Amebis/wxWidgets
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

interpret text invalid in the current encoding (typical for UTF-8 environments) as ISO8859-1, this is wrong but better than eating it completely

Browse Source 
git-svn-id: https://svn.wxwidgets.org/svn/wx/wxWidgets/trunk@44567 c3d73ce0-8a6f-49c7-b76d-6d57e0e08775
This commit is contained in:
Vadim Zeitlin
2007-02-26 09:53:15 +00:00
parent eb41aed90a
commit 404ac4c67d
1 changed files with 17 additions and 39 deletions
Download Patch File Download Diff File Expand all files Collapse all files

56
src/gtk/utilsgtk.cpp
View File

@@ -185,50 +185,28 @@ wxWindow* wxFindWindowAtPoint(const wxPoint& pt)
wxCharBuffer wxConvertToGTK(const wxString& s, wxFontEncoding enc)
{
wxCharBuffer buf;
if ( enc == wxFONTENCODING_UTF8 )
wxWCharBuffer wbuf;
if ( enc == wxFONTENCODING_SYSTEM || enc == wxFONTENCODING_DEFAULT )
{
// no need for conversion at all, but do check that we have a valid
// UTF-8 string because passing invalid UTF-8 to GTK+ is going to
// result in a GTK+ error message and, especially, loss of data which
// was supposed to be shown in the GUI
if ( wxConvUTF8.ToWChar(NULL, 0, s, s.length()) == wxCONV_FAILED )
{
// warn the programmer that something is probably wrong in his code
//
// NB: don't include the string in output because chances are that
// this invalid UTF-8 string could result in more errors itself
// if the application shows logs in the GUI and so we get into
// an infinite loop
wxLogDebug(_T("Invalid UTF-8 string in wxConvertToGTK()"));
// but still try to show at least something on the screen
wxMBConvUTF8 utf8permissive(wxMBConvUTF8::MAP_INVALID_UTF8_TO_OCTAL);
wxWCharBuffer wbuf(utf8permissive.cMB2WC(s));
buf = wxConvUTF8.cWC2MB(wbuf);
}
else // valid UTF-8 string, no need to convert
{
buf = wxCharBuffer(s);
}
wbuf = wxConvUI->cMB2WC(s);
}
else // !UTF-8
else // another encoding, use generic conversion class
{
wxWCharBuffer wbuf;
if ( enc == wxFONTENCODING_SYSTEM || enc == wxFONTENCODING_DEFAULT )
{
wbuf = wxConvUI->cMB2WC(s);
}
else // another encoding, use generic conversion class
{
wbuf = wxCSConv(enc).cMB2WC(s);
}
if ( wbuf )
buf = wxConvUTF8.cWC2MB(wbuf);
wbuf = wxCSConv(enc).cMB2WC(s);
}
return buf;
if ( !wbuf && !s.empty() )
{
// conversion failed, but we still want to show something to the user
// even if it's going to be wrong it is better than nothing
//
// we choose ISO8859-1 here arbitrarily, it's just the most common
// encoding probably and, also importantly here, conversion from it
// never fails as it's done internally by wxCSConv
wbuf = wxCSConv(wxFONTENCODING_ISO8859_1).cMB2WC(s);
}
return wxConvUTF8.cWC2MB(wbuf);
}
#endif // !wxUSE_UNICODE