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 #include "wx/encconv.h"
  22
  23 #include <stdlib.h>
  24
  25 // conversion tables, generated by scripts in $(WXWIN)/misc/unictabl:
  26 #if defined( __BORLANDC__ ) || defined(__DARWIN__)
  27     #include "../common/unictabl.inc"
  28 #else
  29     #include "unictabl.inc"
  30 #endif
  31
  32 #if wxUSE_WCHAR_T
  33     typedef wchar_t tchar;
  34 #else
  35     typedef char tchar;
  36 #endif
  37
  38 #ifdef __WXMAC__
  39     #include <ATSUnicode.h>
  40     #include <TextCommon.h>
  41     #include <TextEncodingConverter.h>
  42
  43     #include "wx/fontutil.h"
  44     #include "wx/mac/private.h"  // includes mac headers
  45
  46     wxUint16 gMacEncodings[wxFONTENCODING_MACMAX-wxFONTENCODING_MACMIN+1][128] ;
  47     bool gMacEncodingsInited[wxFONTENCODING_MACMAX-wxFONTENCODING_MACMIN+1] ;
  48 #endif
  49
  50 #ifdef __WXWINCE__
  51     #include "wx/msw/wince/missing.h"       // for bsearch()
  52 #endif
  53
  54 static wxUint16* GetEncTable(wxFontEncoding enc)
  55 {
  56 #ifdef __WXMAC__
  57     if( enc >= wxFONTENCODING_MACMIN && enc <= wxFONTENCODING_MACMAX )
  58     {
  59         int i = enc-wxFONTENCODING_MACMIN ;
  60         if ( gMacEncodingsInited[i] == false )
  61         {
  62             TECObjectRef converter ;
  63             TextEncodingBase code = wxMacGetSystemEncFromFontEnc( enc ) ;
  64             TextEncodingBase unicode = CreateTextEncoding(kTextEncodingUnicodeDefault,0,kUnicode16BitFormat) ;
  65             OSStatus status = TECCreateConverter(&converter,code,unicode);
  66             char s[2] ;
  67             s[1] = 0 ;
  68             ByteCount byteInLen, byteOutLen ;
  69             for( unsigned char c = 255 ; c >= 128 ; --c )
  70             {
  71                 s[0] = c ;
  72                 status = TECConvertText(converter, (ConstTextPtr) &s , 1, &byteInLen,
  73                 (TextPtr) &gMacEncodings[i][c-128] , 2, &byteOutLen);
  74             }
  75             status = TECDisposeConverter(converter);
  76             gMacEncodingsInited[i]=true;
  77         }
  78         return gMacEncodings[i] ;
  79     }
  80 #endif
  81
  82     for (int i = 0; encodings_list[i].table != NULL; i++)
  83     {
  84         if (encodings_list[i].encoding == enc)
  85             return encodings_list[i].table;
  86     }
  87     return NULL;
  88 }
  89
  90 typedef struct {
  91     wxUint16 u;
  92     wxUint8  c;
  93 } CharsetItem;
  94
  95 extern "C" int wxCMPFUNC_CONV
  96 CompareCharsetItems(const void *i1, const void *i2)
  97 {
  98     return ( ((CharsetItem*)i1) -> u - ((CharsetItem*)i2) -> u );
  99 }
 100
 101
 102 static CharsetItem* BuildReverseTable(wxUint16 *tbl)
 103 {
 104     CharsetItem *rev = new CharsetItem[128];
 105
 106     for (int i = 0; i < 128; i++)
 107         rev[i].c = wxUint8(128 + i), rev[i].u = tbl[i];
 108
 109     qsort(rev, 128, sizeof(CharsetItem), CompareCharsetItems);
 110
 111     return rev;
 112 }
 113
 114
 115
 116 wxEncodingConverter::wxEncodingConverter()
 117 {
 118     m_Table = NULL;
 119     m_UnicodeInput = m_UnicodeOutput = false;
 120     m_JustCopy = false;
 121 }
 122
 123
 124
 125 bool wxEncodingConverter::Init(wxFontEncoding input_enc, wxFontEncoding output_enc, int method)
 126 {
 127     unsigned i;
 128     wxUint16 *in_tbl, *out_tbl = NULL;
 129
 130     if (m_Table) {delete[] m_Table; m_Table = NULL;}
 131
 132 #if !wxUSE_WCHAR_T
 133     if (input_enc == wxFONTENCODING_UNICODE || output_enc == wxFONTENCODING_UNICODE) return false;
 134 #endif
 135
 136     if (input_enc == output_enc) {m_JustCopy = true; return true;}
 137
 138     m_UnicodeOutput = (output_enc == wxFONTENCODING_UNICODE);
 139     m_JustCopy = false;
 140
 141     if (input_enc == wxFONTENCODING_UNICODE)
 142     {
 143         if ((out_tbl = GetEncTable(output_enc)) == NULL) return false;
 144
 145         m_Table = new tchar[65536];
 146         for (i = 0; i < 128; i++)  m_Table[i] = (tchar)i; // 7bit ASCII
 147         for (i = 128; i < 65536; i++)  m_Table[i] = (tchar)0;
 148
 149         if (method == wxCONVERT_SUBSTITUTE)
 150         {
 151             for (i = 0; i < encoding_unicode_fallback_count; i++)
 152                 m_Table[encoding_unicode_fallback[i].c] = (tchar) encoding_unicode_fallback[i].s;
 153         }
 154
 155         for (i = 0; i < 128; i++)
 156             m_Table[out_tbl[i]] = (tchar)(128 + i);
 157
 158         m_UnicodeInput = true;
 159     }
 160     else // input !Unicode
 161     {
 162         if ((in_tbl = GetEncTable(input_enc)) == NULL) return false;
 163         if (output_enc != wxFONTENCODING_UNICODE)
 164             if ((out_tbl = GetEncTable(output_enc)) == NULL) return false;
 165
 166         m_UnicodeInput = false;
 167
 168         m_Table = new tchar[256];
 169         for (i = 0; i < 128; i++)  m_Table[i] = (tchar)i; // 7bit ASCII
 170
 171         if (output_enc == wxFONTENCODING_UNICODE)
 172         {
 173             for (i = 0; i < 128; i++)  m_Table[128 + i] = (tchar)in_tbl[i];
 174             return true;
 175         }
 176         else // output !Unicode
 177         {
 178             CharsetItem *rev = BuildReverseTable(out_tbl);
 179             CharsetItem *item;
 180             CharsetItem key;
 181
 182             for (i = 0; i < 128; i++)
 183             {
 184                 key.u = in_tbl[i];
 185                 item = (CharsetItem*) bsearch(&key, rev, 128, sizeof(CharsetItem), CompareCharsetItems);
 186                 if (item == NULL && method == wxCONVERT_SUBSTITUTE)
 187                     item = (CharsetItem*) bsearch(&key, encoding_unicode_fallback,
 188                                 encoding_unicode_fallback_count, sizeof(CharsetItem), CompareCharsetItems);
 189                 if (item)
 190                     m_Table[128 + i] = (tchar)item -> c;
 191                 else
 192 #if wxUSE_WCHAR_T
 193                     m_Table[128 + i] = (wchar_t)(128 + i);
 194 #else
 195                     m_Table[128 + i] = (char)(128 + i);
 196 #endif
 197             }
 198
 199             delete[] rev;
 200         }
 201     }
 202
 203     return true;
 204 }
 205
 206
 207 #define REPLACEMENT_CHAR  ((tchar)'?')
 208
 209 inline tchar GetTableValue(const tchar *table, tchar value, bool& repl)
 210 {
 211     tchar r = table[value];
 212     if (r == 0 && value != 0)
 213     {
 214         r = REPLACEMENT_CHAR;
 215         repl = true;
 216     }
 217     return r;
 218 }
 219
 220
 221 bool wxEncodingConverter::Convert(const char* input, char* output) const
 222 {
 223     wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
 224     wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
 225
 226     const char *i;
 227     char *o;
 228
 229     if (m_JustCopy)
 230     {
 231         strcpy(output, input);
 232         return true;
 233     }
 234
 235     wxCHECK_MSG(m_Table != NULL, false,
 236                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 237
 238     bool replaced = false;
 239
 240     for (i = input, o = output; *i != 0;)
 241         *(o++) = (char)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
 242     *o = 0;
 243
 244     return !replaced;
 245 }
 246
 247
 248 #if wxUSE_WCHAR_T
 249
 250 bool wxEncodingConverter::Convert(const char* input, wchar_t* output) const
 251 {
 252     wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
 253     wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
 254
 255     const char *i;
 256     wchar_t *o;
 257
 258     if (m_JustCopy)
 259     {
 260         for (i = input, o = output; *i != 0;)
 261             *(o++) = (wchar_t)(*(i++));
 262         *o = 0;
 263         return true;
 264     }
 265
 266     wxCHECK_MSG(m_Table != NULL, false,
 267                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 268
 269     bool replaced = false;
 270
 271     for (i = input, o = output; *i != 0;)
 272         *(o++) = (wchar_t)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
 273     *o = 0;
 274
 275     return !replaced;
 276 }
 277
 278
 279
 280 bool wxEncodingConverter::Convert(const wchar_t* input, char* output) const
 281 {
 282     wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
 283     wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
 284
 285     const wchar_t *i;
 286     char *o;
 287
 288     if (m_JustCopy)
 289     {
 290         for (i = input, o = output; *i != 0;)
 291             *(o++) = (char)(*(i++));
 292         *o = 0;
 293         return true;
 294     }
 295
 296     wxCHECK_MSG(m_Table != NULL, false,
 297                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 298
 299     bool replaced = false;
 300
 301     for (i = input, o = output; *i != 0;)
 302         *(o++) = (char)(GetTableValue(m_Table, (wxUint16)*(i++), replaced));
 303     *o = 0;
 304
 305     return !replaced;
 306 }
 307
 308
 309
 310 bool wxEncodingConverter::Convert(const wchar_t* input, wchar_t* output) const
 311 {
 312     wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
 313     wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
 314
 315     const wchar_t *i;
 316     wchar_t *o;
 317
 318     if (m_JustCopy)
 319     {
 320         // wcscpy() is not guaranteed to exist
 321         for (i = input, o = output; *i != 0;)
 322             *(o++) = (*(i++));
 323         *o = 0;
 324         return true;
 325     }
 326
 327     wxCHECK_MSG(m_Table != NULL, false,
 328                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 329
 330     bool replaced = false;
 331
 332     for (i = input, o = output; *i != 0;)
 333         *(o++) = (wchar_t)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
 334     *o = 0;
 335
 336     return !replaced;
 337 }
 338
 339 #endif // wxUSE_WCHAR_T
 340
 341
 342 wxString wxEncodingConverter::Convert(const wxString& input) const
 343 {
 344     if (m_JustCopy) return input;
 345
 346     wxString s;
 347     const wxChar *i;
 348
 349     wxCHECK_MSG(m_Table != NULL, s,
 350                 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
 351
 352     if (m_UnicodeInput)
 353     {
 354         for (i = input.c_str(); *i != 0; i++)
 355             s << (wxChar)(m_Table[(wxUint16)*i]);
 356     }
 357     else
 358     {
 359         for (i = input.c_str(); *i != 0; i++)
 360             s << (wxChar)(m_Table[(wxUint8)*i]);
 361     }
 362
 363     return s;
 364 }
 365
 366
 367
 368
 369
 370
 371
 372 // Following tables describe classes of encoding equivalence.
 373 //
 374
 375 #define STOP wxFONTENCODING_SYSTEM
 376
 377 #define NUM_OF_PLATFORMS  4 /*must conform to enum wxPLATFORM_XXXX !!!*/
 378 #define ENC_PER_PLATFORM  5
 379            // max no. of encodings for one language used on one platform
 380            // Anybody thinks 5 is not enough? ;-)
 381
 382 static wxFontEncoding
 383     EquivalentEncodings[][NUM_OF_PLATFORMS][ENC_PER_PLATFORM+1] = {
 384
 385     // *** Please put more common encodings as first! ***
 386
 387     // Western European
 388     {
 389         /* unix    */ {wxFONTENCODING_ISO8859_1, wxFONTENCODING_ISO8859_15, STOP},
 390         /* windows */ {wxFONTENCODING_CP1252, STOP},
 391         /* os2     */ {STOP},
 392         /* mac     */ {wxFONTENCODING_MACROMAN, STOP}
 393     },
 394
 395     // Central European
 396     {
 397         /* unix    */ {wxFONTENCODING_ISO8859_2, STOP},
 398         /* windows */ {wxFONTENCODING_CP1250, STOP},
 399         /* os2     */ {STOP},
 400         /* mac     */ {wxFONTENCODING_MACCENTRALEUR, STOP}
 401     },
 402
 403     // Baltic
 404     {
 405         /* unix    */ {wxFONTENCODING_ISO8859_13, wxFONTENCODING_ISO8859_4, STOP},
 406         /* windows */ {wxFONTENCODING_CP1257, STOP},
 407         /* os2     */ {STOP},
 408         /* mac     */ {STOP}
 409     },
 410
 411     // Hebrew
 412     {
 413         /* unix    */ {wxFONTENCODING_ISO8859_8, STOP},
 414         /* windows */ {wxFONTENCODING_CP1255, STOP},
 415         /* os2     */ {STOP},
 416         /* mac     */ {wxFONTENCODING_MACHEBREW, STOP}
 417     },
 418
 419     // Greek
 420     {
 421         /* unix    */ {wxFONTENCODING_ISO8859_7, STOP},
 422         /* windows */ {wxFONTENCODING_CP1253, STOP},
 423         /* os2     */ {STOP},
 424         /* mac     */ {wxFONTENCODING_MACGREEK, STOP}
 425     },
 426
 427     // Arabic
 428     {
 429         /* unix    */ {wxFONTENCODING_ISO8859_6, STOP},
 430         /* windows */ {wxFONTENCODING_CP1256, STOP},
 431         /* os2     */ {STOP},
 432         /* mac     */ {wxFONTENCODING_MACARABIC, STOP}
 433     },
 434
 435     // Turkish
 436     {
 437         /* unix    */ {wxFONTENCODING_ISO8859_9, STOP},
 438         /* windows */ {wxFONTENCODING_CP1254, STOP},
 439         /* os2     */ {STOP},
 440         /* mac     */ {wxFONTENCODING_MACTURKISH, STOP}
 441     },
 442
 443     // Cyrillic
 444     {
 445         /* unix    */ {wxFONTENCODING_KOI8, wxFONTENCODING_KOI8_U, wxFONTENCODING_ISO8859_5, STOP},
 446         /* windows */ {wxFONTENCODING_CP1251, STOP},
 447         /* os2     */ {STOP},
 448         /* mac     */ {wxFONTENCODING_MACCYRILLIC, STOP}
 449     },
 450
 451     {{STOP},{STOP},{STOP},{STOP}} /* Terminator */
 452     /* no, _not_ Arnold! */
 453 };
 454
 455
 456 static bool FindEncoding(const wxFontEncodingArray& arr, wxFontEncoding f)
 457 {
 458     for (wxFontEncodingArray::const_iterator it = arr.begin(), en = arr.end();
 459          it != en; ++it)
 460         if (*it == f)
 461             return true;
 462     return false;
 463 }
 464
 465 wxFontEncodingArray wxEncodingConverter::GetPlatformEquivalents(wxFontEncoding enc, int platform)
 466 {
 467     if (platform == wxPLATFORM_CURRENT)
 468     {
 469 #if defined(__WXMSW__)
 470         platform = wxPLATFORM_WINDOWS;
 471 #elif defined(__WXGTK__) || defined(__WXMOTIF__)
 472         platform = wxPLATFORM_UNIX;
 473 #elif defined(__WXOS2__)
 474         platform = wxPLATFORM_OS2;
 475 #elif defined(__WXMAC__)
 476         platform = wxPLATFORM_MAC;
 477 #endif
 478     }
 479
 480     int i, clas, e ;
 481     wxFontEncoding *f;
 482     wxFontEncodingArray arr;
 483
 484     clas = 0;
 485     while (EquivalentEncodings[clas][0][0] != STOP)
 486     {
 487         for (i = 0; i < NUM_OF_PLATFORMS; i++)
 488             for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
 489                 if (EquivalentEncodings[clas][i][e] == enc)
 490                 {
 491                     for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
 492                         if (*f == enc) arr.push_back(enc);
 493                     for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
 494                         if (!FindEncoding(arr, *f)) arr.push_back(*f);
 495                     i = NUM_OF_PLATFORMS/*hack*/; break;
 496                 }
 497         clas++;
 498     }
 499
 500     return arr;
 501 }
 502
 503
 504
 505 wxFontEncodingArray wxEncodingConverter::GetAllEquivalents(wxFontEncoding enc)
 506 {
 507     int i, clas, e, j ;
 508     wxFontEncoding *f;
 509     wxFontEncodingArray arr;
 510
 511     arr = GetPlatformEquivalents(enc); // we want them to be first items in array
 512
 513     clas = 0;
 514     while (EquivalentEncodings[clas][0][0] != STOP)
 515     {
 516         for (i = 0; i < NUM_OF_PLATFORMS; i++)
 517             for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
 518                 if (EquivalentEncodings[clas][i][e] == enc)
 519                 {
 520                     for (j = 0; j < NUM_OF_PLATFORMS; j++)
 521                         for (f = EquivalentEncodings[clas][j]; *f != STOP; f++)
 522                             if (!FindEncoding(arr, *f)) arr.push_back(*f);
 523                     i = NUM_OF_PLATFORMS/*hack*/; break;
 524                 }
 525         clas++;
 526     }
 527
 528     return arr;
 529 }
 530