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 const 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(const 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     const wxUint16 *in_tbl;
 128     const wxUint16 *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  3
 379            // max no. of encodings for one language used on one platform
 380            // Anybody thinks 5 is not enough? ;-)
 381
 382            // It is too much. Using maximum of everything at the current moment to not
 383            // make the library larger than necessary. Make larger only if necessary - MR
 384
 385 static const wxFontEncoding
 386     EquivalentEncodings[][NUM_OF_PLATFORMS][ENC_PER_PLATFORM+1] = {
 387
 388     // *** Please put more common encodings as first! ***
 389
 390     // Western European
 391     {
 392         /* unix    */ {wxFONTENCODING_ISO8859_1, wxFONTENCODING_ISO8859_15, STOP},
 393         /* windows */ {wxFONTENCODING_CP1252, STOP},
 394         /* os2     */ {STOP},
 395         /* mac     */ {wxFONTENCODING_MACROMAN, STOP}
 396     },
 397
 398     // Central European
 399     {
 400         /* unix    */ {wxFONTENCODING_ISO8859_2, STOP},
 401         /* windows */ {wxFONTENCODING_CP1250, STOP},
 402         /* os2     */ {STOP},
 403         /* mac     */ {wxFONTENCODING_MACCENTRALEUR, STOP}
 404     },
 405
 406     // Baltic
 407     {
 408         /* unix    */ {wxFONTENCODING_ISO8859_13, wxFONTENCODING_ISO8859_4, STOP},
 409         /* windows */ {wxFONTENCODING_CP1257, STOP},
 410         /* os2     */ {STOP},
 411         /* mac     */ {STOP}
 412     },
 413
 414     // Hebrew
 415     {
 416         /* unix    */ {wxFONTENCODING_ISO8859_8, STOP},
 417         /* windows */ {wxFONTENCODING_CP1255, STOP},
 418         /* os2     */ {STOP},
 419         /* mac     */ {wxFONTENCODING_MACHEBREW, STOP}
 420     },
 421
 422     // Greek
 423     {
 424         /* unix    */ {wxFONTENCODING_ISO8859_7, STOP},
 425         /* windows */ {wxFONTENCODING_CP1253, STOP},
 426         /* os2     */ {STOP},
 427         /* mac     */ {wxFONTENCODING_MACGREEK, STOP}
 428     },
 429
 430     // Arabic
 431     {
 432         /* unix    */ {wxFONTENCODING_ISO8859_6, STOP},
 433         /* windows */ {wxFONTENCODING_CP1256, STOP},
 434         /* os2     */ {STOP},
 435         /* mac     */ {wxFONTENCODING_MACARABIC, STOP}
 436     },
 437
 438     // Turkish
 439     {
 440         /* unix    */ {wxFONTENCODING_ISO8859_9, STOP},
 441         /* windows */ {wxFONTENCODING_CP1254, STOP},
 442         /* os2     */ {STOP},
 443         /* mac     */ {wxFONTENCODING_MACTURKISH, STOP}
 444     },
 445
 446     // Cyrillic
 447     {
 448         /* unix    */ {wxFONTENCODING_KOI8, wxFONTENCODING_KOI8_U, wxFONTENCODING_ISO8859_5, STOP},
 449         /* windows */ {wxFONTENCODING_CP1251, STOP},
 450         /* os2     */ {STOP},
 451         /* mac     */ {wxFONTENCODING_MACCYRILLIC, STOP}
 452     },
 453
 454     {{STOP},{STOP},{STOP},{STOP}} /* Terminator */
 455     /* no, _not_ Arnold! */
 456 };
 457
 458
 459 static bool FindEncoding(const wxFontEncodingArray& arr, wxFontEncoding f)
 460 {
 461     for (wxFontEncodingArray::const_iterator it = arr.begin(), en = arr.end();
 462          it != en; ++it)
 463         if (*it == f)
 464             return true;
 465     return false;
 466 }
 467
 468 wxFontEncodingArray wxEncodingConverter::GetPlatformEquivalents(wxFontEncoding enc, int platform)
 469 {
 470     if (platform == wxPLATFORM_CURRENT)
 471     {
 472 #if defined(__WXMSW__)
 473         platform = wxPLATFORM_WINDOWS;
 474 #elif defined(__WXGTK__) || defined(__WXMOTIF__)
 475         platform = wxPLATFORM_UNIX;
 476 #elif defined(__WXPM__)
 477         platform = wxPLATFORM_OS2;
 478 #elif defined(__WXMAC__)
 479         platform = wxPLATFORM_MAC;
 480 #endif
 481     }
 482
 483     int i, clas, e ;
 484     const wxFontEncoding *f;
 485     wxFontEncodingArray arr;
 486
 487     clas = 0;
 488     while (EquivalentEncodings[clas][0][0] != STOP)
 489     {
 490         for (i = 0; i < NUM_OF_PLATFORMS; i++)
 491             for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
 492                 if (EquivalentEncodings[clas][i][e] == enc)
 493                 {
 494                     for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
 495                         if (*f == enc) arr.push_back(enc);
 496                     for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
 497                         if (!FindEncoding(arr, *f)) arr.push_back(*f);
 498                     i = NUM_OF_PLATFORMS/*hack*/; break;
 499                 }
 500         clas++;
 501     }
 502
 503     return arr;
 504 }
 505
 506
 507
 508 wxFontEncodingArray wxEncodingConverter::GetAllEquivalents(wxFontEncoding enc)
 509 {
 510     int i, clas, e, j ;
 511     const wxFontEncoding *f;
 512     wxFontEncodingArray arr;
 513
 514     arr = GetPlatformEquivalents(enc); // we want them to be first items in array
 515
 516     clas = 0;
 517     while (EquivalentEncodings[clas][0][0] != STOP)
 518     {
 519         for (i = 0; i < NUM_OF_PLATFORMS; i++)
 520             for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
 521                 if (EquivalentEncodings[clas][i][e] == enc)
 522                 {
 523                     for (j = 0; j < NUM_OF_PLATFORMS; j++)
 524                         for (f = EquivalentEncodings[clas][j]; *f != STOP; f++)
 525                             if (!FindEncoding(arr, *f)) arr.push_back(*f);
 526                     i = NUM_OF_PLATFORMS/*hack*/; break;
 527                 }
 528         clas++;
 529     }
 530
 531     return arr;
 532 }
 533