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 // For compilers that support precompilation, includes "wx.h".
  11 #include "wx/wxprec.h"
  12
  13 #ifdef __BORLANDC__
  14   #pragma hdrstop
  15 #endif
  16
  17 #include "wx/encconv.h"
  18
  19 #include <stdlib.h>
  20
  21 // conversion tables, generated by scripts in $(WXWIN)/misc/unictabl:
  22 #if defined( __BORLANDC__ ) || defined(__DARWIN__)
  23     #include "../common/unictabl.inc"
  24 #else
  25     #include "unictabl.inc"
  26 #endif
  27
  28 #if wxUSE_WCHAR_T
  29     typedef wchar_t tchar;
  30 #else
  31     typedef char tchar;
  32 #endif
  33
  34 #ifdef __WXMAC__
  35     #include "wx/osx/core/cfstring.h"
  36     #include <CoreFoundation/CFStringEncodingExt.h>
  37
  38     wxUint16 gMacEncodings[wxFONTENCODING_MACMAX-wxFONTENCODING_MACMIN+1][128] ;
  39     bool gMacEncodingsInited[wxFONTENCODING_MACMAX-wxFONTENCODING_MACMIN+1] ;
  40 #endif
  41
  42 #ifdef __WXWINCE__
  43     #include "wx/msw/wince/missing.h"       // for bsearch()
  44 #endif
  45
  46 static const wxUint16* GetEncTable(wxFontEncoding enc)
  47 {
  48 #ifdef __WXMAC__
  49     if( enc >= wxFONTENCODING_MACMIN && enc <= wxFONTENCODING_MACMAX )
  50     {
  51         int i = enc-wxFONTENCODING_MACMIN ;
  52         if ( gMacEncodingsInited[i] == false )
  53         {
  54             // create
  55             CFStringEncoding cfencoding = wxMacGetSystemEncFromFontEnc( enc ) ;
  56             if( !CFStringIsEncodingAvailable( cfencoding ) )
  57                 return NULL;
  58
  59             memset( gMacEncodings[i] , 0 , 128 * 2 );
  60             char s[2] = { 0 , 0 };
  61             CFRange firstchar = CFRangeMake( 0, 1 );
  62             for( unsigned char c = 255 ; c >= 128 ; --c )
  63             {
  64                 s[0] = c ;
  65                 wxCFStringRef cfref( CFStringCreateWithCStringNoCopy( NULL, s, cfencoding , kCFAllocatorNull ) );
  66                 CFStringGetCharacters( cfref, firstchar, (UniChar*)  &gMacEncodings[i][c-128] );
  67             }
  68             gMacEncodingsInited[i]=true;
  69         }
  70         return gMacEncodings[i] ;
  71     }
  72 #endif
  73
  74     for (int i = 0; encodings_list[i].table != NULL; i++)
  75     {
  76         if (encodings_list[i].encoding == enc)
  77             return encodings_list[i].table;
  78     }
  79     return NULL;
  80 }
  81
  82 typedef struct {
  83     wxUint16 u;
  84     wxUint8  c;
  85 } CharsetItem;
  86
  87 extern "C" int wxCMPFUNC_CONV
  88 CompareCharsetItems(const void *i1, const void *i2)
  89 {
  90     return ( ((CharsetItem*)i1) -> u - ((CharsetItem*)i2) -> u );
  91 }
  92
  93
  94 static CharsetItem* BuildReverseTable(const wxUint16 *tbl)
  95 {
  96     CharsetItem *rev = new CharsetItem[128];
  97
  98     for (int i = 0; i < 128; i++)
  99         rev[i].c = wxUint8(128 + i), rev[i].u = tbl[i];
 100
 101     qsort(rev, 128, sizeof(CharsetItem), CompareCharsetItems);
 102
 103     return rev;
 104 }
 105
 106
 107
 108 wxEncodingConverter::wxEncodingConverter()
 109 {
 110     m_Table = NULL;
 111     m_UnicodeInput = m_UnicodeOutput = false;
 112     m_JustCopy = false;
 113 }
 114
 115
 116
 117 bool wxEncodingConverter::Init(wxFontEncoding input_enc, wxFontEncoding output_enc, int method)
 118 {
 119     unsigned i;
 120     const wxUint16 *in_tbl;
 121     const wxUint16 *out_tbl = NULL;
 122
 123     if (m_Table) {delete[] m_Table; m_Table = NULL;}
 124
 125 #if !wxUSE_WCHAR_T
 126     if (input_enc == wxFONTENCODING_UNICODE || output_enc == wxFONTENCODING_UNICODE) return false;
 127 #endif
 128
 129     if (input_enc == output_enc) {m_JustCopy = true; return true;}
 130
 131     m_UnicodeOutput = (output_enc == wxFONTENCODING_UNICODE);
 132     m_JustCopy = false;
 133
 134     if (input_enc == wxFONTENCODING_UNICODE)
 135     {
 136         if ((out_tbl = GetEncTable(output_enc)) == NULL) return false;
 137
 138         m_Table = new tchar[65536];
 139         for (i = 0; i < 128; i++)  m_Table[i] = (tchar)i; // 7bit ASCII
 140         for (i = 128; i < 65536; i++)  m_Table[i] = (tchar)0;
 141
 142         if (method == wxCONVERT_SUBSTITUTE)
 143         {
 144             for (i = 0; i < encoding_unicode_fallback_count; i++)
 145                 m_Table[encoding_unicode_fallback[i].c] = (tchar) encoding_unicode_fallback[i].s;
 146         }
 147
 148         for (i = 0; i < 128; i++)
 149             m_Table[out_tbl[i]] = (tchar)(128 + i);
 150
 151         m_UnicodeInput = true;
 152     }
 153     else // input !Unicode
 154     {
 155         if ((in_tbl = GetEncTable(input_enc)) == NULL) return false;
 156         if (output_enc != wxFONTENCODING_UNICODE)
 157             if ((out_tbl = GetEncTable(output_enc)) == NULL) return false;
 158
 159         m_UnicodeInput = false;
 160
 161         m_Table = new tchar[256];
 162         for (i = 0; i < 128; i++)  m_Table[i] = (tchar)i; // 7bit ASCII
 163
 164         if (output_enc == wxFONTENCODING_UNICODE)
 165         {
 166             for (i = 0; i < 128; i++)  m_Table[128 + i] = (tchar)in_tbl[i];
 167             return true;
 168         }
 169         else // output !Unicode
 170         {
 171             CharsetItem *rev = BuildReverseTable(out_tbl);
 172             CharsetItem *item;
 173             CharsetItem key;
 174
 175             for (i = 0; i < 128; i++)
 176             {
 177                 key.u = in_tbl[i];
 178                 item = (CharsetItem*) bsearch(&key, rev, 128, sizeof(CharsetItem), CompareCharsetItems);
 179                 if (item == NULL && method == wxCONVERT_SUBSTITUTE)
 180                     item = (CharsetItem*) bsearch(&key, encoding_unicode_fallback,
 181                                 encoding_unicode_fallback_count, sizeof(CharsetItem), CompareCharsetItems);
 182                 if (item)
 183                     m_Table[128 + i] = (tchar)item -> c;
 184                 else
 185 #if wxUSE_WCHAR_T
 186                     m_Table[128 + i] = (wchar_t)(128 + i);
 187 #else
 188                     m_Table[128 + i] = (char)(128 + i);
 189 #endif
 190             }
 191
 192             delete[] rev;
 193         }
 194     }
 195
 196     return true;
 197 }
 198
 199
 200 #define REPLACEMENT_CHAR  ((tchar)'?')
 201
 202 inline tchar GetTableValue(const tchar *table, tchar value, bool& repl)
 203 {
 204     tchar r = table[value];
 205     if (r == 0 && value != 0)
 206     {
 207         r = REPLACEMENT_CHAR;
 208         repl = true;
 209     }
 210     return r;
 211 }
 212
 213
 214 bool wxEncodingConverter::Convert(const char* input, char* output) const
 215 {
 216     wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
 217     wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
 218
 219     const char *i;
 220     char *o;
 221
 222     if (m_JustCopy)
 223     {
 224         strcpy(output, input);
 225         return true;
 226     }
 227
 228     wxCHECK_MSG(m_Table != NULL, false,
 229                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 230
 231     bool replaced = false;
 232
 233     for (i = input, o = output; *i != 0;)
 234         *(o++) = (char)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
 235     *o = 0;
 236
 237     return !replaced;
 238 }
 239
 240
 241 #if wxUSE_WCHAR_T
 242
 243 bool wxEncodingConverter::Convert(const char* input, wchar_t* output) const
 244 {
 245     wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
 246     wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
 247
 248     const char *i;
 249     wchar_t *o;
 250
 251     if (m_JustCopy)
 252     {
 253         for (i = input, o = output; *i != 0;)
 254             *(o++) = (wchar_t)(*(i++));
 255         *o = 0;
 256         return true;
 257     }
 258
 259     wxCHECK_MSG(m_Table != NULL, false,
 260                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 261
 262     bool replaced = false;
 263
 264     for (i = input, o = output; *i != 0;)
 265         *(o++) = (wchar_t)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
 266     *o = 0;
 267
 268     return !replaced;
 269 }
 270
 271
 272
 273 bool wxEncodingConverter::Convert(const wchar_t* input, char* output) const
 274 {
 275     wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
 276     wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
 277
 278     const wchar_t *i;
 279     char *o;
 280
 281     if (m_JustCopy)
 282     {
 283         for (i = input, o = output; *i != 0;)
 284             *(o++) = (char)(*(i++));
 285         *o = 0;
 286         return true;
 287     }
 288
 289     wxCHECK_MSG(m_Table != NULL, false,
 290                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 291
 292     bool replaced = false;
 293
 294     for (i = input, o = output; *i != 0;)
 295         *(o++) = (char)(GetTableValue(m_Table, (wxUint16)*(i++), replaced));
 296     *o = 0;
 297
 298     return !replaced;
 299 }
 300
 301
 302
 303 bool wxEncodingConverter::Convert(const wchar_t* input, wchar_t* output) const
 304 {
 305     wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
 306     wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
 307
 308     const wchar_t *i;
 309     wchar_t *o;
 310
 311     if (m_JustCopy)
 312     {
 313         // wcscpy() is not guaranteed to exist
 314         for (i = input, o = output; *i != 0;)
 315             *(o++) = (*(i++));
 316         *o = 0;
 317         return true;
 318     }
 319
 320     wxCHECK_MSG(m_Table != NULL, false,
 321                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 322
 323     bool replaced = false;
 324
 325     for (i = input, o = output; *i != 0;)
 326         *(o++) = (wchar_t)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
 327     *o = 0;
 328
 329     return !replaced;
 330 }
 331
 332 #endif // wxUSE_WCHAR_T
 333
 334
 335 wxString wxEncodingConverter::Convert(const wxString& input) const
 336 {
 337     if (m_JustCopy) return input;
 338
 339     wxString s;
 340     const wxChar *i;
 341
 342     wxCHECK_MSG(m_Table != NULL, s,
 343                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 344
 345     if (m_UnicodeInput)
 346     {
 347         for (i = input.c_str(); *i != 0; i++)
 348             s << (wxChar)(m_Table[(wxUint16)*i]);
 349     }
 350     else
 351     {
 352         for (i = input.c_str(); *i != 0; i++)
 353             s << (wxChar)(m_Table[(wxUint8)*i]);
 354     }
 355
 356     return s;
 357 }
 358
 359
 360
 361
 362
 363
 364
 365 // Following tables describe classes of encoding equivalence.
 366 //
 367
 368 #define STOP wxFONTENCODING_SYSTEM
 369
 370 #define NUM_OF_PLATFORMS  4 /*must conform to enum wxPLATFORM_XXXX !!!*/
 371 #define ENC_PER_PLATFORM  3
 372            // max no. of encodings for one language used on one platform.
 373            // Using maximum of everything at the current moment to not make the
 374            // library larger than necessary. Make larger only if necessary - MR
 375
 376 static const wxFontEncoding
 377     EquivalentEncodings[][NUM_OF_PLATFORMS][ENC_PER_PLATFORM+1] = {
 378
 379     // *** Please put more common encodings as first! ***
 380
 381     // Western European
 382     {
 383         /* unix    */ {wxFONTENCODING_ISO8859_1, wxFONTENCODING_ISO8859_15, STOP},
 384         /* windows */ {wxFONTENCODING_CP1252, STOP},
 385         /* os2     */ {STOP},
 386         /* mac     */ {wxFONTENCODING_MACROMAN, STOP}
 387     },
 388
 389     // Central European
 390     {
 391         /* unix    */ {wxFONTENCODING_ISO8859_2, STOP},
 392         /* windows */ {wxFONTENCODING_CP1250, STOP},
 393         /* os2     */ {STOP},
 394         /* mac     */ {wxFONTENCODING_MACCENTRALEUR, STOP}
 395     },
 396
 397     // Baltic
 398     {
 399         /* unix    */ {wxFONTENCODING_ISO8859_13, wxFONTENCODING_ISO8859_4, STOP},
 400         /* windows */ {wxFONTENCODING_CP1257, STOP},
 401         /* os2     */ {STOP},
 402         /* mac     */ {STOP}
 403     },
 404
 405     // Hebrew
 406     {
 407         /* unix    */ {wxFONTENCODING_ISO8859_8, STOP},
 408         /* windows */ {wxFONTENCODING_CP1255, STOP},
 409         /* os2     */ {STOP},
 410         /* mac     */ {wxFONTENCODING_MACHEBREW, STOP}
 411     },
 412
 413     // Greek
 414     {
 415         /* unix    */ {wxFONTENCODING_ISO8859_7, STOP},
 416         /* windows */ {wxFONTENCODING_CP1253, STOP},
 417         /* os2     */ {STOP},
 418         /* mac     */ {wxFONTENCODING_MACGREEK, STOP}
 419     },
 420
 421     // Arabic
 422     {
 423         /* unix    */ {wxFONTENCODING_ISO8859_6, STOP},
 424         /* windows */ {wxFONTENCODING_CP1256, STOP},
 425         /* os2     */ {STOP},
 426         /* mac     */ {wxFONTENCODING_MACARABIC, STOP}
 427     },
 428
 429     // Turkish
 430     {
 431         /* unix    */ {wxFONTENCODING_ISO8859_9, STOP},
 432         /* windows */ {wxFONTENCODING_CP1254, STOP},
 433         /* os2     */ {STOP},
 434         /* mac     */ {wxFONTENCODING_MACTURKISH, STOP}
 435     },
 436
 437     // Cyrillic
 438     {
 439         /* unix    */ {wxFONTENCODING_KOI8, wxFONTENCODING_KOI8_U, wxFONTENCODING_ISO8859_5, STOP},
 440         /* windows */ {wxFONTENCODING_CP1251, STOP},
 441         /* os2     */ {STOP},
 442         /* mac     */ {wxFONTENCODING_MACCYRILLIC, STOP}
 443     },
 444
 445     {{STOP},{STOP},{STOP},{STOP}} /* Terminator */
 446     /* no, _not_ Arnold! */
 447 };
 448
 449
 450 static bool FindEncoding(const wxFontEncodingArray& arr, wxFontEncoding f)
 451 {
 452     for (wxFontEncodingArray::const_iterator it = arr.begin(), en = arr.end();
 453          it != en; ++it)
 454         if (*it == f)
 455             return true;
 456     return false;
 457 }
 458
 459 wxFontEncodingArray wxEncodingConverter::GetPlatformEquivalents(wxFontEncoding enc, int platform)
 460 {
 461     if (platform == wxPLATFORM_CURRENT)
 462     {
 463 #if defined(__WXMSW__)
 464         platform = wxPLATFORM_WINDOWS;
 465 #elif defined(__WXGTK__) || defined(__WXMOTIF__)
 466         platform = wxPLATFORM_UNIX;
 467 #elif defined(__WXPM__)
 468         platform = wxPLATFORM_OS2;
 469 #elif defined(__WXMAC__)
 470         platform = wxPLATFORM_MAC;
 471 #endif
 472     }
 473
 474     int i, clas, e ;
 475     const wxFontEncoding *f;
 476     wxFontEncodingArray arr;
 477
 478     clas = 0;
 479     while (EquivalentEncodings[clas][0][0] != STOP)
 480     {
 481         for (i = 0; i < NUM_OF_PLATFORMS; i++)
 482             for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
 483                 if (EquivalentEncodings[clas][i][e] == enc)
 484                 {
 485                     for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
 486                         if (*f == enc) arr.push_back(enc);
 487                     for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
 488                         if (!FindEncoding(arr, *f)) arr.push_back(*f);
 489                     i = NUM_OF_PLATFORMS/*hack*/; break;
 490                 }
 491         clas++;
 492     }
 493
 494     return arr;
 495 }
 496
 497
 498
 499 wxFontEncodingArray wxEncodingConverter::GetAllEquivalents(wxFontEncoding enc)
 500 {
 501     int i, clas, e, j ;
 502     const wxFontEncoding *f;
 503     wxFontEncodingArray arr;
 504
 505     arr = GetPlatformEquivalents(enc); // we want them to be first items in array
 506
 507     clas = 0;
 508     while (EquivalentEncodings[clas][0][0] != STOP)
 509     {
 510         for (i = 0; i < NUM_OF_PLATFORMS; i++)
 511             for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
 512                 if (EquivalentEncodings[clas][i][e] == enc)
 513                 {
 514                     for (j = 0; j < NUM_OF_PLATFORMS; j++)
 515                         for (f = EquivalentEncodings[clas][j]; *f != STOP; f++)
 516                             if (!FindEncoding(arr, *f)) arr.push_back(*f);
 517                     i = NUM_OF_PLATFORMS/*hack*/; break;
 518                 }
 519         clas++;
 520     }
 521
 522     return arr;
 523 }
 524