]> git.saurik.com Git - wxWidgets.git/blob - src/common/encconv.cpp
d679bd9b3f10826dc1ed2c452373d2e44a7ca6c6
[wxWidgets.git] / 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 #if wxUSE_FONTMAP
22
23 #include "wx/encconv.h"
24
25 #include <stdlib.h>
26
27 // conversion tables, generated by scripts in $(WXWIN)/misc/unictabl:
28 #if defined( __BORLANDC__ ) || defined(__DARWIN__)
29 #include "../common/unictabl.inc"
30 #else
31 #include "unictabl.inc"
32 #endif
33
34 #if wxUSE_WCHAR_T
35 typedef wchar_t tchar;
36 #else
37 typedef char tchar;
38 #endif
39
40 #ifdef __WXMAC__
41 #include <ATSUnicode.h>
42 #include <TextCommon.h>
43 #include <TextEncodingConverter.h>
44
45 #include "wx/fontutil.h"
46 #include "wx/mac/private.h" // includes mac headers
47
48 wxUint16 gMacEncodings[wxFONTENCODING_MACMAX-wxFONTENCODING_MACMIN+1][128] ;
49 bool gMacEncodingsInited[wxFONTENCODING_MACMAX-wxFONTENCODING_MACMIN+1] ;
50 #endif
51
52 #ifdef __WXWINCE__
53 #include "wx/msw/wince/missing.h" // for bsearch()
54 #endif
55
56 static wxUint16* GetEncTable(wxFontEncoding enc)
57 {
58 #ifdef __WXMAC__
59 if( enc >= wxFONTENCODING_MACMIN && enc <= wxFONTENCODING_MACMAX )
60 {
61 int i = enc-wxFONTENCODING_MACMIN ;
62 if ( gMacEncodingsInited[i] == false )
63 {
64 TECObjectRef converter ;
65 TextEncodingBase code = wxMacGetSystemEncFromFontEnc( enc ) ;
66 TextEncodingBase unicode = CreateTextEncoding(kTextEncodingUnicodeDefault,0,kUnicode16BitFormat) ;
67 OSStatus status = TECCreateConverter(&converter,code,unicode);
68 char s[2] ;
69 s[1] = 0 ;
70 ByteCount byteInLen, byteOutLen ;
71 for( unsigned char c = 255 ; c >= 128 ; --c )
72 {
73 s[0] = c ;
74 status = TECConvertText(converter, (ConstTextPtr) &s , 1, &byteInLen,
75 (TextPtr) &gMacEncodings[i][c-128] , 2, &byteOutLen);
76 }
77 status = TECDisposeConverter(converter);
78 gMacEncodingsInited[i]=true;
79 }
80 return gMacEncodings[i] ;
81 }
82 #endif
83
84 for (int i = 0; encodings_list[i].table != NULL; i++)
85 {
86 if (encodings_list[i].encoding == enc)
87 return encodings_list[i].table;
88 }
89 return NULL;
90 }
91
92 typedef struct {
93 wxUint16 u;
94 wxUint8 c;
95 } CharsetItem;
96
97 extern "C" int wxCMPFUNC_CONV
98 CompareCharsetItems(const void *i1, const void *i2)
99 {
100 return ( ((CharsetItem*)i1) -> u - ((CharsetItem*)i2) -> u );
101 }
102
103
104 static CharsetItem* BuildReverseTable(wxUint16 *tbl)
105 {
106 CharsetItem *rev = new CharsetItem[128];
107
108 for (int i = 0; i < 128; i++)
109 rev[i].c = 128 + i, rev[i].u = tbl[i];
110
111 qsort(rev, 128, sizeof(CharsetItem), CompareCharsetItems);
112
113 return rev;
114 }
115
116
117
118 wxEncodingConverter::wxEncodingConverter()
119 {
120 m_Table = NULL;
121 m_UnicodeInput = m_UnicodeOutput = FALSE;
122 m_JustCopy = FALSE;
123 }
124
125
126
127 bool wxEncodingConverter::Init(wxFontEncoding input_enc, wxFontEncoding output_enc, int method)
128 {
129 unsigned i;
130 wxUint16 *in_tbl, *out_tbl = NULL;
131
132 if (m_Table) {delete[] m_Table; m_Table = NULL;}
133
134 #if !wxUSE_WCHAR_T
135 if (input_enc == wxFONTENCODING_UNICODE || output_enc == wxFONTENCODING_UNICODE) return FALSE;
136 #endif
137
138 if (input_enc == output_enc) {m_JustCopy = TRUE; return TRUE;}
139
140 m_UnicodeOutput = (output_enc == wxFONTENCODING_UNICODE);
141 m_JustCopy = FALSE;
142
143 if (input_enc == wxFONTENCODING_UNICODE)
144 {
145 if ((out_tbl = GetEncTable(output_enc)) == NULL) return FALSE;
146
147 m_Table = new tchar[65536];
148 for (i = 0; i < 128; i++) m_Table[i] = (tchar)i; // 7bit ASCII
149 for (i = 128; i < 65536; i++) m_Table[i] = (tchar)'?';
150 // FIXME - this should be character that means `unicode to charset' impossible, not '?'
151
152 if (method == wxCONVERT_SUBSTITUTE)
153 {
154 for (i = 0; i < encoding_unicode_fallback_count; i++)
155 m_Table[encoding_unicode_fallback[i].c] = (tchar) encoding_unicode_fallback[i].s;
156 }
157
158 for (i = 0; i < 128; i++)
159 m_Table[out_tbl[i]] = (tchar)(128 + i);
160
161 m_UnicodeInput = TRUE;
162 }
163 else // input !Unicode
164 {
165 if ((in_tbl = GetEncTable(input_enc)) == NULL) return FALSE;
166 if (output_enc != wxFONTENCODING_UNICODE)
167 if ((out_tbl = GetEncTable(output_enc)) == NULL) return FALSE;
168
169 m_UnicodeInput = FALSE;
170
171 m_Table = new tchar[256];
172 for (i = 0; i < 128; i++) m_Table[i] = (tchar)i; // 7bit ASCII
173
174 if (output_enc == wxFONTENCODING_UNICODE)
175 {
176 for (i = 0; i < 128; i++) m_Table[128 + i] = (tchar)in_tbl[i];
177 return TRUE;
178 }
179 else // output !Unicode
180 {
181 CharsetItem *rev = BuildReverseTable(out_tbl);
182 CharsetItem *item;
183 CharsetItem key;
184
185 for (i = 0; i < 128; i++)
186 {
187 key.u = in_tbl[i];
188 item = (CharsetItem*) bsearch(&key, rev, 128, sizeof(CharsetItem), CompareCharsetItems);
189 if (item == NULL && method == wxCONVERT_SUBSTITUTE)
190 item = (CharsetItem*) bsearch(&key, encoding_unicode_fallback,
191 encoding_unicode_fallback_count, sizeof(CharsetItem), CompareCharsetItems);
192 if (item)
193 m_Table[128 + i] = (tchar)item -> c;
194 else
195 #if wxUSE_WCHAR_T
196 m_Table[128 + i] = (wchar_t)(128 + i);
197 #else
198 m_Table[128 + i] = (char)(128 + i);
199 #endif
200 }
201
202 delete[] rev;
203 }
204 }
205
206 return TRUE;
207 }
208
209
210
211 void wxEncodingConverter::Convert(const char* input, char* output) const
212 {
213 wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
214 wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
215
216 const char *i;
217 char *o;
218
219 if (m_JustCopy)
220 {
221 strcpy(output, input);
222 return;
223 }
224
225 wxCHECK_RET(m_Table != NULL, wxT("You must call wxEncodingConverter::Init() before actually converting!"));
226
227 for (i = input, o = output; *i != 0;)
228 *(o++) = (char)(m_Table[(wxUint8)*(i++)]);
229 *o = 0;
230 }
231
232
233 #if wxUSE_WCHAR_T
234
235 void wxEncodingConverter::Convert(const char* input, wchar_t* output) const
236 {
237 wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
238 wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
239
240 const char *i;
241 wchar_t *o;
242
243 if (m_JustCopy)
244 {
245 for (i = input, o = output; *i != 0;)
246 *(o++) = (wchar_t)(*(i++));
247 *o = 0;
248 return;
249 }
250
251 wxCHECK_RET(m_Table != NULL, wxT("You must call wxEncodingConverter::Init() before actually converting!"));
252
253 for (i = input, o = output; *i != 0;)
254 *(o++) = (wchar_t)(m_Table[(wxUint8)*(i++)]);
255 *o = 0;
256 }
257
258
259
260 void wxEncodingConverter::Convert(const wchar_t* input, char* output) const
261 {
262 wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
263 wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
264
265 const wchar_t *i;
266 char *o;
267
268 if (m_JustCopy)
269 {
270 for (i = input, o = output; *i != 0;)
271 *(o++) = (char)(*(i++));
272 *o = 0;
273 return;
274 }
275
276 wxCHECK_RET(m_Table != NULL, wxT("You must call wxEncodingConverter::Init() before actually converting!"));
277
278 for (i = input, o = output; *i != 0;)
279 *(o++) = (char)(m_Table[(wxUint16)*(i++)]);
280 *o = 0;
281 }
282
283
284
285 void wxEncodingConverter::Convert(const wchar_t* input, wchar_t* output) const
286 {
287 wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
288 wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
289
290 const wchar_t *i;
291 wchar_t *o;
292
293 if (m_JustCopy)
294 {
295 // wcscpy() is not guaranteed to exist
296 for (i = input, o = output; *i != 0;)
297 *(o++) = (*(i++));
298 *o = 0;
299 return;
300 }
301
302 wxCHECK_RET(m_Table != NULL, wxT("You must call wxEncodingConverter::Init() before actually converting!"));
303
304 for (i = input, o = output; *i != 0;)
305 *(o++) = (wchar_t)(m_Table[(wxUint8)*(i++)]);
306 *o = 0;
307 }
308
309 #endif // wxUSE_WCHAR_T
310
311
312 wxString wxEncodingConverter::Convert(const wxString& input) const
313 {
314 if (m_JustCopy) return input;
315
316 wxString s;
317 const wxChar *i;
318
319 wxCHECK_MSG(m_Table != NULL, s,
320 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
321
322 if (m_UnicodeInput)
323 {
324 for (i = input.c_str(); *i != 0; i++)
325 s << (wxChar)(m_Table[(wxUint16)*i]);
326 }
327 else
328 {
329 for (i = input.c_str(); *i != 0; i++)
330 s << (wxChar)(m_Table[(wxUint8)*i]);
331 }
332
333 return s;
334 }
335
336
337
338
339
340
341
342 // Following tables describe classes of encoding equivalence.
343 //
344
345 #define STOP wxFONTENCODING_SYSTEM
346
347 #define NUM_OF_PLATFORMS 4 /*must conform to enum wxPLATFORM_XXXX !!!*/
348 #define ENC_PER_PLATFORM 5
349 // max no. of encodings for one language used on one platform
350 // Anybody thinks 5 is not enough? ;-)
351
352 static wxFontEncoding
353 EquivalentEncodings[][NUM_OF_PLATFORMS][ENC_PER_PLATFORM+1] = {
354
355 // *** Please put more common encodings as first! ***
356
357 // Western European
358 {
359 /* unix */ {wxFONTENCODING_ISO8859_1, wxFONTENCODING_ISO8859_15, STOP},
360 /* windows */ {wxFONTENCODING_CP1252, STOP},
361 /* os2 */ {STOP},
362 /* mac */ {wxFONTENCODING_MACROMAN, STOP}
363 },
364
365 // Central European
366 {
367 /* unix */ {wxFONTENCODING_ISO8859_2, STOP},
368 /* windows */ {wxFONTENCODING_CP1250, STOP},
369 /* os2 */ {STOP},
370 /* mac */ {wxFONTENCODING_MACCENTRALEUR, STOP}
371 },
372
373 // Baltic
374 {
375 /* unix */ {wxFONTENCODING_ISO8859_13, wxFONTENCODING_ISO8859_4, STOP},
376 /* windows */ {wxFONTENCODING_CP1257, STOP},
377 /* os2 */ {STOP},
378 /* mac */ {STOP}
379 },
380
381 // Hebrew
382 {
383 /* unix */ {wxFONTENCODING_ISO8859_8, STOP},
384 /* windows */ {wxFONTENCODING_CP1255, STOP},
385 /* os2 */ {STOP},
386 /* mac */ {wxFONTENCODING_MACHEBREW, STOP}
387 },
388
389 // Greek
390 {
391 /* unix */ {wxFONTENCODING_ISO8859_7, STOP},
392 /* windows */ {wxFONTENCODING_CP1253, STOP},
393 /* os2 */ {STOP},
394 /* mac */ {wxFONTENCODING_MACGREEK, STOP}
395 },
396
397 // Arabic
398 {
399 /* unix */ {wxFONTENCODING_ISO8859_6, STOP},
400 /* windows */ {wxFONTENCODING_CP1256, STOP},
401 /* os2 */ {STOP},
402 /* mac */ {wxFONTENCODING_MACARABIC, STOP}
403 },
404
405 // Turkish
406 {
407 /* unix */ {wxFONTENCODING_ISO8859_9, STOP},
408 /* windows */ {wxFONTENCODING_CP1254, STOP},
409 /* os2 */ {STOP},
410 /* mac */ {wxFONTENCODING_MACTURKISH, STOP}
411 },
412
413 // Cyrillic
414 {
415 /* unix */ {wxFONTENCODING_KOI8, wxFONTENCODING_KOI8_U, wxFONTENCODING_ISO8859_5, STOP},
416 /* windows */ {wxFONTENCODING_CP1251, STOP},
417 /* os2 */ {STOP},
418 /* mac */ {wxFONTENCODING_MACCYRILLIC, STOP}
419 },
420
421 {{STOP},{STOP},{STOP},{STOP}} /* Terminator */
422 /* no, _not_ Arnold! */
423 };
424
425
426 static bool FindEncoding(const wxFontEncodingArray& arr, wxFontEncoding f)
427 {
428 for (wxFontEncodingArray::const_iterator it = arr.begin(), en = arr.end();
429 it != en; ++it)
430 if (*it == f)
431 return true;
432 return false;
433 }
434
435 wxFontEncodingArray wxEncodingConverter::GetPlatformEquivalents(wxFontEncoding enc, int platform)
436 {
437 if (platform == wxPLATFORM_CURRENT)
438 {
439 #if defined(__WXMSW__)
440 platform = wxPLATFORM_WINDOWS;
441 #elif defined(__WXGTK__) || defined(__WXMOTIF__)
442 platform = wxPLATFORM_UNIX;
443 #elif defined(__WXOS2__)
444 platform = wxPLATFORM_OS2;
445 #elif defined(__WXMAC__)
446 platform = wxPLATFORM_MAC;
447 #endif
448 }
449
450 int i, clas, e ;
451 wxFontEncoding *f;
452 wxFontEncodingArray arr;
453
454 clas = 0;
455 while (EquivalentEncodings[clas][0][0] != STOP)
456 {
457 for (i = 0; i < NUM_OF_PLATFORMS; i++)
458 for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
459 if (EquivalentEncodings[clas][i][e] == enc)
460 {
461 for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
462 if (*f == enc) arr.push_back(enc);
463 for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
464 if (!FindEncoding(arr, *f)) arr.push_back(*f);
465 i = NUM_OF_PLATFORMS/*hack*/; break;
466 }
467 clas++;
468 }
469
470 return arr;
471 }
472
473
474
475 wxFontEncodingArray wxEncodingConverter::GetAllEquivalents(wxFontEncoding enc)
476 {
477 int i, clas, e, j ;
478 wxFontEncoding *f;
479 wxFontEncodingArray arr;
480
481 arr = GetPlatformEquivalents(enc); // we want them to be first items in array
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 (j = 0; j < NUM_OF_PLATFORMS; j++)
491 for (f = EquivalentEncodings[clas][j]; *f != STOP; f++)
492 if (!FindEncoding(arr, *f)) arr.push_back(*f);
493 i = NUM_OF_PLATFORMS/*hack*/; break;
494 }
495 clas++;
496 }
497
498 return arr;
499 }
500
501 #endif // wxUSE_FONTMAP