tests/strings/unicode.cpp

   1 ///////////////////////////////////////////////////////////////////////////////
   2 // Name:        tests/strings/unicode.cpp
   3 // Purpose:     Unicode unit test
   4 // Author:      Vadim Zeitlin, Wlodzimierz ABX Skiba
   5 // Created:     2004-04-28
   6 // RCS-ID:      $Id$
   7 // Copyright:   (c) 2004 Vadim Zeitlin, Wlodzimierz Skiba
   8 ///////////////////////////////////////////////////////////////////////////////
   9
  10 // ----------------------------------------------------------------------------
  11 // headers
  12 // ----------------------------------------------------------------------------
  13
  14 #include "testprec.h"
  15
  16 #ifdef __BORLANDC__
  17     #pragma hdrstop
  18 #endif
  19
  20 #ifndef WX_PRECOMP
  21     #include "wx/wx.h"
  22 #endif // WX_PRECOMP
  23
  24 // helper class holding the matching MB and WC strings
  25 //
  26 // either str or wcs (but not both) may be NULL, this means that the conversion
  27 // to it should fail
  28 struct StringConversionData
  29 {
  30     const char *str;
  31     const wchar_t *wcs;
  32
  33     enum
  34     {
  35         TEST_BOTH  = 0, // test both str -> wcs and wcs -> str
  36         ONLY_MB2WC = 1  // only test str -> wcs conversion
  37     };
  38
  39     int flags;
  40
  41     // test that the conversion between str and wcs (subject to flags) succeeds
  42     //
  43     // the first argument is the index in the test array and is used solely for
  44     // diagnostics
  45     void Test(size_t n, wxMBConv& conv) const
  46     {
  47         if ( str )
  48         {
  49             wxWCharBuffer wbuf = conv.cMB2WC(str);
  50
  51             if ( wcs )
  52             {
  53                 CPPUNIT_ASSERT_MESSAGE
  54                 (
  55                     Message(n, "MB2WC failed"),
  56                     wbuf.data()
  57                 );
  58
  59                 CPPUNIT_ASSERT_MESSAGE
  60                 (
  61                     Message(n, "MB2WC", wbuf, wcs),
  62                     wxStrcmp(wbuf, wcs) == 0
  63                 );
  64             }
  65             else // conversion is supposed to fail
  66             {
  67                 CPPUNIT_ASSERT_MESSAGE
  68                 (
  69                     Message(n, "MB2WC succeeded"),
  70                     !wbuf.data()
  71                 );
  72             }
  73         }
  74
  75         if ( wcs && !(flags & ONLY_MB2WC) )
  76         {
  77             wxCharBuffer buf = conv.cWC2MB(wcs);
  78
  79             if ( str )
  80             {
  81                 CPPUNIT_ASSERT_MESSAGE
  82                 (
  83                     Message(n, "WC2MB failed"),
  84                     buf.data()
  85                 );
  86
  87                 CPPUNIT_ASSERT_MESSAGE
  88                 (
  89                     Message(n, "WC2MB", buf, str),
  90                     strcmp(buf, str) == 0
  91                 );
  92             }
  93             else
  94             {
  95                 CPPUNIT_ASSERT_MESSAGE
  96                 (
  97                     Message(n, "WC2MB succeeded"),
  98                     !buf.data()
  99                 );
 100             }
 101         }
 102     }
 103
 104 private:
 105     static std::string
 106     Message(size_t n, const wxString& msg)
 107     {
 108         return std::string(wxString::Format("#%lu: %s", (unsigned long)n, msg));
 109     }
 110
 111     template <typename T>
 112     static std::string
 113     Message(size_t n,
 114             const char *func,
 115             const wxCharTypeBuffer<T>& actual,
 116             const T *expected)
 117     {
 118         return Message(n,
 119                        wxString::Format("%s returned \"%s\", expected \"%s\"",
 120                                         func, actual.data(), expected));
 121     }
 122 };
 123
 124 // ----------------------------------------------------------------------------
 125 // test class
 126 // ----------------------------------------------------------------------------
 127
 128 class UnicodeTestCase : public CppUnit::TestCase
 129 {
 130 public:
 131     UnicodeTestCase();
 132
 133 private:
 134     CPPUNIT_TEST_SUITE( UnicodeTestCase );
 135         CPPUNIT_TEST( ToFromAscii );
 136         CPPUNIT_TEST( ConstructorsWithConversion );
 137         CPPUNIT_TEST( ConversionEmpty );
 138         CPPUNIT_TEST( ConversionWithNULs );
 139         CPPUNIT_TEST( ConversionUTF7 );
 140         CPPUNIT_TEST( ConversionUTF8 );
 141         CPPUNIT_TEST( ConversionUTF16 );
 142         CPPUNIT_TEST( ConversionUTF32 );
 143         CPPUNIT_TEST( IsConvOk );
 144 #if wxUSE_UNICODE
 145         CPPUNIT_TEST( Iteration );
 146 #endif
 147     CPPUNIT_TEST_SUITE_END();
 148
 149     void ToFromAscii();
 150     void ConstructorsWithConversion();
 151     void ConversionEmpty();
 152     void ConversionWithNULs();
 153     void ConversionUTF7();
 154     void ConversionUTF8();
 155     void ConversionUTF16();
 156     void ConversionUTF32();
 157     void IsConvOk();
 158 #if wxUSE_UNICODE
 159     void Iteration();
 160 #endif
 161
 162     DECLARE_NO_COPY_CLASS(UnicodeTestCase)
 163 };
 164
 165 // register in the unnamed registry so that these tests are run by default
 166 CPPUNIT_TEST_SUITE_REGISTRATION( UnicodeTestCase );
 167
 168 // also include in it's own registry so that these tests can be run alone
 169 CPPUNIT_TEST_SUITE_NAMED_REGISTRATION( UnicodeTestCase, "UnicodeTestCase" );
 170
 171 UnicodeTestCase::UnicodeTestCase()
 172 {
 173 }
 174
 175 void UnicodeTestCase::ToFromAscii()
 176 {
 177
 178 #define TEST_TO_FROM_ASCII(txt)                              \
 179     {                                                        \
 180         static const char *msg = txt;                        \
 181         wxString s = wxString::FromAscii(msg);               \
 182         CPPUNIT_ASSERT( strcmp( s.ToAscii() , msg ) == 0 );  \
 183     }
 184
 185     TEST_TO_FROM_ASCII( "Hello, world!" );
 186     TEST_TO_FROM_ASCII( "additional \" special \t test \\ component \n :-)" );
 187 }
 188
 189 void UnicodeTestCase::ConstructorsWithConversion()
 190 {
 191     // the string "Déjà" in UTF-8 and wchar_t:
 192     const unsigned char utf8Buf[] = {0x44,0xC3,0xA9,0x6A,0xC3,0xA0,0};
 193     const unsigned char utf8subBuf[] = {0x44,0xC3,0xA9,0x6A,0}; // just "Déj"
 194     const char *utf8 = (char *)utf8Buf;
 195     const char *utf8sub = (char *)utf8subBuf;
 196
 197     wxString s1(utf8, wxConvUTF8);
 198
 199 #if wxUSE_UNICODE
 200     const wchar_t wchar[] = {0x44,0xE9,0x6A,0xE0,0};
 201     WX_ASSERT_STR_EQUAL( wchar, s1 );
 202
 203     wxString s2(wchar);
 204     WX_ASSERT_STR_EQUAL( wchar, s2 );
 205     WX_ASSERT_STR_EQUAL( utf8, s2 );
 206 #else
 207     WX_ASSERT_STR_EQUAL( utf8, s1 );
 208 #endif
 209
 210     wxString sub(utf8sub, wxConvUTF8); // "Dej" substring
 211     wxString s3(utf8, wxConvUTF8, 4);
 212     CPPUNIT_ASSERT_EQUAL( sub, s3 );
 213
 214 #if wxUSE_UNICODE
 215     wxString s4(wchar, wxConvUTF8, 3);
 216     CPPUNIT_ASSERT_EQUAL( sub, s4 );
 217
 218     // conversion should stop with failure at pos 35
 219     wxString s("\t[pl]open.format.Sformatuj dyskietkê=gfloppy %f", wxConvUTF8);
 220     CPPUNIT_ASSERT( s.empty() );
 221 #endif // wxUSE_UNICODE
 222
 223
 224     // test using Unicode strings together with char* strings (this must work
 225     // in ANSI mode as well, of course):
 226     wxString s5("ascii");
 227     WX_ASSERT_STR_EQUAL( "ascii", s5 );
 228
 229     s5 += " value";
 230
 231     CPPUNIT_ASSERT( strcmp(s5.mb_str(), "ascii value") == 0 );
 232     WX_ASSERT_STR_EQUAL( "ascii value", s5 );
 233     CPPUNIT_ASSERT( s5 != "SomethingElse" );
 234 }
 235
 236 void UnicodeTestCase::ConversionEmpty()
 237 {
 238     size_t len;
 239
 240 #if wxUSE_UNICODE
 241     wxCharBuffer buf = wxConvLibc.cWC2MB(L"", 0, &len);
 242 #else // !wxUSE_UNICODE
 243     wxWCharBuffer wbuf = wxConvLibc.cMB2WC("", 0, &len);
 244 #endif // wxUSE_UNICODE/!wxUSE_UNICODE
 245
 246     CPPUNIT_ASSERT(len == 0);
 247 }
 248
 249 void UnicodeTestCase::ConversionWithNULs()
 250 {
 251 #if wxUSE_UNICODE
 252     static const size_t lenNulString = 10;
 253
 254     wxString szTheString(L"The\0String", wxConvLibc, lenNulString);
 255     wxCharBuffer theBuffer = szTheString.mb_str();
 256
 257     CPPUNIT_ASSERT( memcmp(theBuffer.data(), "The\0String",
 258                     lenNulString + 1) == 0 );
 259
 260     wxString szTheString2("The\0String", wxConvLocal, lenNulString);
 261     CPPUNIT_ASSERT_EQUAL( lenNulString, szTheString2.length() );
 262     CPPUNIT_ASSERT( wxTmemcmp(szTheString2.c_str(), L"The\0String",
 263                     lenNulString + 1) == 0 );
 264 #else // !wxUSE_UNICODE
 265     wxString szTheString("TheString");
 266     szTheString.insert(3, 1, '\0');
 267     wxWCharBuffer theBuffer = szTheString.wc_str(wxConvLibc);
 268
 269     CPPUNIT_ASSERT( memcmp(theBuffer.data(), L"The\0String", 11 * sizeof(wchar_t)) == 0 );
 270
 271     wxString szLocalTheString("TheString");
 272     szLocalTheString.insert(3, 1, '\0');
 273     wxWCharBuffer theLocalBuffer = szLocalTheString.wc_str(wxConvLocal);
 274
 275     CPPUNIT_ASSERT( memcmp(theLocalBuffer.data(), L"The\0String", 11 * sizeof(wchar_t)) == 0 );
 276 #endif // wxUSE_UNICODE/!wxUSE_UNICODE
 277 }
 278
 279 void UnicodeTestCase::ConversionUTF7()
 280 {
 281     static const StringConversionData utf7data[] =
 282     {
 283         // normal fragments
 284         { "+AKM-", L"\xa3" },
 285         { "+AOk-t+AOk-", L"\xe9t\xe9" },
 286
 287         // this one is an alternative valid encoding of the same string
 288         { "+AOk-t+AOk", L"\xe9t\xe9", StringConversionData::ONLY_MB2WC },
 289
 290         // some special cases
 291         { "+-", L"+" },
 292         { "+--", L"+-" },
 293
 294         // the following are invalid UTF-7 sequences
 295         { "\xa3", NULL },
 296         { "+", NULL },
 297         { "+~", NULL },
 298         { "a+", NULL },
 299     };
 300
 301     for ( size_t n = 0; n < WXSIZEOF(utf7data); n++ )
 302     {
 303         const StringConversionData& d = utf7data[n];
 304
 305         // converting to/from UTF-7 using iconv() currently doesn't work
 306         // because of several problems:
 307         //  - GetMBNulLen() doesn't return correct result (iconv converts L'\0'
 308         //    to an incomplete and anyhow nonsensical "+AA" string)
 309         //  - iconv refuses to convert "+-" (although it converts "+-\n" just
 310         //    fine, go figure)
 311         //
 312         // I have no idea how to fix this so just disable the test for now
 313 #if 0
 314         d.Test(n, wxCSConv("utf-7"));
 315 #endif
 316         d.Test(n, wxConvUTF7);
 317     }
 318 }
 319
 320 void UnicodeTestCase::ConversionUTF8()
 321 {
 322     static const StringConversionData utf8data[] =
 323     {
 324 #ifdef wxHAVE_U_ESCAPE
 325         { "\xc2\xa3", L"\u00a3" },
 326 #endif
 327         { "\xc2", NULL },
 328     };
 329
 330     wxCSConv conv(_T("utf-8"));
 331     for ( size_t n = 0; n < WXSIZEOF(utf8data); n++ )
 332     {
 333         const StringConversionData& d = utf8data[n];
 334         d.Test(n, conv);
 335         d.Test(n, wxConvUTF8);
 336     }
 337 }
 338
 339 void UnicodeTestCase::ConversionUTF16()
 340 {
 341     static const StringConversionData utf16data[] =
 342     {
 343 #ifdef wxHAVE_U_ESCAPE
 344         { "\x04\x1f\x04\x40\x04\x38\x04\x32\x04\x35\x04\x42\0\0",
 345           L"\u041f\u0440\u0438\u0432\u0435\u0442" },
 346         { "\x01\0\0b\x01\0\0a\x01\0\0r\0\0", L"\u0100b\u0100a\u0100r" },
 347 #endif
 348         { "\0f\0o\0o\0\0", L"foo" },
 349     };
 350
 351     wxCSConv conv(wxFONTENCODING_UTF16BE);
 352     for ( size_t n = 0; n < WXSIZEOF(utf16data); n++ )
 353     {
 354         const StringConversionData& d = utf16data[n];
 355         d.Test(n, conv);
 356     }
 357
 358     // special case: this string has consecutive NULs inside it which don't
 359     // terminate the string, this exposed a bug in our conversion code which
 360     // got confused in this case
 361     size_t len;
 362     wxWCharBuffer wbuf(conv.cMB2WC("\x01\0\0B\0C" /* A macron BC */, 6, &len));
 363     CPPUNIT_ASSERT_EQUAL( (size_t)3, len );
 364 }
 365
 366 void UnicodeTestCase::ConversionUTF32()
 367 {
 368     static const StringConversionData utf32data[] =
 369     {
 370 #ifdef wxHAVE_U_ESCAPE
 371         {
 372             "\0\0\x04\x1f\0\0\x04\x40\0\0\x04\x38\0\0\x04\x32\0\0\x04\x35\0\0\x04\x42\0\0\0\0",
 373           L"\u041f\u0440\u0438\u0432\u0435\u0442" },
 374 #endif
 375         { "\0\0\0f\0\0\0o\0\0\0o\0\0\0\0", L"foo" },
 376     };
 377
 378     wxCSConv conv(wxFONTENCODING_UTF32BE);
 379     for ( size_t n = 0; n < WXSIZEOF(utf32data); n++ )
 380     {
 381         const StringConversionData& d = utf32data[n];
 382         d.Test(n, conv);
 383     }
 384
 385     size_t len;
 386     wxWCharBuffer wbuf(conv.cMB2WC("\0\0\x01\0\0\0\0B\0\0\0C" /* A macron BC */,
 387                                    12, &len));
 388     CPPUNIT_ASSERT_EQUAL( (size_t)3, len );
 389 }
 390
 391 void UnicodeTestCase::IsConvOk()
 392 {
 393     CPPUNIT_ASSERT( wxCSConv(wxFONTENCODING_SYSTEM).IsOk() );
 394     CPPUNIT_ASSERT( wxCSConv(_T("UTF-8")).IsOk() );
 395     CPPUNIT_ASSERT( !wxCSConv(_T("NoSuchConversion")).IsOk() );
 396
 397 #ifdef __WINDOWS__
 398     CPPUNIT_ASSERT( wxCSConv(_T("WINDOWS-437")).IsOk() );
 399 #endif
 400 }
 401
 402 #if wxUSE_UNICODE
 403 void UnicodeTestCase::Iteration()
 404 {
 405     // "czech" in Czech ("cestina"):
 406     static const char *textUTF8 = "\304\215e\305\241tina";
 407     static const wchar_t textUTF16[] = {0x10D, 0x65, 0x161, 0x74, 0x69, 0x6E, 0x61, 0};
 408
 409     wxString text(wxString::FromUTF8(textUTF8));
 410     CPPUNIT_ASSERT( wxStrcmp(text.wc_str(), textUTF16) == 0 );
 411
 412     // verify the string was decoded correctly:
 413     {
 414         size_t idx = 0;
 415         for ( wxString::const_iterator i = text.begin(); i != text.end(); ++i, ++idx )
 416         {
 417             CPPUNIT_ASSERT( *i == textUTF16[idx] );
 418         }
 419     }
 420
 421     // overwrite the string with something that is shorter in UTF-8:
 422     {
 423         for ( wxString::iterator i = text.begin(); i != text.end(); ++i )
 424             *i = 'x';
 425     }
 426
 427     // restore the original text now:
 428     {
 429         wxString::iterator end1 = text.end();
 430         wxString::const_iterator end2 = text.end();
 431
 432         size_t idx = 0;
 433         for ( wxString::iterator i = text.begin(); i != text.end(); ++i, ++idx )
 434         {
 435             *i = textUTF16[idx];
 436
 437             CPPUNIT_ASSERT( end1 == text.end() );
 438             CPPUNIT_ASSERT( end2 == text.end() );
 439         }
 440
 441         CPPUNIT_ASSERT( end1 == text.end() );
 442         CPPUNIT_ASSERT( end2 == text.end() );
 443     }
 444
 445     // and verify it again:
 446     {
 447         size_t idx = 0;
 448         for ( wxString::const_iterator i = text.begin(); i != text.end(); ++i, ++idx )
 449         {
 450             CPPUNIT_ASSERT( *i == textUTF16[idx] );
 451         }
 452     }
 453 }
 454 #endif // wxUSE_UNICODE