Allow calling wxStrchr() with a narrow string and wide character.

Calls to wxStrchr(char-string, wide-char) would previously fail if wide
character couldn't be converted to a single character in the current locale
encoding. Change it to simply return NULL in this case as it's a safe and
useful generalization: a narrow string will never contain a wide character not
representable in the current locale.

Add wxUniChar::GetAsChar() to help with implementing this.

Closes #11487.

git-svn-id: https://svn.wxwidgets.org/svn/wx/wxWidgets/trunk@62738 c3d73ce0-8a6f-49c7-b76d-6d57e0e08775

include/wx/unichar.h

@@ -68,6 +68,27 @@ public:
     // Returns true if the character is an ASCII character:
     bool IsAscii() const { return m_value < 0x80; }
 
+    // Returns true if the character is representable as a single byte in the
+    // current locale encoding and return this byte in output argument c (which
+    // must be non-NULL)
+    bool GetAsChar(char *c) const
+    {
+#if wxUSE_UNICODE
+        if ( !IsAscii() )
+        {
+#if !wxUSE_UTF8_LOCALE_ONLY
+            if ( GetAsHi8bit(m_value, c) )
+                return true;
+#endif // !wxUSE_UTF8_LOCALE_ONLY
+
+            return false;
+        }
+#endif // wxUSE_UNICODE
+
+        *c = wx_truncate_cast(char, m_value);
+        return true;
+    }
+
     // Conversions to char and wchar_t types: all of those are needed to be
     // able to pass wxUniChars to verious standard narrow and wide character
     // functions
@@ -165,7 +186,8 @@ private:
 
     // helpers of the functions above called to deal with non-ASCII chars
     static value_type FromHi8bit(char c);
-    static char ToHi8bit(value_type c);
+    static char ToHi8bit(value_type v);
+    static bool GetAsHi8bit(value_type v, char *c);
 
 private:
     value_type m_value;
@@ -209,6 +231,7 @@ public:
 #endif // wxUSE_UNICODE_UTF8
 
     bool IsAscii() const { return UniChar().IsAscii(); }
+    bool GetAsChar(char *c) const { return UniChar().GetAsChar(c); }
 
     // Assignment operators:
 #if wxUSE_UNICODE_UTF8

include/wx/wxcrt.h

@@ -611,20 +611,20 @@ inline const char *wxStrrchr(const char *s, char c)
     { return wxCRT_StrrchrA(s, c); }
 inline const wchar_t *wxStrrchr(const wchar_t *s, wchar_t c)
     { return wxCRT_StrrchrW(s, c); }
-inline const char *wxStrchr(const char *s, const wxUniChar& c)
+inline const char *wxStrchr(const char *s, const wxUniChar& uc)
-    { return wxCRT_StrchrA(s, (char)c); }
+    { char c; return uc.GetAsChar(&c) ? wxCRT_StrchrA(s, c) : NULL; }
 inline const wchar_t *wxStrchr(const wchar_t *s, const wxUniChar& c)
     { return wxCRT_StrchrW(s, (wchar_t)c); }
-inline const char *wxStrrchr(const char *s, const wxUniChar& c)
+inline const char *wxStrrchr(const char *s, const wxUniChar& uc)
-    { return wxCRT_StrrchrA(s, (char)c); }
+    { char c; return uc.GetAsChar(&c) ? wxCRT_StrrchrA(s, c) : NULL; }
 inline const wchar_t *wxStrrchr(const wchar_t *s, const wxUniChar& c)
     { return wxCRT_StrrchrW(s, (wchar_t)c); }
-inline const char *wxStrchr(const char *s, const wxUniCharRef& c)
+inline const char *wxStrchr(const char *s, const wxUniCharRef& uc)
-    { return wxCRT_StrchrA(s, (char)c); }
+    { char c; return uc.GetAsChar(&c) ? wxCRT_StrchrA(s, c) : NULL; }
 inline const wchar_t *wxStrchr(const wchar_t *s, const wxUniCharRef& c)
     { return wxCRT_StrchrW(s, (wchar_t)c); }
-inline const char *wxStrrchr(const char *s, const wxUniCharRef& c)
+inline const char *wxStrrchr(const char *s, const wxUniCharRef& uc)
-    { return wxCRT_StrrchrA(s, (char)c); }
+    { char c; return uc.GetAsChar(&c) ? wxCRT_StrrchrA(s, c) : NULL; }
 inline const wchar_t *wxStrrchr(const wchar_t *s, const wxUniCharRef& c)
     { return wxCRT_StrrchrW(s, (wchar_t)c); }
 template<typename T>
@@ -655,14 +655,14 @@ inline const char* wxStrchr(const wxString& s, int c)
     { return wxCRT_StrchrA((const char*)s.c_str(), c); }
 inline const char* wxStrrchr(const wxString& s, int c)
     { return wxCRT_StrrchrA((const char*)s.c_str(), c); }
-inline const char* wxStrchr(const wxString& s, const wxUniChar& c)
+inline const char* wxStrchr(const wxString& s, const wxUniChar& uc)
-    { return wxCRT_StrchrA((const char*)s.c_str(), (char)c); }
+    { char c; return uc.GetAsChar(&c) ? wxCRT_StrchrA(s, c) : NULL; }
-inline const char* wxStrrchr(const wxString& s, const wxUniChar& c)
+inline const char* wxStrrchr(const wxString& s, const wxUniChar& uc)
-    { return wxCRT_StrrchrA((const char*)s.c_str(), (char)c); }
+    { char c; return uc.GetAsChar(&c) ? wxCRT_StrrchrA(s, c) : NULL; }
-inline const char* wxStrchr(const wxString& s, const wxUniCharRef& c)
+inline const char* wxStrchr(const wxString& s, const wxUniCharRef& uc)
-    { return wxCRT_StrchrA((const char*)s.c_str(), (char)c); }
+    { char c; return uc.GetAsChar(&c) ? wxCRT_StrchrA(s, c) : NULL; }
-inline const char* wxStrrchr(const wxString& s, const wxUniCharRef& c)
+inline const char* wxStrrchr(const wxString& s, const wxUniCharRef& uc)
-    { return wxCRT_StrrchrA((const char*)s.c_str(), (char)c); }
+    { char c; return uc.GetAsChar(&c) ? wxCRT_StrrchrA(s, c) : NULL; }
 inline const wchar_t* wxStrchr(const wxString& s, wchar_t c)
     { return wxCRT_StrchrW((const wchar_t*)s.c_str(), c); }
 inline const wchar_t* wxStrrchr(const wxString& s, wchar_t c)
@@ -675,14 +675,14 @@ inline const char* wxStrchr(const wxCStrData& s, int c)
     { return wxCRT_StrchrA(s.AsChar(), c); }
 inline const char* wxStrrchr(const wxCStrData& s, int c)
     { return wxCRT_StrrchrA(s.AsChar(), c); }
-inline const char* wxStrchr(const wxCStrData& s, const wxUniChar& c)
+inline const char* wxStrchr(const wxCStrData& s, const wxUniChar& uc)
-    { return wxCRT_StrchrA(s.AsChar(), (char)c); }
+    { char c; return uc.GetAsChar(&c) ? wxCRT_StrchrA(s, c) : NULL; }
-inline const char* wxStrrchr(const wxCStrData& s, const wxUniChar& c)
+inline const char* wxStrrchr(const wxCStrData& s, const wxUniChar& uc)
-    { return wxCRT_StrrchrA(s.AsChar(), (char)c); }
+    { char c; return uc.GetAsChar(&c) ? wxCRT_StrrchrA(s, c) : NULL; }
-inline const char* wxStrchr(const wxCStrData& s, const wxUniCharRef& c)
+inline const char* wxStrchr(const wxCStrData& s, const wxUniCharRef& uc)
-    { return wxCRT_StrchrA(s.AsChar(), (char)c); }
+    { char c; return uc.GetAsChar(&c) ? wxCRT_StrchrA(s, c) : NULL; }
-inline const char* wxStrrchr(const wxCStrData& s, const wxUniCharRef& c)
+inline const char* wxStrrchr(const wxCStrData& s, const wxUniCharRef& uc)
-    { return wxCRT_StrrchrA(s.AsChar(), (char)c); }
+    { char c; return uc.GetAsChar(&c) ? wxCRT_StrrchrA(s, c) : NULL; }
 inline const wchar_t* wxStrchr(const wxCStrData& s, wchar_t c)
     { return wxCRT_StrchrW(s.AsWChar(), c); }
 inline const wchar_t* wxStrrchr(const wxCStrData& s, wchar_t c)

interface/wx/unichar.h

@@ -58,6 +58,30 @@ public:
     */
     bool IsAscii() const;
 
+    /**
+        Returns true if the character is representable as a single byte in the
+        current locale encoding.
+
+        This function only returns true if the character can be converted in
+        exactly one byte, e.g. it only returns true for 7 bit ASCII characters
+        when the encoding used is UTF-8.
+
+        It is mostly useful to test if the character can be passed to functions
+        taking a char and is used by wxWidgets itself for this purpose.
+
+        @param c
+            An output pointer to the value of this Unicode character as a @c
+            char. Must be non-@NULL.
+        @return
+            @true if the object is an 8 bit char and @a c was filled with its
+            value as char or @false otherwise (@a c won't be modified then).
+
+        @see IsAscii()
+
+        @since 2.9.1
+     */
+    bool GetAsChar(char *c) const;
+
     //@{
     /**
         Conversions to char and wchar_t types: all of those are needed to be

src/common/unichar.cpp

@@ -58,27 +58,31 @@ wxUniChar::value_type wxUniChar::FromHi8bit(char c)
 }
 
 /* static */
-char wxUniChar::ToHi8bit(wxUniChar::value_type c)
+char wxUniChar::ToHi8bit(wxUniChar::value_type v)
 {
-#if wxUSE_UTF8_LOCALE_ONLY
+    char c;
-    wxFAIL_MSG( "character cannot be converted to single UTF-8 byte" );
+    if ( !GetAsHi8bit(v, &c) )
-    wxUnusedVar(c);
+    {
+        wxFAIL_MSG( "character cannot be converted to single byte" );
+        c = '?'; // FIXME-UTF8: what to use as failure character?
+    }
 
-    return '?'; // FIXME-UTF8: what to use as failure character?
+    return c;
-#else
+}
+
+/* static */
+bool wxUniChar::GetAsHi8bit(value_type v, char *c)
+{
     wchar_t wbuf[2];
-    wbuf[0] = c;
+    wbuf[0] = v;
     wbuf[1] = L'\0';
     char cbuf[2];
     if ( wxConvLibc.FromWChar(cbuf, 2, wbuf, 2) != 2 )
-    {
+        return false;
-        wxFAIL_MSG( "character cannot be converted to single byte" );
-        return '?'; // FIXME-UTF8: what to use as failure character?
-    }
-    return cbuf[0];
-#endif
-}
 
+    *c = cbuf[0];
+    return true;
+}
 
 // ---------------------------------------------------------------------------
 // wxUniCharRef

tests/strings/crt.cpp

@@ -44,6 +44,7 @@ public:
 private:
     CPPUNIT_TEST_SUITE( CrtTestCase );
         CPPUNIT_TEST( SetGetEnv );
+        CPPUNIT_TEST( Strchr );
         CPPUNIT_TEST( Strcmp );
         CPPUNIT_TEST( Strspn );
         CPPUNIT_TEST( Strcspn );
@@ -52,6 +53,7 @@ private:
     CPPUNIT_TEST_SUITE_END();
 
     void SetGetEnv();
+    void Strchr();
     void Strcmp();
     void Strspn();
     void Strcspn();
@@ -89,6 +91,19 @@ void CrtTestCase::SetGetEnv()
 #undef TESTVAR_NAME
 }
 
+void CrtTestCase::Strchr()
+{
+    // test that searching for a wide character in a narrow string simply
+    // doesn't find it but doesn't fail with an assert (#11487)
+    const wxUniChar smiley = *wxString::FromUTF8("\xe2\x98\xba").begin();
+
+    CPPUNIT_ASSERT( !wxStrchr("hello", smiley) );
+
+    // but searching for an explicitly wide character does find it
+    CPPUNIT_ASSERT( wxStrchr(wxString::FromUTF8(":-) == \xe2\x98\xba"),
+                    static_cast<wchar_t>(smiley)) );
+}
+
 void CrtTestCase::Strcmp()
 {
     // this code tests if all possible ways of calling wxStrcmp() compile: