X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/8899b155a1e4fa5f4b90e1f3bebe28088ea46bc9..56caea07f8ab1ae3013617695e0613ad0bec061a:/tests/strings/strings.cpp diff --git a/tests/strings/strings.cpp b/tests/strings/strings.cpp index 8b1c6235b6..aafd0dba57 100644 --- a/tests/strings/strings.cpp +++ b/tests/strings/strings.cpp @@ -42,7 +42,8 @@ private: CPPUNIT_TEST( ConstructorsWithConversion ); CPPUNIT_TEST( Conversion ); CPPUNIT_TEST( ConversionUTF7 ); -#endif + CPPUNIT_TEST( ConversionUTF8 ); +#endif // wxUSE_WCHAR_T CPPUNIT_TEST( Extraction ); CPPUNIT_TEST( Find ); CPPUNIT_TEST( Tokenizer ); @@ -62,7 +63,8 @@ private: void ConstructorsWithConversion(); void Conversion(); void ConversionUTF7(); -#endif + void ConversionUTF8(); +#endif // wxUSE_WCHAR_T void Extraction(); void Find(); void SingleTokenizerTest( wxChar *str, wxChar *delims, size_t count , wxStringTokenizerMode mode ); @@ -74,6 +76,14 @@ private: void Compare(); void CompareNoCase(); +#if wxUSE_WCHAR_T + // test if converting s using the given encoding gives ws and vice versa + // + // if either of the first 2 arguments is NULL, the conversion is supposed + // to fail + void DoTestConversion(const char *s, const wchar_t *w, wxCSConv& conv); +#endif // wxUSE_WCHAR_T + DECLARE_NO_COPY_CLASS(StringTestCase) }; @@ -195,6 +205,7 @@ void StringTestCase::Conversion() CPPUNIT_ASSERT( memcmp(theBuffer.data(), "The\0String", 11) == 0 ); wxString szTheString2("The\0String", wxConvLocal, 10); + CPPUNIT_ASSERT( szTheString2.length() == 11 ); CPPUNIT_ASSERT( wxTmemcmp(szTheString2.c_str(), L"The\0String", 11) == 0 ); #else wxString szTheString(wxT("TheString")); @@ -211,40 +222,87 @@ void StringTestCase::Conversion() #endif } -void StringTestCase::ConversionUTF7() +void +StringTestCase::DoTestConversion(const char *s, + const wchar_t *ws, + wxCSConv& conv) { - const wchar_t data[] = { 0x00A3, 0x00A3, 0x00A3, 0x00A3, 0 }; // pound signs - - //utf7 and utf7alt are equivelent - const char *utf7 = "+AKM-+AKM-+AKM-+AKM-"; - const char *utf7alt = "+AKMAowCjAKM-"; - #if wxUSE_UNICODE - wxString str(data); - - wxCSConv conv(_T("utf-7")); - - wxCharBuffer theBuffer = str.mb_str(conv); - - CPPUNIT_ASSERT( strcmp(theBuffer, utf7) == 0 || strcmp(theBuffer, utf7alt) == 0); -#else //ANSI - wxString str(utf7); + if ( ws ) + { + wxCharBuffer buf(wxString(ws).mb_str(conv)); - wxCSConv conv(_T("utf-7")); + CPPUNIT_ASSERT( strcmp(buf, s) == 0 ); + } +#else // wxUSE_UNICODE + if ( s ) + { + wxWCharBuffer wbuf(wxString(s).wc_str(conv)); - wxWCharBuffer theWBuffer = str.wc_str(conv); + if ( ws ) + CPPUNIT_ASSERT( wcscmp(wbuf, ws) == 0 ); + else + CPPUNIT_ASSERT( !*wbuf ); + } +#endif // wxUSE_UNICODE/!wxUSE_UNICODE +} - CPPUNIT_ASSERT( wxWcslen(theWBuffer) == wxWcslen(data) ); - CPPUNIT_ASSERT( memcmp(theWBuffer, data, wxWcslen(data) * sizeof(wchar_t)) == 0 ); +struct StringConversionData +{ + const char *str; + const wchar_t *wcs; +}; - wxString stralt(utf7alt); +void StringTestCase::ConversionUTF7() +{ + static const StringConversionData utf7data[] = + { + { "+-", L"+" }, + { "+--", L"+-" }, + //\u isn't recognized on MSVC 6 +#if !defined(_MSC_VER) +#if !defined(__GNUC__) || (__GNUC__ >= 3) + { "+AKM-", L"\u00a3" }, +#endif +#endif + // Windows accepts invalid UTF-7 strings and so does our UTF-7 + // conversion code -- this is wrong IMO but the way it is for now + // + // notice that converting "+" still behaves as expected because the + // result is just an empty string, i.e. the same as if there were an + // error, but converting "a+" results in "a" while it really should + // fail + { "+", NULL }, + { "a+", L"a" }, + }; - wxWCharBuffer theWBufferAlt = stralt.wc_str(conv); + wxCSConv conv(_T("utf-7")); + for ( size_t n = 0; n < WXSIZEOF(utf7data); n++ ) + { + const StringConversionData& d = utf7data[n]; + DoTestConversion(d.str, d.wcs, conv); + } +} - CPPUNIT_ASSERT( wxWcslen(theWBufferAlt) == wxWcslen(data) ); - CPPUNIT_ASSERT( memcmp(theWBufferAlt, data, wxWcslen(data) * sizeof(wchar_t)) == 0 ); +void StringTestCase::ConversionUTF8() +{ + static const StringConversionData utf8data[] = + { + //\u isn't recognized on MSVC 6 +#if !defined(_MSC_VER) +#if !defined(__GNUC__) || (__GNUC__ >= 3) + { "\xc2\xa3", L"\u00a3" }, +#endif +#endif + { "\xc2", NULL }, + }; -#endif // wxUSE_UNICODE + wxCSConv conv(_T("utf-8")); + for ( size_t n = 0; n < WXSIZEOF(utf8data); n++ ) + { + const StringConversionData& d = utf8data[n]; + DoTestConversion(d.str, d.wcs, conv); + } } #endif // wxUSE_WCHAR_T @@ -350,8 +408,8 @@ void StringTestCase::Tokenizer() // call this with the string to tokenize, delimeters to use and the expected // positions (i.e. results of GetPosition()) after each GetNextToken() call, // terminate positions with 0 -static void -DoTokenizerGetPosition(const wxChar *s, const wxChar *delims, int pos, ...) +static void DoTokenizerGetPosition(const wxChar *s, + const wxChar *delims, int pos, ...) { wxStringTokenizer tkz(s, delims); @@ -400,6 +458,32 @@ void StringTestCase::Replace() TEST_REPLACE( _T("foobar"), 3, 0, _T("-"), _T("foo-bar") ); TEST_REPLACE( _T("barfoo"), 0, 6, _T("foobar"), _T("foobar") ); + + #define TEST_NULLCHARREPLACE( o , olen, pos , len , replacement , r, rlen ) \ + { \ + wxString s(o,olen); \ + s.replace( pos , len , replacement ); \ + CPPUNIT_ASSERT( s == wxString(r,rlen) ); \ + } + + TEST_NULLCHARREPLACE( _T("null\0char"), 9, 5, 1, _T("d"), + _T("null\0dhar"), 9 ); + + #define TEST_WXREPLACE( o , olen, olds, news, all, r, rlen ) \ + { \ + wxString s(o,olen); \ + s.Replace( olds, news, all ); \ + CPPUNIT_ASSERT( s == wxString(r,rlen) ); \ + } + + TEST_WXREPLACE( _T("null\0char"), 9, _T("c"), _T("de"), true, + _T("null\0dehar"), 10 ); + + TEST_WXREPLACE( _T("null\0dehar"), 10, _T("de"), _T("c"), true, + _T("null\0char"), 9 ); + + #undef TEST_WXREPLACE + #undef TEST_NULLCHARREPLACE #undef TEST_REPLACE }