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"
  88 {
  89 static int wxCMPFUNC_CONV
  90 CompareCharsetItems(const void *i1, const void *i2)
  91 {
  92     return ( ((CharsetItem*)i1) -> u - ((CharsetItem*)i2) -> u );
  93 }
  94 }
  95
  96 static CharsetItem* BuildReverseTable(const wxUint16 *tbl)
  97 {
  98     CharsetItem *rev = new CharsetItem[128];
  99
 100     for (int i = 0; i < 128; i++)
 101         rev[i].c = wxUint8(128 + i), rev[i].u = tbl[i];
 102
 103     qsort(rev, 128, sizeof(CharsetItem), CompareCharsetItems);
 104
 105     return rev;
 106 }
 107
 108
 109
 110 wxEncodingConverter::wxEncodingConverter()
 111 {
 112     m_Table = NULL;
 113     m_UnicodeInput = m_UnicodeOutput = false;
 114     m_JustCopy = false;
 115 }
 116
 117
 118
 119 bool wxEncodingConverter::Init(wxFontEncoding input_enc, wxFontEncoding output_enc, int method)
 120 {
 121     unsigned i;
 122     const wxUint16 *in_tbl;
 123     const wxUint16 *out_tbl = NULL;
 124
 125     if (m_Table) {delete[] m_Table; m_Table = NULL;}
 126
 127 #if !wxUSE_WCHAR_T
 128     if (input_enc == wxFONTENCODING_UNICODE || output_enc == wxFONTENCODING_UNICODE) return false;
 129 #endif
 130
 131     if (input_enc == output_enc) {m_JustCopy = true; return true;}
 132
 133     m_UnicodeOutput = (output_enc == wxFONTENCODING_UNICODE);
 134     m_JustCopy = false;
 135
 136     if (input_enc == wxFONTENCODING_UNICODE)
 137     {
 138         if ((out_tbl = GetEncTable(output_enc)) == NULL) return false;
 139
 140         m_Table = new tchar[65536];
 141         for (i = 0; i < 128; i++)  m_Table[i] = (tchar)i; // 7bit ASCII
 142         for (i = 128; i < 65536; i++)  m_Table[i] = (tchar)0;
 143
 144         if (method == wxCONVERT_SUBSTITUTE)
 145         {
 146             for (i = 0; i < encoding_unicode_fallback_count; i++)
 147                 m_Table[encoding_unicode_fallback[i].c] = (tchar) encoding_unicode_fallback[i].s;
 148         }
 149
 150         for (i = 0; i < 128; i++)
 151             m_Table[out_tbl[i]] = (tchar)(128 + i);
 152
 153         m_UnicodeInput = true;
 154     }
 155     else // input !Unicode
 156     {
 157         if ((in_tbl = GetEncTable(input_enc)) == NULL) return false;
 158         if (output_enc != wxFONTENCODING_UNICODE)
 159             if ((out_tbl = GetEncTable(output_enc)) == NULL) return false;
 160
 161         m_UnicodeInput = false;
 162
 163         m_Table = new tchar[256];
 164         for (i = 0; i < 128; i++)  m_Table[i] = (tchar)i; // 7bit ASCII
 165
 166         if (output_enc == wxFONTENCODING_UNICODE)
 167         {
 168             for (i = 0; i < 128; i++)  m_Table[128 + i] = (tchar)in_tbl[i];
 169             return true;
 170         }
 171         else // output !Unicode
 172         {
 173             CharsetItem *rev = BuildReverseTable(out_tbl);
 174             CharsetItem *item;
 175             CharsetItem key;
 176
 177             for (i = 0; i < 128; i++)
 178             {
 179                 key.u = in_tbl[i];
 180                 item = (CharsetItem*) bsearch(&key, rev, 128, sizeof(CharsetItem), CompareCharsetItems);
 181                 if (item == NULL && method == wxCONVERT_SUBSTITUTE)
 182                     item = (CharsetItem*) bsearch(&key, encoding_unicode_fallback,
 183                                 encoding_unicode_fallback_count, sizeof(CharsetItem), CompareCharsetItems);
 184                 if (item)
 185                     m_Table[128 + i] = (tchar)item -> c;
 186                 else
 187 #if wxUSE_WCHAR_T
 188                     m_Table[128 + i] = (wchar_t)(128 + i);
 189 #else
 190                     m_Table[128 + i] = (char)(128 + i);
 191 #endif
 192             }
 193
 194             delete[] rev;
 195         }
 196     }
 197
 198     return true;
 199 }
 200
 201
 202 #define REPLACEMENT_CHAR  ((tchar)'?')
 203
 204 inline tchar GetTableValue(const tchar *table, tchar value, bool& repl)
 205 {
 206     tchar r = table[value];
 207     if (r == 0 && value != 0)
 208     {
 209         r = REPLACEMENT_CHAR;
 210         repl = true;
 211     }
 212     return r;
 213 }
 214
 215
 216 bool wxEncodingConverter::Convert(const char* input, char* output) const
 217 {
 218     wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
 219     wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
 220
 221     const char *i;
 222     char *o;
 223
 224     if (m_JustCopy)
 225     {
 226         strcpy(output, input);
 227         return true;
 228     }
 229
 230     wxCHECK_MSG(m_Table != NULL, false,
 231                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 232
 233     bool replaced = false;
 234
 235     for (i = input, o = output; *i != 0;)
 236         *(o++) = (char)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
 237     *o = 0;
 238
 239     return !replaced;
 240 }
 241
 242
 243 #if wxUSE_WCHAR_T
 244
 245 bool wxEncodingConverter::Convert(const char* input, wchar_t* output) const
 246 {
 247     wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
 248     wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
 249
 250     const char *i;
 251     wchar_t *o;
 252
 253     if (m_JustCopy)
 254     {
 255         for (i = input, o = output; *i != 0;)
 256             *(o++) = (wchar_t)(*(i++));
 257         *o = 0;
 258         return true;
 259     }
 260
 261     wxCHECK_MSG(m_Table != NULL, false,
 262                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 263
 264     bool replaced = false;
 265
 266     for (i = input, o = output; *i != 0;)
 267         *(o++) = (wchar_t)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
 268     *o = 0;
 269
 270     return !replaced;
 271 }
 272
 273
 274
 275 bool wxEncodingConverter::Convert(const wchar_t* input, char* output) const
 276 {
 277     wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
 278     wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
 279
 280     const wchar_t *i;
 281     char *o;
 282
 283     if (m_JustCopy)
 284     {
 285         for (i = input, o = output; *i != 0;)
 286             *(o++) = (char)(*(i++));
 287         *o = 0;
 288         return true;
 289     }
 290
 291     wxCHECK_MSG(m_Table != NULL, false,
 292                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 293
 294     bool replaced = false;
 295
 296     for (i = input, o = output; *i != 0;)
 297         *(o++) = (char)(GetTableValue(m_Table, (wxUint16)*(i++), replaced));
 298     *o = 0;
 299
 300     return !replaced;
 301 }
 302
 303
 304
 305 bool wxEncodingConverter::Convert(const wchar_t* input, wchar_t* output) const
 306 {
 307     wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
 308     wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
 309
 310     const wchar_t *i;
 311     wchar_t *o;
 312
 313     if (m_JustCopy)
 314     {
 315         // wcscpy() is not guaranteed to exist
 316         for (i = input, o = output; *i != 0;)
 317             *(o++) = (*(i++));
 318         *o = 0;
 319         return true;
 320     }
 321
 322     wxCHECK_MSG(m_Table != NULL, false,
 323                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 324
 325     bool replaced = false;
 326
 327     for (i = input, o = output; *i != 0;)
 328         *(o++) = (wchar_t)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
 329     *o = 0;
 330
 331     return !replaced;
 332 }
 333
 334 #endif // wxUSE_WCHAR_T
 335
 336
 337 wxString wxEncodingConverter::Convert(const wxString& input) const
 338 {
 339     if (m_JustCopy) return input;
 340
 341     wxString s;
 342     const wxChar *i;
 343
 344     wxCHECK_MSG(m_Table != NULL, s,
 345                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 346
 347     if (m_UnicodeInput)
 348     {
 349         for (i = input.c_str(); *i != 0; i++)
 350             s << (wxChar)(m_Table[(wxUint16)*i]);
 351     }
 352     else
 353     {
 354         for (i = input.c_str(); *i != 0; i++)
 355             s << (wxChar)(m_Table[(wxUint8)*i]);
 356     }
 357
 358     return s;
 359 }
 360
 361
 362
 363
 364
 365
 366
 367 // Following tables describe classes of encoding equivalence.
 368 //
 369
 370 #define STOP wxFONTENCODING_SYSTEM
 371
 372 #define NUM_OF_PLATFORMS  4 /*must conform to enum wxPLATFORM_XXXX !!!*/
 373 #define ENC_PER_PLATFORM  3
 374            // max no. of encodings for one language used on one platform.
 375            // Using maximum of everything at the current moment to not make the
 376            // library larger than necessary. Make larger only if necessary - MR
 377
 378 static const wxFontEncoding
 379     EquivalentEncodings[][NUM_OF_PLATFORMS][ENC_PER_PLATFORM+1] = {
 380
 381     // *** Please put more common encodings as first! ***
 382
 383     // Western European
 384     {
 385         /* unix    */ {wxFONTENCODING_ISO8859_1, wxFONTENCODING_ISO8859_15, STOP},
 386         /* windows */ {wxFONTENCODING_CP1252, STOP},
 387         /* os2     */ {STOP},
 388         /* mac     */ {wxFONTENCODING_MACROMAN, STOP}
 389     },
 390
 391     // Central European
 392     {
 393         /* unix    */ {wxFONTENCODING_ISO8859_2, STOP},
 394         /* windows */ {wxFONTENCODING_CP1250, STOP},
 395         /* os2     */ {STOP},
 396         /* mac     */ {wxFONTENCODING_MACCENTRALEUR, STOP}
 397     },
 398
 399     // Baltic
 400     {
 401         /* unix    */ {wxFONTENCODING_ISO8859_13, wxFONTENCODING_ISO8859_4, STOP},
 402         /* windows */ {wxFONTENCODING_CP1257, STOP},
 403         /* os2     */ {STOP},
 404         /* mac     */ {STOP}
 405     },
 406
 407     // Hebrew
 408     {
 409         /* unix    */ {wxFONTENCODING_ISO8859_8, STOP},
 410         /* windows */ {wxFONTENCODING_CP1255, STOP},
 411         /* os2     */ {STOP},
 412         /* mac     */ {wxFONTENCODING_MACHEBREW, STOP}
 413     },
 414
 415     // Greek
 416     {
 417         /* unix    */ {wxFONTENCODING_ISO8859_7, STOP},
 418         /* windows */ {wxFONTENCODING_CP1253, STOP},
 419         /* os2     */ {STOP},
 420         /* mac     */ {wxFONTENCODING_MACGREEK, STOP}
 421     },
 422
 423     // Arabic
 424     {
 425         /* unix    */ {wxFONTENCODING_ISO8859_6, STOP},
 426         /* windows */ {wxFONTENCODING_CP1256, STOP},
 427         /* os2     */ {STOP},
 428         /* mac     */ {wxFONTENCODING_MACARABIC, STOP}
 429     },
 430
 431     // Turkish
 432     {
 433         /* unix    */ {wxFONTENCODING_ISO8859_9, STOP},
 434         /* windows */ {wxFONTENCODING_CP1254, STOP},
 435         /* os2     */ {STOP},
 436         /* mac     */ {wxFONTENCODING_MACTURKISH, STOP}
 437     },
 438
 439     // Cyrillic
 440     {
 441         /* unix    */ {wxFONTENCODING_KOI8, wxFONTENCODING_KOI8_U, wxFONTENCODING_ISO8859_5, STOP},
 442         /* windows */ {wxFONTENCODING_CP1251, STOP},
 443         /* os2     */ {STOP},
 444         /* mac     */ {wxFONTENCODING_MACCYRILLIC, STOP}
 445     },
 446
 447     {{STOP},{STOP},{STOP},{STOP}} /* Terminator */
 448     /* no, _not_ Arnold! */
 449 };
 450
 451
 452 static bool FindEncoding(const wxFontEncodingArray& arr, wxFontEncoding f)
 453 {
 454     for (wxFontEncodingArray::const_iterator it = arr.begin(), en = arr.end();
 455          it != en; ++it)
 456         if (*it == f)
 457             return true;
 458     return false;
 459 }
 460
 461 wxFontEncodingArray wxEncodingConverter::GetPlatformEquivalents(wxFontEncoding enc, int platform)
 462 {
 463     if (platform == wxPLATFORM_CURRENT)
 464     {
 465 #if defined(__WXMSW__)
 466         platform = wxPLATFORM_WINDOWS;
 467 #elif defined(__WXGTK__) || defined(__WXMOTIF__)
 468         platform = wxPLATFORM_UNIX;
 469 #elif defined(__WXPM__)
 470         platform = wxPLATFORM_OS2;
 471 #elif defined(__WXMAC__)
 472         platform = wxPLATFORM_MAC;
 473 #endif
 474     }
 475
 476     int i, clas, e ;
 477     const wxFontEncoding *f;
 478     wxFontEncodingArray arr;
 479
 480     clas = 0;
 481     while (EquivalentEncodings[clas][0][0] != STOP)
 482     {
 483         for (i = 0; i < NUM_OF_PLATFORMS; i++)
 484             for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
 485                 if (EquivalentEncodings[clas][i][e] == enc)
 486                 {
 487                     for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
 488                         if (*f == enc) arr.push_back(enc);
 489                     for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
 490                         if (!FindEncoding(arr, *f)) arr.push_back(*f);
 491                     i = NUM_OF_PLATFORMS/*hack*/; break;
 492                 }
 493         clas++;
 494     }
 495
 496     return arr;
 497 }
 498
 499
 500
 501 wxFontEncodingArray wxEncodingConverter::GetAllEquivalents(wxFontEncoding enc)
 502 {
 503     int i, clas, e, j ;
 504     const wxFontEncoding *f;
 505     wxFontEncodingArray arr;
 506
 507     arr = GetPlatformEquivalents(enc); // we want them to be first items in array
 508
 509     clas = 0;
 510     while (EquivalentEncodings[clas][0][0] != STOP)
 511     {
 512         for (i = 0; i < NUM_OF_PLATFORMS; i++)
 513             for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
 514                 if (EquivalentEncodings[clas][i][e] == enc)
 515                 {
 516                     for (j = 0; j < NUM_OF_PLATFORMS; j++)
 517                         for (f = EquivalentEncodings[clas][j]; *f != STOP; f++)
 518                             if (!FindEncoding(arr, *f)) arr.push_back(*f);
 519                     i = NUM_OF_PLATFORMS/*hack*/; break;
 520                 }
 521         clas++;
 522     }
 523
 524     return arr;
 525 }
 526