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 #endif // WX_PRECOMP
  22
  23 // ----------------------------------------------------------------------------
  24 // local functions
  25 // ----------------------------------------------------------------------------
  26
  27 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
  28
  29 // in case wcscmp is missing
  30 static int wx_wcscmp(const wchar_t *s1, const wchar_t *s2)
  31 {
  32     while (*s1 == *s2 && *s1 != 0)
  33     {
  34         s1++;
  35         s2++;
  36     }
  37     return *s1 - *s2;
  38 }
  39
  40 #endif // wxUSE_WCHAR_T && !wxUSE_UNICODE
  41
  42 // ----------------------------------------------------------------------------
  43 // test class
  44 // ----------------------------------------------------------------------------
  45
  46 class UnicodeTestCase : public CppUnit::TestCase
  47 {
  48 public:
  49     UnicodeTestCase();
  50
  51 private:
  52     CPPUNIT_TEST_SUITE( UnicodeTestCase );
  53         CPPUNIT_TEST( ToFromAscii );
  54 #if wxUSE_WCHAR_T
  55         CPPUNIT_TEST( ConstructorsWithConversion );
  56         CPPUNIT_TEST( ConversionEmpty );
  57         CPPUNIT_TEST( ConversionWithNULs );
  58         CPPUNIT_TEST( ConversionUTF7 );
  59         CPPUNIT_TEST( ConversionUTF8 );
  60         CPPUNIT_TEST( ConversionUTF16 );
  61         CPPUNIT_TEST( ConversionUTF32 );
  62         CPPUNIT_TEST( IsConvOk );
  63 #endif // wxUSE_WCHAR_T
  64     CPPUNIT_TEST_SUITE_END();
  65
  66     void ToFromAscii();
  67 #if wxUSE_WCHAR_T
  68     void ConstructorsWithConversion();
  69     void ConversionEmpty();
  70     void ConversionWithNULs();
  71     void ConversionUTF7();
  72     void ConversionUTF8();
  73     void ConversionUTF16();
  74     void ConversionUTF32();
  75     void IsConvOk();
  76
  77     // test if converting s using the given encoding gives ws and vice versa
  78     //
  79     // if either of the first 2 arguments is NULL, the conversion is supposed
  80     // to fail
  81     void DoTestConversion(const char *s, const wchar_t *w, wxMBConv& conv);
  82 #endif // wxUSE_WCHAR_T
  83
  84
  85     DECLARE_NO_COPY_CLASS(UnicodeTestCase)
  86 };
  87
  88 // register in the unnamed registry so that these tests are run by default
  89 CPPUNIT_TEST_SUITE_REGISTRATION( UnicodeTestCase );
  90
  91 // also include in it's own registry so that these tests can be run alone
  92 CPPUNIT_TEST_SUITE_NAMED_REGISTRATION( UnicodeTestCase, "Unicode" );
  93
  94 UnicodeTestCase::UnicodeTestCase()
  95 {
  96 }
  97
  98 void UnicodeTestCase::ToFromAscii()
  99 {
 100
 101 #define TEST_TO_FROM_ASCII(txt)                              \
 102     {                                                        \
 103         static const char *msg = txt;                        \
 104         wxString s = wxString::FromAscii(msg);               \
 105         CPPUNIT_ASSERT( strcmp( s.ToAscii() , msg ) == 0 );  \
 106     }
 107
 108     TEST_TO_FROM_ASCII( "Hello, world!" );
 109     TEST_TO_FROM_ASCII( "additional \" special \t test \\ component \n :-)" );
 110 }
 111
 112 #if wxUSE_WCHAR_T
 113 void UnicodeTestCase::ConstructorsWithConversion()
 114 {
 115     // the string "Déjà" in UTF-8 and wchar_t:
 116     const unsigned char utf8Buf[] = {0x44,0xC3,0xA9,0x6A,0xC3,0xA0,0};
 117     const wchar_t wchar[] = {0x44,0xE9,0x6A,0xE0,0};
 118     const unsigned char utf8subBuf[] = {0x44,0xC3,0xA9,0x6A,0}; // just "Déj"
 119     const char *utf8 = (char *)utf8Buf;
 120     const char *utf8sub = (char *)utf8subBuf;
 121
 122     wxString s1(utf8, wxConvUTF8);
 123     wxString s2(wchar, wxConvUTF8);
 124
 125 #if wxUSE_UNICODE
 126     CPPUNIT_ASSERT( s1 == wchar );
 127     CPPUNIT_ASSERT( s2 == wchar );
 128 #else
 129     CPPUNIT_ASSERT( s1 == utf8 );
 130     CPPUNIT_ASSERT( s2 == utf8 );
 131 #endif
 132
 133     wxString sub(utf8sub, wxConvUTF8); // "Dej" substring
 134     wxString s3(utf8, wxConvUTF8, 4);
 135     wxString s4(wchar, wxConvUTF8, 3);
 136
 137     CPPUNIT_ASSERT( s3 == sub );
 138     CPPUNIT_ASSERT( s4 == sub );
 139
 140 #if wxUSE_UNICODE
 141     CPPUNIT_ASSERT ( wxString("\t[pl]open.format.Sformatuj dyskietkê=gfloppy %f",
 142                                wxConvUTF8) == wxT("") ); //should stop at pos 35
 143 #endif
 144 }
 145
 146 void UnicodeTestCase::ConversionEmpty()
 147 {
 148     size_t len;
 149
 150 #if wxUSE_UNICODE
 151     wxCharBuffer buf = wxConvLibc.cWC2MB(L"", 0, &len);
 152 #else // !wxUSE_UNICODE
 153     wxWCharBuffer wbuf = wxConvLibc.cMB2WC("", 0, &len);
 154 #endif // wxUSE_UNICODE/!wxUSE_UNICODE
 155
 156     CPPUNIT_ASSERT(len == 0);
 157 }
 158
 159 void UnicodeTestCase::ConversionWithNULs()
 160 {
 161 #if wxUSE_UNICODE
 162     static const size_t lenNulString = 10;
 163
 164     wxString szTheString(L"The\0String", wxConvLibc, lenNulString);
 165     wxCharBuffer theBuffer = szTheString.mb_str();
 166
 167     CPPUNIT_ASSERT( memcmp(theBuffer.data(), "The\0String",
 168                     lenNulString + 1) == 0 );
 169
 170     wxString szTheString2("The\0String", wxConvLocal, lenNulString);
 171     CPPUNIT_ASSERT_EQUAL( lenNulString, szTheString2.length() );
 172     CPPUNIT_ASSERT( wxTmemcmp(szTheString2.c_str(), L"The\0String",
 173                     lenNulString + 1) == 0 );
 174 #else // !wxUSE_UNICODE
 175     wxString szTheString(wxT("TheString"));
 176     szTheString.insert(3, 1, '\0');
 177     wxWCharBuffer theBuffer = szTheString.wc_str(wxConvLibc);
 178
 179     CPPUNIT_ASSERT( memcmp(theBuffer.data(), L"The\0String", 11 * sizeof(wchar_t)) == 0 );
 180
 181     wxString szLocalTheString(wxT("TheString"));
 182     szLocalTheString.insert(3, 1, '\0');
 183     wxWCharBuffer theLocalBuffer = szLocalTheString.wc_str(wxConvLocal);
 184
 185     CPPUNIT_ASSERT( memcmp(theLocalBuffer.data(), L"The\0String", 11 * sizeof(wchar_t)) == 0 );
 186 #endif // wxUSE_UNICODE/!wxUSE_UNICODE
 187 }
 188
 189 void
 190 UnicodeTestCase::DoTestConversion(const char *s,
 191                                   const wchar_t *ws,
 192                                   wxMBConv& conv)
 193 {
 194 #if wxUSE_UNICODE
 195     if ( ws )
 196     {
 197         wxCharBuffer buf = conv.cWC2MB(ws, (size_t)-1, NULL);
 198
 199         CPPUNIT_ASSERT( strcmp(buf, s) == 0 );
 200     }
 201 #else // wxUSE_UNICODE
 202     if ( s )
 203     {
 204         wxWCharBuffer wbuf = conv.cMB2WC(s, (size_t)-1, NULL);
 205
 206         if ( ws )
 207         {
 208             CPPUNIT_ASSERT( wbuf.data() );
 209             CPPUNIT_ASSERT( wx_wcscmp(wbuf, ws) == 0 );
 210         }
 211         else // conversion is supposed to fail
 212         {
 213             CPPUNIT_ASSERT_EQUAL( (wchar_t *)NULL, wbuf.data() );
 214         }
 215     }
 216 #endif // wxUSE_UNICODE/!wxUSE_UNICODE
 217 }
 218
 219 struct StringConversionData
 220 {
 221     const char *str;
 222     const wchar_t *wcs;
 223 };
 224
 225 void UnicodeTestCase::ConversionUTF7()
 226 {
 227     static const StringConversionData utf7data[] =
 228     {
 229         { "+-", L"+" },
 230         { "+--", L"+-" },
 231
 232 #ifdef wxHAVE_U_ESCAPE
 233         { "+AKM-", L"\u00a3" },
 234 #endif // wxHAVE_U_ESCAPE
 235
 236         // the following are invalid UTF-7 sequences
 237         { "+", NULL },
 238         { "+~", NULL },
 239         { "a+", NULL },
 240     };
 241
 242     wxCSConv conv(_T("utf-7"));
 243     for ( size_t n = 0; n < WXSIZEOF(utf7data); n++ )
 244     {
 245         const StringConversionData& d = utf7data[n];
 246
 247         // converting to/from UTF-7 using iconv() currently doesn't work
 248         // because of several problems:
 249         //  - GetMBNulLen() doesn't return correct result (iconv converts L'\0'
 250         //    to an incomplete and anyhow nonsensical "+AA" string)
 251         //  - iconv refuses to convert "+-" (although it converts "+-\n" just
 252         //    fine, go figure)
 253         //
 254         // I have no idea how to fix this so just disable the test for now
 255 #if 0
 256         DoTestConversion(d.str, d.wcs, conv);
 257 #endif
 258         DoTestConversion(d.str, d.wcs, wxConvUTF7);
 259     }
 260 }
 261
 262 void UnicodeTestCase::ConversionUTF8()
 263 {
 264     static const StringConversionData utf8data[] =
 265     {
 266 #ifdef wxHAVE_U_ESCAPE
 267         { "\xc2\xa3", L"\u00a3" },
 268 #endif
 269         { "\xc2", NULL },
 270     };
 271
 272     wxCSConv conv(_T("utf-8"));
 273     for ( size_t n = 0; n < WXSIZEOF(utf8data); n++ )
 274     {
 275         const StringConversionData& d = utf8data[n];
 276         DoTestConversion(d.str, d.wcs, conv);
 277         DoTestConversion(d.str, d.wcs, wxConvUTF8);
 278     }
 279 }
 280
 281 void UnicodeTestCase::ConversionUTF16()
 282 {
 283     static const StringConversionData utf16data[] =
 284     {
 285 #ifdef wxHAVE_U_ESCAPE
 286         { "\x04\x1f\x04\x40\x04\x38\x04\x32\x04\x35\x04\x42\0\0",
 287           L"\u041f\u0440\u0438\u0432\u0435\u0442" },
 288         { "\x01\0\0b\x01\0\0a\x01\0\0r\0\0", L"\u0100b\u0100a\u0100r" },
 289 #endif
 290         { "\0f\0o\0o\0\0", L"foo" },
 291     };
 292
 293     wxCSConv conv(wxFONTENCODING_UTF16BE);
 294     for ( size_t n = 0; n < WXSIZEOF(utf16data); n++ )
 295     {
 296         const StringConversionData& d = utf16data[n];
 297         DoTestConversion(d.str, d.wcs, conv);
 298     }
 299
 300     // special case: this string has consecutive NULs inside it which don't
 301     // terminate the string, this exposed a bug in our conversion code which
 302     // got confused in this case
 303     size_t len;
 304     wxWCharBuffer wbuf(conv.cMB2WC("\x01\0\0B\0C" /* A macron BC */, 6, &len));
 305     CPPUNIT_ASSERT_EQUAL( (size_t)3, len );
 306 }
 307
 308 void UnicodeTestCase::ConversionUTF32()
 309 {
 310     static const StringConversionData utf32data[] =
 311     {
 312 #ifdef wxHAVE_U_ESCAPE
 313         {
 314             "\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",
 315           L"\u041f\u0440\u0438\u0432\u0435\u0442" },
 316 #endif
 317         { "\0\0\0f\0\0\0o\0\0\0o\0\0\0\0", L"foo" },
 318     };
 319
 320     wxCSConv conv(wxFONTENCODING_UTF32BE);
 321     for ( size_t n = 0; n < WXSIZEOF(utf32data); n++ )
 322     {
 323         const StringConversionData& d = utf32data[n];
 324         DoTestConversion(d.str, d.wcs, conv);
 325     }
 326
 327     size_t len;
 328     wxWCharBuffer wbuf(conv.cMB2WC("\0\0\x01\0\0\0\0B\0\0\0C" /* A macron BC */,
 329                                    12, &len));
 330     CPPUNIT_ASSERT_EQUAL( (size_t)3, len );
 331 }
 332
 333 void UnicodeTestCase::IsConvOk()
 334 {
 335     CPPUNIT_ASSERT( wxCSConv(wxFONTENCODING_SYSTEM).IsOk() );
 336     CPPUNIT_ASSERT( wxCSConv(_T("UTF-8")).IsOk() );
 337     CPPUNIT_ASSERT( !wxCSConv(_T("NoSuchConversion")).IsOk() );
 338
 339 #ifdef __WINDOWS__
 340     CPPUNIT_ASSERT( wxCSConv(_T("WINDOWS-437")).IsOk() );
 341 #endif
 342 }
 343
 344 #endif // wxUSE_WCHAR_T
 345