X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/c12b7f79a7ae3d36ec38213c1b872b051c420e5a..b9efe021b554fa3967d1442cf758435c5cd5ae8f:/src/common/strconv.cpp diff --git a/src/common/strconv.cpp b/src/common/strconv.cpp index e40f0f39f7..aea1e1a18a 100644 --- a/src/common/strconv.cpp +++ b/src/common/strconv.cpp @@ -55,9 +55,6 @@ #include #include #include -#ifdef HAVE_LANGINFO_H - #include -#endif #if defined(__WIN32__) && !defined(__WXMICROWIN__) #define wxHAVE_WIN32_MB2WC @@ -155,7 +152,7 @@ static size_t decode_utf16(const wxUint16* input, wxUint32& output) output = *input; return 1; } - else if ((input[1]<0xdc00) || (input[1]>=0xdfff)) + else if ((input[1]<0xdc00) || (input[1]>0xdfff)) { output = *input; return (size_t)-1; @@ -273,9 +270,9 @@ const wxWCharBuffer wxMBConv::cMB2WC(const char *szString, size_t nStringLen, si } //Increment to next (sub)string - //Note that we have to use strlen here instead of nLen - //here because XX2XX gives us the size of the output buffer, - //not neccessarly the length of the string + //Note that we have to use strlen instead of nLen here + //because XX2XX gives us the size of the output buffer, + //which is not necessarily the length of the string szPos += strlen(szPos) + 1; } @@ -335,9 +332,9 @@ const wxCharBuffer wxMBConv::cWC2MB(const wchar_t *szString, size_t nStringLen, } //Increment to next (sub)string - //Note that we have to use wxWcslen here instead of nLen - //here because XX2XX gives us the size of the output buffer, - //not neccessarly the length of the string + //Note that we have to use wxWcslen instead of nLen here + //because XX2XX gives us the size of the output buffer, + //which is not necessarily the length of the string szPos += wxWcslen(szPos) + 1; } @@ -360,57 +357,19 @@ size_t wxMBConvLibc::WC2MB(char *buf, const wchar_t *psz, size_t n) const return wxWC2MB(buf, psz, n); } -#ifdef __WXGTK20__ +#ifdef __UNIX__ // ---------------------------------------------------------------------------- -// wxConvBrokenFileNames is made for GTK2 in Unicode mode when -// files are accidentally written in an encoding which is not -// the system encoding. Typically, the system encoding will be -// UTF8 but there might be files stored in ISO8859-1 on disk. +// wxConvBrokenFileNames // ---------------------------------------------------------------------------- -class wxConvBrokenFileNames : public wxMBConv +wxConvBrokenFileNames::wxConvBrokenFileNames(const wxChar *charset) { -public: - wxConvBrokenFileNames(); - virtual ~wxConvBrokenFileNames() { delete m_conv; } - - virtual size_t MB2WC(wchar_t *outputBuf, const char *psz, size_t outputSize) const; - virtual size_t WC2MB(char *outputBuf, const wchar_t *psz, size_t outputSize) const; - -private: - // the conversion object we forward to - wxMBConv *m_conv; -}; - -wxConvBrokenFileNames::wxConvBrokenFileNames() -{ - // decide which conversion to use for the file names - - // (1) this variable exists for the sole purpose of specifying the encoding - // of the filenames for GTK+ programs, so use it if it is set - const wxChar *encName = wxGetenv(_T("G_FILENAME_ENCODING")); - if ( encName ) - { - m_conv = new wxCSConv(encName); - } - else // no G_FILENAME_ENCODING - { - // (2) if a non default locale is set, assume that the user wants his - // filenames in this locale too - switch ( wxLocale::GetSystemEncoding() ) - { - default: - m_conv = new wxMBConvLibc; - break; - - // (3) finally use UTF-8 by default - case wxFONTENCODING_SYSTEM: - case wxFONTENCODING_UTF8: - m_conv = new wxMBConvUTF8(wxMBConvUTF8::MAP_INVALID_UTF8_TO_OCTAL); - break; - } - } + if ( !charset || wxStricmp(charset, _T("UTF-8")) == 0 + || wxStricmp(charset, _T("UTF8")) == 0 ) + m_conv = new wxMBConvUTF8(wxMBConvUTF8::MAP_INVALID_UTF8_TO_OCTAL); + else + m_conv = new wxCSConv(charset); } size_t @@ -429,7 +388,7 @@ wxConvBrokenFileNames::WC2MB(char *outputBuf, return m_conv->WC2MB( outputBuf, psz, outputSize ); } -#endif // __WXGTK20__ +#endif // ---------------------------------------------------------------------------- // UTF-7 @@ -585,7 +544,7 @@ size_t wxMBConvUTF7::WC2MB(char *buf, const wchar_t *psz, size_t n) const } #ifndef WC_UTF16 else if (((wxUint32)cc) > 0xffff) - { + { // no surrogate pair generation (yet?) return (size_t)-1; } @@ -665,6 +624,15 @@ size_t wxMBConvUTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const if (buf) *buf++ = cc; len++; + + // escape the escape character for octal escapes + if ((m_options & MAP_INVALID_UTF8_TO_OCTAL) + && cc == '\\' && (!buf || len < n)) + { + if (buf) + *buf++ = cc; + len++; + } } else { @@ -755,9 +723,9 @@ size_t wxMBConvUTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const { unsigned char n = *opsz; *buf++ = L'\\'; - *buf++ = L'0' + n / 0100; - *buf++ = L'0' + (n % 0100) / 010; - *buf++ = L'0' + n % 010; + *buf++ = (wchar_t)( L'0' + n / 0100 ); + *buf++ = (wchar_t)( L'0' + (n % 0100) / 010 ); + *buf++ = (wchar_t)( L'0' + n % 010 ); } opsz++; len += 4; @@ -802,15 +770,23 @@ size_t wxMBConvUTF8::WC2MB(char *buf, const wchar_t *psz, size_t n) const *buf++ = (char)(cc - wxUnicodePUA); len++; } + else if ( (m_options & MAP_INVALID_UTF8_TO_OCTAL) + && cc == L'\\' && psz[0] == L'\\' ) + { + if (buf) + *buf++ = (char)cc; + psz++; + len++; + } else if ( (m_options & MAP_INVALID_UTF8_TO_OCTAL) && cc == L'\\' && isoctal(psz[0]) && isoctal(psz[1]) && isoctal(psz[2]) ) { if (buf) { - *buf++ = (char) (psz[0] - L'0')*0100 + - (psz[1] - L'0')*010 + - (psz[2] - L'0'); + *buf++ = (char) ((psz[0] - L'0')*0100 + + (psz[1] - L'0')*010 + + (psz[2] - L'0')); } psz += 3; @@ -2140,7 +2116,7 @@ public: UniChar* szUniBuffer = (UniChar*) szUnConv; #if SIZEOF_WCHAR_T == 4 - wxMBConvUTF16BE converter ; + wxMBConvUTF16 converter ; nBufSize = converter.WC2MB( NULL , szUnConv , 0 ); szUniBuffer = new UniChar[ (nBufSize / sizeof(UniChar)) + 1] ; converter.WC2MB( (char*) szUniBuffer , szUnConv, nBufSize + sizeof(UniChar)) ; @@ -2276,7 +2252,7 @@ public: // we have to terminate here, because n might be larger for the trailing zero, and if UniChar // is not properly terminated we get random characters at the end ubuf[byteOutLen / sizeof( UniChar ) ] = 0 ; - wxMBConvUTF16BE converter ; + wxMBConvUTF16 converter ; res = converter.MB2WC( (buf ? buf : tbuf) , (const char*)ubuf , n ) ; free( ubuf ) ; #else @@ -2309,7 +2285,7 @@ public: ByteCount byteBufferLen = n ; UniChar* ubuf = NULL ; #if SIZEOF_WCHAR_T == 4 - wxMBConvUTF16BE converter ; + wxMBConvUTF16 converter ; size_t unicharlen = converter.WC2MB( NULL , psz , 0 ) ; byteInLen = unicharlen ; ubuf = (UniChar*) malloc( byteInLen + 2 ) ; @@ -2764,10 +2740,6 @@ static wxCSConv wxConvISO8859_1Obj(wxFONTENCODING_ISO8859_1); static wxMBConvUTF7 wxConvUTF7Obj; static wxMBConvUTF8 wxConvUTF8Obj; -#ifdef __WXGTK20__ - static wxConvBrokenFileNames wxConvBrokenFileNamesObj; -#endif - WXDLLIMPEXP_DATA_BASE(wxMBConv&) wxConvLibc = wxConvLibcObj; WXDLLIMPEXP_DATA_BASE(wxCSConv&) wxConvLocal = wxConvLocalObj; WXDLLIMPEXP_DATA_BASE(wxCSConv&) wxConvISO8859_1 = wxConvISO8859_1Obj; @@ -2777,8 +2749,6 @@ WXDLLIMPEXP_DATA_BASE(wxMBConv *) wxConvCurrent = &wxConvLibcObj; WXDLLIMPEXP_DATA_BASE(wxMBConv *) wxConvFileName = & #ifdef __WXOSX__ wxConvUTF8Obj; -#elif __WXGTK20__ - wxConvBrokenFileNamesObj; #else wxConvLibcObj; #endif