]> git.saurik.com Git - wxWidgets.git/blob - src/common/encconv.cpp
UNICODE-capatable UTF8 implementation of wxStringXXXStream
[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 = wxUint8(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)0;
150
151 if (method == wxCONVERT_SUBSTITUTE)
152 {
153 for (i = 0; i < encoding_unicode_fallback_count; i++)
154 m_Table[encoding_unicode_fallback[i].c] = (tchar) encoding_unicode_fallback[i].s;
155 }
156
157 for (i = 0; i < 128; i++)
158 m_Table[out_tbl[i]] = (tchar)(128 + i);
159
160 m_UnicodeInput = true;
161 }
162 else // input !Unicode
163 {
164 if ((in_tbl = GetEncTable(input_enc)) == NULL) return false;
165 if (output_enc != wxFONTENCODING_UNICODE)
166 if ((out_tbl = GetEncTable(output_enc)) == NULL) return false;
167
168 m_UnicodeInput = false;
169
170 m_Table = new tchar[256];
171 for (i = 0; i < 128; i++) m_Table[i] = (tchar)i; // 7bit ASCII
172
173 if (output_enc == wxFONTENCODING_UNICODE)
174 {
175 for (i = 0; i < 128; i++) m_Table[128 + i] = (tchar)in_tbl[i];
176 return true;
177 }
178 else // output !Unicode
179 {
180 CharsetItem *rev = BuildReverseTable(out_tbl);
181 CharsetItem *item;
182 CharsetItem key;
183
184 for (i = 0; i < 128; i++)
185 {
186 key.u = in_tbl[i];
187 item = (CharsetItem*) bsearch(&key, rev, 128, sizeof(CharsetItem), CompareCharsetItems);
188 if (item == NULL && method == wxCONVERT_SUBSTITUTE)
189 item = (CharsetItem*) bsearch(&key, encoding_unicode_fallback,
190 encoding_unicode_fallback_count, sizeof(CharsetItem), CompareCharsetItems);
191 if (item)
192 m_Table[128 + i] = (tchar)item -> c;
193 else
194 #if wxUSE_WCHAR_T
195 m_Table[128 + i] = (wchar_t)(128 + i);
196 #else
197 m_Table[128 + i] = (char)(128 + i);
198 #endif
199 }
200
201 delete[] rev;
202 }
203 }
204
205 return true;
206 }
207
208
209 #define REPLACEMENT_CHAR ((tchar)'?')
210
211 inline tchar GetTableValue(const tchar *table, tchar value, bool& repl)
212 {
213 tchar r = table[value];
214 if (r == 0 && value != 0)
215 {
216 r = REPLACEMENT_CHAR;
217 repl = true;
218 }
219 return r;
220 }
221
222
223 bool wxEncodingConverter::Convert(const char* input, char* output) const
224 {
225 wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
226 wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
227
228 const char *i;
229 char *o;
230
231 if (m_JustCopy)
232 {
233 strcpy(output, input);
234 return true;
235 }
236
237 wxCHECK_MSG(m_Table != NULL, false,
238 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
239
240 bool replaced = false;
241
242 for (i = input, o = output; *i != 0;)
243 *(o++) = (char)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
244 *o = 0;
245
246 return !replaced;
247 }
248
249
250 #if wxUSE_WCHAR_T
251
252 bool wxEncodingConverter::Convert(const char* input, wchar_t* output) const
253 {
254 wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
255 wxASSERT_MSG(!m_UnicodeInput, wxT("You cannot convert from unicode if input is const char*!"));
256
257 const char *i;
258 wchar_t *o;
259
260 if (m_JustCopy)
261 {
262 for (i = input, o = output; *i != 0;)
263 *(o++) = (wchar_t)(*(i++));
264 *o = 0;
265 return true;
266 }
267
268 wxCHECK_MSG(m_Table != NULL, false,
269 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
270
271 bool replaced = false;
272
273 for (i = input, o = output; *i != 0;)
274 *(o++) = (wchar_t)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
275 *o = 0;
276
277 return !replaced;
278 }
279
280
281
282 bool wxEncodingConverter::Convert(const wchar_t* input, char* output) const
283 {
284 wxASSERT_MSG(!m_UnicodeOutput, wxT("You cannot convert to unicode if output is const char*!"));
285 wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
286
287 const wchar_t *i;
288 char *o;
289
290 if (m_JustCopy)
291 {
292 for (i = input, o = output; *i != 0;)
293 *(o++) = (char)(*(i++));
294 *o = 0;
295 return true;
296 }
297
298 wxCHECK_MSG(m_Table != NULL, false,
299 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
300
301 bool replaced = false;
302
303 for (i = input, o = output; *i != 0;)
304 *(o++) = (char)(GetTableValue(m_Table, (wxUint16)*(i++), replaced));
305 *o = 0;
306
307 return !replaced;
308 }
309
310
311
312 bool wxEncodingConverter::Convert(const wchar_t* input, wchar_t* output) const
313 {
314 wxASSERT_MSG(m_UnicodeOutput, wxT("You cannot convert to 8-bit if output is const wchar_t*!"));
315 wxASSERT_MSG(m_UnicodeInput, wxT("You cannot convert from 8-bit if input is const wchar_t*!"));
316
317 const wchar_t *i;
318 wchar_t *o;
319
320 if (m_JustCopy)
321 {
322 // wcscpy() is not guaranteed to exist
323 for (i = input, o = output; *i != 0;)
324 *(o++) = (*(i++));
325 *o = 0;
326 return true;
327 }
328
329 wxCHECK_MSG(m_Table != NULL, false,
330 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
331
332 bool replaced = false;
333
334 for (i = input, o = output; *i != 0;)
335 *(o++) = (wchar_t)(GetTableValue(m_Table, (wxUint8)*(i++), replaced));
336 *o = 0;
337
338 return !replaced;
339 }
340
341 #endif // wxUSE_WCHAR_T
342
343
344 wxString wxEncodingConverter::Convert(const wxString& input) const
345 {
346 if (m_JustCopy) return input;
347
348 wxString s;
349 const wxChar *i;
350
351 wxCHECK_MSG(m_Table != NULL, s,
352 wxT("You must call wxEncodingConverter::Init() before actually converting!"));
353
354 if (m_UnicodeInput)
355 {
356 for (i = input.c_str(); *i != 0; i++)
357 s << (wxChar)(m_Table[(wxUint16)*i]);
358 }
359 else
360 {
361 for (i = input.c_str(); *i != 0; i++)
362 s << (wxChar)(m_Table[(wxUint8)*i]);
363 }
364
365 return s;
366 }
367
368
369
370
371
372
373
374 // Following tables describe classes of encoding equivalence.
375 //
376
377 #define STOP wxFONTENCODING_SYSTEM
378
379 #define NUM_OF_PLATFORMS 4 /*must conform to enum wxPLATFORM_XXXX !!!*/
380 #define ENC_PER_PLATFORM 5
381 // max no. of encodings for one language used on one platform
382 // Anybody thinks 5 is not enough? ;-)
383
384 static wxFontEncoding
385 EquivalentEncodings[][NUM_OF_PLATFORMS][ENC_PER_PLATFORM+1] = {
386
387 // *** Please put more common encodings as first! ***
388
389 // Western European
390 {
391 /* unix */ {wxFONTENCODING_ISO8859_1, wxFONTENCODING_ISO8859_15, STOP},
392 /* windows */ {wxFONTENCODING_CP1252, STOP},
393 /* os2 */ {STOP},
394 /* mac */ {wxFONTENCODING_MACROMAN, STOP}
395 },
396
397 // Central European
398 {
399 /* unix */ {wxFONTENCODING_ISO8859_2, STOP},
400 /* windows */ {wxFONTENCODING_CP1250, STOP},
401 /* os2 */ {STOP},
402 /* mac */ {wxFONTENCODING_MACCENTRALEUR, STOP}
403 },
404
405 // Baltic
406 {
407 /* unix */ {wxFONTENCODING_ISO8859_13, wxFONTENCODING_ISO8859_4, STOP},
408 /* windows */ {wxFONTENCODING_CP1257, STOP},
409 /* os2 */ {STOP},
410 /* mac */ {STOP}
411 },
412
413 // Hebrew
414 {
415 /* unix */ {wxFONTENCODING_ISO8859_8, STOP},
416 /* windows */ {wxFONTENCODING_CP1255, STOP},
417 /* os2 */ {STOP},
418 /* mac */ {wxFONTENCODING_MACHEBREW, STOP}
419 },
420
421 // Greek
422 {
423 /* unix */ {wxFONTENCODING_ISO8859_7, STOP},
424 /* windows */ {wxFONTENCODING_CP1253, STOP},
425 /* os2 */ {STOP},
426 /* mac */ {wxFONTENCODING_MACGREEK, STOP}
427 },
428
429 // Arabic
430 {
431 /* unix */ {wxFONTENCODING_ISO8859_6, STOP},
432 /* windows */ {wxFONTENCODING_CP1256, STOP},
433 /* os2 */ {STOP},
434 /* mac */ {wxFONTENCODING_MACARABIC, STOP}
435 },
436
437 // Turkish
438 {
439 /* unix */ {wxFONTENCODING_ISO8859_9, STOP},
440 /* windows */ {wxFONTENCODING_CP1254, STOP},
441 /* os2 */ {STOP},
442 /* mac */ {wxFONTENCODING_MACTURKISH, STOP}
443 },
444
445 // Cyrillic
446 {
447 /* unix */ {wxFONTENCODING_KOI8, wxFONTENCODING_KOI8_U, wxFONTENCODING_ISO8859_5, STOP},
448 /* windows */ {wxFONTENCODING_CP1251, STOP},
449 /* os2 */ {STOP},
450 /* mac */ {wxFONTENCODING_MACCYRILLIC, STOP}
451 },
452
453 {{STOP},{STOP},{STOP},{STOP}} /* Terminator */
454 /* no, _not_ Arnold! */
455 };
456
457
458 static bool FindEncoding(const wxFontEncodingArray& arr, wxFontEncoding f)
459 {
460 for (wxFontEncodingArray::const_iterator it = arr.begin(), en = arr.end();
461 it != en; ++it)
462 if (*it == f)
463 return true;
464 return false;
465 }
466
467 wxFontEncodingArray wxEncodingConverter::GetPlatformEquivalents(wxFontEncoding enc, int platform)
468 {
469 if (platform == wxPLATFORM_CURRENT)
470 {
471 #if defined(__WXMSW__)
472 platform = wxPLATFORM_WINDOWS;
473 #elif defined(__WXGTK__) || defined(__WXMOTIF__)
474 platform = wxPLATFORM_UNIX;
475 #elif defined(__WXOS2__)
476 platform = wxPLATFORM_OS2;
477 #elif defined(__WXMAC__)
478 platform = wxPLATFORM_MAC;
479 #endif
480 }
481
482 int i, clas, e ;
483 wxFontEncoding *f;
484 wxFontEncodingArray arr;
485
486 clas = 0;
487 while (EquivalentEncodings[clas][0][0] != STOP)
488 {
489 for (i = 0; i < NUM_OF_PLATFORMS; i++)
490 for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
491 if (EquivalentEncodings[clas][i][e] == enc)
492 {
493 for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
494 if (*f == enc) arr.push_back(enc);
495 for (f = EquivalentEncodings[clas][platform]; *f != STOP; f++)
496 if (!FindEncoding(arr, *f)) arr.push_back(*f);
497 i = NUM_OF_PLATFORMS/*hack*/; break;
498 }
499 clas++;
500 }
501
502 return arr;
503 }
504
505
506
507 wxFontEncodingArray wxEncodingConverter::GetAllEquivalents(wxFontEncoding enc)
508 {
509 int i, clas, e, j ;
510 wxFontEncoding *f;
511 wxFontEncodingArray arr;
512
513 arr = GetPlatformEquivalents(enc); // we want them to be first items in array
514
515 clas = 0;
516 while (EquivalentEncodings[clas][0][0] != STOP)
517 {
518 for (i = 0; i < NUM_OF_PLATFORMS; i++)
519 for (e = 0; EquivalentEncodings[clas][i][e] != STOP; e++)
520 if (EquivalentEncodings[clas][i][e] == enc)
521 {
522 for (j = 0; j < NUM_OF_PLATFORMS; j++)
523 for (f = EquivalentEncodings[clas][j]; *f != STOP; f++)
524 if (!FindEncoding(arr, *f)) arr.push_back(*f);
525 i = NUM_OF_PLATFORMS/*hack*/; break;
526 }
527 clas++;
528 }
529
530 return arr;
531 }
532
533 #endif // wxUSE_FONTMAP