X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/bdd8f35a58b2793fafd5020d2f45ff9e39865f64..2db1b1c0becfeb9d667670a7aab0a0873de6d5a2:/tests/strings/strings.cpp

diff --git a/tests/strings/strings.cpp b/tests/strings/strings.cpp
index afb6207ae1..aafd0dba57 100644
--- a/tests/strings/strings.cpp
+++ b/tests/strings/strings.cpp
@@ -11,7 +11,7 @@
 // headers
 // ----------------------------------------------------------------------------
 
-#include "wx/wxprec.h"
+#include "testprec.h"
 
 #ifdef __BORLANDC__
     #pragma hdrstop
@@ -23,8 +23,6 @@
 
 #include "wx/tokenzr.h"
 
-#include "wx/cppunit.h"
-
 // ----------------------------------------------------------------------------
 // test class
 // ----------------------------------------------------------------------------
@@ -43,10 +41,9 @@ private:
 #if wxUSE_WCHAR_T
         CPPUNIT_TEST( ConstructorsWithConversion );
         CPPUNIT_TEST( Conversion );
-#endif
-#if wxUSE_UNICODE
         CPPUNIT_TEST( ConversionUTF7 );
-#endif
+        CPPUNIT_TEST( ConversionUTF8 );
+#endif // wxUSE_WCHAR_T
         CPPUNIT_TEST( Extraction );
         CPPUNIT_TEST( Find );
         CPPUNIT_TEST( Tokenizer );
@@ -65,10 +62,9 @@ private:
 #if wxUSE_WCHAR_T
     void ConstructorsWithConversion();
     void Conversion();
-#endif
-#if wxUSE_UNICODE
     void ConversionUTF7();
-#endif
+    void ConversionUTF8();
+#endif // wxUSE_WCHAR_T
     void Extraction();
     void Find();
     void SingleTokenizerTest( wxChar *str, wxChar *delims, size_t count , wxStringTokenizerMode mode );
@@ -80,6 +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)
 };
 
@@ -188,10 +192,7 @@ void StringTestCase::ConstructorsWithConversion()
 
 #if wxUSE_UNICODE
     CPPUNIT_ASSERT ( wxString("\t[pl]open.format.Sformatuj dyskietkê=gfloppy %f", 
-                               wxConvUTF8) == wxT("") ); //Pos 35 (funky e) is invalid UTF8
-#else
-    CPPUNIT_ASSERT ( wxString(L"\t[pl]open.format.Sformatuj dyskietkê=gfloppy %f", 
-                               wxConvUTF8) == wxT("") ); //Pos 35 (funky e) is invalid UTF8
+                               wxConvUTF8) == wxT("") ); //should stop at pos 35 
 #endif
 }
 
@@ -204,7 +205,8 @@ void StringTestCase::Conversion()
         CPPUNIT_ASSERT( memcmp(theBuffer.data(), "The\0String", 11) == 0 );
 
         wxString szTheString2("The\0String", wxConvLocal, 10);
-        CPPUNIT_ASSERT( wxMemcmp(szTheString2.c_str(), L"The\0String", 11) == 0 );
+        CPPUNIT_ASSERT( szTheString2.length() == 11 );
+        CPPUNIT_ASSERT( wxTmemcmp(szTheString2.c_str(), L"The\0String", 11) == 0 );
 #else
         wxString szTheString(wxT("TheString"));
         szTheString.insert(3, 1, '\0');
@@ -219,19 +221,92 @@ void StringTestCase::Conversion()
         CPPUNIT_ASSERT( memcmp(theLocalBuffer.data(), L"The\0String", 11 * sizeof(wchar_t)) == 0 );
 #endif
 }
-#endif // wxUSE_WCHAR_T
 
+void
+StringTestCase::DoTestConversion(const char *s,
+                                 const wchar_t *ws,
+                                 wxCSConv& conv)
+{
 #if wxUSE_UNICODE
+    if ( ws )
+    {
+        wxCharBuffer buf(wxString(ws).mb_str(conv));
+
+        CPPUNIT_ASSERT( strcmp(buf, s) == 0 );
+    }
+#else // wxUSE_UNICODE
+    if ( s )
+    {
+        wxWCharBuffer wbuf(wxString(s).wc_str(conv));
+
+        if ( ws )
+            CPPUNIT_ASSERT( wcscmp(wbuf, ws) == 0 );
+        else
+            CPPUNIT_ASSERT( !*wbuf );
+    }
+#endif // wxUSE_UNICODE/!wxUSE_UNICODE
+}
+
+struct StringConversionData
+{
+    const char *str;
+    const wchar_t *wcs;
+};
+
 void StringTestCase::ConversionUTF7()
 {
-    const wxChar wdata[] = { 0x00A3, 0x00A3, 0x00A3, 0x00A3, 0 }; // pound signs
-    const char *utf7 = "+AKM-+AKM-+AKM-+AKM-";
-    wxString str(wdata);
+    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" },
+    };
 
     wxCSConv conv(_T("utf-7"));
-    CPPUNIT_ASSERT( strcmp(str.mb_str(conv), utf7) == 0 );
+    for ( size_t n = 0; n < WXSIZEOF(utf7data); n++ )
+    {
+        const StringConversionData& d = utf7data[n];
+        DoTestConversion(d.str, d.wcs, conv);
+    }
 }
-#endif // wxUSE_UNICODE
+
+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 },
+    };
+
+    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
+
 
 void StringTestCase::Extraction()
 {
@@ -333,8 +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);
 
@@ -383,6 +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
 }