src/common/strconv.cpp

   1 /////////////////////////////////////////////////////////////////////////////
   2 // Name:        src/common/strconv.cpp
   3 // Purpose:     Unicode conversion classes
   4 // Author:      Ove Kaaven, Robert Roebling, Vadim Zeitlin, Vaclav Slavik,
   5 //              Ryan Norton, Fredrik Roubert (UTF7)
   6 // Modified by:
   7 // Created:     29/01/98
   8 // RCS-ID:      $Id$
   9 // Copyright:   (c) 1999 Ove Kaaven, Robert Roebling, Vaclav Slavik
  10 //              (c) 2000-2003 Vadim Zeitlin
  11 //              (c) 2004 Ryan Norton, Fredrik Roubert
  12 // Licence:     wxWindows licence
  13 /////////////////////////////////////////////////////////////////////////////
  14
  15 // ============================================================================
  16 // declarations
  17 // ============================================================================
  18
  19 // ----------------------------------------------------------------------------
  20 // headers
  21 // ----------------------------------------------------------------------------
  22
  23 // For compilers that support precompilation, includes "wx.h".
  24 #include "wx/wxprec.h"
  25
  26 #ifdef __BORLANDC__
  27   #pragma hdrstop
  28 #endif
  29
  30 #ifndef WX_PRECOMP
  31     #include "wx/intl.h"
  32     #include "wx/log.h"
  33 #endif // WX_PRECOMP
  34
  35 #include "wx/strconv.h"
  36
  37 #if wxUSE_WCHAR_T
  38
  39 #ifdef __WINDOWS__
  40     #include "wx/msw/private.h"
  41     #include "wx/msw/missing.h"
  42 #endif
  43
  44 #ifndef __WXWINCE__
  45 #include <errno.h>
  46 #endif
  47
  48 #include <ctype.h>
  49 #include <string.h>
  50 #include <stdlib.h>
  51
  52 #if defined(__WIN32__) && !defined(__WXMICROWIN__)
  53     #define wxHAVE_WIN32_MB2WC
  54 #endif // __WIN32__ but !__WXMICROWIN__
  55
  56 #ifdef __SALFORDC__
  57     #include <clib.h>
  58 #endif
  59
  60 #ifdef HAVE_ICONV
  61     #include <iconv.h>
  62     #include "wx/thread.h"
  63 #endif
  64
  65 #include "wx/encconv.h"
  66 #include "wx/fontmap.h"
  67 #include "wx/utils.h"
  68
  69 #ifdef __WXMAC__
  70 #ifndef __DARWIN__
  71 #include <ATSUnicode.h>
  72 #include <TextCommon.h>
  73 #include <TextEncodingConverter.h>
  74 #endif
  75
  76 #include  "wx/mac/private.h"  // includes mac headers
  77 #endif
  78
  79 #define TRACE_STRCONV _T("strconv")
  80
  81 #if SIZEOF_WCHAR_T == 2
  82     #define WC_UTF16
  83 #endif
  84
  85 // ============================================================================
  86 // implementation
  87 // ============================================================================
  88
  89 // ----------------------------------------------------------------------------
  90 // UTF-16 en/decoding to/from UCS-4
  91 // ----------------------------------------------------------------------------
  92
  93
  94 static size_t encode_utf16(wxUint32 input, wxUint16 *output)
  95 {
  96     if (input<=0xffff)
  97     {
  98         if (output)
  99             *output = (wxUint16) input;
 100         return 1;
 101     }
 102     else if (input>=0x110000)
 103     {
 104         return (size_t)-1;
 105     }
 106     else
 107     {
 108         if (output)
 109         {
 110             *output++ = (wxUint16) ((input >> 10)+0xd7c0);
 111             *output = (wxUint16) ((input&0x3ff)+0xdc00);
 112         }
 113         return 2;
 114     }
 115 }
 116
 117 static size_t decode_utf16(const wxUint16* input, wxUint32& output)
 118 {
 119     if ((*input<0xd800) || (*input>0xdfff))
 120     {
 121         output = *input;
 122         return 1;
 123     }
 124     else if ((input[1]<0xdc00) || (input[1]>0xdfff))
 125     {
 126         output = *input;
 127         return (size_t)-1;
 128     }
 129     else
 130     {
 131         output = ((input[0] - 0xd7c0) << 10) + (input[1] - 0xdc00);
 132         return 2;
 133     }
 134 }
 135
 136
 137 // ----------------------------------------------------------------------------
 138 // wxMBConv
 139 // ----------------------------------------------------------------------------
 140
 141 wxMBConv::~wxMBConv()
 142 {
 143     // nothing to do here (necessary for Darwin linking probably)
 144 }
 145
 146 const wxWCharBuffer wxMBConv::cMB2WC(const char *psz) const
 147 {
 148     if ( psz )
 149     {
 150         // calculate the length of the buffer needed first
 151         size_t nLen = MB2WC(NULL, psz, 0);
 152         if ( nLen != (size_t)-1 )
 153         {
 154             // now do the actual conversion
 155             wxWCharBuffer buf(nLen);
 156             nLen = MB2WC(buf.data(), psz, nLen + 1); // with the trailing NULL
 157             if ( nLen != (size_t)-1 )
 158             {
 159                 return buf;
 160             }
 161         }
 162     }
 163
 164     wxWCharBuffer buf((wchar_t *)NULL);
 165
 166     return buf;
 167 }
 168
 169 const wxCharBuffer wxMBConv::cWC2MB(const wchar_t *pwz) const
 170 {
 171     if ( pwz )
 172     {
 173         size_t nLen = WC2MB(NULL, pwz, 0);
 174         if ( nLen != (size_t)-1 )
 175         {
 176             wxCharBuffer buf(nLen+3);       // space for a wxUint32 trailing zero
 177             nLen = WC2MB(buf.data(), pwz, nLen + 4);
 178             if ( nLen != (size_t)-1 )
 179             {
 180                 return buf;
 181             }
 182         }
 183     }
 184
 185     wxCharBuffer buf((char *)NULL);
 186
 187     return buf;
 188 }
 189
 190 const wxWCharBuffer wxMBConv::cMB2WC(const char *szString, size_t nStringLen, size_t* pOutSize) const
 191 {
 192     wxASSERT(pOutSize != NULL);
 193
 194     const char* szEnd = szString + nStringLen + 1;
 195     const char* szPos = szString;
 196     const char* szStart = szPos;
 197
 198     size_t nActualLength = 0;
 199     size_t nCurrentSize = nStringLen; //try normal size first (should never resize?)
 200
 201     wxWCharBuffer theBuffer(nCurrentSize);
 202
 203     //Convert the string until the length() is reached, continuing the
 204     //loop every time a null character is reached
 205     while(szPos != szEnd)
 206     {
 207         wxASSERT(szPos < szEnd); //something is _really_ screwed up if this rings true
 208
 209         //Get the length of the current (sub)string
 210         size_t nLen = MB2WC(NULL, szPos, 0);
 211
 212         //Invalid conversion?
 213         if( nLen == (size_t)-1 )
 214         {
 215             *pOutSize = 0;
 216             theBuffer.data()[0u] = wxT('\0');
 217             return theBuffer;
 218         }
 219
 220
 221         //Increase the actual length (+1 for current null character)
 222         nActualLength += nLen + 1;
 223
 224         //if buffer too big, realloc the buffer
 225         if (nActualLength > (nCurrentSize+1))
 226         {
 227             wxWCharBuffer theNewBuffer(nCurrentSize << 1);
 228             memcpy(theNewBuffer.data(), theBuffer.data(), nCurrentSize * sizeof(wchar_t));
 229             theBuffer = theNewBuffer;
 230             nCurrentSize <<= 1;
 231         }
 232
 233         //Convert the current (sub)string
 234         if ( MB2WC(&theBuffer.data()[szPos - szStart], szPos, nLen + 1) == (size_t)-1 )
 235         {
 236             *pOutSize = 0;
 237             theBuffer.data()[0u] = wxT('\0');
 238             return theBuffer;
 239         }
 240
 241         //Increment to next (sub)string
 242         //Note that we have to use strlen instead of nLen here
 243         //because XX2XX gives us the size of the output buffer,
 244         //which is not necessarily the length of the string
 245         szPos += strlen(szPos) + 1;
 246     }
 247
 248     //success - return actual length and the buffer
 249     *pOutSize = nActualLength;
 250     return theBuffer;
 251 }
 252
 253 const wxCharBuffer wxMBConv::cWC2MB(const wchar_t *szString, size_t nStringLen, size_t* pOutSize) const
 254 {
 255     wxASSERT(pOutSize != NULL);
 256
 257     const wchar_t* szEnd = szString + nStringLen + 1;
 258     const wchar_t* szPos = szString;
 259     const wchar_t* szStart = szPos;
 260
 261     size_t nActualLength = 0;
 262     size_t nCurrentSize = nStringLen << 2; //try * 4 first
 263
 264     wxCharBuffer theBuffer(nCurrentSize);
 265
 266     //Convert the string until the length() is reached, continuing the
 267     //loop every time a null character is reached
 268     while(szPos != szEnd)
 269     {
 270         wxASSERT(szPos < szEnd); //something is _really_ screwed up if this rings true
 271
 272         //Get the length of the current (sub)string
 273         size_t nLen = WC2MB(NULL, szPos, 0);
 274
 275         //Invalid conversion?
 276         if( nLen == (size_t)-1 )
 277         {
 278             *pOutSize = 0;
 279             theBuffer.data()[0u] = wxT('\0');
 280             return theBuffer;
 281         }
 282
 283         //Increase the actual length (+1 for current null character)
 284         nActualLength += nLen + 1;
 285
 286         //if buffer too big, realloc the buffer
 287         if (nActualLength > (nCurrentSize+1))
 288         {
 289             wxCharBuffer theNewBuffer(nCurrentSize << 1);
 290             memcpy(theNewBuffer.data(), theBuffer.data(), nCurrentSize);
 291             theBuffer = theNewBuffer;
 292             nCurrentSize <<= 1;
 293         }
 294
 295         //Convert the current (sub)string
 296         if(WC2MB(&theBuffer.data()[szPos - szStart], szPos, nLen + 1) == (size_t)-1 )
 297         {
 298             *pOutSize = 0;
 299             theBuffer.data()[0u] = wxT('\0');
 300             return theBuffer;
 301         }
 302
 303         //Increment to next (sub)string
 304         //Note that we have to use wxWcslen instead of nLen here
 305         //because XX2XX gives us the size of the output buffer,
 306         //which is not necessarily the length of the string
 307         szPos += wxWcslen(szPos) + 1;
 308     }
 309
 310     //success - return actual length and the buffer
 311     *pOutSize = nActualLength;
 312     return theBuffer;
 313 }
 314
 315 // ----------------------------------------------------------------------------
 316 // wxMBConvLibc
 317 // ----------------------------------------------------------------------------
 318
 319 size_t wxMBConvLibc::MB2WC(wchar_t *buf, const char *psz, size_t n) const
 320 {
 321     return wxMB2WC(buf, psz, n);
 322 }
 323
 324 size_t wxMBConvLibc::WC2MB(char *buf, const wchar_t *psz, size_t n) const
 325 {
 326     return wxWC2MB(buf, psz, n);
 327 }
 328
 329 #ifdef __UNIX__
 330
 331 // ----------------------------------------------------------------------------
 332 // wxConvBrokenFileNames
 333 // ----------------------------------------------------------------------------
 334
 335 wxConvBrokenFileNames::wxConvBrokenFileNames(const wxChar *charset)
 336 {
 337     if ( !charset || wxStricmp(charset, _T("UTF-8")) == 0
 338                   || wxStricmp(charset, _T("UTF8")) == 0  )
 339         m_conv = new wxMBConvUTF8(wxMBConvUTF8::MAP_INVALID_UTF8_TO_OCTAL);
 340     else
 341         m_conv = new wxCSConv(charset);
 342 }
 343
 344 size_t
 345 wxConvBrokenFileNames::MB2WC(wchar_t *outputBuf,
 346                              const char *psz,
 347                              size_t outputSize) const
 348 {
 349     return m_conv->MB2WC( outputBuf, psz, outputSize );
 350 }
 351
 352 size_t
 353 wxConvBrokenFileNames::WC2MB(char *outputBuf,
 354                              const wchar_t *psz,
 355                              size_t outputSize) const
 356 {
 357     return m_conv->WC2MB( outputBuf, psz, outputSize );
 358 }
 359
 360 #endif
 361
 362 // ----------------------------------------------------------------------------
 363 // UTF-7
 364 // ----------------------------------------------------------------------------
 365
 366 // Implementation (C) 2004 Fredrik Roubert
 367
 368 //
 369 // BASE64 decoding table
 370 //
 371 static const unsigned char utf7unb64[] =
 372 {
 373     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 374     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 375     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 376     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 377     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 378     0xff, 0xff, 0xff, 0x3e, 0xff, 0xff, 0xff, 0x3f,
 379     0x34, 0x35, 0x36, 0x37, 0x38, 0x39, 0x3a, 0x3b,
 380     0x3c, 0x3d, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 381     0xff, 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06,
 382     0x07, 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e,
 383     0x0f, 0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16,
 384     0x17, 0x18, 0x19, 0xff, 0xff, 0xff, 0xff, 0xff,
 385     0xff, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f, 0x20,
 386     0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27, 0x28,
 387     0x29, 0x2a, 0x2b, 0x2c, 0x2d, 0x2e, 0x2f, 0x30,
 388     0x31, 0x32, 0x33, 0xff, 0xff, 0xff, 0xff, 0xff,
 389     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 390     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 391     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 392     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 393     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 394     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 395     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 396     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 397     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 398     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 399     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 400     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 401     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 402     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 403     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
 404     0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff
 405 };
 406
 407 size_t wxMBConvUTF7::MB2WC(wchar_t *buf, const char *psz, size_t n) const
 408 {
 409     size_t len = 0;
 410
 411     while (*psz && ((!buf) || (len < n)))
 412     {
 413         unsigned char cc = *psz++;
 414         if (cc != '+')
 415         {
 416             // plain ASCII char
 417             if (buf)
 418                 *buf++ = cc;
 419             len++;
 420         }
 421         else if (*psz == '-')
 422         {
 423             // encoded plus sign
 424             if (buf)
 425                 *buf++ = cc;
 426             len++;
 427             psz++;
 428         }
 429         else
 430         {
 431             // BASE64 encoded string
 432             bool lsb;
 433             unsigned char c;
 434             unsigned int d, l;
 435             for (lsb = false, d = 0, l = 0;
 436                 (cc = utf7unb64[(unsigned char)*psz]) != 0xff; psz++)
 437             {
 438                 d <<= 6;
 439                 d += cc;
 440                 for (l += 6; l >= 8; lsb = !lsb)
 441                 {
 442                     c = (unsigned char)((d >> (l -= 8)) % 256);
 443                     if (lsb)
 444                     {
 445                         if (buf)
 446                             *buf++ |= c;
 447                         len ++;
 448                     }
 449                     else
 450                         if (buf)
 451                             *buf = (wchar_t)(c << 8);
 452                 }
 453             }
 454             if (*psz == '-')
 455                 psz++;
 456         }
 457     }
 458     if (buf && (len < n))
 459         *buf = 0;
 460     return len;
 461 }
 462
 463 //
 464 // BASE64 encoding table
 465 //
 466 static const unsigned char utf7enb64[] =
 467 {
 468     'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H',
 469     'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P',
 470     'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X',
 471     'Y', 'Z', 'a', 'b', 'c', 'd', 'e', 'f',
 472     'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n',
 473     'o', 'p', 'q', 'r', 's', 't', 'u', 'v',
 474     'w', 'x', 'y', 'z', '0', '1', '2', '3',
 475     '4', '5', '6', '7', '8', '9', '+', '/'
 476 };
 477
 478 //
 479 // UTF-7 encoding table
 480 //
 481 // 0 - Set D (directly encoded characters)
 482 // 1 - Set O (optional direct characters)
 483 // 2 - whitespace characters (optional)
 484 // 3 - special characters
 485 //
 486 static const unsigned char utf7encode[128] =
 487 {
 488     3, 3, 3, 3, 3, 3, 3, 3, 3, 2, 2, 3, 3, 2, 3, 3,
 489     3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
 490     2, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 3, 0, 0, 0, 3,
 491     0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0,
 492     1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
 493     0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 3, 1, 1, 1,
 494     1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
 495     0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 3, 3
 496 };
 497
 498 size_t wxMBConvUTF7::WC2MB(char *buf, const wchar_t *psz, size_t n) const
 499 {
 500
 501
 502     size_t len = 0;
 503
 504     while (*psz && ((!buf) || (len < n)))
 505     {
 506         wchar_t cc = *psz++;
 507         if (cc < 0x80 && utf7encode[cc] < 1)
 508         {
 509             // plain ASCII char
 510             if (buf)
 511                 *buf++ = (char)cc;
 512             len++;
 513         }
 514 #ifndef WC_UTF16
 515         else if (((wxUint32)cc) > 0xffff)
 516         {
 517             // no surrogate pair generation (yet?)
 518             return (size_t)-1;
 519         }
 520 #endif
 521         else
 522         {
 523             if (buf)
 524                 *buf++ = '+';
 525             len++;
 526             if (cc != '+')
 527             {
 528                 // BASE64 encode string
 529                 unsigned int lsb, d, l;
 530                 for (d = 0, l = 0; /*nothing*/; psz++)
 531                 {
 532                     for (lsb = 0; lsb < 2; lsb ++)
 533                     {
 534                         d <<= 8;
 535                         d += lsb ? cc & 0xff : (cc & 0xff00) >> 8;
 536
 537                         for (l += 8; l >= 6; )
 538                         {
 539                             l -= 6;
 540                             if (buf)
 541                                 *buf++ = utf7enb64[(d >> l) % 64];
 542                             len++;
 543                         }
 544                     }
 545                     cc = *psz;
 546                     if (!(cc) || (cc < 0x80 && utf7encode[cc] < 1))
 547                         break;
 548                 }
 549                 if (l != 0)
 550                 {
 551                     if (buf)
 552                         *buf++ = utf7enb64[((d % 16) << (6 - l)) % 64];
 553                     len++;
 554                 }
 555             }
 556             if (buf)
 557                 *buf++ = '-';
 558             len++;
 559         }
 560     }
 561     if (buf && (len < n))
 562         *buf = 0;
 563     return len;
 564 }
 565
 566 // ----------------------------------------------------------------------------
 567 // UTF-8
 568 // ----------------------------------------------------------------------------
 569
 570 static wxUint32 utf8_max[]=
 571     { 0x7f, 0x7ff, 0xffff, 0x1fffff, 0x3ffffff, 0x7fffffff, 0xffffffff };
 572
 573 // boundaries of the private use area we use to (temporarily) remap invalid
 574 // characters invalid in a UTF-8 encoded string
 575 const wxUint32 wxUnicodePUA = 0x100000;
 576 const wxUint32 wxUnicodePUAEnd = wxUnicodePUA + 256;
 577
 578 size_t wxMBConvUTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const
 579 {
 580     size_t len = 0;
 581
 582     while (*psz && ((!buf) || (len < n)))
 583     {
 584         const char *opsz = psz;
 585         bool invalid = false;
 586         unsigned char cc = *psz++, fc = cc;
 587         unsigned cnt;
 588         for (cnt = 0; fc & 0x80; cnt++)
 589             fc <<= 1;
 590         if (!cnt)
 591         {
 592             // plain ASCII char
 593             if (buf)
 594                 *buf++ = cc;
 595             len++;
 596
 597             // escape the escape character for octal escapes
 598             if ((m_options & MAP_INVALID_UTF8_TO_OCTAL)
 599                     && cc == '\\' && (!buf || len < n))
 600             {
 601                 if (buf)
 602                     *buf++ = cc;
 603                 len++;
 604             }
 605         }
 606         else
 607         {
 608             cnt--;
 609             if (!cnt)
 610             {
 611                 // invalid UTF-8 sequence
 612                 invalid = true;
 613             }
 614             else
 615             {
 616                 unsigned ocnt = cnt - 1;
 617                 wxUint32 res = cc & (0x3f >> cnt);
 618                 while (cnt--)
 619                 {
 620                     cc = *psz;
 621                     if ((cc & 0xC0) != 0x80)
 622                     {
 623                         // invalid UTF-8 sequence
 624                         invalid = true;
 625                         break;
 626                     }
 627                     psz++;
 628                     res = (res << 6) | (cc & 0x3f);
 629                 }
 630                 if (invalid || res <= utf8_max[ocnt])
 631                 {
 632                     // illegal UTF-8 encoding
 633                     invalid = true;
 634                 }
 635                 else if ((m_options & MAP_INVALID_UTF8_TO_PUA) &&
 636                         res >= wxUnicodePUA && res < wxUnicodePUAEnd)
 637                 {
 638                     // if one of our PUA characters turns up externally
 639                     // it must also be treated as an illegal sequence
 640                     // (a bit like you have to escape an escape character)
 641                     invalid = true;
 642                 }
 643                 else
 644                 {
 645 #ifdef WC_UTF16
 646                     // cast is ok because wchar_t == wxUuint16 if WC_UTF16
 647                     size_t pa = encode_utf16(res, (wxUint16 *)buf);
 648                     if (pa == (size_t)-1)
 649                     {
 650                         invalid = true;
 651                     }
 652                     else
 653                     {
 654                         if (buf)
 655                             buf += pa;
 656                         len += pa;
 657                     }
 658 #else // !WC_UTF16
 659                     if (buf)
 660                         *buf++ = (wchar_t)res;
 661                     len++;
 662 #endif // WC_UTF16/!WC_UTF16
 663                 }
 664             }
 665             if (invalid)
 666             {
 667                 if (m_options & MAP_INVALID_UTF8_TO_PUA)
 668                 {
 669                     while (opsz < psz && (!buf || len < n))
 670                     {
 671 #ifdef WC_UTF16
 672                         // cast is ok because wchar_t == wxUuint16 if WC_UTF16
 673                         size_t pa = encode_utf16((unsigned char)*opsz + wxUnicodePUA, (wxUint16 *)buf);
 674                         wxASSERT(pa != (size_t)-1);
 675                         if (buf)
 676                             buf += pa;
 677                         opsz++;
 678                         len += pa;
 679 #else
 680                         if (buf)
 681                             *buf++ = (wchar_t)(wxUnicodePUA + (unsigned char)*opsz);
 682                         opsz++;
 683                         len++;
 684 #endif
 685                     }
 686                 }
 687                 else if (m_options & MAP_INVALID_UTF8_TO_OCTAL)
 688                 {
 689                     while (opsz < psz && (!buf || len < n))
 690                     {
 691                         if ( buf && len + 3 < n )
 692                         {
 693                             unsigned char on = *opsz;
 694                             *buf++ = L'\\';
 695                             *buf++ = (wchar_t)( L'0' + on / 0100 );
 696                             *buf++ = (wchar_t)( L'0' + (on % 0100) / 010 );
 697                             *buf++ = (wchar_t)( L'0' + on % 010 );
 698                         }
 699                         opsz++;
 700                         len += 4;
 701                     }
 702                 }
 703                 else // MAP_INVALID_UTF8_NOT
 704                 {
 705                     return (size_t)-1;
 706                 }
 707             }
 708         }
 709     }
 710     if (buf && (len < n))
 711         *buf = 0;
 712     return len;
 713 }
 714
 715 static inline bool isoctal(wchar_t wch)
 716 {
 717     return L'0' <= wch && wch <= L'7';
 718 }
 719
 720 size_t wxMBConvUTF8::WC2MB(char *buf, const wchar_t *psz, size_t n) const
 721 {
 722     size_t len = 0;
 723
 724     while (*psz && ((!buf) || (len < n)))
 725     {
 726         wxUint32 cc;
 727 #ifdef WC_UTF16
 728         // cast is ok for WC_UTF16
 729         size_t pa = decode_utf16((const wxUint16 *)psz, cc);
 730         psz += (pa == (size_t)-1) ? 1 : pa;
 731 #else
 732         cc=(*psz++) & 0x7fffffff;
 733 #endif
 734
 735         if ( (m_options & MAP_INVALID_UTF8_TO_PUA)
 736                 && cc >= wxUnicodePUA && cc < wxUnicodePUAEnd )
 737         {
 738             if (buf)
 739                 *buf++ = (char)(cc - wxUnicodePUA);
 740             len++;
 741         }
 742         else if ( (m_options & MAP_INVALID_UTF8_TO_OCTAL)
 743                     && cc == L'\\' && psz[0] == L'\\' )
 744         {
 745             if (buf)
 746                 *buf++ = (char)cc;
 747             psz++;
 748             len++;
 749         }
 750         else if ( (m_options & MAP_INVALID_UTF8_TO_OCTAL) &&
 751                     cc == L'\\' &&
 752                         isoctal(psz[0]) && isoctal(psz[1]) && isoctal(psz[2]) )
 753         {
 754             if (buf)
 755             {
 756                 *buf++ = (char) ((psz[0] - L'0')*0100 +
 757                                  (psz[1] - L'0')*010 +
 758                                  (psz[2] - L'0'));
 759             }
 760
 761             psz += 3;
 762             len++;
 763         }
 764         else
 765         {
 766             unsigned cnt;
 767             for (cnt = 0; cc > utf8_max[cnt]; cnt++) {}
 768             if (!cnt)
 769             {
 770                 // plain ASCII char
 771                 if (buf)
 772                     *buf++ = (char) cc;
 773                 len++;
 774             }
 775
 776             else
 777             {
 778                 len += cnt + 1;
 779                 if (buf)
 780                 {
 781                     *buf++ = (char) ((-128 >> cnt) | ((cc >> (cnt * 6)) & (0x3f >> cnt)));
 782                     while (cnt--)
 783                         *buf++ = (char) (0x80 | ((cc >> (cnt * 6)) & 0x3f));
 784                 }
 785             }
 786         }
 787     }
 788
 789     if (buf && (len<n))
 790         *buf = 0;
 791
 792     return len;
 793 }
 794
 795 // ----------------------------------------------------------------------------
 796 // UTF-16
 797 // ----------------------------------------------------------------------------
 798
 799 #ifdef WORDS_BIGENDIAN
 800     #define wxMBConvUTF16straight wxMBConvUTF16BE
 801     #define wxMBConvUTF16swap     wxMBConvUTF16LE
 802 #else
 803     #define wxMBConvUTF16swap     wxMBConvUTF16BE
 804     #define wxMBConvUTF16straight wxMBConvUTF16LE
 805 #endif
 806
 807
 808 #ifdef WC_UTF16
 809
 810 // copy 16bit MB to 16bit String
 811 size_t wxMBConvUTF16straight::MB2WC(wchar_t *buf, const char *psz, size_t n) const
 812 {
 813     size_t len=0;
 814
 815     while (*(wxUint16*)psz && (!buf || len < n))
 816     {
 817         if (buf)
 818             *buf++ = *(wxUint16*)psz;
 819         len++;
 820
 821         psz += sizeof(wxUint16);
 822     }
 823     if (buf && len<n)   *buf=0;
 824
 825     return len;
 826 }
 827
 828
 829 // copy 16bit String to 16bit MB
 830 size_t wxMBConvUTF16straight::WC2MB(char *buf, const wchar_t *psz, size_t n) const
 831 {
 832     size_t len=0;
 833
 834     while (*psz && (!buf || len < n))
 835     {
 836         if (buf)
 837         {
 838             *(wxUint16*)buf = *psz;
 839             buf += sizeof(wxUint16);
 840         }
 841         len += sizeof(wxUint16);
 842         psz++;
 843     }
 844     if (buf && len<=n-sizeof(wxUint16))   *(wxUint16*)buf=0;
 845
 846     return len;
 847 }
 848
 849
 850 // swap 16bit MB to 16bit String
 851 size_t wxMBConvUTF16swap::MB2WC(wchar_t *buf, const char *psz, size_t n) const
 852 {
 853     size_t len=0;
 854
 855     while (*(wxUint16*)psz && (!buf || len < n))
 856     {
 857         if (buf)
 858         {
 859             ((char *)buf)[0] = psz[1];
 860             ((char *)buf)[1] = psz[0];
 861             buf++;
 862         }
 863         len++;
 864         psz += sizeof(wxUint16);
 865     }
 866     if (buf && len<n)   *buf=0;
 867
 868     return len;
 869 }
 870
 871
 872 // swap 16bit MB to 16bit String
 873 size_t wxMBConvUTF16swap::WC2MB(char *buf, const wchar_t *psz, size_t n) const
 874 {
 875     size_t len=0;
 876
 877     while (*psz && (!buf || len < n))
 878     {
 879         if (buf)
 880         {
 881             *buf++ = ((char*)psz)[1];
 882             *buf++ = ((char*)psz)[0];
 883         }
 884         len += sizeof(wxUint16);
 885         psz++;
 886     }
 887     if (buf && len<=n-sizeof(wxUint16))   *(wxUint16*)buf=0;
 888
 889     return len;
 890 }
 891
 892
 893 #else // WC_UTF16
 894
 895
 896 // copy 16bit MB to 32bit String
 897 size_t wxMBConvUTF16straight::MB2WC(wchar_t *buf, const char *psz, size_t n) const
 898 {
 899     size_t len=0;
 900
 901     while (*(wxUint16*)psz && (!buf || len < n))
 902     {
 903         wxUint32 cc;
 904         size_t pa=decode_utf16((wxUint16*)psz, cc);
 905         if (pa == (size_t)-1)
 906             return pa;
 907
 908         if (buf)
 909             *buf++ = (wchar_t)cc;
 910         len++;
 911         psz += pa * sizeof(wxUint16);
 912     }
 913     if (buf && len<n)   *buf=0;
 914
 915     return len;
 916 }
 917
 918
 919 // copy 32bit String to 16bit MB
 920 size_t wxMBConvUTF16straight::WC2MB(char *buf, const wchar_t *psz, size_t n) const
 921 {
 922     size_t len=0;
 923
 924     while (*psz && (!buf || len < n))
 925     {
 926         wxUint16 cc[2];
 927         size_t pa=encode_utf16(*psz, cc);
 928
 929         if (pa == (size_t)-1)
 930             return pa;
 931
 932         if (buf)
 933         {
 934             *(wxUint16*)buf = cc[0];
 935             buf += sizeof(wxUint16);
 936             if (pa > 1)
 937             {
 938                 *(wxUint16*)buf = cc[1];
 939                 buf += sizeof(wxUint16);
 940             }
 941         }
 942
 943         len += pa*sizeof(wxUint16);
 944         psz++;
 945     }
 946     if (buf && len<=n-sizeof(wxUint16))   *(wxUint16*)buf=0;
 947
 948     return len;
 949 }
 950
 951
 952 // swap 16bit MB to 32bit String
 953 size_t wxMBConvUTF16swap::MB2WC(wchar_t *buf, const char *psz, size_t n) const
 954 {
 955     size_t len=0;
 956
 957     while (*(wxUint16*)psz && (!buf || len < n))
 958     {
 959         wxUint32 cc;
 960         char tmp[4];
 961         tmp[0]=psz[1];  tmp[1]=psz[0];
 962         tmp[2]=psz[3];  tmp[3]=psz[2];
 963
 964         size_t pa=decode_utf16((wxUint16*)tmp, cc);
 965         if (pa == (size_t)-1)
 966             return pa;
 967
 968         if (buf)
 969             *buf++ = (wchar_t)cc;
 970
 971         len++;
 972         psz += pa * sizeof(wxUint16);
 973     }
 974     if (buf && len<n)   *buf=0;
 975
 976     return len;
 977 }
 978
 979
 980 // swap 32bit String to 16bit MB
 981 size_t wxMBConvUTF16swap::WC2MB(char *buf, const wchar_t *psz, size_t n) const
 982 {
 983     size_t len=0;
 984
 985     while (*psz && (!buf || len < n))
 986     {
 987         wxUint16 cc[2];
 988         size_t pa=encode_utf16(*psz, cc);
 989
 990         if (pa == (size_t)-1)
 991             return pa;
 992
 993         if (buf)
 994         {
 995             *buf++ = ((char*)cc)[1];
 996             *buf++ = ((char*)cc)[0];
 997             if (pa > 1)
 998             {
 999                 *buf++ = ((char*)cc)[3];
1000                 *buf++ = ((char*)cc)[2];
1001             }
1002         }
1003
1004         len += pa*sizeof(wxUint16);
1005         psz++;
1006     }
1007     if (buf && len<=n-sizeof(wxUint16))   *(wxUint16*)buf=0;
1008
1009     return len;
1010 }
1011
1012 #endif // WC_UTF16
1013
1014
1015 // ----------------------------------------------------------------------------
1016 // UTF-32
1017 // ----------------------------------------------------------------------------
1018
1019 #ifdef WORDS_BIGENDIAN
1020 #define wxMBConvUTF32straight  wxMBConvUTF32BE
1021 #define wxMBConvUTF32swap      wxMBConvUTF32LE
1022 #else
1023 #define wxMBConvUTF32swap      wxMBConvUTF32BE
1024 #define wxMBConvUTF32straight  wxMBConvUTF32LE
1025 #endif
1026
1027
1028 WXDLLIMPEXP_DATA_BASE(wxMBConvUTF32LE) wxConvUTF32LE;
1029 WXDLLIMPEXP_DATA_BASE(wxMBConvUTF32BE) wxConvUTF32BE;
1030
1031
1032 #ifdef WC_UTF16
1033
1034 // copy 32bit MB to 16bit String
1035 size_t wxMBConvUTF32straight::MB2WC(wchar_t *buf, const char *psz, size_t n) const
1036 {
1037     size_t len=0;
1038
1039     while (*(wxUint32*)psz && (!buf || len < n))
1040     {
1041         wxUint16 cc[2];
1042
1043         size_t pa=encode_utf16(*(wxUint32*)psz, cc);
1044         if (pa == (size_t)-1)
1045             return pa;
1046
1047         if (buf)
1048         {
1049             *buf++ = cc[0];
1050             if (pa > 1)
1051                 *buf++ = cc[1];
1052         }
1053         len += pa;
1054         psz += sizeof(wxUint32);
1055     }
1056     if (buf && len<n)   *buf=0;
1057
1058     return len;
1059 }
1060
1061
1062 // copy 16bit String to 32bit MB
1063 size_t wxMBConvUTF32straight::WC2MB(char *buf, const wchar_t *psz, size_t n) const
1064 {
1065     size_t len=0;
1066
1067     while (*psz && (!buf || len < n))
1068     {
1069         wxUint32 cc;
1070
1071         // cast is ok for WC_UTF16
1072         size_t pa = decode_utf16((const wxUint16 *)psz, cc);
1073         if (pa == (size_t)-1)
1074             return pa;
1075
1076         if (buf)
1077         {
1078             *(wxUint32*)buf = cc;
1079             buf += sizeof(wxUint32);
1080         }
1081         len += sizeof(wxUint32);
1082         psz += pa;
1083     }
1084
1085     if (buf && len<=n-sizeof(wxUint32))
1086         *(wxUint32*)buf=0;
1087
1088     return len;
1089 }
1090
1091
1092
1093 // swap 32bit MB to 16bit String
1094 size_t wxMBConvUTF32swap::MB2WC(wchar_t *buf, const char *psz, size_t n) const
1095 {
1096     size_t len=0;
1097
1098     while (*(wxUint32*)psz && (!buf || len < n))
1099     {
1100         char tmp[4];
1101         tmp[0] = psz[3];   tmp[1] = psz[2];
1102         tmp[2] = psz[1];   tmp[3] = psz[0];
1103
1104
1105         wxUint16 cc[2];
1106
1107         size_t pa=encode_utf16(*(wxUint32*)tmp, cc);
1108         if (pa == (size_t)-1)
1109             return pa;
1110
1111         if (buf)
1112         {
1113             *buf++ = cc[0];
1114             if (pa > 1)
1115                 *buf++ = cc[1];
1116         }
1117         len += pa;
1118         psz += sizeof(wxUint32);
1119     }
1120
1121     if (buf && len<n)
1122         *buf=0;
1123
1124     return len;
1125 }
1126
1127
1128 // swap 16bit String to 32bit MB
1129 size_t wxMBConvUTF32swap::WC2MB(char *buf, const wchar_t *psz, size_t n) const
1130 {
1131     size_t len=0;
1132
1133     while (*psz && (!buf || len < n))
1134     {
1135         char cc[4];
1136
1137         // cast is ok for WC_UTF16
1138         size_t pa=decode_utf16((const wxUint16 *)psz, *(wxUint32*)cc);
1139         if (pa == (size_t)-1)
1140             return pa;
1141
1142         if (buf)
1143         {
1144             *buf++ = cc[3];
1145             *buf++ = cc[2];
1146             *buf++ = cc[1];
1147             *buf++ = cc[0];
1148         }
1149         len += sizeof(wxUint32);
1150         psz += pa;
1151     }
1152
1153     if (buf && len<=n-sizeof(wxUint32))
1154         *(wxUint32*)buf=0;
1155
1156     return len;
1157 }
1158
1159 #else // WC_UTF16
1160
1161
1162 // copy 32bit MB to 32bit String
1163 size_t wxMBConvUTF32straight::MB2WC(wchar_t *buf, const char *psz, size_t n) const
1164 {
1165     size_t len=0;
1166
1167     while (*(wxUint32*)psz && (!buf || len < n))
1168     {
1169         if (buf)
1170             *buf++ = (wchar_t)(*(wxUint32*)psz);
1171         len++;
1172         psz += sizeof(wxUint32);
1173     }
1174
1175     if (buf && len<n)
1176         *buf=0;
1177
1178     return len;
1179 }
1180
1181
1182 // copy 32bit String to 32bit MB
1183 size_t wxMBConvUTF32straight::WC2MB(char *buf, const wchar_t *psz, size_t n) const
1184 {
1185     size_t len=0;
1186
1187     while (*psz && (!buf || len < n))
1188     {
1189         if (buf)
1190         {
1191             *(wxUint32*)buf = *psz;
1192             buf += sizeof(wxUint32);
1193         }
1194
1195         len += sizeof(wxUint32);
1196         psz++;
1197     }
1198
1199     if (buf && len<=n-sizeof(wxUint32))
1200         *(wxUint32*)buf=0;
1201
1202     return len;
1203 }
1204
1205
1206 // swap 32bit MB to 32bit String
1207 size_t wxMBConvUTF32swap::MB2WC(wchar_t *buf, const char *psz, size_t n) const
1208 {
1209     size_t len=0;
1210
1211     while (*(wxUint32*)psz && (!buf || len < n))
1212     {
1213         if (buf)
1214         {
1215             ((char *)buf)[0] = psz[3];
1216             ((char *)buf)[1] = psz[2];
1217             ((char *)buf)[2] = psz[1];
1218             ((char *)buf)[3] = psz[0];
1219             buf++;
1220         }
1221         len++;
1222         psz += sizeof(wxUint32);
1223     }
1224
1225     if (buf && len<n)
1226         *buf=0;
1227
1228     return len;
1229 }
1230
1231
1232 // swap 32bit String to 32bit MB
1233 size_t wxMBConvUTF32swap::WC2MB(char *buf, const wchar_t *psz, size_t n) const
1234 {
1235     size_t len=0;
1236
1237     while (*psz && (!buf || len < n))
1238     {
1239         if (buf)
1240         {
1241             *buf++ = ((char *)psz)[3];
1242             *buf++ = ((char *)psz)[2];
1243             *buf++ = ((char *)psz)[1];
1244             *buf++ = ((char *)psz)[0];
1245         }
1246         len += sizeof(wxUint32);
1247         psz++;
1248     }
1249
1250     if (buf && len<=n-sizeof(wxUint32))
1251         *(wxUint32*)buf=0;
1252
1253     return len;
1254 }
1255
1256
1257 #endif // WC_UTF16
1258
1259
1260 // ============================================================================
1261 // The classes doing conversion using the iconv_xxx() functions
1262 // ============================================================================
1263
1264 #ifdef HAVE_ICONV
1265
1266 // VS: glibc 2.1.3 is broken in that iconv() conversion to/from UCS4 fails with
1267 //     E2BIG if output buffer is _exactly_ as big as needed. Such case is
1268 //     (unless there's yet another bug in glibc) the only case when iconv()
1269 //     returns with (size_t)-1 (which means error) and says there are 0 bytes
1270 //     left in the input buffer -- when _real_ error occurs,
1271 //     bytes-left-in-input buffer is non-zero. Hence, this alternative test for
1272 //     iconv() failure.
1273 //     [This bug does not appear in glibc 2.2.]
1274 #if defined(__GLIBC__) && __GLIBC__ == 2 && __GLIBC_MINOR__ <= 1
1275 #define ICONV_FAILED(cres, bufLeft) ((cres == (size_t)-1) && \
1276                                      (errno != E2BIG || bufLeft != 0))
1277 #else
1278 #define ICONV_FAILED(cres, bufLeft)  (cres == (size_t)-1)
1279 #endif
1280
1281 #define ICONV_CHAR_CAST(x)  ((ICONV_CONST char **)(x))
1282
1283 #define ICONV_T_INVALID ((iconv_t)-1)
1284
1285 #if SIZEOF_WCHAR_T == 4
1286     #define WC_BSWAP    wxUINT32_SWAP_ALWAYS
1287     #define WC_ENC      wxFONTENCODING_UTF32
1288 #elif SIZEOF_WCHAR_T == 2
1289     #define WC_BSWAP    wxUINT16_SWAP_ALWAYS
1290     #define WC_ENC      wxFONTENCODING_UTF16
1291 #else // sizeof(wchar_t) != 2 nor 4
1292     // does this ever happen?
1293     #error "Unknown sizeof(wchar_t): please report this to wx-dev@lists.wxwindows.org"
1294 #endif
1295
1296 // ----------------------------------------------------------------------------
1297 // wxMBConv_iconv: encapsulates an iconv character set
1298 // ----------------------------------------------------------------------------
1299
1300 class wxMBConv_iconv : public wxMBConv
1301 {
1302 public:
1303     wxMBConv_iconv(const wxChar *name);
1304     virtual ~wxMBConv_iconv();
1305
1306     virtual size_t MB2WC(wchar_t *buf, const char *psz, size_t n) const;
1307     virtual size_t WC2MB(char *buf, const wchar_t *psz, size_t n) const;
1308
1309     bool IsOk() const
1310         { return (m2w != ICONV_T_INVALID) && (w2m != ICONV_T_INVALID); }
1311
1312 protected:
1313     // the iconv handlers used to translate from multibyte to wide char and in
1314     // the other direction
1315     iconv_t m2w,
1316             w2m;
1317 #if wxUSE_THREADS
1318     // guards access to m2w and w2m objects
1319     wxMutex m_iconvMutex;
1320 #endif
1321
1322 private:
1323     // the name (for iconv_open()) of a wide char charset -- if none is
1324     // available on this machine, it will remain NULL
1325     static wxString ms_wcCharsetName;
1326
1327     // true if the wide char encoding we use (i.e. ms_wcCharsetName) has
1328     // different endian-ness than the native one
1329     static bool ms_wcNeedsSwap;
1330 };
1331
1332 // make the constructor available for unit testing
1333 WXDLLIMPEXP_BASE wxMBConv* new_wxMBConv_iconv( const wxChar* name )
1334 {
1335     wxMBConv_iconv* result = new wxMBConv_iconv( name );
1336     if ( !result->IsOk() )
1337     {
1338         delete result;
1339         return 0;
1340     }
1341     return result;
1342 }
1343
1344 wxString wxMBConv_iconv::ms_wcCharsetName;
1345 bool wxMBConv_iconv::ms_wcNeedsSwap = false;
1346
1347 wxMBConv_iconv::wxMBConv_iconv(const wxChar *name)
1348 {
1349     // iconv operates with chars, not wxChars, but luckily it uses only ASCII
1350     // names for the charsets
1351     const wxCharBuffer cname(wxString(name).ToAscii());
1352
1353     // check for charset that represents wchar_t:
1354     if ( ms_wcCharsetName.empty() )
1355     {
1356         wxLogTrace(TRACE_STRCONV, _T("Looking for wide char codeset:"));
1357
1358 #if wxUSE_FONTMAP
1359         const wxChar **names = wxFontMapperBase::GetAllEncodingNames(WC_ENC);
1360 #else // !wxUSE_FONTMAP
1361         static const wxChar *names[] =
1362         {
1363 #if SIZEOF_WCHAR_T == 4
1364             _T("UCS-4"),
1365 #elif SIZEOF_WCHAR_T = 2
1366             _T("UCS-2"),
1367 #endif
1368             NULL
1369         };
1370 #endif // wxUSE_FONTMAP/!wxUSE_FONTMAP
1371
1372         for ( ; *names && ms_wcCharsetName.empty(); ++names )
1373         {
1374             const wxString nameCS(*names);
1375
1376             // first try charset with explicit bytesex info (e.g. "UCS-4LE"):
1377             wxString nameXE(nameCS);
1378             #ifdef WORDS_BIGENDIAN
1379                 nameXE += _T("BE");
1380             #else // little endian
1381                 nameXE += _T("LE");
1382             #endif
1383
1384             wxLogTrace(TRACE_STRCONV, _T("  trying charset \"%s\""),
1385                        nameXE.c_str());
1386
1387             m2w = iconv_open(nameXE.ToAscii(), cname);
1388             if ( m2w == ICONV_T_INVALID )
1389             {
1390                 // try charset w/o bytesex info (e.g. "UCS4")
1391                 wxLogTrace(TRACE_STRCONV, _T("  trying charset \"%s\""),
1392                            nameCS.c_str());
1393                 m2w = iconv_open(nameCS.ToAscii(), cname);
1394
1395                 // and check for bytesex ourselves:
1396                 if ( m2w != ICONV_T_INVALID )
1397                 {
1398                     char    buf[2], *bufPtr;
1399                     wchar_t wbuf[2], *wbufPtr;
1400                     size_t  insz, outsz;
1401                     size_t  res;
1402
1403                     buf[0] = 'A';
1404                     buf[1] = 0;
1405                     wbuf[0] = 0;
1406                     insz = 2;
1407                     outsz = SIZEOF_WCHAR_T * 2;
1408                     wbufPtr = wbuf;
1409                     bufPtr = buf;
1410
1411                     res = iconv(m2w, ICONV_CHAR_CAST(&bufPtr), &insz,
1412                                 (char**)&wbufPtr, &outsz);
1413
1414                     if (ICONV_FAILED(res, insz))
1415                     {
1416                         wxLogLastError(wxT("iconv"));
1417                         wxLogError(_("Conversion to charset '%s' doesn't work."),
1418                                    nameCS.c_str());
1419                     }
1420                     else // ok, can convert to this encoding, remember it
1421                     {
1422                         ms_wcCharsetName = nameCS;
1423                         ms_wcNeedsSwap = wbuf[0] != (wchar_t)buf[0];
1424                     }
1425                 }
1426             }
1427             else // use charset not requiring byte swapping
1428             {
1429                 ms_wcCharsetName = nameXE;
1430             }
1431         }
1432
1433         wxLogTrace(TRACE_STRCONV,
1434                    wxT("iconv wchar_t charset is \"%s\"%s"),
1435                    ms_wcCharsetName.empty() ? _T("<none>")
1436                                             : ms_wcCharsetName.c_str(),
1437                    ms_wcNeedsSwap ? _T(" (needs swap)")
1438                                   : _T(""));
1439     }
1440     else // we already have ms_wcCharsetName
1441     {
1442         m2w = iconv_open(ms_wcCharsetName.ToAscii(), cname);
1443     }
1444
1445     if ( ms_wcCharsetName.empty() )
1446     {
1447         w2m = ICONV_T_INVALID;
1448     }
1449     else
1450     {
1451         w2m = iconv_open(cname, ms_wcCharsetName.ToAscii());
1452         if ( w2m == ICONV_T_INVALID )
1453         {
1454             wxLogTrace(TRACE_STRCONV,
1455                        wxT("\"%s\" -> \"%s\" works but not the converse!?"),
1456                        ms_wcCharsetName.c_str(), cname.data());
1457         }
1458     }
1459 }
1460
1461 wxMBConv_iconv::~wxMBConv_iconv()
1462 {
1463     if ( m2w != ICONV_T_INVALID )
1464         iconv_close(m2w);
1465     if ( w2m != ICONV_T_INVALID )
1466         iconv_close(w2m);
1467 }
1468
1469 size_t wxMBConv_iconv::MB2WC(wchar_t *buf, const char *psz, size_t n) const
1470 {
1471 #if wxUSE_THREADS
1472     // NB: iconv() is MT-safe, but each thread must use it's own iconv_t handle.
1473     //     Unfortunately there is a couple of global wxCSConv objects such as
1474     //     wxConvLocal that are used all over wx code, so we have to make sure
1475     //     the handle is used by at most one thread at the time. Otherwise
1476     //     only a few wx classes would be safe to use from non-main threads
1477     //     as MB<->WC conversion would fail "randomly".
1478     wxMutexLocker lock(wxConstCast(this, wxMBConv_iconv)->m_iconvMutex);
1479 #endif
1480
1481     size_t inbuf = strlen(psz);
1482     size_t outbuf = n * SIZEOF_WCHAR_T;
1483     size_t res, cres;
1484     // VS: Use these instead of psz, buf because iconv() modifies its arguments:
1485     wchar_t *bufPtr = buf;
1486     const char *pszPtr = psz;
1487
1488     if (buf)
1489     {
1490         // have destination buffer, convert there
1491         cres = iconv(m2w,
1492                      ICONV_CHAR_CAST(&pszPtr), &inbuf,
1493                      (char**)&bufPtr, &outbuf);
1494         res = n - (outbuf / SIZEOF_WCHAR_T);
1495
1496         if (ms_wcNeedsSwap)
1497         {
1498             // convert to native endianness
1499             for ( unsigned i = 0; i < res; i++ )
1500                 buf[n] = WC_BSWAP(buf[i]);
1501         }
1502
1503         // NB: iconv was given only strlen(psz) characters on input, and so
1504         //     it couldn't convert the trailing zero. Let's do it ourselves
1505         //     if there's some room left for it in the output buffer.
1506         if (res < n)
1507             buf[res] = 0;
1508     }
1509     else
1510     {
1511         // no destination buffer... convert using temp buffer
1512         // to calculate destination buffer requirement
1513         wchar_t tbuf[8];
1514         res = 0;
1515         do {
1516             bufPtr = tbuf;
1517             outbuf = 8*SIZEOF_WCHAR_T;
1518
1519             cres = iconv(m2w,
1520                          ICONV_CHAR_CAST(&pszPtr), &inbuf,
1521                          (char**)&bufPtr, &outbuf );
1522
1523             res += 8-(outbuf/SIZEOF_WCHAR_T);
1524         } while ((cres==(size_t)-1) && (errno==E2BIG));
1525     }
1526
1527     if (ICONV_FAILED(cres, inbuf))
1528     {
1529         //VS: it is ok if iconv fails, hence trace only
1530         wxLogTrace(TRACE_STRCONV, wxT("iconv failed: %s"), wxSysErrorMsg(wxSysErrorCode()));
1531         return (size_t)-1;
1532     }
1533
1534     return res;
1535 }
1536
1537 size_t wxMBConv_iconv::WC2MB(char *buf, const wchar_t *psz, size_t n) const
1538 {
1539 #if wxUSE_THREADS
1540     // NB: explained in MB2WC
1541     wxMutexLocker lock(wxConstCast(this, wxMBConv_iconv)->m_iconvMutex);
1542 #endif
1543
1544     size_t inlen = wxWcslen(psz);
1545     size_t inbuf = inlen * SIZEOF_WCHAR_T;
1546     size_t outbuf = n;
1547     size_t res, cres;
1548
1549     wchar_t *tmpbuf = 0;
1550
1551     if (ms_wcNeedsSwap)
1552     {
1553         // need to copy to temp buffer to switch endianness
1554         // (doing WC_BSWAP twice on the original buffer won't help, as it
1555         //  could be in read-only memory, or be accessed in some other thread)
1556         tmpbuf = (wchar_t *)malloc(inbuf + SIZEOF_WCHAR_T);
1557         for ( size_t i = 0; i < inlen; i++ )
1558             tmpbuf[n] = WC_BSWAP(psz[i]);
1559         tmpbuf[inlen] = L'\0';
1560         psz = tmpbuf;
1561     }
1562
1563     if (buf)
1564     {
1565         // have destination buffer, convert there
1566         cres = iconv( w2m, ICONV_CHAR_CAST(&psz), &inbuf, &buf, &outbuf );
1567
1568         res = n-outbuf;
1569
1570         // NB: iconv was given only wcslen(psz) characters on input, and so
1571         //     it couldn't convert the trailing zero. Let's do it ourselves
1572         //     if there's some room left for it in the output buffer.
1573         if (res < n)
1574             buf[0] = 0;
1575     }
1576     else
1577     {
1578         // no destination buffer... convert using temp buffer
1579         // to calculate destination buffer requirement
1580         char tbuf[16];
1581         res = 0;
1582         do {
1583             buf = tbuf; outbuf = 16;
1584
1585             cres = iconv( w2m, ICONV_CHAR_CAST(&psz), &inbuf, &buf, &outbuf );
1586
1587             res += 16 - outbuf;
1588         } while ((cres==(size_t)-1) && (errno==E2BIG));
1589     }
1590
1591     if (ms_wcNeedsSwap)
1592     {
1593         free(tmpbuf);
1594     }
1595
1596     if (ICONV_FAILED(cres, inbuf))
1597     {
1598         wxLogTrace(TRACE_STRCONV, wxT("iconv failed: %s"), wxSysErrorMsg(wxSysErrorCode()));
1599         return (size_t)-1;
1600     }
1601
1602     return res;
1603 }
1604
1605 #endif // HAVE_ICONV
1606
1607
1608 // ============================================================================
1609 // Win32 conversion classes
1610 // ============================================================================
1611
1612 #ifdef wxHAVE_WIN32_MB2WC
1613
1614 // from utils.cpp
1615 #if wxUSE_FONTMAP
1616 extern WXDLLIMPEXP_BASE long wxCharsetToCodepage(const wxChar *charset);
1617 extern WXDLLIMPEXP_BASE long wxEncodingToCodepage(wxFontEncoding encoding);
1618 #endif
1619
1620 class wxMBConv_win32 : public wxMBConv
1621 {
1622 public:
1623     wxMBConv_win32()
1624     {
1625         m_CodePage = CP_ACP;
1626     }
1627
1628 #if wxUSE_FONTMAP
1629     wxMBConv_win32(const wxChar* name)
1630     {
1631         m_CodePage = wxCharsetToCodepage(name);
1632     }
1633
1634     wxMBConv_win32(wxFontEncoding encoding)
1635     {
1636         m_CodePage = wxEncodingToCodepage(encoding);
1637     }
1638 #endif
1639
1640     size_t MB2WC(wchar_t *buf, const char *psz, size_t n) const
1641     {
1642         // note that we have to use MB_ERR_INVALID_CHARS flag as it without it
1643         // the behaviour is not compatible with the Unix version (using iconv)
1644         // and break the library itself, e.g. wxTextInputStream::NextChar()
1645         // wouldn't work if reading an incomplete MB char didn't result in an
1646         // error
1647         //
1648         // note however that using MB_ERR_INVALID_CHARS with CP_UTF7 results in
1649         // an error (tested under Windows Server 2003) and apparently it is
1650         // done on purpose, i.e. the function accepts any input in this case
1651         // and although I'd prefer to return error on ill-formed output, our
1652         // own wxMBConvUTF7 doesn't detect errors (e.g. lone "+" which is
1653         // explicitly ill-formed according to RFC 2152) neither so we don't
1654         // even have any fallback here...
1655         int flags = m_CodePage == CP_UTF7 ? 0 : MB_ERR_INVALID_CHARS;
1656
1657         const size_t len = ::MultiByteToWideChar
1658                              (
1659                                 m_CodePage,     // code page
1660                                 flags,          // flags: fall on error
1661                                 psz,            // input string
1662                                 -1,             // its length (NUL-terminated)
1663                                 buf,            // output string
1664                                 buf ? n : 0     // size of output buffer
1665                              );
1666
1667         // note that it returns count of written chars for buf != NULL and size
1668         // of the needed buffer for buf == NULL so in either case the length of
1669         // the string (which never includes the terminating NUL) is one less
1670         return len ? len - 1 : (size_t)-1;
1671     }
1672
1673     size_t WC2MB(char *buf, const wchar_t *pwz, size_t n) const
1674     {
1675         /*
1676             we have a problem here: by default, WideCharToMultiByte() may
1677             replace characters unrepresentable in the target code page with bad
1678             quality approximations such as turning "1/2" symbol (U+00BD) into
1679             "1" for the code pages which don't have it and we, obviously, want
1680             to avoid this at any price
1681
1682             the trouble is that this function does it _silently_, i.e. it won't
1683             even tell us whether it did or not... Win98/2000 and higher provide
1684             WC_NO_BEST_FIT_CHARS but it doesn't work for the older systems and
1685             we have to resort to a round trip, i.e. check that converting back
1686             results in the same string -- this is, of course, expensive but
1687             otherwise we simply can't be sure to not garble the data.
1688          */
1689
1690         // determine if we can rely on WC_NO_BEST_FIT_CHARS: according to MSDN
1691         // it doesn't work with CJK encodings (which we test for rather roughly
1692         // here...) nor with UTF-7/8 nor, of course, with Windows versions not
1693         // supporting it
1694         BOOL usedDef wxDUMMY_INITIALIZE(false);
1695         BOOL *pUsedDef;
1696         int flags;
1697         if ( CanUseNoBestFit() && m_CodePage < 50000 )
1698         {
1699             // it's our lucky day
1700             flags = WC_NO_BEST_FIT_CHARS;
1701             pUsedDef = &usedDef;
1702         }
1703         else // old system or unsupported encoding
1704         {
1705             flags = 0;
1706             pUsedDef = NULL;
1707         }
1708
1709         const size_t len = ::WideCharToMultiByte
1710                              (
1711                                 m_CodePage,     // code page
1712                                 flags,          // either none or no best fit
1713                                 pwz,            // input string
1714                                 -1,             // it is (wide) NUL-terminated
1715                                 buf,            // output buffer
1716                                 buf ? n : 0,    // and its size
1717                                 NULL,           // default "replacement" char
1718                                 pUsedDef        // [out] was it used?
1719                              );
1720
1721         if ( !len )
1722         {
1723             // function totally failed
1724             return (size_t)-1;
1725         }
1726
1727         // if we were really converting, check if we succeeded
1728         if ( buf )
1729         {
1730             if ( flags )
1731             {
1732                 // check if the conversion failed, i.e. if any replacements
1733                 // were done
1734                 if ( usedDef )
1735                     return (size_t)-1;
1736             }
1737             else // we must resort to double tripping...
1738             {
1739                 wxWCharBuffer wcBuf(n);
1740                 if ( MB2WC(wcBuf.data(), buf, n) == (size_t)-1 ||
1741                         wcscmp(wcBuf, pwz) != 0 )
1742                 {
1743                     // we didn't obtain the same thing we started from, hence
1744                     // the conversion was lossy and we consider that it failed
1745                     return (size_t)-1;
1746                 }
1747             }
1748         }
1749
1750         // see the comment above for the reason of "len - 1"
1751         return len - 1;
1752     }
1753
1754     bool IsOk() const { return m_CodePage != -1; }
1755
1756 private:
1757     static bool CanUseNoBestFit()
1758     {
1759         static int s_isWin98Or2k = -1;
1760
1761         if ( s_isWin98Or2k == -1 )
1762         {
1763             int verMaj, verMin;
1764             switch ( wxGetOsVersion(&verMaj, &verMin) )
1765             {
1766                 case wxWIN95:
1767                     s_isWin98Or2k = verMaj >= 4 && verMin >= 10;
1768                     break;
1769
1770                 case wxWINDOWS_NT:
1771                     s_isWin98Or2k = verMaj >= 5;
1772                     break;
1773
1774                 default:
1775                     // unknown, be conseravtive by default
1776                     s_isWin98Or2k = 0;
1777             }
1778
1779             wxASSERT_MSG( s_isWin98Or2k != -1, _T("should be set above") );
1780         }
1781
1782         return s_isWin98Or2k == 1;
1783     }
1784
1785     long m_CodePage;
1786 };
1787
1788 #endif // wxHAVE_WIN32_MB2WC
1789
1790 // ============================================================================
1791 // Cocoa conversion classes
1792 // ============================================================================
1793
1794 #if defined(__WXCOCOA__)
1795
1796 // RN:  There is no UTF-32 support in either Core Foundation or
1797 // Cocoa.  Strangely enough, internally Core Foundation uses
1798 // UTF 32 internally quite a bit - its just not public (yet).
1799
1800 #include <CoreFoundation/CFString.h>
1801 #include <CoreFoundation/CFStringEncodingExt.h>
1802
1803 CFStringEncoding wxCFStringEncFromFontEnc(wxFontEncoding encoding)
1804 {
1805     CFStringEncoding enc = kCFStringEncodingInvalidId ;
1806     if ( encoding == wxFONTENCODING_DEFAULT )
1807     {
1808         enc = CFStringGetSystemEncoding();
1809     }
1810     else switch( encoding)
1811     {
1812         case wxFONTENCODING_ISO8859_1 :
1813             enc = kCFStringEncodingISOLatin1 ;
1814             break ;
1815         case wxFONTENCODING_ISO8859_2 :
1816             enc = kCFStringEncodingISOLatin2;
1817             break ;
1818         case wxFONTENCODING_ISO8859_3 :
1819             enc = kCFStringEncodingISOLatin3 ;
1820             break ;
1821         case wxFONTENCODING_ISO8859_4 :
1822             enc = kCFStringEncodingISOLatin4;
1823             break ;
1824         case wxFONTENCODING_ISO8859_5 :
1825             enc = kCFStringEncodingISOLatinCyrillic;
1826             break ;
1827         case wxFONTENCODING_ISO8859_6 :
1828             enc = kCFStringEncodingISOLatinArabic;
1829             break ;
1830         case wxFONTENCODING_ISO8859_7 :
1831             enc = kCFStringEncodingISOLatinGreek;
1832             break ;
1833         case wxFONTENCODING_ISO8859_8 :
1834             enc = kCFStringEncodingISOLatinHebrew;
1835             break ;
1836         case wxFONTENCODING_ISO8859_9 :
1837             enc = kCFStringEncodingISOLatin5;
1838             break ;
1839         case wxFONTENCODING_ISO8859_10 :
1840             enc = kCFStringEncodingISOLatin6;
1841             break ;
1842         case wxFONTENCODING_ISO8859_11 :
1843             enc = kCFStringEncodingISOLatinThai;
1844             break ;
1845         case wxFONTENCODING_ISO8859_13 :
1846             enc = kCFStringEncodingISOLatin7;
1847             break ;
1848         case wxFONTENCODING_ISO8859_14 :
1849             enc = kCFStringEncodingISOLatin8;
1850             break ;
1851         case wxFONTENCODING_ISO8859_15 :
1852             enc = kCFStringEncodingISOLatin9;
1853             break ;
1854
1855         case wxFONTENCODING_KOI8 :
1856             enc = kCFStringEncodingKOI8_R;
1857             break ;
1858         case wxFONTENCODING_ALTERNATIVE : // MS-DOS CP866
1859             enc = kCFStringEncodingDOSRussian;
1860             break ;
1861
1862 //      case wxFONTENCODING_BULGARIAN :
1863 //          enc = ;
1864 //          break ;
1865
1866         case wxFONTENCODING_CP437 :
1867             enc =kCFStringEncodingDOSLatinUS ;
1868             break ;
1869         case wxFONTENCODING_CP850 :
1870             enc = kCFStringEncodingDOSLatin1;
1871             break ;
1872         case wxFONTENCODING_CP852 :
1873             enc = kCFStringEncodingDOSLatin2;
1874             break ;
1875         case wxFONTENCODING_CP855 :
1876             enc = kCFStringEncodingDOSCyrillic;
1877             break ;
1878         case wxFONTENCODING_CP866 :
1879             enc =kCFStringEncodingDOSRussian ;
1880             break ;
1881         case wxFONTENCODING_CP874 :
1882             enc = kCFStringEncodingDOSThai;
1883             break ;
1884         case wxFONTENCODING_CP932 :
1885             enc = kCFStringEncodingDOSJapanese;
1886             break ;
1887         case wxFONTENCODING_CP936 :
1888             enc =kCFStringEncodingDOSChineseSimplif ;
1889             break ;
1890         case wxFONTENCODING_CP949 :
1891             enc = kCFStringEncodingDOSKorean;
1892             break ;
1893         case wxFONTENCODING_CP950 :
1894             enc = kCFStringEncodingDOSChineseTrad;
1895             break ;
1896         case wxFONTENCODING_CP1250 :
1897             enc = kCFStringEncodingWindowsLatin2;
1898             break ;
1899         case wxFONTENCODING_CP1251 :
1900             enc =kCFStringEncodingWindowsCyrillic ;
1901             break ;
1902         case wxFONTENCODING_CP1252 :
1903             enc =kCFStringEncodingWindowsLatin1 ;
1904             break ;
1905         case wxFONTENCODING_CP1253 :
1906             enc = kCFStringEncodingWindowsGreek;
1907             break ;
1908         case wxFONTENCODING_CP1254 :
1909             enc = kCFStringEncodingWindowsLatin5;
1910             break ;
1911         case wxFONTENCODING_CP1255 :
1912             enc =kCFStringEncodingWindowsHebrew ;
1913             break ;
1914         case wxFONTENCODING_CP1256 :
1915             enc =kCFStringEncodingWindowsArabic ;
1916             break ;
1917         case wxFONTENCODING_CP1257 :
1918             enc = kCFStringEncodingWindowsBalticRim;
1919             break ;
1920 //   This only really encodes to UTF7 (if that) evidently
1921 //        case wxFONTENCODING_UTF7 :
1922 //            enc = kCFStringEncodingNonLossyASCII ;
1923 //            break ;
1924         case wxFONTENCODING_UTF8 :
1925             enc = kCFStringEncodingUTF8 ;
1926             break ;
1927         case wxFONTENCODING_EUC_JP :
1928             enc = kCFStringEncodingEUC_JP;
1929             break ;
1930         case wxFONTENCODING_UTF16 :
1931             enc = kCFStringEncodingUnicode ;
1932             break ;
1933         case wxFONTENCODING_MACROMAN :
1934             enc = kCFStringEncodingMacRoman ;
1935             break ;
1936         case wxFONTENCODING_MACJAPANESE :
1937             enc = kCFStringEncodingMacJapanese ;
1938             break ;
1939         case wxFONTENCODING_MACCHINESETRAD :
1940             enc = kCFStringEncodingMacChineseTrad ;
1941             break ;
1942         case wxFONTENCODING_MACKOREAN :
1943             enc = kCFStringEncodingMacKorean ;
1944             break ;
1945         case wxFONTENCODING_MACARABIC :
1946             enc = kCFStringEncodingMacArabic ;
1947             break ;
1948         case wxFONTENCODING_MACHEBREW :
1949             enc = kCFStringEncodingMacHebrew ;
1950             break ;
1951         case wxFONTENCODING_MACGREEK :
1952             enc = kCFStringEncodingMacGreek ;
1953             break ;
1954         case wxFONTENCODING_MACCYRILLIC :
1955             enc = kCFStringEncodingMacCyrillic ;
1956             break ;
1957         case wxFONTENCODING_MACDEVANAGARI :
1958             enc = kCFStringEncodingMacDevanagari ;
1959             break ;
1960         case wxFONTENCODING_MACGURMUKHI :
1961             enc = kCFStringEncodingMacGurmukhi ;
1962             break ;
1963         case wxFONTENCODING_MACGUJARATI :
1964             enc = kCFStringEncodingMacGujarati ;
1965             break ;
1966         case wxFONTENCODING_MACORIYA :
1967             enc = kCFStringEncodingMacOriya ;
1968             break ;
1969         case wxFONTENCODING_MACBENGALI :
1970             enc = kCFStringEncodingMacBengali ;
1971             break ;
1972         case wxFONTENCODING_MACTAMIL :
1973             enc = kCFStringEncodingMacTamil ;
1974             break ;
1975         case wxFONTENCODING_MACTELUGU :
1976             enc = kCFStringEncodingMacTelugu ;
1977             break ;
1978         case wxFONTENCODING_MACKANNADA :
1979             enc = kCFStringEncodingMacKannada ;
1980             break ;
1981         case wxFONTENCODING_MACMALAJALAM :
1982             enc = kCFStringEncodingMacMalayalam ;
1983             break ;
1984         case wxFONTENCODING_MACSINHALESE :
1985             enc = kCFStringEncodingMacSinhalese ;
1986             break ;
1987         case wxFONTENCODING_MACBURMESE :
1988             enc = kCFStringEncodingMacBurmese ;
1989             break ;
1990         case wxFONTENCODING_MACKHMER :
1991             enc = kCFStringEncodingMacKhmer ;
1992             break ;
1993         case wxFONTENCODING_MACTHAI :
1994             enc = kCFStringEncodingMacThai ;
1995             break ;
1996         case wxFONTENCODING_MACLAOTIAN :
1997             enc = kCFStringEncodingMacLaotian ;
1998             break ;
1999         case wxFONTENCODING_MACGEORGIAN :
2000             enc = kCFStringEncodingMacGeorgian ;
2001             break ;
2002         case wxFONTENCODING_MACARMENIAN :
2003             enc = kCFStringEncodingMacArmenian ;
2004             break ;
2005         case wxFONTENCODING_MACCHINESESIMP :
2006             enc = kCFStringEncodingMacChineseSimp ;
2007             break ;
2008         case wxFONTENCODING_MACTIBETAN :
2009             enc = kCFStringEncodingMacTibetan ;
2010             break ;
2011         case wxFONTENCODING_MACMONGOLIAN :
2012             enc = kCFStringEncodingMacMongolian ;
2013             break ;
2014         case wxFONTENCODING_MACETHIOPIC :
2015             enc = kCFStringEncodingMacEthiopic ;
2016             break ;
2017         case wxFONTENCODING_MACCENTRALEUR :
2018             enc = kCFStringEncodingMacCentralEurRoman ;
2019             break ;
2020         case wxFONTENCODING_MACVIATNAMESE :
2021             enc = kCFStringEncodingMacVietnamese ;
2022             break ;
2023         case wxFONTENCODING_MACARABICEXT :
2024             enc = kCFStringEncodingMacExtArabic ;
2025             break ;
2026         case wxFONTENCODING_MACSYMBOL :
2027             enc = kCFStringEncodingMacSymbol ;
2028             break ;
2029         case wxFONTENCODING_MACDINGBATS :
2030             enc = kCFStringEncodingMacDingbats ;
2031             break ;
2032         case wxFONTENCODING_MACTURKISH :
2033             enc = kCFStringEncodingMacTurkish ;
2034             break ;
2035         case wxFONTENCODING_MACCROATIAN :
2036             enc = kCFStringEncodingMacCroatian ;
2037             break ;
2038         case wxFONTENCODING_MACICELANDIC :
2039             enc = kCFStringEncodingMacIcelandic ;
2040             break ;
2041         case wxFONTENCODING_MACROMANIAN :
2042             enc = kCFStringEncodingMacRomanian ;
2043             break ;
2044         case wxFONTENCODING_MACCELTIC :
2045             enc = kCFStringEncodingMacCeltic ;
2046             break ;
2047         case wxFONTENCODING_MACGAELIC :
2048             enc = kCFStringEncodingMacGaelic ;
2049             break ;
2050 //      case wxFONTENCODING_MACKEYBOARD :
2051 //          enc = kCFStringEncodingMacKeyboardGlyphs ;
2052 //          break ;
2053         default :
2054             // because gcc is picky
2055             break ;
2056     } ;
2057     return enc ;
2058 }
2059
2060 class wxMBConv_cocoa : public wxMBConv
2061 {
2062 public:
2063     wxMBConv_cocoa()
2064     {
2065         Init(CFStringGetSystemEncoding()) ;
2066     }
2067
2068 #if wxUSE_FONTMAP
2069     wxMBConv_cocoa(const wxChar* name)
2070     {
2071         Init( wxCFStringEncFromFontEnc(wxFontMapperBase::Get()->CharsetToEncoding(name, false) ) ) ;
2072     }
2073 #endif
2074
2075     wxMBConv_cocoa(wxFontEncoding encoding)
2076     {
2077         Init( wxCFStringEncFromFontEnc(encoding) );
2078     }
2079
2080     ~wxMBConv_cocoa()
2081     {
2082     }
2083
2084     void Init( CFStringEncoding encoding)
2085     {
2086         m_encoding = encoding ;
2087     }
2088
2089     size_t MB2WC(wchar_t * szOut, const char * szUnConv, size_t nOutSize) const
2090     {
2091         wxASSERT(szUnConv);
2092
2093         CFStringRef theString = CFStringCreateWithBytes (
2094                                                 NULL, //the allocator
2095                                                 (const UInt8*)szUnConv,
2096                                                 strlen(szUnConv),
2097                                                 m_encoding,
2098                                                 false //no BOM/external representation
2099                                                 );
2100
2101         wxASSERT(theString);
2102
2103         size_t nOutLength = CFStringGetLength(theString);
2104
2105         if (szOut == NULL)
2106         {
2107             CFRelease(theString);
2108             return nOutLength;
2109         }
2110
2111         CFRange theRange = { 0, nOutSize };
2112
2113 #if SIZEOF_WCHAR_T == 4
2114         UniChar* szUniCharBuffer = new UniChar[nOutSize];
2115 #endif
2116
2117         CFStringGetCharacters(theString, theRange, szUniCharBuffer);
2118
2119         CFRelease(theString);
2120
2121         szUniCharBuffer[nOutLength] = '\0' ;
2122
2123 #if SIZEOF_WCHAR_T == 4
2124         wxMBConvUTF16 converter ;
2125         converter.MB2WC(szOut, (const char*)szUniCharBuffer , nOutSize ) ;
2126         delete[] szUniCharBuffer;
2127 #endif
2128
2129         return nOutLength;
2130     }
2131
2132     size_t WC2MB(char *szOut, const wchar_t *szUnConv, size_t nOutSize) const
2133     {
2134         wxASSERT(szUnConv);
2135
2136         size_t nRealOutSize;
2137         size_t nBufSize = wxWcslen(szUnConv);
2138         UniChar* szUniBuffer = (UniChar*) szUnConv;
2139
2140 #if SIZEOF_WCHAR_T == 4
2141         wxMBConvUTF16 converter ;
2142         nBufSize = converter.WC2MB( NULL , szUnConv , 0 );
2143         szUniBuffer = new UniChar[ (nBufSize / sizeof(UniChar)) + 1] ;
2144         converter.WC2MB( (char*) szUniBuffer , szUnConv, nBufSize + sizeof(UniChar)) ;
2145         nBufSize /= sizeof(UniChar);
2146 #endif
2147
2148         CFStringRef theString = CFStringCreateWithCharactersNoCopy(
2149                                 NULL, //allocator
2150                                 szUniBuffer,
2151                                 nBufSize,
2152                                 kCFAllocatorNull //deallocator - we want to deallocate it ourselves
2153                             );
2154
2155         wxASSERT(theString);
2156
2157         //Note that CER puts a BOM when converting to unicode
2158         //so we  check and use getchars instead in that case
2159         if (m_encoding == kCFStringEncodingUnicode)
2160         {
2161             if (szOut != NULL)
2162                 CFStringGetCharacters(theString, CFRangeMake(0, nOutSize - 1), (UniChar*) szOut);
2163
2164             nRealOutSize = CFStringGetLength(theString) + 1;
2165         }
2166         else
2167         {
2168             CFStringGetBytes(
2169                 theString,
2170                 CFRangeMake(0, CFStringGetLength(theString)),
2171                 m_encoding,
2172                 0, //what to put in characters that can't be converted -
2173                     //0 tells CFString to return NULL if it meets such a character
2174                 false, //not an external representation
2175                 (UInt8*) szOut,
2176                 nOutSize,
2177                 (CFIndex*) &nRealOutSize
2178                         );
2179         }
2180
2181         CFRelease(theString);
2182
2183 #if SIZEOF_WCHAR_T == 4
2184         delete[] szUniBuffer;
2185 #endif
2186
2187         return  nRealOutSize - 1;
2188     }
2189
2190     bool IsOk() const
2191     {
2192         return m_encoding != kCFStringEncodingInvalidId &&
2193               CFStringIsEncodingAvailable(m_encoding);
2194     }
2195
2196 private:
2197     CFStringEncoding m_encoding ;
2198 };
2199
2200 #endif // defined(__WXCOCOA__)
2201
2202 // ============================================================================
2203 // Mac conversion classes
2204 // ============================================================================
2205
2206 #if defined(__WXMAC__) && defined(TARGET_CARBON)
2207
2208 class wxMBConv_mac : public wxMBConv
2209 {
2210 public:
2211     wxMBConv_mac()
2212     {
2213         Init(CFStringGetSystemEncoding()) ;
2214     }
2215
2216 #if wxUSE_FONTMAP
2217     wxMBConv_mac(const wxChar* name)
2218     {
2219         Init( wxMacGetSystemEncFromFontEnc(wxFontMapperBase::Get()->CharsetToEncoding(name, false) ) ) ;
2220     }
2221 #endif
2222
2223     wxMBConv_mac(wxFontEncoding encoding)
2224     {
2225         Init( wxMacGetSystemEncFromFontEnc(encoding) );
2226     }
2227
2228     ~wxMBConv_mac()
2229     {
2230         OSStatus status = noErr ;
2231         status = TECDisposeConverter(m_MB2WC_converter);
2232         status = TECDisposeConverter(m_WC2MB_converter);
2233     }
2234
2235
2236     void Init( TextEncodingBase encoding)
2237     {
2238         OSStatus status = noErr ;
2239         m_char_encoding = encoding ;
2240         m_unicode_encoding = CreateTextEncoding(kTextEncodingUnicodeDefault,0,kUnicode16BitFormat) ;
2241
2242         status = TECCreateConverter(&m_MB2WC_converter,
2243                                     m_char_encoding,
2244                                     m_unicode_encoding);
2245         status = TECCreateConverter(&m_WC2MB_converter,
2246                                     m_unicode_encoding,
2247                                     m_char_encoding);
2248     }
2249
2250     size_t MB2WC(wchar_t *buf, const char *psz, size_t n) const
2251     {
2252         OSStatus status = noErr ;
2253         ByteCount byteOutLen ;
2254         ByteCount byteInLen = strlen(psz) ;
2255         wchar_t *tbuf = NULL ;
2256         UniChar* ubuf = NULL ;
2257         size_t res = 0 ;
2258
2259         if (buf == NULL)
2260         {
2261             //apple specs say at least 32
2262             n = wxMax( 32 , byteInLen ) ;
2263             tbuf = (wchar_t*) malloc( n * SIZEOF_WCHAR_T) ;
2264         }
2265         ByteCount byteBufferLen = n * sizeof( UniChar ) ;
2266 #if SIZEOF_WCHAR_T == 4
2267         ubuf = (UniChar*) malloc( byteBufferLen + 2 ) ;
2268 #else
2269         ubuf = (UniChar*) (buf ? buf : tbuf) ;
2270 #endif
2271         status = TECConvertText(m_MB2WC_converter, (ConstTextPtr) psz , byteInLen, &byteInLen,
2272           (TextPtr) ubuf , byteBufferLen, &byteOutLen);
2273 #if SIZEOF_WCHAR_T == 4
2274         // we have to terminate here, because n might be larger for the trailing zero, and if UniChar
2275         // is not properly terminated we get random characters at the end
2276         ubuf[byteOutLen / sizeof( UniChar ) ] = 0 ;
2277         wxMBConvUTF16 converter ;
2278         res = converter.MB2WC( (buf ? buf : tbuf) , (const char*)ubuf , n ) ;
2279         free( ubuf ) ;
2280 #else
2281         res = byteOutLen / sizeof( UniChar ) ;
2282 #endif
2283         if ( buf == NULL )
2284              free(tbuf) ;
2285
2286         if ( buf  && res < n)
2287             buf[res] = 0;
2288
2289         return res ;
2290     }
2291
2292     size_t WC2MB(char *buf, const wchar_t *psz, size_t n) const
2293     {
2294         OSStatus status = noErr ;
2295         ByteCount byteOutLen ;
2296         ByteCount byteInLen = wxWcslen(psz) * SIZEOF_WCHAR_T ;
2297
2298         char *tbuf = NULL ;
2299
2300         if (buf == NULL)
2301         {
2302             //apple specs say at least 32
2303             n = wxMax( 32 , ((byteInLen / SIZEOF_WCHAR_T) * 8) + SIZEOF_WCHAR_T );
2304             tbuf = (char*) malloc( n ) ;
2305         }
2306
2307         ByteCount byteBufferLen = n ;
2308         UniChar* ubuf = NULL ;
2309 #if SIZEOF_WCHAR_T == 4
2310         wxMBConvUTF16 converter ;
2311         size_t unicharlen = converter.WC2MB( NULL , psz , 0 ) ;
2312         byteInLen = unicharlen ;
2313         ubuf = (UniChar*) malloc( byteInLen + 2 ) ;
2314         converter.WC2MB( (char*) ubuf , psz, unicharlen + 2 ) ;
2315 #else
2316         ubuf = (UniChar*) psz ;
2317 #endif
2318         status = TECConvertText(m_WC2MB_converter, (ConstTextPtr) ubuf , byteInLen, &byteInLen,
2319             (TextPtr) (buf ? buf : tbuf) , byteBufferLen, &byteOutLen);
2320 #if SIZEOF_WCHAR_T == 4
2321         free( ubuf ) ;
2322 #endif
2323         if ( buf == NULL )
2324             free(tbuf) ;
2325
2326         size_t res = byteOutLen ;
2327         if ( buf  && res < n)
2328         {
2329             buf[res] = 0;
2330
2331             //we need to double-trip to verify it didn't insert any ? in place
2332             //of bogus characters
2333             wxWCharBuffer wcBuf(n);
2334             size_t pszlen = wxWcslen(psz);
2335             if ( MB2WC(wcBuf.data(), buf, n) == (size_t)-1 ||
2336                         wxWcslen(wcBuf) != pszlen ||
2337                         memcmp(wcBuf, psz, pszlen * sizeof(wchar_t)) != 0 )
2338             {
2339                 // we didn't obtain the same thing we started from, hence
2340                 // the conversion was lossy and we consider that it failed
2341                 return (size_t)-1;
2342             }
2343         }
2344
2345         return res ;
2346     }
2347
2348     bool IsOk() const
2349         { return m_MB2WC_converter !=  NULL && m_WC2MB_converter != NULL  ; }
2350
2351 private:
2352     TECObjectRef m_MB2WC_converter ;
2353     TECObjectRef m_WC2MB_converter ;
2354
2355     TextEncodingBase m_char_encoding ;
2356     TextEncodingBase m_unicode_encoding ;
2357 };
2358
2359 #endif // defined(__WXMAC__) && defined(TARGET_CARBON)
2360
2361 // ============================================================================
2362 // wxEncodingConverter based conversion classes
2363 // ============================================================================
2364
2365 #if wxUSE_FONTMAP
2366
2367 class wxMBConv_wxwin : public wxMBConv
2368 {
2369 private:
2370     void Init()
2371     {
2372         m_ok = m2w.Init(m_enc, wxFONTENCODING_UNICODE) &&
2373                w2m.Init(wxFONTENCODING_UNICODE, m_enc);
2374     }
2375
2376 public:
2377     // temporarily just use wxEncodingConverter stuff,
2378     // so that it works while a better implementation is built
2379     wxMBConv_wxwin(const wxChar* name)
2380     {
2381         if (name)
2382             m_enc = wxFontMapperBase::Get()->CharsetToEncoding(name, false);
2383         else
2384             m_enc = wxFONTENCODING_SYSTEM;
2385
2386         Init();
2387     }
2388
2389     wxMBConv_wxwin(wxFontEncoding enc)
2390     {
2391         m_enc = enc;
2392
2393         Init();
2394     }
2395
2396     size_t MB2WC(wchar_t *buf, const char *psz, size_t WXUNUSED(n)) const
2397     {
2398         size_t inbuf = strlen(psz);
2399         if (buf)
2400         {
2401             if (!m2w.Convert(psz,buf))
2402                 return (size_t)-1;
2403         }
2404         return inbuf;
2405     }
2406
2407     size_t WC2MB(char *buf, const wchar_t *psz, size_t WXUNUSED(n)) const
2408     {
2409         const size_t inbuf = wxWcslen(psz);
2410         if (buf)
2411         {
2412             if (!w2m.Convert(psz,buf))
2413                 return (size_t)-1;
2414         }
2415
2416         return inbuf;
2417     }
2418
2419     bool IsOk() const { return m_ok; }
2420
2421 public:
2422     wxFontEncoding m_enc;
2423     wxEncodingConverter m2w, w2m;
2424
2425     // were we initialized successfully?
2426     bool m_ok;
2427
2428     DECLARE_NO_COPY_CLASS(wxMBConv_wxwin)
2429 };
2430
2431 // make the constructors available for unit testing
2432 WXDLLIMPEXP_BASE wxMBConv* new_wxMBConv_wxwin( const wxChar* name )
2433 {
2434     wxMBConv_wxwin* result = new wxMBConv_wxwin( name );
2435     if ( !result->IsOk() )
2436     {
2437         delete result;
2438         return 0;
2439     }
2440     return result;
2441 }
2442
2443 #endif // wxUSE_FONTMAP
2444
2445 // ============================================================================
2446 // wxCSConv implementation
2447 // ============================================================================
2448
2449 void wxCSConv::Init()
2450 {
2451     m_name = NULL;
2452     m_convReal =  NULL;
2453     m_deferred = true;
2454 }
2455
2456 wxCSConv::wxCSConv(const wxChar *charset)
2457 {
2458     Init();
2459
2460     if ( charset )
2461     {
2462         SetName(charset);
2463     }
2464
2465 #if wxUSE_FONTMAP
2466     m_encoding = wxFontMapperBase::GetEncodingFromName(charset);
2467 #else
2468     m_encoding = wxFONTENCODING_SYSTEM;
2469 #endif
2470 }
2471
2472 wxCSConv::wxCSConv(wxFontEncoding encoding)
2473 {
2474     if ( encoding == wxFONTENCODING_MAX || encoding == wxFONTENCODING_DEFAULT )
2475     {
2476         wxFAIL_MSG( _T("invalid encoding value in wxCSConv ctor") );
2477
2478         encoding = wxFONTENCODING_SYSTEM;
2479     }
2480
2481     Init();
2482
2483     m_encoding = encoding;
2484 }
2485
2486 wxCSConv::~wxCSConv()
2487 {
2488     Clear();
2489 }
2490
2491 wxCSConv::wxCSConv(const wxCSConv& conv)
2492         : wxMBConv()
2493 {
2494     Init();
2495
2496     SetName(conv.m_name);
2497     m_encoding = conv.m_encoding;
2498 }
2499
2500 wxCSConv& wxCSConv::operator=(const wxCSConv& conv)
2501 {
2502     Clear();
2503
2504     SetName(conv.m_name);
2505     m_encoding = conv.m_encoding;
2506
2507     return *this;
2508 }
2509
2510 void wxCSConv::Clear()
2511 {
2512     free(m_name);
2513     delete m_convReal;
2514
2515     m_name = NULL;
2516     m_convReal = NULL;
2517 }
2518
2519 void wxCSConv::SetName(const wxChar *charset)
2520 {
2521     if (charset)
2522     {
2523         m_name = wxStrdup(charset);
2524         m_deferred = true;
2525     }
2526 }
2527
2528 #if wxUSE_FONTMAP
2529 #include "wx/hashmap.h"
2530
2531 WX_DECLARE_HASH_MAP( wxFontEncoding, wxString, wxIntegerHash, wxIntegerEqual,
2532                      wxEncodingNameCache );
2533
2534 static wxEncodingNameCache gs_nameCache;
2535 #endif
2536
2537 wxMBConv *wxCSConv::DoCreate() const
2538 {
2539 #if wxUSE_FONTMAP
2540     wxLogTrace(TRACE_STRCONV,
2541                wxT("creating conversion for %s"),
2542                (m_name ? m_name
2543                        : wxFontMapperBase::GetEncodingName(m_encoding).c_str()));
2544 #endif // wxUSE_FONTMAP
2545
2546     // check for the special case of ASCII or ISO8859-1 charset: as we have
2547     // special knowledge of it anyhow, we don't need to create a special
2548     // conversion object
2549     if ( m_encoding == wxFONTENCODING_ISO8859_1 ||
2550             m_encoding == wxFONTENCODING_DEFAULT )
2551     {
2552         // don't convert at all
2553         return NULL;
2554     }
2555
2556     // we trust OS to do conversion better than we can so try external
2557     // conversion methods first
2558     //
2559     // the full order is:
2560     //      1. OS conversion (iconv() under Unix or Win32 API)
2561     //      2. hard coded conversions for UTF
2562     //      3. wxEncodingConverter as fall back
2563
2564     // step (1)
2565 #ifdef HAVE_ICONV
2566 #if !wxUSE_FONTMAP
2567     if ( m_name )
2568 #endif // !wxUSE_FONTMAP
2569     {
2570         wxString name(m_name);
2571         wxFontEncoding encoding(m_encoding);
2572
2573         if ( !name.empty() )
2574         {
2575             wxMBConv_iconv *conv = new wxMBConv_iconv(name);
2576             if ( conv->IsOk() )
2577                 return conv;
2578
2579             delete conv;
2580
2581 #if wxUSE_FONTMAP
2582             encoding =
2583                 wxFontMapperBase::Get()->CharsetToEncoding(name, false);
2584 #endif // wxUSE_FONTMAP
2585         }
2586 #if wxUSE_FONTMAP
2587         {
2588             const wxEncodingNameCache::iterator it = gs_nameCache.find(encoding);
2589             if ( it != gs_nameCache.end() )
2590             {
2591                 if ( it->second.empty() )
2592                     return NULL;
2593
2594                 wxMBConv_iconv *conv = new wxMBConv_iconv(it->second);
2595                 if ( conv->IsOk() )
2596                     return conv;
2597
2598                 delete conv;
2599             }
2600
2601             const wxChar** names = wxFontMapperBase::GetAllEncodingNames(encoding);
2602
2603             for ( ; *names; ++names )
2604             {
2605                 wxMBConv_iconv *conv = new wxMBConv_iconv(*names);
2606                 if ( conv->IsOk() )
2607                 {
2608                     gs_nameCache[encoding] = *names;
2609                     return conv;
2610                 }
2611
2612                 delete conv;
2613             }
2614
2615             gs_nameCache[encoding] = _T(""); // cache the failure
2616         }
2617 #endif // wxUSE_FONTMAP
2618     }
2619 #endif // HAVE_ICONV
2620
2621 #ifdef wxHAVE_WIN32_MB2WC
2622     {
2623 #if wxUSE_FONTMAP
2624         wxMBConv_win32 *conv = m_name ? new wxMBConv_win32(m_name)
2625                                       : new wxMBConv_win32(m_encoding);
2626         if ( conv->IsOk() )
2627             return conv;
2628
2629         delete conv;
2630 #else
2631         return NULL;
2632 #endif
2633     }
2634 #endif // wxHAVE_WIN32_MB2WC
2635 #if defined(__WXMAC__)
2636     {
2637         // leave UTF16 and UTF32 to the built-ins of wx
2638         if ( m_name || ( m_encoding < wxFONTENCODING_UTF16BE ||
2639             ( m_encoding >= wxFONTENCODING_MACMIN && m_encoding <= wxFONTENCODING_MACMAX ) ) )
2640         {
2641
2642 #if wxUSE_FONTMAP
2643             wxMBConv_mac *conv = m_name ? new wxMBConv_mac(m_name)
2644                                         : new wxMBConv_mac(m_encoding);
2645 #else
2646             wxMBConv_mac *conv = new wxMBConv_mac(m_encoding);
2647 #endif
2648             if ( conv->IsOk() )
2649                  return conv;
2650
2651             delete conv;
2652         }
2653     }
2654 #endif
2655 #if defined(__WXCOCOA__)
2656     {
2657         if ( m_name || ( m_encoding <= wxFONTENCODING_UTF16 ) )
2658         {
2659
2660 #if wxUSE_FONTMAP
2661             wxMBConv_cocoa *conv = m_name ? new wxMBConv_cocoa(m_name)
2662                                           : new wxMBConv_cocoa(m_encoding);
2663 #else
2664             wxMBConv_cocoa *conv = new wxMBConv_cocoa(m_encoding);
2665 #endif
2666             if ( conv->IsOk() )
2667                  return conv;
2668
2669             delete conv;
2670         }
2671     }
2672 #endif
2673     // step (2)
2674     wxFontEncoding enc = m_encoding;
2675 #if wxUSE_FONTMAP
2676     if ( enc == wxFONTENCODING_SYSTEM && m_name )
2677     {
2678         // use "false" to suppress interactive dialogs -- we can be called from
2679         // anywhere and popping up a dialog from here is the last thing we want to
2680         // do
2681         enc = wxFontMapperBase::Get()->CharsetToEncoding(m_name, false);
2682     }
2683 #endif // wxUSE_FONTMAP
2684
2685     switch ( enc )
2686     {
2687         case wxFONTENCODING_UTF7:
2688              return new wxMBConvUTF7;
2689
2690         case wxFONTENCODING_UTF8:
2691              return new wxMBConvUTF8;
2692
2693         case wxFONTENCODING_UTF16BE:
2694              return new wxMBConvUTF16BE;
2695
2696         case wxFONTENCODING_UTF16LE:
2697              return new wxMBConvUTF16LE;
2698
2699         case wxFONTENCODING_UTF32BE:
2700              return new wxMBConvUTF32BE;
2701
2702         case wxFONTENCODING_UTF32LE:
2703              return new wxMBConvUTF32LE;
2704
2705         default:
2706              // nothing to do but put here to suppress gcc warnings
2707              ;
2708     }
2709
2710     // step (3)
2711 #if wxUSE_FONTMAP
2712     {
2713         wxMBConv_wxwin *conv = m_name ? new wxMBConv_wxwin(m_name)
2714                                       : new wxMBConv_wxwin(m_encoding);
2715         if ( conv->IsOk() )
2716             return conv;
2717
2718         delete conv;
2719     }
2720 #endif // wxUSE_FONTMAP
2721
2722     // NB: This is a hack to prevent deadlock. What could otherwise happen
2723     //     in Unicode build: wxConvLocal creation ends up being here
2724     //     because of some failure and logs the error. But wxLog will try to
2725     //     attach timestamp, for which it will need wxConvLocal (to convert
2726     //     time to char* and then wchar_t*), but that fails, tries to log
2727     //     error, but wxLog has a (already locked) critical section that
2728     //     guards static buffer.
2729     static bool alreadyLoggingError = false;
2730     if (!alreadyLoggingError)
2731     {
2732         alreadyLoggingError = true;
2733         wxLogError(_("Cannot convert from the charset '%s'!"),
2734                    m_name ? m_name
2735                       :
2736 #if wxUSE_FONTMAP
2737                          wxFontMapperBase::GetEncodingDescription(m_encoding).c_str()
2738 #else // !wxUSE_FONTMAP
2739                          wxString::Format(_("encoding %s"), m_encoding).c_str()
2740 #endif // wxUSE_FONTMAP/!wxUSE_FONTMAP
2741               );
2742         alreadyLoggingError = false;
2743     }
2744
2745     return NULL;
2746 }
2747
2748 void wxCSConv::CreateConvIfNeeded() const
2749 {
2750     if ( m_deferred )
2751     {
2752         wxCSConv *self = (wxCSConv *)this; // const_cast
2753
2754 #if wxUSE_INTL
2755         // if we don't have neither the name nor the encoding, use the default
2756         // encoding for this system
2757         if ( !m_name && m_encoding == wxFONTENCODING_SYSTEM )
2758         {
2759             self->m_name = wxStrdup(wxLocale::GetSystemEncodingName());
2760         }
2761 #endif // wxUSE_INTL
2762
2763         self->m_convReal = DoCreate();
2764         self->m_deferred = false;
2765     }
2766 }
2767
2768 size_t wxCSConv::MB2WC(wchar_t *buf, const char *psz, size_t n) const
2769 {
2770     CreateConvIfNeeded();
2771
2772     if (m_convReal)
2773         return m_convReal->MB2WC(buf, psz, n);
2774
2775     // latin-1 (direct)
2776     size_t len = strlen(psz);
2777
2778     if (buf)
2779     {
2780         for (size_t c = 0; c <= len; c++)
2781             buf[c] = (unsigned char)(psz[c]);
2782     }
2783
2784     return len;
2785 }
2786
2787 size_t wxCSConv::WC2MB(char *buf, const wchar_t *psz, size_t n) const
2788 {
2789     CreateConvIfNeeded();
2790
2791     if (m_convReal)
2792         return m_convReal->WC2MB(buf, psz, n);
2793
2794     // latin-1 (direct)
2795     const size_t len = wxWcslen(psz);
2796     if (buf)
2797     {
2798         for (size_t c = 0; c <= len; c++)
2799         {
2800             if (psz[c] > 0xFF)
2801                 return (size_t)-1;
2802             buf[c] = (char)psz[c];
2803         }
2804     }
2805     else
2806     {
2807         for (size_t c = 0; c <= len; c++)
2808         {
2809             if (psz[c] > 0xFF)
2810                 return (size_t)-1;
2811         }
2812     }
2813
2814     return len;
2815 }
2816
2817 // ----------------------------------------------------------------------------
2818 // globals
2819 // ----------------------------------------------------------------------------
2820
2821 #ifdef __WINDOWS__
2822     static wxMBConv_win32 wxConvLibcObj;
2823 #elif defined(__WXMAC__) && !defined(__MACH__)
2824     static wxMBConv_mac wxConvLibcObj ;
2825 #else
2826     static wxMBConvLibc wxConvLibcObj;
2827 #endif
2828
2829 static wxCSConv wxConvLocalObj(wxFONTENCODING_SYSTEM);
2830 static wxCSConv wxConvISO8859_1Obj(wxFONTENCODING_ISO8859_1);
2831 static wxMBConvUTF7 wxConvUTF7Obj;
2832 static wxMBConvUTF8 wxConvUTF8Obj;
2833
2834 WXDLLIMPEXP_DATA_BASE(wxMBConv&) wxConvLibc = wxConvLibcObj;
2835 WXDLLIMPEXP_DATA_BASE(wxCSConv&) wxConvLocal = wxConvLocalObj;
2836 WXDLLIMPEXP_DATA_BASE(wxCSConv&) wxConvISO8859_1 = wxConvISO8859_1Obj;
2837 WXDLLIMPEXP_DATA_BASE(wxMBConvUTF7&) wxConvUTF7 = wxConvUTF7Obj;
2838 WXDLLIMPEXP_DATA_BASE(wxMBConvUTF8&) wxConvUTF8 = wxConvUTF8Obj;
2839 WXDLLIMPEXP_DATA_BASE(wxMBConv *) wxConvCurrent = &wxConvLibcObj;
2840 WXDLLIMPEXP_DATA_BASE(wxMBConv *) wxConvFileName = &
2841 #ifdef __WXOSX__
2842                                     wxConvUTF8Obj;
2843 #else
2844                                     wxConvLibcObj;
2845 #endif
2846
2847
2848 #else // !wxUSE_WCHAR_T
2849
2850 // stand-ins in absence of wchar_t
2851 WXDLLIMPEXP_DATA_BASE(wxMBConv) wxConvLibc,
2852                                 wxConvISO8859_1,
2853                                 wxConvLocal,
2854                                 wxConvUTF8;
2855
2856 #endif // wxUSE_WCHAR_T/!wxUSE_WCHAR_T