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 #ifdef __DARWIN__
  36 #include <Carbon/Carbon.h>
  37 #else
  38 #include <ATSUnicode.h>
  39 #include <TextCommon.h>
  40 #include <TextEncodingConverter.h>
  41 #endif
  42     #include "wx/fontutil.h"
  43     #include "wx/mac/private.h"  // includes mac headers
  44
  45     wxUint16 gMacEncodings[wxFONTENCODING_MACMAX-wxFONTENCODING_MACMIN+1][128] ;
  46     bool gMacEncodingsInited[wxFONTENCODING_MACMAX-wxFONTENCODING_MACMIN+1] ;
  47 #endif
  48
  49 #ifdef __WXWINCE__
  50     #include "wx/msw/wince/missing.h"       // for bsearch()
  51 #endif
  52
  53 static wxUint16* GetEncTable(wxFontEncoding enc)
  54 {
  55 #ifdef __WXMAC__
  56     if( enc >= wxFONTENCODING_MACMIN && enc <= wxFONTENCODING_MACMAX )
  57     {
  58         int i = enc-wxFONTENCODING_MACMIN ;
  59         if ( gMacEncodingsInited[i] == false )
  60         {
  61             TECObjectRef converter ;
  62             TextEncodingBase code = wxMacGetSystemEncFromFontEnc( enc ) ;
  63             TextEncodingBase unicode = CreateTextEncoding(kTextEncodingUnicodeDefault,0,kUnicode16BitFormat) ;
  64             OSStatus status = TECCreateConverter(&converter,code,unicode);
  65             char s[2] ;
  66             s[1] = 0 ;
  67             ByteCount byteInLen, byteOutLen ;
  68             for( unsigned char c = 255 ; c >= 128 ; --c )
  69             {
  70                 s[0] = c ;
  71                 status = TECConvertText(converter, (ConstTextPtr) &s , 1, &byteInLen,
  72                 (TextPtr) &gMacEncodings[i][c-128] , 2, &byteOutLen);
  73             }
  74             status = TECDisposeConverter(converter);
  75             gMacEncodingsInited[i]=true;
  76         }
  77         return gMacEncodings[i] ;
  78     }
  79 #endif
  80
  81     for (int i = 0; encodings_list[i].table != NULL; i++)
  82     {
  83         if (encodings_list[i].encoding == enc)
  84             return encodings_list[i].table;
  85     }
  86     return NULL;
  87 }
  88
  89 typedef struct {
  90     wxUint16 u;
  91     wxUint8  c;
  92 } CharsetItem;
  93
  94 extern "C" int wxCMPFUNC_CONV
  95 CompareCharsetItems(const void *i1, const void *i2)
  96 {
  97     return ( ((CharsetItem*)i1) -> u - ((CharsetItem*)i2) -> u );
  98 }
  99
 100
 101 static CharsetItem* BuildReverseTable(wxUint16 *tbl)
 102 {
 103     CharsetItem *rev = new CharsetItem[128];
 104
 105     for (int i = 0; i < 128; i++)
 106         rev[i].c = wxUint8(128 + i), rev[i].u = tbl[i];
 107
 108     qsort(rev, 128, sizeof(CharsetItem), CompareCharsetItems);
 109
 110     return rev;
 111 }
 112
 113
 114
 115 wxEncodingConverter::wxEncodingConverter()
 116 {
 117     m_Table = NULL;
 118     m_UnicodeInput = m_UnicodeOutput = false;
 119     m_JustCopy = false;
 120 }
 121
 122
 123
 124 bool wxEncodingConverter::Init(wxFontEncoding input_enc, wxFontEncoding output_enc, int method)
 125 {
 126     unsigned i;
 127     wxUint16 *in_tbl, *out_tbl = NULL;
 128
 129     if (m_Table) {delete[] m_Table; m_Table = NULL;}
 130
 131 #if !wxUSE_WCHAR_T
 132     if (input_enc == wxFONTENCODING_UNICODE || output_enc == wxFONTENCODING_UNICODE) return false;
 133 #endif
 134
 135     if (input_enc == output_enc) {m_JustCopy = true; return true;}
 136
 137     m_UnicodeOutput = (output_enc == wxFONTENCODING_UNICODE);
 138     m_JustCopy = false;
 139
 140     if (input_enc == wxFONTENCODING_UNICODE)
 141     {
 142         if ((out_tbl = GetEncTable(output_enc)) == NULL) return false;
 143
 144         m_Table = new tchar[65536];
 145         for (i = 0; i < 128; i++)  m_Table[i] = (tchar)i; // 7bit ASCII
 146         for (i = 128; i < 65536; i++)  m_Table[i] = (tchar)0;
 147
 148         if (method == wxCONVERT_SUBSTITUTE)
 149         {
 150             for (i = 0; i < encoding_unicode_fallback_count; i++)
 151                 m_Table[encoding_unicode_fallback[i].c] = (tchar) encoding_unicode_fallback[i].s;
 152         }
 153
 154         for (i = 0; i < 128; i++)
 155             m_Table[out_tbl[i]] = (tchar)(128 + i);
 156
 157         m_UnicodeInput = true;
 158     }
 159     else // input !Unicode
 160     {
 161         if ((in_tbl = GetEncTable(input_enc)) == NULL) return false;
 162         if (output_enc != wxFONTENCODING_UNICODE)
 163             if ((out_tbl = GetEncTable(output_enc)) == NULL) return false;
 164
 165         m_UnicodeInput = false;
 166
 167         m_Table = new tchar[256];
 168         for (i = 0; i < 128; i++)  m_Table[i] = (tchar)i; // 7bit ASCII
 169
 170         if (output_enc == wxFONTENCODING_UNICODE)
 171         {
 172             for (i = 0; i < 128; i++)  m_Table[128 + i] = (tchar)in_tbl[i];
 173             return true;
 174         }
 175         else // output !Unicode
 176         {
 177             CharsetItem *rev = BuildReverseTable(out_tbl);
 178             CharsetItem *item;
 179             CharsetItem key;
 180
 181             for (i = 0; i < 128; i++)
 182             {
 183                 key.u = in_tbl[i];
 184                 item = (CharsetItem*) bsearch(&key, rev, 128, sizeof(CharsetItem), CompareCharsetItems);
 185                 if (item == NULL && method == wxCONVERT_SUBSTITUTE)
 186                     item = (CharsetItem*) bsearch(&key, encoding_unicode_fallback,
 187                                 encoding_unicode_fallback_count, sizeof(CharsetItem), CompareCharsetItems);
 188                 if (item)
 189                     m_Table[128 + i] = (tchar)item -> c;
 190                 else
 191 #if wxUSE_WCHAR_T
 192                     m_Table[128 + i] = (wchar_t)(128 + i);
 193 #else
 194                     m_Table[128 + i] = (char)(128 + i);
 195 #endif
 196             }
 197
 198             delete[] rev;
 199         }
 200     }
 201
 202     return true;
 203 }
 204
 205
 206 #define REPLACEMENT_CHAR  ((tchar)'?')
 207
 208 inline tchar GetTableValue(const tchar *table, tchar value, bool& repl)
 209 {
 210     tchar r = table[value];
 211     if (r == 0 && value != 0)
 212     {
 213         r = REPLACEMENT_CHAR;
 214         repl = true;
 215     }
 216     return r;
 217 }
 218
 219
 220 bool wxEncodingConverter::Convert(const char* input, char* output) const
 221 {
 222     wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
 223     wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
 224
 225     const char *i;
 226     char *o;
 227
 228     if (m_JustCopy)
 229     {
 230         strcpy(output, input);
 231         return true;
 232     }
 233
 234     wxCHECK_MSG(m_Table != NULL, false,
 235                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 236
 237     bool replaced = false;
 238
 239     for (i = input, o = output; *i != 0;)
 240         *(o++) = (char)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
 241     *o = 0;
 242
 243     return !replaced;
 244 }
 245
 246
 247 #if wxUSE_WCHAR_T
 248
 249 bool wxEncodingConverter::Convert(const char* input, wchar_t* output) const
 250 {
 251     wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
 252     wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
 253
 254     const char *i;
 255     wchar_t *o;
 256
 257     if (m_JustCopy)
 258     {
 259         for (i = input, o = output; *i != 0;)
 260             *(o++) = (wchar_t)(*(i++));
 261         *o = 0;
 262         return true;
 263     }
 264
 265     wxCHECK_MSG(m_Table != NULL, false,
 266                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 267
 268     bool replaced = false;
 269
 270     for (i = input, o = output; *i != 0;)
 271         *(o++) = (wchar_t)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
 272     *o = 0;
 273
 274     return !replaced;
 275 }
 276
 277
 278
 279 bool wxEncodingConverter::Convert(const wchar_t* input, char* output) const
 280 {
 281     wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
 282     wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
 283
 284     const wchar_t *i;
 285     char *o;
 286
 287     if (m_JustCopy)
 288     {
 289         for (i = input, o = output; *i != 0;)
 290             *(o++) = (char)(*(i++));
 291         *o = 0;
 292         return true;
 293     }
 294
 295     wxCHECK_MSG(m_Table != NULL, false,
 296                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 297
 298     bool replaced = false;
 299
 300     for (i = input, o = output; *i != 0;)
 301         *(o++) = (char)(GetTableValue(m_Table, (wxUint16)*(i++), replaced));
 302     *o = 0;
 303
 304     return !replaced;
 305 }
 306
 307
 308
 309 bool wxEncodingConverter::Convert(const wchar_t* input, wchar_t* output) const
 310 {
 311     wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
 312     wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
 313
 314     const wchar_t *i;
 315     wchar_t *o;
 316
 317     if (m_JustCopy)
 318     {
 319         // wcscpy() is not guaranteed to exist
 320         for (i = input, o = output; *i != 0;)
 321             *(o++) = (*(i++));
 322         *o = 0;
 323         return true;
 324     }
 325
 326     wxCHECK_MSG(m_Table != NULL, false,
 327                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 328
 329     bool replaced = false;
 330
 331     for (i = input, o = output; *i != 0;)
 332         *(o++) = (wchar_t)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
 333     *o = 0;
 334
 335     return !replaced;
 336 }
 337
 338 #endif // wxUSE_WCHAR_T
 339
 340
 341 wxString wxEncodingConverter::Convert(const wxString& input) const
 342 {
 343     if (m_JustCopy) return input;
 344
 345     wxString s;
 346     const wxChar *i;
 347
 348     wxCHECK_MSG(m_Table != NULL, s,
 349                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 350
 351     if (m_UnicodeInput)
 352     {
 353         for (i = input.c_str(); *i != 0; i++)
 354             s << (wxChar)(m_Table[(wxUint16)*i]);
 355     }
 356     else
 357     {
 358         for (i = input.c_str(); *i != 0; i++)
 359             s << (wxChar)(m_Table[(wxUint8)*i]);
 360     }
 361
 362     return s;
 363 }
 364
 365
 366
 367
 368
 369
 370
 371 // Following tables describe classes of encoding equivalence.
 372 //
 373
 374 #define STOP wxFONTENCODING_SYSTEM
 375
 376 #define NUM_OF_PLATFORMS  4 /*must conform to enum wxPLATFORM_XXXX !!!*/
 377 #define ENC_PER_PLATFORM  5
 378            // max no. of encodings for one language used on one platform
 379            // Anybody thinks 5 is not enough? ;-)
 380
 381 static wxFontEncoding
 382     EquivalentEncodings[][NUM_OF_PLATFORMS][ENC_PER_PLATFORM+1] = {
 383
 384     // *** Please put more common encodings as first! ***
 385
 386     // Western European
 387     {
 388         /* unix    */ {wxFONTENCODING_ISO8859_1, wxFONTENCODING_ISO8859_15, STOP},
 389         /* windows */ {wxFONTENCODING_CP1252, STOP},
 390         /* os2     */ {STOP},
 391         /* mac     */ {wxFONTENCODING_MACROMAN, STOP}
 392     },
 393
 394     // Central European
 395     {
 396         /* unix    */ {wxFONTENCODING_ISO8859_2, STOP},
 397         /* windows */ {wxFONTENCODING_CP1250, STOP},
 398         /* os2     */ {STOP},
 399         /* mac     */ {wxFONTENCODING_MACCENTRALEUR, STOP}
 400     },
 401
 402     // Baltic
 403     {
 404         /* unix    */ {wxFONTENCODING_ISO8859_13, wxFONTENCODING_ISO8859_4, STOP},
 405         /* windows */ {wxFONTENCODING_CP1257, STOP},
 406         /* os2     */ {STOP},
 407         /* mac     */ {STOP}
 408     },
 409
 410     // Hebrew
 411     {
 412         /* unix    */ {wxFONTENCODING_ISO8859_8, STOP},
 413         /* windows */ {wxFONTENCODING_CP1255, STOP},
 414         /* os2     */ {STOP},
 415         /* mac     */ {wxFONTENCODING_MACHEBREW, STOP}
 416     },
 417
 418     // Greek
 419     {
 420         /* unix    */ {wxFONTENCODING_ISO8859_7, STOP},
 421         /* windows */ {wxFONTENCODING_CP1253, STOP},
 422         /* os2     */ {STOP},
 423         /* mac     */ {wxFONTENCODING_MACGREEK, STOP}
 424     },
 425
 426     // Arabic
 427     {
 428         /* unix    */ {wxFONTENCODING_ISO8859_6, STOP},
 429         /* windows */ {wxFONTENCODING_CP1256, STOP},
 430         /* os2     */ {STOP},
 431         /* mac     */ {wxFONTENCODING_MACARABIC, STOP}
 432     },
 433
 434     // Turkish
 435     {
 436         /* unix    */ {wxFONTENCODING_ISO8859_9, STOP},
 437         /* windows */ {wxFONTENCODING_CP1254, STOP},
 438         /* os2     */ {STOP},
 439         /* mac     */ {wxFONTENCODING_MACTURKISH, STOP}
 440     },
 441
 442     // Cyrillic
 443     {
 444         /* unix    */ {wxFONTENCODING_KOI8, wxFONTENCODING_KOI8_U, wxFONTENCODING_ISO8859_5, STOP},
 445         /* windows */ {wxFONTENCODING_CP1251, STOP},
 446         /* os2     */ {STOP},
 447         /* mac     */ {wxFONTENCODING_MACCYRILLIC, STOP}
 448     },
 449
 450     {{STOP},{STOP},{STOP},{STOP}} /* Terminator */
 451     /* no, _not_ Arnold! */
 452 };
 453
 454
 455 static bool FindEncoding(const wxFontEncodingArray& arr, wxFontEncoding f)
 456 {
 457     for (wxFontEncodingArray::const_iterator it = arr.begin(), en = arr.end();
 458          it != en; ++it)
 459         if (*it == f)
 460             return true;
 461     return false;
 462 }
 463
 464 wxFontEncodingArray wxEncodingConverter::GetPlatformEquivalents(wxFontEncoding enc, int platform)
 465 {
 466     if (platform == wxPLATFORM_CURRENT)
 467     {
 468 #if defined(__WXMSW__)
 469         platform = wxPLATFORM_WINDOWS;
 470 #elif defined(__WXGTK__) || defined(__WXMOTIF__)
 471         platform = wxPLATFORM_UNIX;
 472 #elif defined(__WXPM__)
 473         platform = wxPLATFORM_OS2;
 474 #elif defined(__WXMAC__)
 475         platform = wxPLATFORM_MAC;
 476 #endif
 477     }
 478
 479     int i, clas, e ;
 480     wxFontEncoding *f;
 481     wxFontEncodingArray arr;
 482
 483     clas = 0;
 484     while (EquivalentEncodings[clas][0][0] != STOP)
 485     {
 486         for (i = 0; i < NUM_OF_PLATFORMS; i++)
 487             for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
 488                 if (EquivalentEncodings[clas][i][e] == enc)
 489                 {
 490                     for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
 491                         if (*f == enc) arr.push_back(enc);
 492                     for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
 493                         if (!FindEncoding(arr, *f)) arr.push_back(*f);
 494                     i = NUM_OF_PLATFORMS/*hack*/; break;
 495                 }
 496         clas++;
 497     }
 498
 499     return arr;
 500 }
 501
 502
 503
 504 wxFontEncodingArray wxEncodingConverter::GetAllEquivalents(wxFontEncoding enc)
 505 {
 506     int i, clas, e, j ;
 507     wxFontEncoding *f;
 508     wxFontEncodingArray arr;
 509
 510     arr = GetPlatformEquivalents(enc); // we want them to be first items in array
 511
 512     clas = 0;
 513     while (EquivalentEncodings[clas][0][0] != STOP)
 514     {
 515         for (i = 0; i < NUM_OF_PLATFORMS; i++)
 516             for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
 517                 if (EquivalentEncodings[clas][i][e] == enc)
 518                 {
 519                     for (j = 0; j < NUM_OF_PLATFORMS; j++)
 520                         for (f = EquivalentEncodings[clas][j]; *f != STOP; f++)
 521                             if (!FindEncoding(arr, *f)) arr.push_back(*f);
 522                     i = NUM_OF_PLATFORMS/*hack*/; break;
 523                 }
 524         clas++;
 525     }
 526
 527     return arr;
 528 }
 529