCENTRALEUR, STOP}
 393     },
 394
 395     // Baltic
 396     {
 397         /* unix    */ {wxFONTENCODING_ISO8859_13, wxFONTENCODING_ISO8859_4, STOP},
 398         /* windows */ {wxFONTENCODING_CP1257, STOP},
 399         /* os2     */ {STOP},
 400         /* mac     */ {STOP}
 401     },
 402
 403     // Hebrew
 404     {
 405         /* unix    */ {wxFONTENCODING_ISO8859_8, STOP},
 406         /* windows */ {wxFONTENCODING_CP1255, STOP},
 407         /* os2     */ {STOP},
 408         /* mac     */ {wxFONTENCODING_MACHEBREW, STOP}
 409     },
 410
 411     // Greek
 412     {
 413         /* unix    */ {wxFONTENCODING_ISO8859_7, STOP},
 414         /* windows */ {wxFONTENCODING_CP1253, STOP},
 415         /* os2     */ {STOP},
 416         /* mac     */ {wxFONTENCODING_MACGREEK, STOP}
 417     },
 418
 419     // Arabic
 420     {
 421         /* unix    */ {wxFONTENCODING_ISO8859_6, STOP},
 422         /* windows */ {wxFONTENCODING_CP1256, STOP},
 423         /* os2     */ {STOP},
 424         /* mac     */ {wxFONTENCODING_MACARABIC, STOP}
 425     },
 426
 427     // Turkish
 428     {
 429         /* unix    */ {wxFONTENCODING_ISO8859_9, STOP},
 430         /* windows */ {wxFONTENCODING_CP1254, STOP},
 431         /* os2     */ {STOP},
 432         /* mac     */ {wxFONTENCODING_MACTURKISH, STOP}
 433     },
 434
 435     // Cyrillic
 436     {
 437         /* unix    */ {wxFONTENCODING_KOI8, wxFONTENCODING_ISO8859_5, STOP},
 438         /* windows */ {wxFONTENCODING_CP1251, STOP},
 439         /* os2     */ {STOP},
 440         /* mac     */ {wxFONTENCODING_MACCYRILLIC, STOP}
 441     },
 442
 443     {{STOP},{STOP},{STOP},{STOP}} /* Terminator */
 444     /* no, _not_ Arnold! */
 445 };
 446
 447
 448 static bool FindEncoding(const wxFontEncodingArray& arr, wxFontEncoding f)
 449 {
 450     for (wxFontEncodingArray::const_iterator it = arr.begin(), en = arr.end();
 451          it != en; ++it)
 452         if (*it == f)
 453             return true;
 454     return false;
 455 }
 456
 457 wxFontEncodingArray wxEncodingConverter::GetPlatformEquivalents(wxFontEncoding enc, int platform)
 458 {
 459     if (platform == wxPLATFORM_CURRENT)
 460     {
 461 #if defined(__WXMSW__)
 462         platform = wxPLATFORM_WINDOWS;
 463 #elif defined(__WXGTK__) || defined(__WXMOTIF__)
 464         platform = wxPLATFORM_UNIX;
 465 #elif defined(__WXOS2__)
 466         platform = wxPLATFORM_OS2;
 467 #elif defined(__WXMAC__)
 468         platform = wxPLATFORM_MAC;
 469 #endif
 470     }
 471
 472     int i, clas, e ;
 473     wxFontEncoding *f;
 474     wxFontEncodingArray arr;
 475
 476     clas = 0;
 477     while (EquivalentEncodings[clas][0][0] != STOP)
 478     {
 479         for (i = 0; i < NUM_OF_PLATFORMS; i++)
 480             for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
 481                 if (EquivalentEncodings[clas][i][e] == enc)
 482                 {
 483                     for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
 484                         if (*f == enc) arr.push_back(enc);
 485                     for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
 486                         if (!FindEncoding(arr, *f)) arr.push_back(*f);
 487                     i = NUM_OF_PLATFORMS/*hack*/; break;
 488                 }
 489         clas++;
 490     }
 491
 492     return arr;
 493 }
 494
 495
 496
 497 wxFontEncodingArray wxEncodingConverter::GetAllEquivalents(wxFontEncoding enc)
 498 {
 499     int i, clas, e, j ;
 500     wxFontEncoding *f;
 501     wxFontEncodingArray arr;
 502
 503     arr = GetPlatformEquivalents(enc); // we want them to be first items in array
 504
 505     clas = 0;
 506     while (EquivalentEncodings[clas][0][0] != STOP)
 507     {
 508         for (i = 0; i < NUM_OF_PLATFORMS; i++)
 509             for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
 510                 if (EquivalentEncodings[clas][i][e] == enc)
 511                 {
 512                     for (j = 0; j < NUM_OF_PLATFORMS; j++)
 513                         for (f = EquivalentEncodings[clas][j]; *f != STOP; f++)
 514                             if (!FindEncoding(arr, *f)) arr.push_back(*f);
 515                     i = NUM_OF_PLATFORMS/*hack*/; break;
 516                 }
 517         clas++;
 518     }
 519
 520     return arr;
 521 }
 522
 523 #endif // wxUSE_FONTMAP