X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/467e04791c0d5914dc8ed8e535af607f7fc1dfe1..137c8bde085d6d5b7c459902d2ea1a198ab48765:/src/common/strconv.cpp diff --git a/src/common/strconv.cpp b/src/common/strconv.cpp index 7b28a13b10..ef07724797 100644 --- a/src/common/strconv.cpp +++ b/src/common/strconv.cpp @@ -12,25 +12,15 @@ // Licence: wxWindows licence ///////////////////////////////////////////////////////////////////////////// -// ============================================================================ -// declarations -// ============================================================================ - -// ---------------------------------------------------------------------------- -// headers -// ---------------------------------------------------------------------------- - // For compilers that support precompilation, includes "wx.h". #include "wx/wxprec.h" -#ifdef __BORLANDC__ - #pragma hdrstop -#endif - #ifndef WX_PRECOMP #include "wx/intl.h" #include "wx/log.h" -#endif // WX_PRECOMP + #include "wx/utils.h" + #include "wx/hashmap.h" +#endif #include "wx/strconv.h" @@ -51,7 +41,7 @@ #if defined(__WIN32__) && !defined(__WXMICROWIN__) #define wxHAVE_WIN32_MB2WC -#endif // __WIN32__ but !__WXMICROWIN__ +#endif #ifdef __SALFORDC__ #include @@ -64,7 +54,6 @@ #include "wx/encconv.h" #include "wx/fontmap.h" -#include "wx/utils.h" #ifdef __WXMAC__ #ifndef __DARWIN__ @@ -73,9 +62,11 @@ #include #endif -#include "wx/mac/private.h" // includes mac headers +// includes Mac headers +#include "wx/mac/private.h" #endif + #define TRACE_STRCONV _T("strconv") // WC_UTF16 is defined only if sizeof(wchar_t) == 2, otherwise it's supposed to @@ -84,6 +75,7 @@ #define WC_UTF16 #endif + // ============================================================================ // implementation // ============================================================================ @@ -103,13 +95,14 @@ static bool NotAllNULs(const char *p, size_t n) static size_t encode_utf16(wxUint32 input, wxUint16 *output) { - if (input<=0xffff) + if (input <= 0xffff) { if (output) *output = (wxUint16) input; + return 1; } - else if (input>=0x110000) + else if (input >= 0x110000) { return wxCONV_FAILED; } @@ -117,21 +110,22 @@ static size_t encode_utf16(wxUint32 input, wxUint16 *output) { if (output) { - *output++ = (wxUint16) ((input >> 10)+0xd7c0); - *output = (wxUint16) ((input&0x3ff)+0xdc00); + *output++ = (wxUint16) ((input >> 10) + 0xd7c0); + *output = (wxUint16) ((input & 0x3ff) + 0xdc00); } + return 2; } } static size_t decode_utf16(const wxUint16* input, wxUint32& output) { - if ((*input<0xd800) || (*input>0xdfff)) + if ((*input < 0xd800) || (*input > 0xdfff)) { output = *input; return 1; } - else if ((input[1]<0xdc00) || (input[1]>0xdfff)) + else if ((input[1] < 0xdc00) || (input[1] > 0xdfff)) { output = *input; return wxCONV_FAILED; @@ -144,16 +138,21 @@ static size_t decode_utf16(const wxUint16* input, wxUint32& output) } #ifdef WC_UTF16 + typedef wchar_t wxDecodeSurrogate_t; +#else // !WC_UTF16 + typedef wxUint16 wxDecodeSurrogate_t; +#endif // WC_UTF16/!WC_UTF16 // returns the next UTF-32 character from the wchar_t buffer and advances the // pointer to the character after this one // // if an invalid character is found, *pSrc is set to NULL, the caller must // check for this -static wxUint32 wxDecodeSurrogate(const wchar_t **pSrc) +static wxUint32 wxDecodeSurrogate(const wxDecodeSurrogate_t **pSrc) { wxUint32 out; - const size_t n = decode_utf16(*pSrc, out); + const size_t + n = decode_utf16(wx_reinterpret_cast(const wxUint16 *, *pSrc), out); if ( n == wxCONV_FAILED ) *pSrc = NULL; else @@ -162,8 +161,6 @@ static wxUint32 wxDecodeSurrogate(const wchar_t **pSrc) return out; } -#endif // WC_UTF16 - // ---------------------------------------------------------------------------- // wxMBConv // ---------------------------------------------------------------------------- @@ -183,7 +180,7 @@ wxMBConv::ToWChar(wchar_t *dst, size_t dstLen, size_t dstWritten = 0; // the number of NULs terminating this string - size_t nulLen wxDUMMY_INITIALIZE(0); + size_t nulLen = 0; // not really needed, but just to avoid warnings // if we were not given the input size we just have to assume that the // string is properly terminated as we have no way of knowing how long it @@ -296,7 +293,7 @@ wxMBConv::FromWChar(char *dst, size_t dstLen, { // make a copy in order to properly NUL-terminate the string bufTmp = wxWCharBuffer(srcLen); - memcpy(bufTmp.data(), src, srcLen*sizeof(wchar_t)); + memcpy(bufTmp.data(), src, srcLen * sizeof(wchar_t)); src = bufTmp; } @@ -329,9 +326,9 @@ wxMBConv::FromWChar(char *dst, size_t dstLen, return dstWritten; } -size_t wxMBConv::MB2WC(wchar_t *out, const char *in, size_t outLen) const +size_t wxMBConv::MB2WC(wchar_t *outBuff, const char *inBuff, size_t outLen) const { - size_t rc = ToWChar(out, outLen, in); + size_t rc = ToWChar(outBuff, outLen, inBuff); if ( rc != wxCONV_FAILED ) { // ToWChar() returns the buffer length, i.e. including the trailing @@ -342,9 +339,9 @@ size_t wxMBConv::MB2WC(wchar_t *out, const char *in, size_t outLen) const return rc; } -size_t wxMBConv::WC2MB(char *out, const wchar_t *in, size_t outLen) const +size_t wxMBConv::WC2MB(char *outBuff, const wchar_t *inBuff, size_t outLen) const { - size_t rc = FromWChar(out, outLen, in); + size_t rc = FromWChar(outBuff, outLen, inBuff); if ( rc != wxCONV_FAILED ) { rc -= GetMBNulLen(); @@ -398,13 +395,13 @@ const wxCharBuffer wxMBConv::cWC2MB(const wchar_t *pwz) const } const wxWCharBuffer -wxMBConv::cMB2WC(const char *in, size_t inLen, size_t *outLen) const +wxMBConv::cMB2WC(const char *inBuff, size_t inLen, size_t *outLen) const { - const size_t dstLen = ToWChar(NULL, 0, in, inLen); + const size_t dstLen = ToWChar(NULL, 0, inBuff, inLen); if ( dstLen != wxCONV_FAILED ) { wxWCharBuffer wbuf(dstLen - 1); - if ( ToWChar(wbuf.data(), dstLen, in, inLen) != wxCONV_FAILED ) + if ( ToWChar(wbuf.data(), dstLen, inBuff, inLen) != wxCONV_FAILED ) { if ( outLen ) { @@ -424,24 +421,27 @@ wxMBConv::cMB2WC(const char *in, size_t inLen, size_t *outLen) const } const wxCharBuffer -wxMBConv::cWC2MB(const wchar_t *in, size_t inLen, size_t *outLen) const +wxMBConv::cWC2MB(const wchar_t *inBuff, size_t inLen, size_t *outLen) const { - const size_t dstLen = FromWChar(NULL, 0, in, inLen); + size_t dstLen = FromWChar(NULL, 0, inBuff, inLen); if ( dstLen != wxCONV_FAILED ) { - wxCharBuffer buf(dstLen - 1); - if ( FromWChar(buf.data(), dstLen, in, inLen) != wxCONV_FAILED ) + // special case of empty input: can't allocate 0 size buffer below as + // wxCharBuffer insists on NUL-terminating it + wxCharBuffer buf(dstLen ? dstLen - 1 : 1); + if ( FromWChar(buf.data(), dstLen, inBuff, inLen) != wxCONV_FAILED ) { if ( outLen ) { *outLen = dstLen; const size_t nulLen = GetMBNulLen(); - if ( !NotAllNULs(buf.data() + dstLen - nulLen, nulLen) ) + if ( dstLen >= nulLen && + !NotAllNULs(buf.data() + dstLen - nulLen, nulLen) ) { // in this case the output is NUL-terminated and we're not // supposed to count NUL - (*outLen) -= nulLen; + *outLen -= nulLen; } } @@ -646,6 +646,7 @@ size_t wxMBConvUTF7::WC2MB(char *buf, const wchar_t *psz, size_t n) const // plain ASCII char if (buf) *buf++ = (char)cc; + len++; } #ifndef WC_UTF16 @@ -659,6 +660,7 @@ size_t wxMBConvUTF7::WC2MB(char *buf, const wchar_t *psz, size_t n) const { if (buf) *buf++ = '+'; + len++; if (cc != '+') { @@ -679,24 +681,30 @@ size_t wxMBConvUTF7::WC2MB(char *buf, const wchar_t *psz, size_t n) const len++; } } + cc = *psz; if (!(cc) || (cc < 0x80 && utf7encode[cc] < 1)) break; } + if (l != 0) { if (buf) *buf++ = utf7enb64[((d % 16) << (6 - l)) % 64]; + len++; } } + if (buf) *buf++ = '-'; len++; } } + if (buf && (len < n)) *buf = 0; + return len; } @@ -724,6 +732,7 @@ size_t wxMBConvUTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const unsigned cnt; for (cnt = 0; fc & 0x80; cnt++) fc <<= 1; + if (!cnt) { // plain ASCII char @@ -761,9 +770,11 @@ size_t wxMBConvUTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const invalid = true; break; } + psz++; res = (res << 6) | (cc & 0x3f); } + if (invalid || res <= utf8_max[ocnt]) { // illegal UTF-8 encoding @@ -799,6 +810,7 @@ size_t wxMBConvUTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const #endif // WC_UTF16/!WC_UTF16 } } + if (invalid) { if (m_options & MAP_INVALID_UTF8_TO_PUA) @@ -833,6 +845,7 @@ size_t wxMBConvUTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const *buf++ = (wchar_t)( L'0' + (on % 0100) / 010 ); *buf++ = (wchar_t)( L'0' + on % 010 ); } + opsz++; len += 4; } @@ -844,8 +857,10 @@ size_t wxMBConvUTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const } } } + if (buf && (len < n)) *buf = 0; + return len; } @@ -861,12 +876,13 @@ size_t wxMBConvUTF8::WC2MB(char *buf, const wchar_t *psz, size_t n) const while (*psz && ((!buf) || (len < n))) { wxUint32 cc; + #ifdef WC_UTF16 // cast is ok for WC_UTF16 size_t pa = decode_utf16((const wxUint16 *)psz, cc); psz += (pa == wxCONV_FAILED) ? 1 : pa; #else - cc=(*psz++) & 0x7fffffff; + cc = (*psz++) & 0x7fffffff; #endif if ( (m_options & MAP_INVALID_UTF8_TO_PUA) @@ -890,8 +906,8 @@ size_t wxMBConvUTF8::WC2MB(char *buf, const wchar_t *psz, size_t n) const { if (buf) { - *buf++ = (char) ((psz[0] - L'0')*0100 + - (psz[1] - L'0')*010 + + *buf++ = (char) ((psz[0] - L'0') * 0100 + + (psz[1] - L'0') * 010 + (psz[2] - L'0')); } @@ -901,7 +917,10 @@ size_t wxMBConvUTF8::WC2MB(char *buf, const wchar_t *psz, size_t n) const else { unsigned cnt; - for (cnt = 0; cc > utf8_max[cnt]; cnt++) {} + for (cnt = 0; cc > utf8_max[cnt]; cnt++) + { + } + if (!cnt) { // plain ASCII char @@ -909,7 +928,6 @@ size_t wxMBConvUTF8::WC2MB(char *buf, const wchar_t *psz, size_t n) const *buf++ = (char) cc; len++; } - else { len += cnt + 1; @@ -923,7 +941,7 @@ size_t wxMBConvUTF8::WC2MB(char *buf, const wchar_t *psz, size_t n) const } } - if (buf && (len dstLen ) + return wxCONV_FAILED; + + *dst++ = ch; } - if (buf && len 1) + if ( outLen > dstLen ) + return wxCONV_FAILED; + + *outBuff++ = cc[0]; + if ( numChars == 2 ) { - *(wxUint16*)buf = cc[1]; - buf += sizeof(wxUint16); + // second character of a surrogate + *outBuff++ = cc[1]; } } - - len += pa*sizeof(wxUint16); - psz++; } - if (buf && len<=n-sizeof(wxUint16)) *(wxUint16*)buf=0; - return len; + return outLen; } // ---------------------------------------------------------------------------- // endian-reversing conversions // ---------------------------------------------------------------------------- -// swap 16bit MB to 32bit String -size_t wxMBConvUTF16swap::MB2WC(wchar_t *buf, const char *psz, size_t n) const +size_t +wxMBConvUTF16swap::ToWChar(wchar_t *dst, size_t dstLen, + const char *src, size_t srcLen) const { - size_t len=0; + srcLen = GetLength(src, srcLen); + if ( srcLen == wxNO_LEN ) + return wxCONV_FAILED; - while (*(wxUint16*)psz && (!buf || len < n)) + const size_t inLen = srcLen / BYTES_PER_CHAR; + if ( !dst ) { - wxUint32 cc; - char tmp[4]; - tmp[0]=psz[1]; tmp[1]=psz[0]; - tmp[2]=psz[3]; tmp[3]=psz[2]; + // optimization: return maximal space which could be needed for this + // string even if the real size could be smaller if the buffer contains + // any surrogates + return inLen; + } - size_t pa=decode_utf16((wxUint16*)tmp, cc); - if (pa == wxCONV_FAILED) - return pa; + size_t outLen = 0; + const wxUint16 *inBuff = wx_reinterpret_cast(const wxUint16 *, src); + for ( const wxUint16 * const inEnd = inBuff + inLen; inBuff < inEnd; ) + { + wxUint32 ch; + wxUint16 tmp[2]; - if (buf) - *buf++ = (wchar_t)cc; + tmp[0] = wxUINT16_SWAP_ALWAYS(*inBuff); + inBuff++; + tmp[1] = wxUINT16_SWAP_ALWAYS(*inBuff); - len++; - psz += pa * sizeof(wxUint16); + const size_t numChars = decode_utf16(tmp, ch); + if ( numChars == wxCONV_FAILED ) + return wxCONV_FAILED; + + if ( numChars == 2 ) + inBuff++; + + if ( ++outLen > dstLen ) + return wxCONV_FAILED; + + *dst++ = ch; } - if (buf && len 1) + if ( outLen > dstLen ) + return wxCONV_FAILED; + + *outBuff++ = wxUINT16_SWAP_ALWAYS(cc[0]); + if ( numChars == 2 ) { - *buf++ = ((char*)cc)[3]; - *buf++ = ((char*)cc)[2]; + // second character of a surrogate + *outBuff++ = wxUINT16_SWAP_ALWAYS(cc[1]); } } - - len += pa*sizeof(wxUint16); - psz++; } - if (buf && len<=n-sizeof(wxUint16)) *(wxUint16*)buf=0; - return len; + return outLen; } #endif // WC_UTF16/!WC_UTF16 -// ---------------------------------------------------------------------------- +// ============================================================================ // UTF-32 -// ---------------------------------------------------------------------------- +// ============================================================================ #ifdef WORDS_BIGENDIAN #define wxMBConvUTF32straight wxMBConvUTF32BE @@ -1214,8 +1266,8 @@ size_t wxMBConvUTF32Base::GetLength(const char *src, size_t srcLen) if ( srcLen == wxNO_LEN ) { // count the number of bytes in input, including the trailing NULs - const wxUint32 *in = wx_reinterpret_cast(const wxUint32 *, src); - for ( srcLen = 1; *in++; srcLen++ ) + const wxUint32 *inBuff = wx_reinterpret_cast(const wxUint32 *, src); + for ( srcLen = 1; *inBuff++; srcLen++ ) ; srcLen *= BYTES_PER_CHAR; @@ -1245,13 +1297,13 @@ wxMBConvUTF32straight::ToWChar(wchar_t *dst, size_t dstLen, if ( srcLen == wxNO_LEN ) return wxCONV_FAILED; - const wxUint32 *in = wx_reinterpret_cast(const wxUint32 *, src); - const size_t inLen = srcLen/BYTES_PER_CHAR; + const wxUint32 *inBuff = wx_reinterpret_cast(const wxUint32 *, src); + const size_t inLen = srcLen / BYTES_PER_CHAR; size_t outLen = 0; for ( size_t n = 0; n < inLen; n++ ) { wxUint16 cc[2]; - const size_t numChars = encode_utf16(*in++, cc); + const size_t numChars = encode_utf16(*inBuff++, cc); if ( numChars == wxCONV_FAILED ) return wxCONV_FAILED; @@ -1289,10 +1341,10 @@ wxMBConvUTF32straight::FromWChar(char *dst, size_t dstLen, // we consider that surrogates are rare enough to make it worthwhile to // avoid running the loop below at the cost of slightly extra memory // consumption - return srcLen*BYTES_PER_CHAR; + return srcLen * BYTES_PER_CHAR; } - wxUint32 *out = wx_reinterpret_cast(wxUint32 *, dst); + wxUint32 *outBuff = wx_reinterpret_cast(wxUint32 *, dst); size_t outLen = 0; for ( const wchar_t * const srcEnd = src + srcLen; src < srcEnd; ) { @@ -1305,7 +1357,7 @@ wxMBConvUTF32straight::FromWChar(char *dst, size_t dstLen, if ( outLen > dstLen ) return wxCONV_FAILED; - *out++ = ch; + *outBuff++ = ch; } return outLen; @@ -1323,13 +1375,13 @@ wxMBConvUTF32swap::ToWChar(wchar_t *dst, size_t dstLen, if ( srcLen == wxNO_LEN ) return wxCONV_FAILED; - const wxUint32 *in = wx_reinterpret_cast(const wxUint32 *, src); - const size_t inLen = srcLen/BYTES_PER_CHAR; + const wxUint32 *inBuff = wx_reinterpret_cast(const wxUint32 *, src); + const size_t inLen = srcLen / BYTES_PER_CHAR; size_t outLen = 0; - for ( size_t n = 0; n < inLen; n++, in++ ) + for ( size_t n = 0; n < inLen; n++, inBuff++ ) { wxUint16 cc[2]; - const size_t numChars = encode_utf16(wxUINT32_SWAP_ALWAYS(*in), cc); + const size_t numChars = encode_utf16(wxUINT32_SWAP_ALWAYS(*inBuff), cc); if ( numChars == wxCONV_FAILED ) return wxCONV_FAILED; @@ -1370,7 +1422,7 @@ wxMBConvUTF32swap::FromWChar(char *dst, size_t dstLen, return srcLen*BYTES_PER_CHAR; } - wxUint32 *out = wx_reinterpret_cast(wxUint32 *, dst); + wxUint32 *outBuff = wx_reinterpret_cast(wxUint32 *, dst); size_t outLen = 0; for ( const wchar_t * const srcEnd = src + srcLen; src < srcEnd; ) { @@ -1383,7 +1435,7 @@ wxMBConvUTF32swap::FromWChar(char *dst, size_t dstLen, if ( outLen > dstLen ) return wxCONV_FAILED; - *out++ = wxUINT32_SWAP_ALWAYS(ch); + *outBuff++ = wxUINT32_SWAP_ALWAYS(ch); } return outLen; @@ -1391,101 +1443,104 @@ wxMBConvUTF32swap::FromWChar(char *dst, size_t dstLen, #else // !WC_UTF16: wchar_t is UTF-32 -// copy 32bit MB to 32bit String -size_t wxMBConvUTF32straight::MB2WC(wchar_t *buf, const char *psz, size_t n) const +// ---------------------------------------------------------------------------- +// conversions without endianness change +// ---------------------------------------------------------------------------- + +size_t +wxMBConvUTF32straight::ToWChar(wchar_t *dst, size_t dstLen, + const char *src, size_t srcLen) const { - size_t len=0; + // use memcpy() as it should be much faster than hand-written loop + srcLen = GetLength(src, srcLen); + if ( srcLen == wxNO_LEN ) + return wxCONV_FAILED; - while (*(wxUint32*)psz && (!buf || len < n)) + const size_t inLen = srcLen/BYTES_PER_CHAR; + if ( dst ) { - if (buf) - *buf++ = (wchar_t)(*(wxUint32*)psz); - len++; - psz += sizeof(wxUint32); - } + if ( dstLen < inLen ) + return wxCONV_FAILED; - if (buf && lenm_iconvMutex); #endif // wxUSE_THREADS - size_t outbuf = n * SIZEOF_WCHAR_T; size_t res, cres; // VS: Use these instead of psz, buf because iconv() modifies its arguments: @@ -1788,16 +1846,19 @@ size_t wxMBConv_iconv::MB2WC(wchar_t *buf, const char *psz, size_t n) const // to calculate destination buffer requirement wchar_t tbuf[8]; res = 0; - do { + + do + { bufPtr = tbuf; - outbuf = 8*SIZEOF_WCHAR_T; + outbuf = 8 * SIZEOF_WCHAR_T; cres = iconv(m2w, ICONV_CHAR_CAST(&pszPtr), &inbuf, (char**)&bufPtr, &outbuf ); - res += 8-(outbuf/SIZEOF_WCHAR_T); - } while ((cres==(size_t)-1) && (errno==E2BIG)); + res += 8 - (outbuf / SIZEOF_WCHAR_T); + } + while ((cres == (size_t)-1) && (errno == E2BIG)); } if (ICONV_FAILED(cres, inbuf)) @@ -1832,6 +1893,7 @@ size_t wxMBConv_iconv::WC2MB(char *buf, const wchar_t *psz, size_t n) const tmpbuf = (wchar_t *)malloc(inbuf + SIZEOF_WCHAR_T); for ( size_t i = 0; i < inlen; i++ ) tmpbuf[n] = WC_BSWAP(psz[i]); + tmpbuf[inlen] = L'\0'; psz = tmpbuf; } @@ -1841,7 +1903,7 @@ size_t wxMBConv_iconv::WC2MB(char *buf, const wchar_t *psz, size_t n) const // have destination buffer, convert there cres = iconv( w2m, ICONV_CHAR_CAST(&psz), &inbuf, &buf, &outbuf ); - res = n-outbuf; + res = n - outbuf; // NB: iconv was given only wcslen(psz) characters on input, and so // it couldn't convert the trailing zero. Let's do it ourselves @@ -1851,17 +1913,20 @@ size_t wxMBConv_iconv::WC2MB(char *buf, const wchar_t *psz, size_t n) const } else { - // no destination buffer... convert using temp buffer + // no destination buffer: convert using temp buffer // to calculate destination buffer requirement char tbuf[16]; res = 0; - do { - buf = tbuf; outbuf = 16; + do + { + buf = tbuf; + outbuf = 16; cres = iconv( w2m, ICONV_CHAR_CAST(&psz), &inbuf, &buf, &outbuf ); res += 16 - outbuf; - } while ((cres==(size_t)-1) && (errno==E2BIG)); + } + while ((cres == (size_t)-1) && (errno == E2BIG)); } if (ms_wcNeedsSwap) @@ -1893,15 +1958,15 @@ size_t wxMBConv_iconv::GetMBNulLen() const char buf[8]; // should be enough for NUL in any encoding size_t inLen = sizeof(wchar_t), outLen = WXSIZEOF(buf); - char *in = (char *)wnul; - char *out = buf; - if ( iconv(w2m, ICONV_CHAR_CAST(&in), &inLen, &out, &outLen) == (size_t)-1 ) + char *inBuff = (char *)wnul; + char *outBuff = buf; + if ( iconv(w2m, ICONV_CHAR_CAST(&inBuff), &inLen, &outBuff, &outLen) == (size_t)-1 ) { self->m_minMBCharWidth = (size_t)-1; } else // ok { - self->m_minMBCharWidth = out - buf; + self->m_minMBCharWidth = outBuff - buf; } } @@ -1933,6 +1998,7 @@ public: } wxMBConv_win32(const wxMBConv_win32& conv) + : wxMBConv() { m_CodePage = conv.m_CodePage; m_minMBCharWidth = conv.m_minMBCharWidth; @@ -2130,7 +2196,8 @@ public: { default: wxLogDebug(_T("Unexpected NUL length %d"), len); - // fall through + self->m_minMBCharWidth = (size_t)-1; + break; case 0: self->m_minMBCharWidth = (size_t)-1; @@ -2170,8 +2237,9 @@ private: break; default: - // unknown, be conseravtive by default + // unknown: be conservative by default s_isWin98Or2k = 0; + break; } wxASSERT_MSG( s_isWin98Or2k != -1, _T("should be set above") ); @@ -2224,9 +2292,9 @@ private: #if defined(__WXCOCOA__) -// RN: There is no UTF-32 support in either Core Foundation or -// Cocoa. Strangely enough, internally Core Foundation uses -// UTF 32 internally quite a bit - its just not public (yet). +// RN: There is no UTF-32 support in either Core Foundation or Cocoa. +// Strangely enough, internally Core Foundation uses +// UTF-32 internally quite a bit - its just not public (yet). #include #include @@ -2234,12 +2302,13 @@ private: CFStringEncoding wxCFStringEncFromFontEnc(wxFontEncoding encoding) { CFStringEncoding enc = kCFStringEncodingInvalidId ; - if ( encoding == wxFONTENCODING_DEFAULT ) - { - enc = CFStringGetSystemEncoding(); - } - else switch( encoding) + + switch (encoding) { + case wxFONTENCODING_DEFAULT : + enc = CFStringGetSystemEncoding(); + break ; + case wxFONTENCODING_ISO8859_1 : enc = kCFStringEncodingISOLatin1 ; break ; @@ -2295,7 +2364,7 @@ CFStringEncoding wxCFStringEncFromFontEnc(wxFontEncoding encoding) // break ; case wxFONTENCODING_CP437 : - enc =kCFStringEncodingDOSLatinUS ; + enc = kCFStringEncodingDOSLatinUS ; break ; case wxFONTENCODING_CP850 : enc = kCFStringEncodingDOSLatin1; @@ -2307,7 +2376,7 @@ CFStringEncoding wxCFStringEncFromFontEnc(wxFontEncoding encoding) enc = kCFStringEncodingDOSCyrillic; break ; case wxFONTENCODING_CP866 : - enc =kCFStringEncodingDOSRussian ; + enc = kCFStringEncodingDOSRussian ; break ; case wxFONTENCODING_CP874 : enc = kCFStringEncodingDOSThai; @@ -2316,7 +2385,7 @@ CFStringEncoding wxCFStringEncFromFontEnc(wxFontEncoding encoding) enc = kCFStringEncodingDOSJapanese; break ; case wxFONTENCODING_CP936 : - enc =kCFStringEncodingDOSChineseSimplif ; + enc = kCFStringEncodingDOSChineseSimplif ; break ; case wxFONTENCODING_CP949 : enc = kCFStringEncodingDOSKorean; @@ -2328,10 +2397,10 @@ CFStringEncoding wxCFStringEncFromFontEnc(wxFontEncoding encoding) enc = kCFStringEncodingWindowsLatin2; break ; case wxFONTENCODING_CP1251 : - enc =kCFStringEncodingWindowsCyrillic ; + enc = kCFStringEncodingWindowsCyrillic ; break ; case wxFONTENCODING_CP1252 : - enc =kCFStringEncodingWindowsLatin1 ; + enc = kCFStringEncodingWindowsLatin1 ; break ; case wxFONTENCODING_CP1253 : enc = kCFStringEncodingWindowsGreek; @@ -2340,10 +2409,10 @@ CFStringEncoding wxCFStringEncFromFontEnc(wxFontEncoding encoding) enc = kCFStringEncodingWindowsLatin5; break ; case wxFONTENCODING_CP1255 : - enc =kCFStringEncodingWindowsHebrew ; + enc = kCFStringEncodingWindowsHebrew ; break ; case wxFONTENCODING_CP1256 : - enc =kCFStringEncodingWindowsArabic ; + enc = kCFStringEncodingWindowsArabic ; break ; case wxFONTENCODING_CP1257 : enc = kCFStringEncodingWindowsBalticRim; @@ -2481,10 +2550,12 @@ CFStringEncoding wxCFStringEncFromFontEnc(wxFontEncoding encoding) // case wxFONTENCODING_MACKEYBOARD : // enc = kCFStringEncodingMacKeyboardGlyphs ; // break ; + default : // because gcc is picky break ; - } ; + } + return enc ; } @@ -2554,12 +2625,12 @@ public: CFRelease(theString); - szUniCharBuffer[nOutLength] = '\0' ; + szUniCharBuffer[nOutLength] = '\0'; #if SIZEOF_WCHAR_T == 4 - wxMBConvUTF16 converter ; - converter.MB2WC(szOut, (const char*)szUniCharBuffer , nOutSize ) ; - delete[] szUniCharBuffer; + wxMBConvUTF16 converter; + converter.MB2WC( szOut, (const char*)szUniCharBuffer, nOutSize ); + delete [] szUniCharBuffer; #endif return nOutLength; @@ -2575,9 +2646,9 @@ public: #if SIZEOF_WCHAR_T == 4 wxMBConvUTF16 converter ; - nBufSize = converter.WC2MB( NULL , szUnConv , 0 ); - szUniBuffer = new UniChar[ (nBufSize / sizeof(UniChar)) + 1] ; - converter.WC2MB( (char*) szUniBuffer , szUnConv, nBufSize + sizeof(UniChar)) ; + nBufSize = converter.WC2MB( NULL, szUnConv, 0 ); + szUniBuffer = new UniChar[ (nBufSize / sizeof(UniChar)) + 1]; + converter.WC2MB( (char*) szUniBuffer, szUnConv, nBufSize + sizeof(UniChar)); nBufSize /= sizeof(UniChar); #endif @@ -2659,7 +2730,7 @@ public: #if wxUSE_FONTMAP wxMBConv_mac(const wxChar* name) { - Init( wxMacGetSystemEncFromFontEnc(wxFontMapperBase::Get()->CharsetToEncoding(name, false) ) ) ; + Init( wxMacGetSystemEncFromFontEnc( wxFontMapperBase::Get()->CharsetToEncoding(name, false) ) ); } #endif @@ -2680,7 +2751,7 @@ public: { OSStatus status = noErr ; m_char_encoding = encoding ; - m_unicode_encoding = CreateTextEncoding(kTextEncodingUnicodeDefault,0,kUnicode16BitFormat) ; + m_unicode_encoding = CreateTextEncoding(kTextEncodingUnicodeDefault, 0, kUnicode16BitFormat) ; status = TECCreateConverter(&m_MB2WC_converter, m_char_encoding, @@ -2694,35 +2765,41 @@ public: { OSStatus status = noErr ; ByteCount byteOutLen ; - ByteCount byteInLen = strlen(psz) ; + ByteCount byteInLen = strlen(psz) + 1; wchar_t *tbuf = NULL ; UniChar* ubuf = NULL ; size_t res = 0 ; if (buf == NULL) { - //apple specs say at least 32 - n = wxMax( 32 , byteInLen ) ; - tbuf = (wchar_t*) malloc( n * SIZEOF_WCHAR_T) ; + // Apple specs say at least 32 + n = wxMax( 32, byteInLen ) ; + tbuf = (wchar_t*) malloc( n * SIZEOF_WCHAR_T ) ; } + ByteCount byteBufferLen = n * sizeof( UniChar ) ; + #if SIZEOF_WCHAR_T == 4 ubuf = (UniChar*) malloc( byteBufferLen + 2 ) ; #else ubuf = (UniChar*) (buf ? buf : tbuf) ; #endif - status = TECConvertText(m_MB2WC_converter, (ConstTextPtr) psz , byteInLen, &byteInLen, - (TextPtr) ubuf , byteBufferLen, &byteOutLen); + + status = TECConvertText( + m_MB2WC_converter, (ConstTextPtr) psz, byteInLen, &byteInLen, + (TextPtr) ubuf, byteBufferLen, &byteOutLen); + #if SIZEOF_WCHAR_T == 4 // 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 ; wxMBConvUTF16 converter ; - res = converter.MB2WC( (buf ? buf : tbuf) , (const char*)ubuf , n ) ; + res = converter.MB2WC( (buf ? buf : tbuf), (const char*)ubuf, n ) ; free( ubuf ) ; #else res = byteOutLen / sizeof( UniChar ) ; #endif + if ( buf == NULL ) free(tbuf) ; @@ -2742,27 +2819,32 @@ public: if (buf == NULL) { - //apple specs say at least 32 - n = wxMax( 32 , ((byteInLen / SIZEOF_WCHAR_T) * 8) + SIZEOF_WCHAR_T ); + // Apple specs say at least 32 + n = wxMax( 32, ((byteInLen / SIZEOF_WCHAR_T) * 8) + SIZEOF_WCHAR_T ); tbuf = (char*) malloc( n ) ; } ByteCount byteBufferLen = n ; UniChar* ubuf = NULL ; + #if SIZEOF_WCHAR_T == 4 wxMBConvUTF16 converter ; - size_t unicharlen = converter.WC2MB( NULL , psz , 0 ) ; + size_t unicharlen = converter.WC2MB( NULL, psz, 0 ) ; byteInLen = unicharlen ; ubuf = (UniChar*) malloc( byteInLen + 2 ) ; - converter.WC2MB( (char*) ubuf , psz, unicharlen + 2 ) ; + converter.WC2MB( (char*) ubuf, psz, unicharlen + 2 ) ; #else ubuf = (UniChar*) psz ; #endif - status = TECConvertText(m_WC2MB_converter, (ConstTextPtr) ubuf , byteInLen, &byteInLen, - (TextPtr) (buf ? buf : tbuf) , byteBufferLen, &byteOutLen); + + status = TECConvertText( + m_WC2MB_converter, (ConstTextPtr) ubuf, byteInLen, &byteInLen, + (TextPtr) (buf ? buf : tbuf), byteBufferLen, &byteOutLen); + #if SIZEOF_WCHAR_T == 4 free( ubuf ) ; #endif + if ( buf == NULL ) free(tbuf) ; @@ -2791,14 +2873,14 @@ public: virtual wxMBConv *Clone() const { return new wxMBConv_mac(*this); } bool IsOk() const - { return m_MB2WC_converter != NULL && m_WC2MB_converter != NULL ; } + { return m_MB2WC_converter != NULL && m_WC2MB_converter != NULL; } private: - TECObjectRef m_MB2WC_converter ; - TECObjectRef m_WC2MB_converter ; + TECObjectRef m_MB2WC_converter; + TECObjectRef m_WC2MB_converter; - TextEncodingBase m_char_encoding ; - TextEncodingBase m_unicode_encoding ; + TextEncodingBase m_char_encoding; + TextEncodingBase m_unicode_encoding; }; #endif // defined(__WXMAC__) && defined(TARGET_CARBON) @@ -2843,7 +2925,7 @@ public: size_t inbuf = strlen(psz); if (buf) { - if (!m2w.Convert(psz,buf)) + if (!m2w.Convert(psz, buf)) return wxCONV_FAILED; } return inbuf; @@ -2854,7 +2936,7 @@ public: const size_t inbuf = wxWcslen(psz); if (buf) { - if (!w2m.Convert(psz,buf)) + if (!w2m.Convert(psz, buf)) return wxCONV_FAILED; } @@ -2902,6 +2984,7 @@ WXDLLIMPEXP_BASE wxMBConv* new_wxMBConv_wxwin( const wxChar* name ) delete result; return 0; } + return result; } @@ -2991,7 +3074,6 @@ void wxCSConv::SetName(const wxChar *charset) } #if wxUSE_FONTMAP -#include "wx/hashmap.h" WX_DECLARE_HASH_MAP( wxFontEncoding, wxString, wxIntegerHash, wxIntegerEqual, wxEncodingNameCache ); @@ -3097,13 +3179,13 @@ wxMBConv *wxCSConv::DoCreate() const #endif } #endif // wxHAVE_WIN32_MB2WC + #if defined(__WXMAC__) { // leave UTF16 and UTF32 to the built-ins of wx if ( m_name || ( m_encoding < wxFONTENCODING_UTF16BE || ( m_encoding >= wxFONTENCODING_MACMIN && m_encoding <= wxFONTENCODING_MACMAX ) ) ) { - #if wxUSE_FONTMAP wxMBConv_mac *conv = m_name ? new wxMBConv_mac(m_name) : new wxMBConv_mac(m_encoding); @@ -3117,17 +3199,18 @@ wxMBConv *wxCSConv::DoCreate() const } } #endif + #if defined(__WXCOCOA__) { if ( m_name || ( m_encoding <= wxFONTENCODING_UTF16 ) ) { - #if wxUSE_FONTMAP wxMBConv_cocoa *conv = m_name ? new wxMBConv_cocoa(m_name) : new wxMBConv_cocoa(m_encoding); #else wxMBConv_cocoa *conv = new wxMBConv_cocoa(m_encoding); #endif + if ( conv->IsOk() ) return conv; @@ -3169,7 +3252,7 @@ wxMBConv *wxCSConv::DoCreate() const default: // nothing to do but put here to suppress gcc warnings - ; + break; } // step (3) @@ -3204,6 +3287,7 @@ wxMBConv *wxCSConv::DoCreate() const wxString::Format(_("encoding %s"), m_encoding).c_str() #endif // wxUSE_FONTMAP/!wxUSE_FONTMAP ); + alreadyLoggingError = false; } @@ -3264,6 +3348,7 @@ size_t wxCSConv::WC2MB(char *buf, const wchar_t *psz, size_t n) const { if (psz[c] > 0xFF) return wxCONV_FAILED; + buf[c] = (char)psz[c]; } } @@ -3314,6 +3399,7 @@ WXDLLIMPEXP_DATA_BASE(wxCSConv&) wxConvISO8859_1 = wxConvISO8859_1Obj; WXDLLIMPEXP_DATA_BASE(wxMBConvUTF7&) wxConvUTF7 = wxConvUTF7Obj; WXDLLIMPEXP_DATA_BASE(wxMBConvUTF8&) wxConvUTF8 = wxConvUTF8Obj; WXDLLIMPEXP_DATA_BASE(wxMBConv *) wxConvCurrent = &wxConvLibcObj; +WXDLLIMPEXP_DATA_BASE(wxMBConv *) wxConvUI = &wxConvLocal; WXDLLIMPEXP_DATA_BASE(wxMBConv *) wxConvFileName = & #ifdef __WXOSX__ wxConvUTF8Obj; @@ -3321,7 +3407,6 @@ WXDLLIMPEXP_DATA_BASE(wxMBConv *) wxConvFileName = & wxConvLibcObj; #endif - #else // !wxUSE_WCHAR_T // stand-ins in absence of wchar_t