src/common/encconv.cpp

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