tests/mbconv/mbconvtest.cpp

   1 ///////////////////////////////////////////////////////////////////////////////
   2 // Name:        tests/mbconv/main.cpp
   3 // Purpose:     wxMBConv unit test
   4 // Author:      Vadim Zeitlin, Mike Wetherell, Vince Harron
   5 // Created:     14.02.04
   6 // Copyright:   (c) 2003 TT-Solutions, (c) 2005 Mike Wetherell, Vince Harron
   7 ///////////////////////////////////////////////////////////////////////////////
   8
   9 // ----------------------------------------------------------------------------
  10 // headers
  11 // ----------------------------------------------------------------------------
  12
  13 #include "testprec.h"
  14
  15 #ifdef __BORLANDC__
  16     #pragma hdrstop
  17 #endif
  18
  19 #ifndef WX_PRECOMP
  20     #include "wx/wx.h"
  21 #endif // WX_PRECOMP
  22
  23 #include "wx/strconv.h"
  24 #include "wx/string.h"
  25 #include "wx/txtstrm.h"
  26 #include "wx/mstream.h"
  27
  28 #if defined wxHAVE_TCHAR_SUPPORT && !defined HAVE_WCHAR_H
  29     #define HAVE_WCHAR_H
  30 #endif
  31
  32 // ----------------------------------------------------------------------------
  33 // Some wide character constants. "\uXXXX" escapes aren't supported by old
  34 // compilers such as VC++ 5 and g++ 2.95.
  35 // ----------------------------------------------------------------------------
  36
  37 wchar_t u41[] = { 0x41, 0 };
  38 wchar_t u7f[] = { 0x7f, 0 };
  39
  40 wchar_t u80[] = { 0x80, 0 };
  41 wchar_t u391[] = { 0x391, 0 };
  42 wchar_t u7ff[] = { 0x7ff, 0 };
  43
  44 wchar_t u800[] = { 0x800, 0 };
  45 wchar_t u2620[] = { 0x2620, 0 };
  46 wchar_t ufffd[] = { 0xfffd, 0 };
  47
  48 #if SIZEOF_WCHAR_T == 4
  49 wchar_t u10000[] = { 0x10000, 0 };
  50 wchar_t u1000a5[] = { 0x1000a5, 0 };
  51 wchar_t u10fffd[] = { 0x10fffd, 0 };
  52 #else
  53 wchar_t u10000[] = { 0xd800, 0xdc00, 0 };
  54 wchar_t u1000a5[] = { 0xdbc0, 0xdca5, 0 };
  55 wchar_t u10fffd[] = { 0xdbff, 0xdffd, 0 };
  56 #endif
  57
  58 // ----------------------------------------------------------------------------
  59 // test class
  60 // ----------------------------------------------------------------------------
  61
  62 class MBConvTestCase : public CppUnit::TestCase
  63 {
  64 public:
  65     MBConvTestCase() { }
  66
  67 private:
  68     CPPUNIT_TEST_SUITE( MBConvTestCase );
  69         CPPUNIT_TEST( UTF32LETests );
  70         CPPUNIT_TEST( UTF32BETests );
  71         CPPUNIT_TEST( WC2CP1250 );
  72         CPPUNIT_TEST( UTF7Tests );
  73         CPPUNIT_TEST( UTF8Tests );
  74         CPPUNIT_TEST( UTF16LETests );
  75         CPPUNIT_TEST( UTF16BETests );
  76         CPPUNIT_TEST( CP932Tests );
  77         CPPUNIT_TEST( CP1252Tests ); // depends on UTF8 Decoder functioning correctly
  78         CPPUNIT_TEST( LibcTests );
  79         CPPUNIT_TEST( IconvTests );
  80         CPPUNIT_TEST( Latin1Tests );
  81         CPPUNIT_TEST( FontmapTests );
  82         CPPUNIT_TEST( BufSize );
  83         CPPUNIT_TEST( FromWCharTests );
  84 #ifdef HAVE_WCHAR_H
  85         CPPUNIT_TEST( UTF8_41 );
  86         CPPUNIT_TEST( UTF8_7f );
  87         CPPUNIT_TEST( UTF8_80 );
  88         CPPUNIT_TEST( UTF8_c2_7f );
  89         CPPUNIT_TEST( UTF8_c2_80 );
  90         CPPUNIT_TEST( UTF8_ce_91 );
  91         CPPUNIT_TEST( UTF8_df_bf );
  92         CPPUNIT_TEST( UTF8_df_c0 );
  93         CPPUNIT_TEST( UTF8_e0_a0_7f );
  94         CPPUNIT_TEST( UTF8_e0_a0_80 );
  95         CPPUNIT_TEST( UTF8_e2_98_a0 );
  96         CPPUNIT_TEST( UTF8_ef_bf_bd );
  97         CPPUNIT_TEST( UTF8_ef_bf_c0 );
  98         CPPUNIT_TEST( UTF8_f0_90_80_7f );
  99         CPPUNIT_TEST( UTF8_f0_90_80_80 );
 100         CPPUNIT_TEST( UTF8_f4_8f_bf_bd );
 101         CPPUNIT_TEST( UTF8PUA_f4_80_82_a5 );
 102         CPPUNIT_TEST( UTF8Octal_backslash245 );
 103 #endif // HAVE_WCHAR_H
 104     CPPUNIT_TEST_SUITE_END();
 105
 106     void WC2CP1250();
 107     void UTF7Tests();
 108     void UTF8Tests();
 109     void UTF16LETests();
 110     void UTF16BETests();
 111     void UTF32LETests();
 112     void UTF32BETests();
 113     void CP932Tests();
 114     void CP1252Tests();
 115     void LibcTests();
 116     void FontmapTests();
 117     void BufSize();
 118     void FromWCharTests();
 119     void IconvTests();
 120     void Latin1Tests();
 121
 122     // verifies that the specified multibyte sequence decodes to the specified wchar_t sequence
 123     void TestDecoder(
 124         const wchar_t* wideBuffer,  // the same character sequence as multiBuffer, encoded as wchar_t
 125         size_t         wideChars,   // the number of wide characters at wideBuffer
 126         const char*    multiBuffer, // a multibyte encoded character sequence that can be decoded by "converter"
 127         size_t         multiBytes,  // the byte length of the multibyte character sequence that can be decoded by "converter"
 128         wxMBConv&      converter,   // the wxMBConv object that can decode multiBuffer into a wide character sequence
 129         int            sizeofNull   // number of bytes occupied by terminating null in this encoding
 130         );
 131
 132     // verifies that the specified wchar_t sequence encodes to the specified multibyte sequence
 133     void TestEncoder(
 134         const wchar_t* wideBuffer,  // the same character sequence as multiBuffer, encoded as wchar_t
 135         size_t         wideChars,   // the number of wide characters at wideBuffer
 136         const char*    multiBuffer, // a multibyte encoded character sequence that can be decoded by "converter"
 137         size_t         multiBytes,  // the byte length of the multibyte character sequence that can be decoded by "converter"
 138         wxMBConv&      converter,   // the wxMBConv object that can decode multiBuffer into a wide character sequence
 139         int            sizeofNull   // number of bytes occupied by terminating null in this encoding
 140         );
 141
 142 #if wxUSE_UNICODE && wxUSE_STREAMS
 143     // use wxTextInputStream to exercise wxMBConv interface
 144     // (this reveals some bugs in certain wxMBConv subclasses)
 145     void TestStreamDecoder(
 146         const wchar_t* wideBuffer,  // the same character sequence as multiBuffer, encoded as wchar_t
 147         size_t         wideChars,   // the number of wide characters at wideBuffer
 148         const char*    multiBuffer, // a multibyte encoded character sequence that can be decoded by "converter"
 149         size_t         multiBytes,  // the byte length of the multibyte character sequence that can be decoded by "converter"
 150         wxMBConv&      converter    // the wxMBConv object that can decode multiBuffer into a wide character sequence
 151         );
 152
 153     // use wxTextOutputStream to exercise wxMBConv interface
 154     // (this reveals some bugs in certain wxMBConv subclasses)
 155     void TestStreamEncoder(
 156         const wchar_t* wideBuffer,  // the same character sequence as multiBuffer, encoded as wchar_t
 157         size_t         wideChars,   // the number of wide characters at wideBuffer
 158         const char*    multiBuffer, // a multibyte encoded character sequence that can be decoded by "converter"
 159         size_t         multiBytes,  // the byte length of the multibyte character sequence that can be decoded by "converter"
 160         wxMBConv&      converter    // the wxMBConv object that can decode multiBuffer into a wide character sequence
 161         );
 162 #endif
 163
 164     // tests the encoding and decoding capability of an wxMBConv object
 165     //
 166     // decodes the utf-8 bytes into wide characters
 167     // encodes the wide characters to compare against input multiBuffer
 168     // decodes the multiBuffer to compare against wide characters
 169     // decodes the multiBuffer into wide characters
 170     void TestCoder(
 171         const char*    multiBuffer, // a multibyte encoded character sequence that can be decoded by "converter"
 172         size_t         multiBytes,  // the byte length of the multibyte character sequence that can be decoded by "converter"
 173         const char*    utf8Buffer,  // the same character sequence as multiBuffer, encoded as UTF-8
 174         size_t         utf8Bytes,   // the byte length of the UTF-8 encoded character sequence
 175         wxMBConv&      converter,   // the wxMBConv object that can decode multiBuffer into a wide character sequence
 176         int            sizeofNull   // the number of bytes occupied by a terminating null in the converter's encoding
 177         );
 178
 179 #ifdef HAVE_WCHAR_H
 180     // UTF-8 tests. Test the first, last and one in the middle for sequences
 181     // of each length
 182     void UTF8_41() { UTF8("\x41", u41); }
 183     void UTF8_7f() { UTF8("\x7f", u7f); }
 184     void UTF8_80() { UTF8("\x80", NULL); }
 185
 186     void UTF8_c2_7f() { UTF8("\xc2\x7f", NULL); }
 187     void UTF8_c2_80() { UTF8("\xc2\x80", u80); }
 188     void UTF8_ce_91() { UTF8("\xce\x91", u391); }
 189     void UTF8_df_bf() { UTF8("\xdf\xbf", u7ff); }
 190     void UTF8_df_c0() { UTF8("\xdf\xc0", NULL); }
 191
 192     void UTF8_e0_a0_7f() { UTF8("\xe0\xa0\x7f", NULL); }
 193     void UTF8_e0_a0_80() { UTF8("\xe0\xa0\x80", u800); }
 194     void UTF8_e2_98_a0() { UTF8("\xe2\x98\xa0", u2620); }
 195     void UTF8_ef_bf_bd() { UTF8("\xef\xbf\xbd", ufffd); }
 196     void UTF8_ef_bf_c0() { UTF8("\xef\xbf\xc0", NULL); }
 197
 198     void UTF8_f0_90_80_7f() { UTF8("\xf0\x90\x80\x7f", NULL); }
 199     void UTF8_f0_90_80_80() { UTF8("\xf0\x90\x80\x80", u10000); }
 200     void UTF8_f4_8f_bf_bd() { UTF8("\xf4\x8f\xbf\xbd", u10fffd); }
 201
 202     // test 'escaping the escape characters' for the two escaping schemes
 203     void UTF8PUA_f4_80_82_a5() { UTF8PUA("\xf4\x80\x82\xa5", u1000a5); }
 204     void UTF8Octal_backslash245() { UTF8Octal("\\245", L"\\245"); }
 205
 206     // implementation for the utf-8 tests (see comments below)
 207     void UTF8(const char *charSequence, const wchar_t *wideSequence);
 208     void UTF8PUA(const char *charSequence, const wchar_t *wideSequence);
 209     void UTF8Octal(const char *charSequence, const wchar_t *wideSequence);
 210     void UTF8(const char *charSequence, const wchar_t *wideSequence, int option);
 211 #endif // HAVE_WCHAR_H
 212
 213     DECLARE_NO_COPY_CLASS(MBConvTestCase)
 214 };
 215
 216 // register in the unnamed registry so that these tests are run by default
 217 CPPUNIT_TEST_SUITE_REGISTRATION( MBConvTestCase );
 218
 219 // also include in its own registry so that these tests can be run alone
 220 CPPUNIT_TEST_SUITE_NAMED_REGISTRATION( MBConvTestCase, "MBConvTestCase" );
 221 CPPUNIT_TEST_SUITE_NAMED_REGISTRATION( MBConvTestCase, "MBConv" );
 222
 223 void MBConvTestCase::WC2CP1250()
 224 {
 225     static const struct Data
 226     {
 227         const wchar_t *wc;
 228         const char *cp1250;
 229     } data[] =
 230     {
 231         { L"hello", "hello" },  // test that it works in simplest case
 232         { L"\xBD of \xBD is \xBC", NULL }, // this should fail as cp1250 doesn't have 1/2
 233     };
 234
 235     wxCSConv cs1250(wxFONTENCODING_CP1250);
 236     for ( size_t n = 0; n < WXSIZEOF(data); n++ )
 237     {
 238         const Data& d = data[n];
 239         if (d.cp1250)
 240         {
 241             CPPUNIT_ASSERT( strcmp(cs1250.cWC2MB(d.wc), d.cp1250) == 0 );
 242         }
 243         else
 244         {
 245             CPPUNIT_ASSERT( (const char*)cs1250.cWC2MB(d.wc) == NULL );
 246         }
 247     }
 248 }
 249
 250 // Print an unsigned character array as a C unsigned character array.
 251 // NB: Please don't remove this function even though it's not used anywhere,
 252 //     it's very useful when debugging a failed test.
 253 wxString CByteArrayFormat( const void* data, size_t len, const wxChar* name )
 254 {
 255     const unsigned char* bytes = (unsigned char*)data;
 256     wxString result;
 257
 258     result.Printf( wxT("static const unsigned char %s[%i] = \n{"), name, (int)len );
 259
 260     for ( size_t i = 0; i < len; i++ )
 261     {
 262         if ( i != 0 )
 263         {
 264             result.append( wxT(",") );
 265         }
 266         if ((i%16)==0)
 267         {
 268             result.append( wxT("\n    ") );
 269         }
 270         wxString byte = wxString::Format( wxT("0x%02x"), bytes[i] );
 271         result.append(byte);
 272     }
 273     result.append( wxT("\n};\n") );
 274     return result;
 275 }
 276
 277 // The following bytes represent the same string, containing Japanese and English
 278 // characters, encoded in several different formats.
 279
 280 // encoded by iconv
 281 static const unsigned char welcome_utf7_iconv[84] =
 282 {
 283     0x57,0x65,0x6c,0x63,0x6f,0x6d,0x65,0x20,0x74,0x6f,0x20,0x6f,0x75,0x72,0x20,0x63,
 284     0x79,0x62,0x65,0x72,0x20,0x73,0x70,0x61,0x63,0x65,0x20,0x66,0x6f,0x72,0x63,0x65,
 285     0x2e,0x20,0x20,0x2b,0x4d,0x46,0x6b,0x77,0x55,0x49,0x74,0x6d,0x57,0x39,0x38,0x77,
 286     0x61,0x35,0x62,0x37,0x69,0x6e,0x45,0x77,0x6b,0x6a,0x42,0x5a,0x4d,0x49,0x73,0x77,
 287     0x65,0x7a,0x42,0x47,0x4d,0x45,0x77,0x77,0x52,0x44,0x42,0x45,0x4d,0x47,0x63,0x77,
 288     0x57,0x54,0x41,0x43
 289 };
 290 // encoded by wxWindows (iconv can decode this successfully)
 291 static const unsigned char welcome_utf7_wx[109] =
 292 {
 293     0x57,0x65,0x6c,0x63,0x6f,0x6d,0x65,0x2b,0x41,0x43,0x41,0x2d,0x74,0x6f,0x2b,0x41,
 294     0x43,0x41,0x2d,0x6f,0x75,0x72,0x2b,0x41,0x43,0x41,0x2d,0x63,0x79,0x62,0x65,0x72,
 295     0x2b,0x41,0x43,0x41,0x2d,0x73,0x70,0x61,0x63,0x65,0x2b,0x41,0x43,0x41,0x2d,0x66,
 296     0x6f,0x72,0x63,0x65,0x2e,0x2b,0x41,0x43,0x41,0x41,0x49,0x44,0x42,0x5a,0x4d,0x46,
 297     0x43,0x4c,0x5a,0x6c,0x76,0x66,0x4d,0x47,0x75,0x57,0x2b,0x34,0x70,0x78,0x4d,0x4a,
 298     0x49,0x77,0x57,0x54,0x43,0x4c,0x4d,0x48,0x73,0x77,0x52,0x6a,0x42,0x4d,0x4d,0x45,
 299     0x51,0x77,0x52,0x44,0x42,0x6e,0x4d,0x46,0x6b,0x77,0x41,0x67,0x2d
 300 };
 301 // encoded by iconv
 302 static const unsigned char welcome_utf8[89] =
 303 {
 304     0x57,0x65,0x6c,0x63,0x6f,0x6d,0x65,0x20,0x74,0x6f,0x20,0x6f,0x75,0x72,0x20,0x63,
 305     0x79,0x62,0x65,0x72,0x20,0x73,0x70,0x61,0x63,0x65,0x20,0x66,0x6f,0x72,0x63,0x65,
 306     0x2e,0x20,0x20,0xe3,0x81,0x99,0xe3,0x81,0x90,0xe8,0xad,0xa6,0xe5,0xaf,0x9f,0xe3,
 307     0x81,0xab,0xe9,0x9b,0xbb,0xe8,0xa9,0xb1,0xe3,0x82,0x92,0xe3,0x81,0x99,0xe3,0x82,
 308     0x8b,0xe3,0x81,0xbb,0xe3,0x81,0x86,0xe3,0x81,0x8c,0xe3,0x81,0x84,0xe3,0x81,0x84,
 309     0xe3,0x81,0xa7,0xe3,0x81,0x99,0xe3,0x80,0x82
 310 };
 311 // encoded by iconv
 312 static const unsigned char welcome_utf16le[106] =
 313 {
 314     0x57,0x00,0x65,0x00,0x6c,0x00,0x63,0x00,0x6f,0x00,0x6d,0x00,0x65,0x00,0x20,0x00,
 315     0x74,0x00,0x6f,0x00,0x20,0x00,0x6f,0x00,0x75,0x00,0x72,0x00,0x20,0x00,0x63,0x00,
 316     0x79,0x00,0x62,0x00,0x65,0x00,0x72,0x00,0x20,0x00,0x73,0x00,0x70,0x00,0x61,0x00,
 317     0x63,0x00,0x65,0x00,0x20,0x00,0x66,0x00,0x6f,0x00,0x72,0x00,0x63,0x00,0x65,0x00,
 318     0x2e,0x00,0x20,0x00,0x20,0x00,0x59,0x30,0x50,0x30,0x66,0x8b,0xdf,0x5b,0x6b,0x30,
 319     0xfb,0x96,0x71,0x8a,0x92,0x30,0x59,0x30,0x8b,0x30,0x7b,0x30,0x46,0x30,0x4c,0x30,
 320     0x44,0x30,0x44,0x30,0x67,0x30,0x59,0x30,0x02,0x30
 321 };
 322 // encoded by iconv
 323 static const unsigned char welcome_utf16be[106] =
 324 {
 325     0x00,0x57,0x00,0x65,0x00,0x6c,0x00,0x63,0x00,0x6f,0x00,0x6d,0x00,0x65,0x00,0x20,
 326     0x00,0x74,0x00,0x6f,0x00,0x20,0x00,0x6f,0x00,0x75,0x00,0x72,0x00,0x20,0x00,0x63,
 327     0x00,0x79,0x00,0x62,0x00,0x65,0x00,0x72,0x00,0x20,0x00,0x73,0x00,0x70,0x00,0x61,
 328     0x00,0x63,0x00,0x65,0x00,0x20,0x00,0x66,0x00,0x6f,0x00,0x72,0x00,0x63,0x00,0x65,
 329     0x00,0x2e,0x00,0x20,0x00,0x20,0x30,0x59,0x30,0x50,0x8b,0x66,0x5b,0xdf,0x30,0x6b,
 330     0x96,0xfb,0x8a,0x71,0x30,0x92,0x30,0x59,0x30,0x8b,0x30,0x7b,0x30,0x46,0x30,0x4c,
 331     0x30,0x44,0x30,0x44,0x30,0x67,0x30,0x59,0x30,0x02
 332 };
 333 // encoded by iconv
 334 static const unsigned char welcome_utf32le[212] =
 335 {
 336     0x57,0x00,0x00,0x00,0x65,0x00,0x00,0x00,0x6c,0x00,0x00,0x00,0x63,0x00,0x00,0x00,
 337     0x6f,0x00,0x00,0x00,0x6d,0x00,0x00,0x00,0x65,0x00,0x00,0x00,0x20,0x00,0x00,0x00,
 338     0x74,0x00,0x00,0x00,0x6f,0x00,0x00,0x00,0x20,0x00,0x00,0x00,0x6f,0x00,0x00,0x00,
 339     0x75,0x00,0x00,0x00,0x72,0x00,0x00,0x00,0x20,0x00,0x00,0x00,0x63,0x00,0x00,0x00,
 340     0x79,0x00,0x00,0x00,0x62,0x00,0x00,0x00,0x65,0x00,0x00,0x00,0x72,0x00,0x00,0x00,
 341     0x20,0x00,0x00,0x00,0x73,0x00,0x00,0x00,0x70,0x00,0x00,0x00,0x61,0x00,0x00,0x00,
 342     0x63,0x00,0x00,0x00,0x65,0x00,0x00,0x00,0x20,0x00,0x00,0x00,0x66,0x00,0x00,0x00,
 343     0x6f,0x00,0x00,0x00,0x72,0x00,0x00,0x00,0x63,0x00,0x00,0x00,0x65,0x00,0x00,0x00,
 344     0x2e,0x00,0x00,0x00,0x20,0x00,0x00,0x00,0x20,0x00,0x00,0x00,0x59,0x30,0x00,0x00,
 345     0x50,0x30,0x00,0x00,0x66,0x8b,0x00,0x00,0xdf,0x5b,0x00,0x00,0x6b,0x30,0x00,0x00,
 346     0xfb,0x96,0x00,0x00,0x71,0x8a,0x00,0x00,0x92,0x30,0x00,0x00,0x59,0x30,0x00,0x00,
 347     0x8b,0x30,0x00,0x00,0x7b,0x30,0x00,0x00,0x46,0x30,0x00,0x00,0x4c,0x30,0x00,0x00,
 348     0x44,0x30,0x00,0x00,0x44,0x30,0x00,0x00,0x67,0x30,0x00,0x00,0x59,0x30,0x00,0x00,
 349     0x02,0x30,0x00,0x00
 350 };
 351 // encoded by iconv
 352 static const unsigned char welcome_utf32be[212] =
 353 {
 354     0x00,0x00,0x00,0x57,0x00,0x00,0x00,0x65,0x00,0x00,0x00,0x6c,0x00,0x00,0x00,0x63,
 355     0x00,0x00,0x00,0x6f,0x00,0x00,0x00,0x6d,0x00,0x00,0x00,0x65,0x00,0x00,0x00,0x20,
 356     0x00,0x00,0x00,0x74,0x00,0x00,0x00,0x6f,0x00,0x00,0x00,0x20,0x00,0x00,0x00,0x6f,
 357     0x00,0x00,0x00,0x75,0x00,0x00,0x00,0x72,0x00,0x00,0x00,0x20,0x00,0x00,0x00,0x63,
 358     0x00,0x00,0x00,0x79,0x00,0x00,0x00,0x62,0x00,0x00,0x00,0x65,0x00,0x00,0x00,0x72,
 359     0x00,0x00,0x00,0x20,0x00,0x00,0x00,0x73,0x00,0x00,0x00,0x70,0x00,0x00,0x00,0x61,
 360     0x00,0x00,0x00,0x63,0x00,0x00,0x00,0x65,0x00,0x00,0x00,0x20,0x00,0x00,0x00,0x66,
 361     0x00,0x00,0x00,0x6f,0x00,0x00,0x00,0x72,0x00,0x00,0x00,0x63,0x00,0x00,0x00,0x65,
 362     0x00,0x00,0x00,0x2e,0x00,0x00,0x00,0x20,0x00,0x00,0x00,0x20,0x00,0x00,0x30,0x59,
 363     0x00,0x00,0x30,0x50,0x00,0x00,0x8b,0x66,0x00,0x00,0x5b,0xdf,0x00,0x00,0x30,0x6b,
 364     0x00,0x00,0x96,0xfb,0x00,0x00,0x8a,0x71,0x00,0x00,0x30,0x92,0x00,0x00,0x30,0x59,
 365     0x00,0x00,0x30,0x8b,0x00,0x00,0x30,0x7b,0x00,0x00,0x30,0x46,0x00,0x00,0x30,0x4c,
 366     0x00,0x00,0x30,0x44,0x00,0x00,0x30,0x44,0x00,0x00,0x30,0x67,0x00,0x00,0x30,0x59,
 367     0x00,0x00,0x30,0x02
 368 };
 369 // encoded by iconv
 370 static const unsigned char welcome_cp932[71] =
 371 {
 372     0x57,0x65,0x6c,0x63,0x6f,0x6d,0x65,0x20,0x74,0x6f,0x20,0x6f,0x75,0x72,0x20,0x63,
 373     0x79,0x62,0x65,0x72,0x20,0x73,0x70,0x61,0x63,0x65,0x20,0x66,0x6f,0x72,0x63,0x65,
 374     0x2e,0x20,0x20,0x82,0xb7,0x82,0xae,0x8c,0x78,0x8e,0x40,0x82,0xc9,0x93,0x64,0x98,
 375     0x62,0x82,0xf0,0x82,0xb7,0x82,0xe9,0x82,0xd9,0x82,0xa4,0x82,0xaa,0x82,0xa2,0x82,
 376     0xa2,0x82,0xc5,0x82,0xb7,0x81,0x42
 377 };
 378
 379 #if wxBYTE_ORDER == wxBIG_ENDIAN
 380     #if SIZEOF_WCHAR_T == 2
 381         #define welcome_wchar_t welcome_utf16be
 382     #elif SIZEOF_WCHAR_T == 4
 383         #define welcome_wchar_t welcome_utf32be
 384     #endif
 385 #elif wxBYTE_ORDER == wxLITTLE_ENDIAN
 386     #if SIZEOF_WCHAR_T == 2
 387         #define welcome_wchar_t welcome_utf16le
 388     #elif SIZEOF_WCHAR_T == 4
 389         #define welcome_wchar_t welcome_utf32le
 390     #endif
 391 #endif
 392
 393 void MBConvTestCase::UTF7Tests()
 394 {
 395 #if 0
 396     wxCSConv convUTF7(wxFONTENCODING_UTF7);
 397 #else
 398     wxMBConvUTF7 convUTF7;
 399 #endif
 400
 401     TestDecoder
 402         (
 403         (const wchar_t*)welcome_wchar_t,
 404         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 405         (const char*)welcome_utf7_iconv,
 406         sizeof(welcome_utf7_iconv),
 407         convUTF7,
 408         1
 409         );
 410     TestDecoder
 411         (
 412         (const wchar_t*)welcome_wchar_t,
 413         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 414         (const char*)welcome_utf7_wx,
 415         sizeof(welcome_utf7_wx),
 416         convUTF7,
 417         1
 418         );
 419 #if 0
 420     // wxWidget's UTF-7 encoder generates different byte sequences than iconv's.
 421     // but both seem to be equally legal.
 422     // This test won't work and that's okay.
 423     TestEncoder
 424         (
 425         (const wchar_t*)welcome_wchar_t,
 426         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 427         (const char*)welcome_utf7_iconv,
 428         sizeof(welcome_utf7_iconv),
 429         convUTF7,
 430         1
 431         );
 432 #endif
 433     TestEncoder
 434         (
 435         (const wchar_t*)welcome_wchar_t,
 436         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 437         (const char*)welcome_utf7_wx,
 438         sizeof(welcome_utf7_wx),
 439         convUTF7,
 440         1
 441         );
 442 }
 443
 444 void MBConvTestCase::UTF8Tests()
 445 {
 446     TestDecoder
 447         (
 448         (const wchar_t*)welcome_wchar_t,
 449         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 450         (const char*)welcome_utf8,
 451         sizeof(welcome_utf8),
 452         wxConvUTF8,
 453         1
 454         );
 455     TestEncoder
 456         (
 457         (const wchar_t*)welcome_wchar_t,
 458         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 459         (const char*)welcome_utf8,
 460         sizeof(welcome_utf8),
 461         wxConvUTF8,
 462         1
 463         );
 464 }
 465
 466 void MBConvTestCase::UTF16LETests()
 467 {
 468     wxMBConvUTF16LE convUTF16LE;
 469     TestDecoder
 470         (
 471         (const wchar_t*)welcome_wchar_t,
 472         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 473         (const char*)welcome_utf16le,
 474         sizeof(welcome_utf16le),
 475         convUTF16LE,
 476         2
 477         );
 478     TestEncoder
 479         (
 480         (const wchar_t*)welcome_wchar_t,
 481         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 482         (const char*)welcome_utf16le,
 483         sizeof(welcome_utf16le),
 484         convUTF16LE,
 485         2
 486         );
 487 }
 488
 489 void MBConvTestCase::UTF16BETests()
 490 {
 491     wxMBConvUTF16BE convUTF16BE;
 492     TestDecoder
 493         (
 494         (const wchar_t*)welcome_wchar_t,
 495         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 496         (const char*)welcome_utf16be,
 497         sizeof(welcome_utf16be),
 498         convUTF16BE,
 499         2
 500         );
 501     TestEncoder
 502         (
 503         (const wchar_t*)welcome_wchar_t,
 504         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 505         (const char*)welcome_utf16be,
 506         sizeof(welcome_utf16be),
 507         convUTF16BE,
 508         2
 509         );
 510 }
 511
 512 void MBConvTestCase::UTF32LETests()
 513 {
 514     wxMBConvUTF32LE convUTF32LE;
 515     TestDecoder
 516         (
 517         (const wchar_t*)welcome_wchar_t,
 518         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 519         (const char*)welcome_utf32le,
 520         sizeof(welcome_utf32le),
 521         convUTF32LE,
 522         4
 523         );
 524     TestEncoder
 525         (
 526         (const wchar_t*)welcome_wchar_t,
 527         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 528         (const char*)welcome_utf32le,
 529         sizeof(welcome_utf32le),
 530         convUTF32LE,
 531         4
 532         );
 533 }
 534
 535 void MBConvTestCase::UTF32BETests()
 536 {
 537     wxMBConvUTF32BE convUTF32BE;
 538     TestDecoder
 539         (
 540         (const wchar_t*)welcome_wchar_t,
 541         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 542         (const char*)welcome_utf32be,
 543         sizeof(welcome_utf32be),
 544         convUTF32BE,
 545         4
 546         );
 547     TestEncoder
 548         (
 549         (const wchar_t*)welcome_wchar_t,
 550         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 551         (const char*)welcome_utf32be,
 552         sizeof(welcome_utf32be),
 553         convUTF32BE,
 554         4
 555         );
 556 }
 557
 558 void MBConvTestCase::CP932Tests()
 559 {
 560     wxCSConv convCP932( wxFONTENCODING_CP932 );
 561     TestDecoder
 562         (
 563         (const wchar_t*)welcome_wchar_t,
 564         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 565         (const char*)welcome_cp932,
 566         sizeof(welcome_cp932),
 567         convCP932,
 568         1
 569         );
 570     TestEncoder
 571         (
 572         (const wchar_t*)welcome_wchar_t,
 573         sizeof(welcome_wchar_t)/sizeof(wchar_t),
 574         (const char*)welcome_cp932,
 575         sizeof(welcome_cp932),
 576         convCP932,
 577         1
 578         );
 579 }
 580
 581 // a character sequence encoded as iso8859-1 (iconv)
 582 static const unsigned char iso8859_1[251] =
 583 {
 584     0x05,0x06,0x07,0x08,0x09,0x0a,0x0b,0x0c,0x0d,0x0e,0x0f,0x10,0x11,0x12,0x13,0x14,
 585     0x15,0x16,0x17,0x18,0x19,0x1a,0x1b,0x1c,0x1d,0x1e,0x1f,0x20,0x21,0x22,0x23,0x24,
 586     0x25,0x26,0x27,0x28,0x29,0x2a,0x2b,0x2c,0x2d,0x2e,0x2f,0x30,0x31,0x32,0x33,0x34,
 587     0x35,0x36,0x37,0x38,0x39,0x3a,0x3b,0x3c,0x3d,0x3e,0x3f,0x40,0x41,0x42,0x43,0x44,
 588     0x45,0x46,0x47,0x48,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f,0x50,0x51,0x52,0x53,0x54,
 589     0x55,0x56,0x57,0x58,0x59,0x5a,0x5b,0x5c,0x5d,0x5e,0x5f,0x60,0x61,0x62,0x63,0x64,
 590     0x65,0x66,0x67,0x68,0x69,0x6a,0x6b,0x6c,0x6d,0x6e,0x6f,0x70,0x71,0x72,0x73,0x74,
 591     0x75,0x76,0x77,0x78,0x79,0x7a,0x7b,0x7c,0x7d,0x7e,0x7f,0x80,0x81,0x82,0x83,0x84,
 592     0x85,0x86,0x87,0x88,0x89,0x8a,0x8b,0x8c,0x8d,0x8e,0x8f,0x90,0x91,0x92,0x93,0x94,
 593     0x95,0x96,0x97,0x98,0x99,0x9a,0x9b,0x9c,0x9d,0x9e,0x9f,0xa0,0xa1,0xa2,0xa3,0xa4,
 594     0xa5,0xa6,0xa7,0xa8,0xa9,0xaa,0xab,0xac,0xad,0xae,0xaf,0xb0,0xb1,0xb2,0xb3,0xb4,
 595     0xb5,0xb6,0xb7,0xb8,0xb9,0xba,0xbb,0xbc,0xbd,0xbe,0xbf,0xc0,0xc1,0xc2,0xc3,0xc4,
 596     0xc5,0xc6,0xc7,0xc8,0xc9,0xca,0xcb,0xcc,0xcd,0xce,0xcf,0xd0,0xd1,0xd2,0xd3,0xd4,
 597     0xd5,0xd6,0xd7,0xd8,0xd9,0xda,0xdb,0xdc,0xdd,0xde,0xdf,0xe0,0xe1,0xe2,0xe3,0xe4,
 598     0xe5,0xe6,0xe7,0xe8,0xe9,0xea,0xeb,0xec,0xed,0xee,0xef,0xf0,0xf1,0xf2,0xf3,0xf4,
 599     0xf5,0xf6,0xf7,0xf8,0xf9,0xfa,0xfb,0xfc,0xfd,0xfe,0xff
 600 };
 601 // the above character sequence encoded as UTF-8 (iconv)
 602 static const unsigned char iso8859_1_utf8[379] =
 603 {
 604     0x05,0x06,0x07,0x08,0x09,0x0a,0x0b,0x0c,0x0d,0x0e,0x0f,0x10,0x11,0x12,0x13,0x14,
 605     0x15,0x16,0x17,0x18,0x19,0x1a,0x1b,0x1c,0x1d,0x1e,0x1f,0x20,0x21,0x22,0x23,0x24,
 606     0x25,0x26,0x27,0x28,0x29,0x2a,0x2b,0x2c,0x2d,0x2e,0x2f,0x30,0x31,0x32,0x33,0x34,
 607     0x35,0x36,0x37,0x38,0x39,0x3a,0x3b,0x3c,0x3d,0x3e,0x3f,0x40,0x41,0x42,0x43,0x44,
 608     0x45,0x46,0x47,0x48,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f,0x50,0x51,0x52,0x53,0x54,
 609     0x55,0x56,0x57,0x58,0x59,0x5a,0x5b,0x5c,0x5d,0x5e,0x5f,0x60,0x61,0x62,0x63,0x64,
 610     0x65,0x66,0x67,0x68,0x69,0x6a,0x6b,0x6c,0x6d,0x6e,0x6f,0x70,0x71,0x72,0x73,0x74,
 611     0x75,0x76,0x77,0x78,0x79,0x7a,0x7b,0x7c,0x7d,0x7e,0x7f,0xc2,0x80,0xc2,0x81,0xc2,
 612     0x82,0xc2,0x83,0xc2,0x84,0xc2,0x85,0xc2,0x86,0xc2,0x87,0xc2,0x88,0xc2,0x89,0xc2,
 613     0x8a,0xc2,0x8b,0xc2,0x8c,0xc2,0x8d,0xc2,0x8e,0xc2,0x8f,0xc2,0x90,0xc2,0x91,0xc2,
 614     0x92,0xc2,0x93,0xc2,0x94,0xc2,0x95,0xc2,0x96,0xc2,0x97,0xc2,0x98,0xc2,0x99,0xc2,
 615     0x9a,0xc2,0x9b,0xc2,0x9c,0xc2,0x9d,0xc2,0x9e,0xc2,0x9f,0xc2,0xa0,0xc2,0xa1,0xc2,
 616     0xa2,0xc2,0xa3,0xc2,0xa4,0xc2,0xa5,0xc2,0xa6,0xc2,0xa7,0xc2,0xa8,0xc2,0xa9,0xc2,
 617     0xaa,0xc2,0xab,0xc2,0xac,0xc2,0xad,0xc2,0xae,0xc2,0xaf,0xc2,0xb0,0xc2,0xb1,0xc2,
 618     0xb2,0xc2,0xb3,0xc2,0xb4,0xc2,0xb5,0xc2,0xb6,0xc2,0xb7,0xc2,0xb8,0xc2,0xb9,0xc2,
 619     0xba,0xc2,0xbb,0xc2,0xbc,0xc2,0xbd,0xc2,0xbe,0xc2,0xbf,0xc3,0x80,0xc3,0x81,0xc3,
 620     0x82,0xc3,0x83,0xc3,0x84,0xc3,0x85,0xc3,0x86,0xc3,0x87,0xc3,0x88,0xc3,0x89,0xc3,
 621     0x8a,0xc3,0x8b,0xc3,0x8c,0xc3,0x8d,0xc3,0x8e,0xc3,0x8f,0xc3,0x90,0xc3,0x91,0xc3,
 622     0x92,0xc3,0x93,0xc3,0x94,0xc3,0x95,0xc3,0x96,0xc3,0x97,0xc3,0x98,0xc3,0x99,0xc3,
 623     0x9a,0xc3,0x9b,0xc3,0x9c,0xc3,0x9d,0xc3,0x9e,0xc3,0x9f,0xc3,0xa0,0xc3,0xa1,0xc3,
 624     0xa2,0xc3,0xa3,0xc3,0xa4,0xc3,0xa5,0xc3,0xa6,0xc3,0xa7,0xc3,0xa8,0xc3,0xa9,0xc3,
 625     0xaa,0xc3,0xab,0xc3,0xac,0xc3,0xad,0xc3,0xae,0xc3,0xaf,0xc3,0xb0,0xc3,0xb1,0xc3,
 626     0xb2,0xc3,0xb3,0xc3,0xb4,0xc3,0xb5,0xc3,0xb6,0xc3,0xb7,0xc3,0xb8,0xc3,0xb9,0xc3,
 627     0xba,0xc3,0xbb,0xc3,0xbc,0xc3,0xbd,0xc3,0xbe,0xc3,0xbf
 628 };
 629
 630 // a character sequence encoded as CP1252 (iconv)
 631 static const unsigned char CP1252[246] =
 632 {
 633     0x05,0x06,0x07,0x08,0x09,0x0a,0x0b,0x0c,0x0d,0x0e,0x0f,0x10,0x11,0x12,0x13,0x14,
 634     0x15,0x16,0x17,0x18,0x19,0x1a,0x1b,0x1c,0x1d,0x1e,0x1f,0x20,0x21,0x22,0x23,0x24,
 635     0x25,0x26,0x27,0x28,0x29,0x2a,0x2b,0x2c,0x2d,0x2e,0x2f,0x30,0x31,0x32,0x33,0x34,
 636     0x35,0x36,0x37,0x38,0x39,0x3a,0x3b,0x3c,0x3d,0x3e,0x3f,0x40,0x41,0x42,0x43,0x44,
 637     0x45,0x46,0x47,0x48,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f,0x50,0x51,0x52,0x53,0x54,
 638     0x55,0x56,0x57,0x58,0x59,0x5a,0x5b,0x5c,0x5d,0x5e,0x5f,0x60,0x61,0x62,0x63,0x64,
 639     0x65,0x66,0x67,0x68,0x69,0x6a,0x6b,0x6c,0x6d,0x6e,0x6f,0x70,0x71,0x72,0x73,0x74,
 640     0x75,0x76,0x77,0x78,0x79,0x7a,0x7b,0x7c,0x7d,0x7e,0x7f,0xa0,0xa1,0xa2,0xa3,0xa4,
 641     0xa5,0xa6,0xa7,0xa8,0xa9,0xaa,0xab,0xac,0xad,0xae,0xaf,0xb0,0xb1,0xb2,0xb3,0xb4,
 642     0xb5,0xb6,0xb7,0xb8,0xb9,0xba,0xbb,0xbc,0xbd,0xbe,0xbf,0xc0,0xc1,0xc2,0xc3,0xc4,
 643     0xc5,0xc6,0xc7,0xc8,0xc9,0xca,0xcb,0xcc,0xcd,0xce,0xcf,0xd0,0xd1,0xd2,0xd3,0xd4,
 644     0xd5,0xd6,0xd7,0xd8,0xd9,0xda,0xdb,0xdc,0xdd,0xde,0xdf,0xe0,0xe1,0xe2,0xe3,0xe4,
 645     0xe5,0xe6,0xe7,0xe8,0xe9,0xea,0xeb,0xec,0xed,0xee,0xef,0xf0,0xf1,0xf2,0xf3,0xf4,
 646     0xf5,0xf6,0xf7,0xf8,0xf9,0xfa,0xfb,0xfc,0xfd,0xfe,0xff,0x8c,0x9c,0x8a,0x9a,0x9f,
 647     0x8e,0x9e,0x83,0x88,0x98,0x96,0x97,0x91,0x92,0x82,0x93,0x94,0x84,0x86,0x87,0x95,
 648     0x85,0x89,0x8b,0x9b,0x80,0x99
 649 };
 650 // the above character sequence encoded as UTF-8 (iconv)
 651 static const unsigned char CP1252_utf8[386] =
 652 {
 653     0x05,0x06,0x07,0x08,0x09,0x0a,0x0b,0x0c,0x0d,0x0e,0x0f,0x10,0x11,0x12,0x13,0x14,
 654     0x15,0x16,0x17,0x18,0x19,0x1a,0x1b,0x1c,0x1d,0x1e,0x1f,0x20,0x21,0x22,0x23,0x24,
 655     0x25,0x26,0x27,0x28,0x29,0x2a,0x2b,0x2c,0x2d,0x2e,0x2f,0x30,0x31,0x32,0x33,0x34,
 656     0x35,0x36,0x37,0x38,0x39,0x3a,0x3b,0x3c,0x3d,0x3e,0x3f,0x40,0x41,0x42,0x43,0x44,
 657     0x45,0x46,0x47,0x48,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f,0x50,0x51,0x52,0x53,0x54,
 658     0x55,0x56,0x57,0x58,0x59,0x5a,0x5b,0x5c,0x5d,0x5e,0x5f,0x60,0x61,0x62,0x63,0x64,
 659     0x65,0x66,0x67,0x68,0x69,0x6a,0x6b,0x6c,0x6d,0x6e,0x6f,0x70,0x71,0x72,0x73,0x74,
 660     0x75,0x76,0x77,0x78,0x79,0x7a,0x7b,0x7c,0x7d,0x7e,0x7f,0xc2,0xa0,0xc2,0xa1,0xc2,
 661     0xa2,0xc2,0xa3,0xc2,0xa4,0xc2,0xa5,0xc2,0xa6,0xc2,0xa7,0xc2,0xa8,0xc2,0xa9,0xc2,
 662     0xaa,0xc2,0xab,0xc2,0xac,0xc2,0xad,0xc2,0xae,0xc2,0xaf,0xc2,0xb0,0xc2,0xb1,0xc2,
 663     0xb2,0xc2,0xb3,0xc2,0xb4,0xc2,0xb5,0xc2,0xb6,0xc2,0xb7,0xc2,0xb8,0xc2,0xb9,0xc2,
 664     0xba,0xc2,0xbb,0xc2,0xbc,0xc2,0xbd,0xc2,0xbe,0xc2,0xbf,0xc3,0x80,0xc3,0x81,0xc3,
 665     0x82,0xc3,0x83,0xc3,0x84,0xc3,0x85,0xc3,0x86,0xc3,0x87,0xc3,0x88,0xc3,0x89,0xc3,
 666     0x8a,0xc3,0x8b,0xc3,0x8c,0xc3,0x8d,0xc3,0x8e,0xc3,0x8f,0xc3,0x90,0xc3,0x91,0xc3,
 667     0x92,0xc3,0x93,0xc3,0x94,0xc3,0x95,0xc3,0x96,0xc3,0x97,0xc3,0x98,0xc3,0x99,0xc3,
 668     0x9a,0xc3,0x9b,0xc3,0x9c,0xc3,0x9d,0xc3,0x9e,0xc3,0x9f,0xc3,0xa0,0xc3,0xa1,0xc3,
 669     0xa2,0xc3,0xa3,0xc3,0xa4,0xc3,0xa5,0xc3,0xa6,0xc3,0xa7,0xc3,0xa8,0xc3,0xa9,0xc3,
 670     0xaa,0xc3,0xab,0xc3,0xac,0xc3,0xad,0xc3,0xae,0xc3,0xaf,0xc3,0xb0,0xc3,0xb1,0xc3,
 671     0xb2,0xc3,0xb3,0xc3,0xb4,0xc3,0xb5,0xc3,0xb6,0xc3,0xb7,0xc3,0xb8,0xc3,0xb9,0xc3,
 672     0xba,0xc3,0xbb,0xc3,0xbc,0xc3,0xbd,0xc3,0xbe,0xc3,0xbf,0xc5,0x92,0xc5,0x93,0xc5,
 673     0xa0,0xc5,0xa1,0xc5,0xb8,0xc5,0xbd,0xc5,0xbe,0xc6,0x92,0xcb,0x86,0xcb,0x9c,0xe2,
 674     0x80,0x93,0xe2,0x80,0x94,0xe2,0x80,0x98,0xe2,0x80,0x99,0xe2,0x80,0x9a,0xe2,0x80,
 675     0x9c,0xe2,0x80,0x9d,0xe2,0x80,0x9e,0xe2,0x80,0xa0,0xe2,0x80,0xa1,0xe2,0x80,0xa2,
 676     0xe2,0x80,0xa6,0xe2,0x80,0xb0,0xe2,0x80,0xb9,0xe2,0x80,0xba,0xe2,0x82,0xac,0xe2,
 677     0x84,0xa2
 678 };
 679
 680 // this is unused currently so avoid warnings about this
 681 #if 0
 682
 683 // a character sequence encoded as iso8859-5 (iconv)
 684 static const unsigned char iso8859_5[251] =
 685 {
 686     0x05,0x06,0x07,0x08,0x09,0x0a,0x0b,0x0c,0x0d,0x0e,0x0f,0x10,0x11,0x12,0x13,0x14,
 687     0x15,0x16,0x17,0x18,0x19,0x1a,0x1b,0x1c,0x1d,0x1e,0x1f,0x20,0x21,0x22,0x23,0x24,
 688     0x25,0x26,0x27,0x28,0x29,0x2a,0x2b,0x2c,0x2d,0x2e,0x2f,0x30,0x31,0x32,0x33,0x34,
 689     0x35,0x36,0x37,0x38,0x39,0x3a,0x3b,0x3c,0x3d,0x3e,0x3f,0x40,0x41,0x42,0x43,0x44,
 690     0x45,0x46,0x47,0x48,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f,0x50,0x51,0x52,0x53,0x54,
 691     0x55,0x56,0x57,0x58,0x59,0x5a,0x5b,0x5c,0x5d,0x5e,0x5f,0x60,0x61,0x62,0x63,0x64,
 692     0x65,0x66,0x67,0x68,0x69,0x6a,0x6b,0x6c,0x6d,0x6e,0x6f,0x70,0x71,0x72,0x73,0x74,
 693     0x75,0x76,0x77,0x78,0x79,0x7a,0x7b,0x7c,0x7d,0x7e,0x7f,0x80,0x81,0x82,0x83,0x84,
 694     0x85,0x86,0x87,0x88,0x89,0x8a,0x8b,0x8c,0x8d,0x8e,0x8f,0x90,0x91,0x92,0x93,0x94,
 695     0x95,0x96,0x97,0x98,0x99,0x9a,0x9b,0x9c,0x9d,0x9e,0x9f,0xa0,0xfd,0xad,0xa1,0xa2,
 696     0xa3,0xa4,0xa5,0xa6,0xa7,0xa8,0xa9,0xaa,0xab,0xac,0xae,0xaf,0xb0,0xb1,0xb2,0xb3,
 697     0xb4,0xb5,0xb6,0xb7,0xb8,0xb9,0xba,0xbb,0xbc,0xbd,0xbe,0xbf,0xc0,0xc1,0xc2,0xc3,
 698     0xc4,0xc5,0xc6,0xc7,0xc8,0xc9,0xca,0xcb,0xcc,0xcd,0xce,0xcf,0xd0,0xd1,0xd2,0xd3,
 699     0xd4,0xd5,0xd6,0xd7,0xd8,0xd9,0xda,0xdb,0xdc,0xdd,0xde,0xdf,0xe0,0xe1,0xe2,0xe3,
 700     0xe4,0xe5,0xe6,0xe7,0xe8,0xe9,0xea,0xeb,0xec,0xed,0xee,0xef,0xf1,0xf2,0xf3,0xf4,
 701     0xf5,0xf6,0xf7,0xf8,0xf9,0xfa,0xfb,0xfc,0xfe,0xff,0xf0
 702 };
 703 // the above character sequence encoded as UTF-8 (iconv)
 704 static const unsigned char iso8859_5_utf8[380] =
 705 {
 706     0x05,0x06,0x07,0x08,0x09,0x0a,0x0b,0x0c,0x0d,0x0e,0x0f,0x10,0x11,0x12,0x13,0x14,
 707     0x15,0x16,0x17,0x18,0x19,0x1a,0x1b,0x1c,0x1d,0x1e,0x1f,0x20,0x21,0x22,0x23,0x24,
 708     0x25,0x26,0x27,0x28,0x29,0x2a,0x2b,0x2c,0x2d,0x2e,0x2f,0x30,0x31,0x32,0x33,0x34,
 709     0x35,0x36,0x37,0x38,0x39,0x3a,0x3b,0x3c,0x3d,0x3e,0x3f,0x40,0x41,0x42,0x43,0x44,
 710     0x45,0x46,0x47,0x48,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f,0x50,0x51,0x52,0x53,0x54,
 711     0x55,0x56,0x57,0x58,0x59,0x5a,0x5b,0x5c,0x5d,0x5e,0x5f,0x60,0x61,0x62,0x63,0x64,
 712     0x65,0x66,0x67,0x68,0x69,0x6a,0x6b,0x6c,0x6d,0x6e,0x6f,0x70,0x71,0x72,0x73,0x74,
 713     0x75,0x76,0x77,0x78,0x79,0x7a,0x7b,0x7c,0x7d,0x7e,0x7f,0xc2,0x80,0xc2,0x81,0xc2,
 714     0x82,0xc2,0x83,0xc2,0x84,0xc2,0x85,0xc2,0x86,0xc2,0x87,0xc2,0x88,0xc2,0x89,0xc2,
 715     0x8a,0xc2,0x8b,0xc2,0x8c,0xc2,0x8d,0xc2,0x8e,0xc2,0x8f,0xc2,0x90,0xc2,0x91,0xc2,
 716     0x92,0xc2,0x93,0xc2,0x94,0xc2,0x95,0xc2,0x96,0xc2,0x97,0xc2,0x98,0xc2,0x99,0xc2,
 717     0x9a,0xc2,0x9b,0xc2,0x9c,0xc2,0x9d,0xc2,0x9e,0xc2,0x9f,0xc2,0xa0,0xc2,0xa7,0xc2,
 718     0xad,0xd0,0x81,0xd0,0x82,0xd0,0x83,0xd0,0x84,0xd0,0x85,0xd0,0x86,0xd0,0x87,0xd0,
 719     0x88,0xd0,0x89,0xd0,0x8a,0xd0,0x8b,0xd0,0x8c,0xd0,0x8e,0xd0,0x8f,0xd0,0x90,0xd0,
 720     0x91,0xd0,0x92,0xd0,0x93,0xd0,0x94,0xd0,0x95,0xd0,0x96,0xd0,0x97,0xd0,0x98,0xd0,
 721     0x99,0xd0,0x9a,0xd0,0x9b,0xd0,0x9c,0xd0,0x9d,0xd0,0x9e,0xd0,0x9f,0xd0,0xa0,0xd0,
 722     0xa1,0xd0,0xa2,0xd0,0xa3,0xd0,0xa4,0xd0,0xa5,0xd0,0xa6,0xd0,0xa7,0xd0,0xa8,0xd0,
 723     0xa9,0xd0,0xaa,0xd0,0xab,0xd0,0xac,0xd0,0xad,0xd0,0xae,0xd0,0xaf,0xd0,0xb0,0xd0,
 724     0xb1,0xd0,0xb2,0xd0,0xb3,0xd0,0xb4,0xd0,0xb5,0xd0,0xb6,0xd0,0xb7,0xd0,0xb8,0xd0,
 725     0xb9,0xd0,0xba,0xd0,0xbb,0xd0,0xbc,0xd0,0xbd,0xd0,0xbe,0xd0,0xbf,0xd1,0x80,0xd1,
 726     0x81,0xd1,0x82,0xd1,0x83,0xd1,0x84,0xd1,0x85,0xd1,0x86,0xd1,0x87,0xd1,0x88,0xd1,
 727     0x89,0xd1,0x8a,0xd1,0x8b,0xd1,0x8c,0xd1,0x8d,0xd1,0x8e,0xd1,0x8f,0xd1,0x91,0xd1,
 728     0x92,0xd1,0x93,0xd1,0x94,0xd1,0x95,0xd1,0x96,0xd1,0x97,0xd1,0x98,0xd1,0x99,0xd1,
 729     0x9a,0xd1,0x9b,0xd1,0x9c,0xd1,0x9e,0xd1,0x9f,0xe2,0x84,0x96
 730 };
 731 #endif // 0
 732
 733 // DecodeUTF8
 734 // decodes the specified *unterminated* UTF-8 byte array
 735 wxWCharBuffer DecodeUTF8(
 736     const void* data, // an unterminated UTF-8 encoded byte array
 737     size_t size       // the byte length of data
 738     )
 739 {
 740     // the decoder requires a null terminated buffer.
 741     // the input data is not null terminated.
 742     // copy to null terminated buffer
 743
 744     wxCharBuffer nullTerminated( size+1 );
 745     memcpy( nullTerminated.data(), data, size );
 746     nullTerminated.data()[size] = 0;
 747     return wxConvUTF8.cMB2WC(nullTerminated.data());
 748 }
 749
 750 // tests the encoding and decoding capability of an wxMBConv object
 751 //
 752 // decodes the utf-8 bytes into wide characters
 753 // encodes the wide characters to compare against input multiBuffer
 754 // decodes the multiBuffer to compare against wide characters
 755 // decodes the multiBuffer into wide characters
 756 void MBConvTestCase::TestCoder(
 757     const char*    multiBuffer, // a multibyte encoded character sequence that can be decoded by "converter"
 758     size_t         multiBytes,  // the byte length of the multibyte character sequence that can be decoded by "converter"
 759     const char*    utf8Buffer,  // the same character sequence as multiBuffer, encoded as UTF-8
 760     size_t         utf8Bytes,   // the byte length of the UTF-8 encoded character sequence
 761     wxMBConv&      converter,   // the wxMBConv object thta can decode multiBuffer into a wide character sequence
 762     int            sizeofNull   // the number of bytes occupied by a terminating null in the converter's encoding
 763     )
 764 {
 765     // wide character size and endian-ess varies from platform to platform
 766     // compiler support for wide character literals varies from compiler to compiler
 767     // so we should store the wide character version as UTF-8 and depend on
 768     // the UTF-8 converter's ability to decode it to platform specific wide characters
 769     // this test is invalid if the UTF-8 converter can't decode
 770     const wxWCharBuffer wideBuffer(DecodeUTF8(utf8Buffer, utf8Bytes));
 771     const size_t wideChars = wxWcslen(wideBuffer);
 772
 773     TestDecoder
 774         (
 775         wideBuffer.data(),
 776         wideChars,
 777         multiBuffer,
 778         multiBytes,
 779         converter,
 780         sizeofNull
 781         );
 782     TestEncoder
 783         (
 784         wideBuffer.data(),
 785         wideChars,
 786         multiBuffer,
 787         multiBytes,
 788         converter,
 789         sizeofNull
 790         );
 791 }
 792
 793
 794 WXDLLIMPEXP_BASE wxMBConv* new_wxMBConv_wxwin( const char* name );
 795
 796 void MBConvTestCase::FontmapTests()
 797 {
 798 #ifdef wxUSE_FONTMAP
 799     wxMBConv* converter = new_wxMBConv_wxwin("CP1252");
 800     if ( !converter )
 801     {
 802         return;
 803     }
 804     TestCoder(
 805         (const char*)CP1252,
 806         sizeof(CP1252),
 807         (const char*)CP1252_utf8,
 808         sizeof(CP1252_utf8),
 809         *converter,
 810         1
 811         );
 812     delete converter;
 813 #endif
 814 }
 815
 816 void MBConvTestCase::BufSize()
 817 {
 818     wxCSConv conv1251(wxT("CP1251"));
 819     CPPUNIT_ASSERT( conv1251.IsOk() );
 820     const char *cp1251text =
 821         "\313\301\326\305\324\323\321 \325\304\301\336\316\331\315";
 822
 823     const size_t lenW = conv1251.MB2WC(NULL, cp1251text, 0);
 824     CPPUNIT_ASSERT_EQUAL( strlen(cp1251text), lenW );
 825     wxWCharBuffer wbuf(lenW + 1); // allocates lenW + 2 characters
 826     wbuf.data()[lenW + 1] = L'!';
 827
 828     // lenW is not enough because it's the length and we need the size
 829     CPPUNIT_ASSERT_EQUAL(
 830         wxCONV_FAILED, conv1251.MB2WC(wbuf.data(), cp1251text, lenW) );
 831
 832     // lenW+1 is just fine
 833     CPPUNIT_ASSERT(
 834         conv1251.MB2WC(wbuf.data(), cp1251text, lenW + 1) != wxCONV_FAILED );
 835
 836     // of course, greater values work too
 837     CPPUNIT_ASSERT(
 838         conv1251.MB2WC(wbuf.data(), cp1251text, lenW + 2) != wxCONV_FAILED );
 839
 840     // but they shouldn't write more stuff to the buffer
 841     CPPUNIT_ASSERT_EQUAL( L'!', wbuf[lenW + 1] );
 842
 843
 844     // test in the other direction too, using an encoding with multibyte NUL
 845     wxCSConv convUTF16(wxT("UTF-16LE"));
 846     CPPUNIT_ASSERT( convUTF16.IsOk() );
 847     const wchar_t *utf16text = L"Hello";
 848
 849     const size_t lenMB = convUTF16.WC2MB(NULL, utf16text, 0);
 850     CPPUNIT_ASSERT_EQUAL( wcslen(utf16text)*2, lenMB );
 851     wxCharBuffer buf(lenMB + 2); // it only adds 1 for NUL on its own, we need 2
 852                                  // for NUL and an extra one for the guard byte
 853     buf.data()[lenMB + 2] = '?';
 854
 855     CPPUNIT_ASSERT_EQUAL(
 856         wxCONV_FAILED, convUTF16.WC2MB(buf.data(), utf16text, lenMB) );
 857     CPPUNIT_ASSERT_EQUAL(
 858         wxCONV_FAILED, convUTF16.WC2MB(buf.data(), utf16text, lenMB + 1) );
 859     CPPUNIT_ASSERT(
 860         convUTF16.WC2MB(buf.data(), utf16text, lenMB + 2) != wxCONV_FAILED );
 861     CPPUNIT_ASSERT(
 862         convUTF16.WC2MB(buf.data(), utf16text, lenMB + 3) != wxCONV_FAILED );
 863     CPPUNIT_ASSERT_EQUAL( '?', buf[lenMB + 2] );
 864 }
 865
 866 void MBConvTestCase::FromWCharTests()
 867 {
 868     wxCSConv conv950("CP950");
 869     char mbuf[10];
 870     // U+4e00 is 2 bytes (0xa4 0x40) in cp950
 871     wchar_t wbuf[] = { 0x4e00, 0, 0x4e00, 0 };
 872
 873     // test simple ASCII text
 874     memset(mbuf, '!', sizeof(mbuf));
 875     CPPUNIT_ASSERT_EQUAL( wxCONV_FAILED, conv950.FromWChar(mbuf, 0, L"a", 1));
 876     CPPUNIT_ASSERT_EQUAL( '!', mbuf[0]);
 877
 878     memset(mbuf, '!', sizeof(mbuf));
 879     CPPUNIT_ASSERT_EQUAL( 1, conv950.FromWChar(mbuf, 1, L"a", 1));
 880     CPPUNIT_ASSERT_EQUAL( 'a', mbuf[0]);
 881     CPPUNIT_ASSERT_EQUAL( '!', mbuf[1]);
 882
 883     memset(mbuf, '!', sizeof(mbuf));
 884     CPPUNIT_ASSERT_EQUAL( wxCONV_FAILED, conv950.FromWChar(mbuf, 1, L"a", 2));
 885
 886     memset(mbuf, '!', sizeof(mbuf));
 887     CPPUNIT_ASSERT_EQUAL( 2, conv950.FromWChar(mbuf, 2, L"a", 2));
 888     CPPUNIT_ASSERT_EQUAL( 'a', mbuf[0]);
 889     CPPUNIT_ASSERT_EQUAL( '\0', mbuf[1]);
 890     CPPUNIT_ASSERT_EQUAL( '!', mbuf[2]);
 891
 892     // test non-ASCII text, 1 wchar -> 2 char
 893     memset(mbuf, '!', sizeof(mbuf));
 894     CPPUNIT_ASSERT_EQUAL( wxCONV_FAILED, conv950.FromWChar(mbuf, 0, wbuf, 1));
 895
 896     memset(mbuf, '!', sizeof(mbuf));
 897     CPPUNIT_ASSERT_EQUAL( wxCONV_FAILED, conv950.FromWChar(mbuf, 1, wbuf, 1));
 898
 899     memset(mbuf, '!', sizeof(mbuf));
 900     CPPUNIT_ASSERT_EQUAL( 2, conv950.FromWChar(mbuf, 2, wbuf, 1));
 901     CPPUNIT_ASSERT_EQUAL( '!', mbuf[2]);
 902
 903     memset(mbuf, '!', sizeof(mbuf));
 904     CPPUNIT_ASSERT_EQUAL( wxCONV_FAILED, conv950.FromWChar(mbuf, 2, wbuf, 2));
 905
 906     memset(mbuf, '!', sizeof(mbuf));
 907     CPPUNIT_ASSERT_EQUAL( 3, conv950.FromWChar(mbuf, 3, wbuf, 2));
 908     CPPUNIT_ASSERT_EQUAL( '\0', mbuf[2]);
 909     CPPUNIT_ASSERT_EQUAL( '!', mbuf[3]);
 910
 911     // test text with embedded NUL-character and srcLen specified
 912     memset(mbuf, '!', sizeof(mbuf));
 913     CPPUNIT_ASSERT_EQUAL( wxCONV_FAILED, conv950.FromWChar(mbuf, 3, wbuf, 3));
 914
 915     memset(mbuf, '!', sizeof(mbuf));
 916     CPPUNIT_ASSERT_EQUAL( wxCONV_FAILED, conv950.FromWChar(mbuf, 4, wbuf, 3));
 917     CPPUNIT_ASSERT_EQUAL( 5, conv950.FromWChar(mbuf, 5, wbuf, 3));
 918     CPPUNIT_ASSERT_EQUAL( '\0', mbuf[2]);
 919     CPPUNIT_ASSERT_EQUAL( '!', mbuf[5]);
 920
 921     memset(mbuf, '!', sizeof(mbuf));
 922     CPPUNIT_ASSERT_EQUAL( wxCONV_FAILED, conv950.FromWChar(mbuf, 5, wbuf, 4));
 923
 924     memset(mbuf, '!', sizeof(mbuf));
 925     CPPUNIT_ASSERT_EQUAL( 6, conv950.FromWChar(mbuf, 6, wbuf, 4));
 926     CPPUNIT_ASSERT_EQUAL( '\0', mbuf[2]);
 927     CPPUNIT_ASSERT_EQUAL( '\0', mbuf[5]);
 928     CPPUNIT_ASSERT_EQUAL( '!', mbuf[6]);
 929 }
 930
 931 WXDLLIMPEXP_BASE wxMBConv* new_wxMBConv_iconv( const char* name );
 932
 933 void MBConvTestCase::IconvTests()
 934 {
 935 #ifdef HAVE_ICONV
 936     wxMBConv* converter = new_wxMBConv_iconv("CP932");
 937     if ( !converter )
 938     {
 939         return;
 940     }
 941     TestCoder(
 942         (const char*)welcome_cp932,
 943         sizeof(welcome_cp932),
 944         (const char*)welcome_utf8,
 945         sizeof(welcome_utf8),
 946         *converter,
 947         1
 948         );
 949     delete converter;
 950 #endif
 951 }
 952
 953 void MBConvTestCase::Latin1Tests()
 954 {
 955     TestCoder(
 956         (const char*)iso8859_1,
 957         sizeof(iso8859_1),
 958         (const char*)iso8859_1_utf8,
 959         sizeof(iso8859_1_utf8),
 960         wxConvISO8859_1,
 961         1
 962         );
 963
 964     static const char nulstr[] = "foo\0bar\0";
 965     static const size_t mbLen = WXSIZEOF(nulstr) - 1;
 966     size_t wcLen;
 967     wxConvISO8859_1.cMB2WC(nulstr, mbLen, &wcLen);
 968     CPPUNIT_ASSERT_EQUAL( mbLen, wcLen );
 969 }
 970
 971 void MBConvTestCase::CP1252Tests()
 972 {
 973     wxCSConv convCP1252( wxFONTENCODING_CP1252 );
 974     TestCoder(
 975         (const char*)CP1252,
 976         sizeof(CP1252),
 977         (const char*)CP1252_utf8,
 978         sizeof(CP1252_utf8),
 979         convCP1252,
 980         1
 981         );
 982 }
 983
 984 void MBConvTestCase::LibcTests()
 985 {
 986     // The locale name are OS-dependent so this test is done only under Windows
 987     // when using MSVC (surprisingly it fails with MinGW, even though it's
 988     // supposed to use the same CRT -- no idea why and unfortunately gdb is too
 989     // flaky to debug it)
 990 #ifdef __VISUALC__
 991     LocaleSetter loc("English_United States.1252");
 992
 993     wxMBConvLibc convLibc;
 994     TestCoder(
 995         (const char*)CP1252,
 996         sizeof(CP1252),
 997         (const char*)CP1252_utf8,
 998         sizeof(CP1252_utf8),
 999         convLibc,
1000         1
1001         );
1002 #endif // __VISUALC__
1003 }
1004
1005 // verifies that the specified mb sequences decode to the specified wc sequence
1006 void MBConvTestCase::TestDecoder(
1007     const wchar_t* wideBuffer,  // the same character sequence as multiBuffer, encoded as wchar_t
1008     size_t         wideChars,   // the number of wide characters at wideBuffer
1009     const char*    multiBuffer, // a multibyte encoded character sequence that can be decoded by "converter"
1010     size_t         multiBytes,  // the byte length of the multibyte character sequence that can be decoded by "converter"
1011     wxMBConv&      converter,   // the wxMBConv object that can decode multiBuffer into a wide character sequence
1012     int            sizeofNull   // number of bytes occupied by terminating null in this encoding
1013     )
1014 {
1015     const unsigned UNINITIALIZED = 0xcd;
1016
1017     // copy the input bytes into a null terminated buffer
1018     wxCharBuffer inputCopy( multiBytes+sizeofNull );
1019     memcpy( inputCopy.data(), multiBuffer, multiBytes );
1020     memset( &inputCopy.data()[multiBytes], 0, sizeofNull );
1021
1022     // calculate the output size
1023     size_t outputWritten = converter.MB2WC
1024         (
1025         0,
1026         (const char*)inputCopy.data(),
1027         0
1028         );
1029     // make sure the correct output length was calculated
1030     WX_ASSERT_EQUAL_MESSAGE
1031     (
1032         ("while converting \"%s\"", multiBuffer),
1033         wideChars,
1034         outputWritten
1035     );
1036
1037     // convert the string
1038     size_t guardChars = 8; // to make sure we're not overrunning the output buffer
1039     size_t nullCharacters = 1;
1040     size_t outputBufferChars = outputWritten + nullCharacters + guardChars;
1041     wxWCharBuffer outputBuffer(outputBufferChars);
1042     memset( outputBuffer.data(), UNINITIALIZED, outputBufferChars*sizeof(wchar_t) );
1043
1044     outputWritten = converter.MB2WC
1045         (
1046         outputBuffer.data(),
1047         (const char*)inputCopy.data(),
1048         outputBufferChars
1049         );
1050     // make sure the correct number of characters were outputs
1051     CPPUNIT_ASSERT_EQUAL( wideChars, outputWritten );
1052
1053     // make sure the characters generated are correct
1054     CPPUNIT_ASSERT( 0 == memcmp( outputBuffer, wideBuffer, wideChars*sizeof(wchar_t) ) );
1055
1056     // the output buffer should be null terminated
1057     CPPUNIT_ASSERT(  outputBuffer[outputWritten] == 0 );
1058
1059     // make sure the rest of the output buffer is untouched
1060     for ( size_t i = (wideChars+1)*sizeof(wchar_t); i < (outputBufferChars*sizeof(wchar_t)); i++ )
1061     {
1062         CPPUNIT_ASSERT( ((unsigned char*)outputBuffer.data())[i] == UNINITIALIZED );
1063     }
1064
1065 #if wxUSE_UNICODE && wxUSE_STREAMS
1066     TestStreamDecoder( wideBuffer, wideChars, multiBuffer, multiBytes, converter );
1067 #endif
1068 }
1069
1070 // verifies that the specified wc sequences encodes to the specified mb sequence
1071 void MBConvTestCase::TestEncoder(
1072     const wchar_t* wideBuffer,  // the same character sequence as multiBuffer, encoded as wchar_t
1073     size_t         wideChars,   // the number of wide characters at wideBuffer
1074     const char*    multiBuffer, // a multibyte encoded character sequence that can be decoded by "converter"
1075     size_t         multiBytes,  // the byte length of the multibyte character sequence that can be decoded by "converter"
1076     wxMBConv&      converter,   // the wxMBConv object that can decode multiBuffer into a wide character sequence
1077     int            sizeofNull   // number of bytes occupied by terminating null in this encoding
1078     )
1079 {
1080     const unsigned UNINITIALIZED = 0xcd;
1081
1082     // copy the input bytes into a null terminated buffer
1083     wxWCharBuffer inputCopy( wideChars + 1 );
1084     memcpy( inputCopy.data(), wideBuffer, (wideChars*sizeof(wchar_t)) );
1085     inputCopy.data()[wideChars] = 0;
1086
1087     // calculate the output size
1088     size_t outputWritten = converter.WC2MB
1089         (
1090         0,
1091         (const wchar_t*)inputCopy.data(),
1092         0
1093         );
1094     // make sure the correct output length was calculated
1095     CPPUNIT_ASSERT_EQUAL( multiBytes, outputWritten );
1096
1097     // convert the string
1098     size_t guardBytes = 8; // to make sure we're not overrunning the output buffer
1099     size_t outputBufferSize = outputWritten + sizeofNull + guardBytes;
1100     wxCharBuffer outputBuffer(outputBufferSize);
1101     memset( outputBuffer.data(), UNINITIALIZED, outputBufferSize );
1102
1103     outputWritten = converter.WC2MB
1104         (
1105         outputBuffer.data(),
1106         (const wchar_t*)inputCopy.data(),
1107         outputBufferSize
1108         );
1109
1110     // make sure the correct number of characters were output
1111     CPPUNIT_ASSERT_EQUAL( multiBytes, outputWritten );
1112
1113     // make sure the characters generated are correct
1114     CPPUNIT_ASSERT( 0 == memcmp( outputBuffer, multiBuffer, multiBytes ) );
1115
1116     size_t i;
1117
1118     // the output buffer should be null terminated
1119     for ( i = multiBytes; i < multiBytes + sizeofNull; i++ )
1120     {
1121         CPPUNIT_ASSERT( ((unsigned char*)outputBuffer.data())[i] == 0 );
1122     }
1123
1124     // make sure the rest of the output buffer is untouched
1125     for ( i = multiBytes + sizeofNull; i < outputBufferSize; i++ )
1126     {
1127         CPPUNIT_ASSERT( ((unsigned char*)outputBuffer.data())[i] == UNINITIALIZED );
1128     }
1129
1130 #if wxUSE_UNICODE && wxUSE_STREAMS
1131     TestStreamEncoder( wideBuffer, wideChars, multiBuffer, multiBytes, converter );
1132 #endif
1133 }
1134
1135 #if wxUSE_UNICODE && wxUSE_STREAMS
1136 // use wxTextInputStream to exercise wxMBConv interface
1137 // (this reveals some bugs in certain wxMBConv subclasses)
1138 void MBConvTestCase::TestStreamDecoder(
1139                                        const wchar_t* wideBuffer,  // the same character sequence as multiBuffer, encoded as wchar_t
1140                                        size_t         wideChars,   // the number of wide characters at wideBuffer
1141                                        const char*    multiBuffer, // a multibyte encoded character sequence that can be decoded by "converter"
1142                                        size_t         multiBytes,  // the byte length of the multibyte character sequence that can be decoded by "converter"
1143                                        wxMBConv&      converter    // the wxMBConv object that can decode multiBuffer into a wide character sequence
1144                                        )
1145 {
1146     // this isn't meant to test wxMemoryInputStream or wxTextInputStream
1147     // it's meant to test the way wxTextInputStream uses wxMBConv
1148     // (which has exposed some problems with wxMBConv)
1149     wxMemoryInputStream memoryInputStream( multiBuffer, multiBytes );
1150     wxTextInputStream textInputStream( memoryInputStream, wxT(""), converter );
1151     for ( size_t i = 0; i < wideChars; i++ )
1152     {
1153         wxChar wc = textInputStream.GetChar();
1154         CPPUNIT_ASSERT_EQUAL_MESSAGE(
1155             std::string(wxString::Format("At position %lu", (unsigned long)i)),
1156             wideBuffer[i],
1157             wc
1158         );
1159     }
1160     CPPUNIT_ASSERT( 0 == textInputStream.GetChar() );
1161     CPPUNIT_ASSERT( memoryInputStream.Eof() );
1162 }
1163 #endif
1164
1165 #if wxUSE_UNICODE && wxUSE_STREAMS
1166 // use wxTextInputStream to exercise wxMBConv interface
1167 // (this reveals some bugs in certain wxMBConv subclasses)
1168 void MBConvTestCase::TestStreamEncoder(
1169     const wchar_t* wideBuffer,  // the same character sequence as multiBuffer, encoded as wchar_t
1170     size_t         wideChars,   // the number of wide characters at wideBuffer
1171     const char*    multiBuffer, // a multibyte encoded character sequence that can be decoded by "converter"
1172     size_t         multiBytes,  // the byte length of the multibyte character sequence that can be decoded by "converter"
1173     wxMBConv&      converter    // the wxMBConv object that can decode multiBuffer into a wide character sequence
1174     )
1175 {
1176     // this isn't meant to test wxMemoryOutputStream or wxTextOutputStream
1177     // it's meant to test the way wxTextOutputStream uses wxMBConv
1178     // (which has exposed some problems with wxMBConv)
1179     wxMemoryOutputStream memoryOutputStream;
1180     // wxEOL_UNIX will pass \n \r unchanged
1181     wxTextOutputStream textOutputStream( memoryOutputStream, wxEOL_UNIX, converter );
1182     for ( size_t i = 0; i < wideChars; i++ )
1183     {
1184         textOutputStream.PutChar( wideBuffer[i] );
1185     }
1186
1187     textOutputStream.Flush();
1188
1189     CPPUNIT_ASSERT_EQUAL( multiBytes, size_t(memoryOutputStream.TellO()) );
1190     wxCharBuffer copy( memoryOutputStream.TellO() );
1191     memoryOutputStream.CopyTo( copy.data(), memoryOutputStream.TellO());
1192     CPPUNIT_ASSERT_EQUAL( 0, memcmp( copy.data(), multiBuffer, multiBytes ) );
1193 }
1194 #endif
1195
1196
1197 // ----------------------------------------------------------------------------
1198 // UTF-8 tests
1199 // ----------------------------------------------------------------------------
1200
1201 #ifdef HAVE_WCHAR_H
1202
1203 // Check that 'charSequence' translates to 'wideSequence' and back.
1204 // Invalid sequences can be tested by giving NULL for 'wideSequence'. Even
1205 // invalid sequences should roundtrip when an option is given and this is
1206 // checked.
1207 //
1208 void MBConvTestCase::UTF8(const char *charSequence,
1209                           const wchar_t *wideSequence)
1210 {
1211     UTF8(charSequence, wideSequence, wxMBConvUTF8::MAP_INVALID_UTF8_NOT);
1212     UTF8(charSequence, wideSequence, wxMBConvUTF8::MAP_INVALID_UTF8_TO_PUA);
1213     UTF8(charSequence, wideSequence, wxMBConvUTF8::MAP_INVALID_UTF8_TO_OCTAL);
1214 }
1215
1216 // Use this alternative when 'charSequence' contains a PUA character. Such
1217 // sequences should still roundtrip ok, and this is checked.
1218 //
1219 void MBConvTestCase::UTF8PUA(const char *charSequence,
1220                              const wchar_t *wideSequence)
1221 {
1222     UTF8(charSequence, wideSequence, wxMBConvUTF8::MAP_INVALID_UTF8_NOT);
1223     UTF8(charSequence, NULL, wxMBConvUTF8::MAP_INVALID_UTF8_TO_PUA);
1224     UTF8(charSequence, wideSequence, wxMBConvUTF8::MAP_INVALID_UTF8_TO_OCTAL);
1225 }
1226
1227 // Use this alternative when 'charSequence' contains an octal escape sequence.
1228 // Such sequences should still roundtrip ok, and this is checked.
1229 //
1230 void MBConvTestCase::UTF8Octal(const char *charSequence,
1231                                const wchar_t *wideSequence)
1232 {
1233     UTF8(charSequence, wideSequence, wxMBConvUTF8::MAP_INVALID_UTF8_NOT);
1234     UTF8(charSequence, wideSequence, wxMBConvUTF8::MAP_INVALID_UTF8_TO_PUA);
1235     UTF8(charSequence, NULL, wxMBConvUTF8::MAP_INVALID_UTF8_TO_OCTAL);
1236 }
1237
1238 // in case wcscpy is missing
1239 //
1240 static wchar_t *wx_wcscpy(wchar_t *dest, const wchar_t *src)
1241 {
1242     wchar_t *d = dest;
1243     while ((*d++ = *src++) != 0)
1244         ;
1245     return dest;
1246 }
1247
1248 // in case wcscat is missing
1249 //
1250 static wchar_t *wx_wcscat(wchar_t *dest, const wchar_t *src)
1251 {
1252     wchar_t *d = dest;
1253     while (*d)
1254         d++;
1255     while ((*d++ = *src++) != 0)
1256         ;
1257     return dest;
1258 }
1259
1260 // in case wcscmp is missing
1261 //
1262 static int wx_wcscmp(const wchar_t *s1, const wchar_t *s2)
1263 {
1264     while (*s1 == *s2 && *s1 != 0)
1265     {
1266         s1++;
1267         s2++;
1268     }
1269     return *s1 - *s2;
1270 }
1271
1272 // in case wcslen is missing
1273 //
1274 static size_t wx_wcslen(const wchar_t *s)
1275 {
1276     const wchar_t *t = s;
1277     while (*t != 0)
1278         t++;
1279     return t - s;
1280 }
1281
1282 // include the option in the error messages so it's possible to see which
1283 // test failed
1284 #define UTF8ASSERT(expr) CPPUNIT_ASSERT_MESSAGE(#expr + errmsg,  expr)
1285
1286 // The test implementation
1287 //
1288 void MBConvTestCase::UTF8(const char *charSequence,
1289                           const wchar_t *wideSequence,
1290                           int option)
1291 {
1292     const size_t BUFSIZE = 128;
1293     wxASSERT(strlen(charSequence) * 3 + 10 < BUFSIZE);
1294     char bytes[BUFSIZE];
1295
1296     // include the option in the error messages so it's possible to see
1297     // which test failed
1298     sprintf(bytes, " (with option == %d)", option);
1299     std::string errmsg(bytes);
1300
1301     // put the charSequence at the start, middle and end of a string
1302     strcpy(bytes, charSequence);
1303     strcat(bytes, "ABC");
1304     strcat(bytes, charSequence);
1305     strcat(bytes, "XYZ");
1306     strcat(bytes, charSequence);
1307
1308     // translate it into wide characters
1309     wxMBConvUTF8 utf8(option);
1310     wchar_t widechars[BUFSIZE];
1311     size_t lenResult = utf8.MB2WC(NULL, bytes, 0);
1312     size_t result = utf8.MB2WC(widechars, bytes, BUFSIZE);
1313     UTF8ASSERT(result == lenResult);
1314
1315     // check we got the expected result
1316     if (wideSequence) {
1317         UTF8ASSERT(result != (size_t)-1);
1318         wxASSERT(result < BUFSIZE);
1319
1320         wchar_t expected[BUFSIZE];
1321         wx_wcscpy(expected, wideSequence);
1322         wx_wcscat(expected, L"ABC");
1323         wx_wcscat(expected, wideSequence);
1324         wx_wcscat(expected, L"XYZ");
1325         wx_wcscat(expected, wideSequence);
1326
1327         UTF8ASSERT(wx_wcscmp(widechars, expected) == 0);
1328         UTF8ASSERT(wx_wcslen(widechars) == result);
1329     }
1330     else {
1331         // If 'wideSequence' is NULL, then the result is expected to be
1332         // invalid.  Normally that is as far as we can go, but if there is an
1333         // option then the conversion should succeed anyway, and it should be
1334         // possible to translate back to the original
1335         if (!option) {
1336             UTF8ASSERT(result == (size_t)-1);
1337             return;
1338         }
1339         else {
1340             UTF8ASSERT(result != (size_t)-1);
1341         }
1342     }
1343
1344     // translate it back and check we get the original
1345     char bytesAgain[BUFSIZE];
1346     size_t lenResultAgain = utf8.WC2MB(NULL, widechars, 0);
1347     size_t resultAgain = utf8.WC2MB(bytesAgain, widechars, BUFSIZE);
1348     UTF8ASSERT(resultAgain == lenResultAgain);
1349     UTF8ASSERT(resultAgain != (size_t)-1);
1350     wxASSERT(resultAgain < BUFSIZE);
1351
1352     UTF8ASSERT(strcmp(bytes, bytesAgain) == 0);
1353     UTF8ASSERT(strlen(bytesAgain) == resultAgain);
1354 }
1355
1356 #endif // HAVE_WCHAR_H