X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/95c8801cdf222d27d92350a0f67938554fea8a63..3b01b1fe4cb656f39290bf7f3891667855abe601:/src/common/strconv.cpp diff --git a/src/common/strconv.cpp b/src/common/strconv.cpp index 95fc0d9c07..83c06de9c6 100644 --- a/src/common/strconv.cpp +++ b/src/common/strconv.cpp @@ -1,12 +1,15 @@ ///////////////////////////////////////////////////////////////////////////// // Name: strconv.cpp // Purpose: Unicode conversion classes -// Author: Ove Kaaven, Robert Roebling, Vadim Zeitlin +// Author: Ove Kaaven, Robert Roebling, Vadim Zeitlin, Vaclav Slavik, +// Ryan Norton, Fredrik Roubert (UTF7) // Modified by: // Created: 29/01/98 // RCS-ID: $Id$ -// Copyright: (c) 1999 Ove Kaaven, Robert Roebling, Vadim Zeitlin -// Licence: wxWindows license +// Copyright: (c) 1999 Ove Kaaven, Robert Roebling, Vaclav Slavik +// (c) 2000-2003 Vadim Zeitlin +// (c) 2004 Ryan Norton, Fredrik Roubert +// Licence: wxWindows licence ///////////////////////////////////////////////////////////////////////////// // ============================================================================ @@ -17,7 +20,7 @@ // headers // ---------------------------------------------------------------------------- -#ifdef __GNUG__ +#if defined(__GNUG__) && !defined(NO_GCC_PRAGMA) #pragma implementation "strconv.h" #endif @@ -28,642 +31,2409 @@ #pragma hdrstop #endif +#ifndef WX_PRECOMP + #include "wx/intl.h" + #include "wx/log.h" +#endif // WX_PRECOMP + +#include "wx/strconv.h" + +#if wxUSE_WCHAR_T + #ifdef __WXMSW__ - #include "wx/msw/private.h" + #include "wx/msw/private.h" +#endif + +#ifdef __WINDOWS__ + #include "wx/msw/missing.h" #endif +#ifndef __WXWINCE__ #include +#endif + #include #include #include +#if defined(__WIN32__) && !defined(__WXMICROWIN__) + #define wxHAVE_WIN32_MB2WC +#endif // __WIN32__ but !__WXMICROWIN__ + +// ---------------------------------------------------------------------------- +// headers +// ---------------------------------------------------------------------------- + #ifdef __SALFORDC__ - #include + #include #endif -#ifdef HAVE_ICONV_H - #include -#endif -#ifdef HAVE_LANGINFO_H - #include +#ifdef HAVE_ICONV + #include #endif -#ifdef __WXMSW__ - #include -#endif +#include "wx/encconv.h" +#include "wx/fontmap.h" +#include "wx/utils.h" -#include "wx/debug.h" -#include "wx/strconv.h" +#ifdef __WXMAC__ +#include +#include +#include -#if defined(WORDS_BIGENDIAN) || defined(__STDC_ISO_10646__) -#define BSWAP_UCS4(str, len) -#define BSWAP_UCS2(str, len) -#else -#define BSWAP_UCS4(str, len) { unsigned _c; for (_c=0; _c=0x110000) { - return (size_t)-1; - } else { - if (output) { - *output++ = (input >> 10)+0xd7c0; - *output++ = (input&0x3ff)+0xdc00; + if (input<=0xffff) + { + if (output) + *output = (wxUint16) input; + return 1; + } + else if (input>=0x110000) + { + return (size_t)-1; + } + else + { + if (output) + { + *output++ = (wxUint16) ((input >> 10)+0xd7c0); + *output = (wxUint16) ((input&0x3ff)+0xdc00); + } + return 2; } - return 2; - } } -static size_t decode_utf16(wxUint16*input,wxUint32&output) +static size_t decode_utf16(const wxUint16* input, wxUint32& output) { - if ((*input<0xd800) || (*input>0xdfff)) { - output = *input; - return 1; - } else - if ((input[1]<0xdc00) || (input[1]>=0xdfff)) { - output = *input; - return (size_t)-1; - } else { - output = ((input[0] - 0xd7c0) << 10) + (input[1] - 0xdc00); - return 2; - } + if ((*input<0xd800) || (*input>0xdfff)) + { + output = *input; + return 1; + } + else if ((input[1]<0xdc00) || (input[1]>=0xdfff)) + { + output = *input; + return (size_t)-1; + } + else + { + output = ((input[0] - 0xd7c0) << 10) + (input[1] - 0xdc00); + return 2; + } } -#endif // WC_UTF16 // ---------------------------------------------------------------------------- // wxMBConv // ---------------------------------------------------------------------------- -WXDLLEXPORT_DATA(wxMBConv) wxConvLibc; - -size_t wxMBConv::MB2WC(wchar_t *buf, const char *psz, size_t n) const -{ - return wxMB2WC(buf, psz, n); -} - -size_t wxMBConv::WC2MB(char *buf, const wchar_t *psz, size_t n) const +wxMBConv::~wxMBConv() { - return wxWC2MB(buf, psz, n); + // nothing to do here (necessary for Darwin linking probably) } const wxWCharBuffer wxMBConv::cMB2WC(const char *psz) const { - if (psz) + if ( psz ) { - size_t nLen = MB2WC((wchar_t *) NULL, psz, 0); - if (nLen == (size_t)-1) - return wxWCharBuffer((wchar_t *) NULL); - wxWCharBuffer buf(nLen); - MB2WC((wchar_t *)(const wchar_t *) buf, psz, nLen); - return buf; + // calculate the length of the buffer needed first + size_t nLen = MB2WC(NULL, psz, 0); + if ( nLen != (size_t)-1 ) + { + // now do the actual conversion + wxWCharBuffer buf(nLen); + nLen = MB2WC(buf.data(), psz, nLen + 1); // with the trailing NULL + if ( nLen != (size_t)-1 ) + { + return buf; + } + } } - else - return wxWCharBuffer((wchar_t *) NULL); + + wxWCharBuffer buf((wchar_t *)NULL); + + return buf; } -const wxCharBuffer wxMBConv::cWC2MB(const wchar_t *psz) const +const wxCharBuffer wxMBConv::cWC2MB(const wchar_t *pwz) const { - if (psz) + if ( pwz ) { - size_t nLen = WC2MB((char *) NULL, psz, 0); - if (nLen == (size_t)-1) - return wxCharBuffer((char *) NULL); - wxCharBuffer buf(nLen); - WC2MB((char *)(const char *) buf, psz, nLen); - return buf; + size_t nLen = WC2MB(NULL, pwz, 0); + if ( nLen != (size_t)-1 ) + { + wxCharBuffer buf(nLen+3); // space for a wxUint32 trailing zero + nLen = WC2MB(buf.data(), pwz, nLen + 4); + if ( nLen != (size_t)-1 ) + { + return buf; + } + } } - else - return wxCharBuffer((char *) NULL); + + wxCharBuffer buf((char *)NULL); + + return buf; } // ---------------------------------------------------------------------------- -// standard file conversion +// wxMBConvLibc // ---------------------------------------------------------------------------- -WXDLLEXPORT_DATA(wxMBConvFile) wxConvFile; - -// just use the libc conversion for now -size_t wxMBConvFile::MB2WC(wchar_t *buf, const char *psz, size_t n) const +size_t wxMBConvLibc::MB2WC(wchar_t *buf, const char *psz, size_t n) const { return wxMB2WC(buf, psz, n); } -size_t wxMBConvFile::WC2MB(char *buf, const wchar_t *psz, size_t n) const +size_t wxMBConvLibc::WC2MB(char *buf, const wchar_t *psz, size_t n) const { return wxWC2MB(buf, psz, n); } - // ---------------------------------------------------------------------------- -// standard gdk conversion +// UTF-7 // ---------------------------------------------------------------------------- -#ifdef __WXGTK12__ - -WXDLLEXPORT_DATA(wxMBConvGdk) wxConvGdk; +// Implementation (C) 2004 Fredrik Roubert -#include +// +// BASE64 decoding table +// +static const unsigned char utf7unb64[] = +{ + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0x3e, 0xff, 0xff, 0xff, 0x3f, + 0x34, 0x35, 0x36, 0x37, 0x38, 0x39, 0x3a, 0x3b, + 0x3c, 0x3d, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, + 0x07, 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, + 0x0f, 0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, + 0x17, 0x18, 0x19, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f, 0x20, + 0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27, 0x28, + 0x29, 0x2a, 0x2b, 0x2c, 0x2d, 0x2e, 0x2f, 0x30, + 0x31, 0x32, 0x33, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff +}; -size_t wxMBConvGdk::MB2WC(wchar_t *buf, const char *psz, size_t n) const +size_t wxMBConvUTF7::MB2WC(wchar_t *buf, const char *psz, size_t n) const { - if (buf) { - return gdk_mbstowcs((GdkWChar *)buf, psz, n); - } else { - GdkWChar *nbuf = new GdkWChar[n=strlen(psz)]; - size_t len = gdk_mbstowcs(nbuf, psz, n); - delete [] nbuf; + + size_t len = 0; + + while (*psz && ((!buf) || (len < n))) + { + unsigned char cc = *psz++; + if (cc != '+') + { + // plain ASCII char + if (buf) + *buf++ = cc; + len++; + } + else if (*psz == '-') + { + // encoded plus sign + if (buf) + *buf++ = cc; + len++; + psz++; + } + else + { + // BASE64 encoded string + bool lsb; + unsigned char c; + unsigned int d, l; + for (lsb = false, d = 0, l = 0; + (cc = utf7unb64[(unsigned char)*psz]) != 0xff; psz++) + { + d <<= 6; + d += cc; + for (l += 6; l >= 8; lsb = !lsb) + { + c = (d >> (l -= 8)) % 256; + if (lsb) + { + if (buf) + *buf++ |= c; + len ++; + } + else + if (buf) + *buf = c << 8; + } + } + if (*psz == '-') + psz++; + } + } + if (buf && (len < n)) + *buf = 0; return len; - } } -size_t wxMBConvGdk::WC2MB(char *buf, const wchar_t *psz, size_t n) const +// +// BASE64 encoding table +// +static const unsigned char utf7enb64[] = { - char *mbstr = gdk_wcstombs((GdkWChar *)psz); - size_t len = mbstr ? strlen(mbstr) : 0; - if (buf) { - if (len > n) len = n; - memcpy(buf, psz, len); - if (len < n) buf[len] = 0; - } - return len; -} - -#endif // GTK > 1.0 + 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', + 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', + 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', + 'Y', 'Z', 'a', 'b', 'c', 'd', 'e', 'f', + 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', + 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', + 'w', 'x', 'y', 'z', '0', '1', '2', '3', + '4', '5', '6', '7', '8', '9', '+', '/' +}; -// ---------------------------------------------------------------------------- -// UTF-7 -// ---------------------------------------------------------------------------- +// +// UTF-7 encoding table +// +// 0 - Set D (directly encoded characters) +// 1 - Set O (optional direct characters) +// 2 - whitespace characters (optional) +// 3 - special characters +// +static const unsigned char utf7encode[128] = +{ + 3, 3, 3, 3, 3, 3, 3, 3, 3, 2, 2, 3, 3, 2, 3, 3, + 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, + 2, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 3, 0, 0, 0, 3, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, + 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 3, 1, 1, 1, + 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 3, 3 +}; -WXDLLEXPORT_DATA(wxMBConvUTF7) wxConvUTF7; +size_t wxMBConvUTF7::WC2MB(char *buf, const wchar_t +*psz, size_t n) const +{ -#if 0 -static char utf7_setD[]="ABCDEFGHIJKLMNOPQRSTUVWXYZ" - "abcdefghijklmnopqrstuvwxyz" - "0123456789'(),-./:?"; -static char utf7_setO[]="!\"#$%&*;<=>@[]^_`{|}"; -static char utf7_setB[]="ABCDEFGHIJKLMNOPQRSTUVWXYZ" - "abcdefghijklmnopqrstuvwxyz" - "0123456789+/"; -#endif -// TODO: write actual implementations of UTF-7 here -size_t wxMBConvUTF7::MB2WC(wchar_t * WXUNUSED(buf), - const char * WXUNUSED(psz), - size_t WXUNUSED(n)) const -{ - return 0; -} + size_t len = 0; -size_t wxMBConvUTF7::WC2MB(char * WXUNUSED(buf), - const wchar_t * WXUNUSED(psz), - size_t WXUNUSED(n)) const -{ - return 0; + while (*psz && ((!buf) || (len < n))) + { + wchar_t cc = *psz++; + if (cc < 0x80 && utf7encode[cc] < 1) + { + // plain ASCII char + if (buf) + *buf++ = (char)cc; + len++; + } +#ifndef WC_UTF16 + else if (cc > 0xffff) + { + // no surrogate pair generation (yet?) + return (size_t)-1; + } +#endif + else + { + if (buf) + *buf++ = '+'; + len++; + if (cc != '+') + { + // BASE64 encode string + unsigned int lsb, d, l; + for (d = 0, l = 0;; psz++) + { + for (lsb = 0; lsb < 2; lsb ++) + { + d <<= 8; + d += lsb ? cc & 0xff : (cc & 0xff00) >> 8; + + for (l += 8; l >= 6; ) + { + l -= 6; + if (buf) + *buf++ = utf7enb64[(d >> l) % 64]; + 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; } // ---------------------------------------------------------------------------- // UTF-8 // ---------------------------------------------------------------------------- -WXDLLEXPORT_DATA(wxMBConvUTF8) wxConvUTF8; - -static wxUint32 utf8_max[]={0x7f,0x7ff,0xffff,0x1fffff,0x3ffffff,0x7fffffff,0xffffffff}; +static wxUint32 utf8_max[]= + { 0x7f, 0x7ff, 0xffff, 0x1fffff, 0x3ffffff, 0x7fffffff, 0xffffffff }; size_t wxMBConvUTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const { - size_t len = 0; - - while (*psz && ((!buf) || (len>cnt); - while (cnt--) { - cc = *psz++; - if ((cc&0xC0)!=0x80) { - // invalid UTF-8 sequence - return (size_t)-1; - } - res=(res<<6)|(cc&0x3f); - } - if (res<=utf8_max[ocnt]) { - // illegal UTF-8 encoding - return (size_t)-1; + size_t len = 0; + + while (*psz && ((!buf) || (len < n))) + { + unsigned char cc = *psz++, fc = cc; + unsigned cnt; + for (cnt = 0; fc & 0x80; cnt++) + fc <<= 1; + if (!cnt) + { + // plain ASCII char + if (buf) + *buf++ = cc; + len++; } + else + { + cnt--; + if (!cnt) + { + // invalid UTF-8 sequence + return (size_t)-1; + } + else + { + unsigned ocnt = cnt - 1; + wxUint32 res = cc & (0x3f >> cnt); + while (cnt--) + { + cc = *psz++; + if ((cc & 0xC0) != 0x80) + { + // invalid UTF-8 sequence + return (size_t)-1; + } + res = (res << 6) | (cc & 0x3f); + } + if (res <= utf8_max[ocnt]) + { + // illegal UTF-8 encoding + return (size_t)-1; + } #ifdef WC_UTF16 - size_t pa = encode_utf16(res, buf); - if (pa == (size_t)-1) - return (size_t)-1; - if (buf) buf+=pa; - len+=pa; -#else - if (buf) *buf++=res; - len++; -#endif - } + // cast is ok because wchar_t == wxUuint16 if WC_UTF16 + size_t pa = encode_utf16(res, (wxUint16 *)buf); + if (pa == (size_t)-1) + return (size_t)-1; + if (buf) + buf += pa; + len += pa; +#else // !WC_UTF16 + if (buf) + *buf++ = res; + len++; +#endif // WC_UTF16/!WC_UTF16 + } + } } - } - if (buf && (lenutf8_max[cnt]; cnt++); - if (!cnt) { - // plain ASCII char - if (buf) *buf++=cc; - len++; - } else { - len+=cnt+1; - if (buf) { - *buf++=(-128>>cnt)|((cc>>(cnt*6))&(0x3f>>cnt)); - while (cnt--) - *buf++=0x80|((cc>>(cnt*6))&0x3f); - } - } - } - if (buf && (len utf8_max[cnt]; cnt++) {} + if (!cnt) + { + // plain ASCII char + if (buf) + *buf++ = (char) cc; + len++; + } + + else + { + len += cnt + 1; + if (buf) + { + *buf++ = (char) ((-128 >> cnt) | ((cc >> (cnt * 6)) & (0x3f >> cnt))); + while (cnt--) + *buf++ = (char) (0x80 | ((cc >> (cnt * 6)) & 0x3f)); + } + } + } + + if (buf && (len wxConvUTF8) -// - move wxEncodingConverter meat in here - -#ifdef __WIN32__ -#include "wx/msw/registry.h" -// this should work if M$ Internet Exploiter is installed -static long CharsetToCodepage(const wxChar *name) -{ - if (!name) - return GetACP(); - - long CP=-1; - - wxString cn(name); - do { - wxString path( wxT("MIME\\Database\\Charset\\") ); - path += cn; - wxRegKey key( wxRegKey::HKCR, path ); - - /* two cases: either there's an AliasForCharset string, - * or there are Codepage and InternetEncoding dwords. - * The InternetEncoding gives us the actual encoding, - * the Codepage just says which Windows character set to - * use when displaying the data. - */ - if (key.QueryValue( wxT("InternetEncoding"), &CP )) break; - - // no encoding, see if it's an alias - if (!key.QueryValue( wxT("AliasForCharset"), cn )) break; - } while (1); - - return CP; -} +// ---------------------------------------------------------------------------- +// UTF-16 +// ---------------------------------------------------------------------------- + +#ifdef WORDS_BIGENDIAN + #define wxMBConvUTF16straight wxMBConvUTF16BE + #define wxMBConvUTF16swap wxMBConvUTF16LE +#else + #define wxMBConvUTF16swap wxMBConvUTF16BE + #define wxMBConvUTF16straight wxMBConvUTF16LE #endif -class wxCharacterSet -{ -public: - wxCharacterSet(const wxChar*name) - : cname(name) {} - virtual ~wxCharacterSet() - {} - virtual size_t MB2WC(wchar_t*buf, const char*psz, size_t n) - { return (size_t)-1; } - virtual size_t WC2MB(char*buf, const wchar_t*psz, size_t n) - { return (size_t)-1; } - virtual bool usable() - { return FALSE; } -public: - const wxChar*cname; -}; -class ID_CharSet : public wxCharacterSet -{ -public: - ID_CharSet(const wxChar*name,wxMBConv*cnv) - : wxCharacterSet(name), work(cnv) {} - - size_t MB2WC(wchar_t*buf, const char*psz, size_t n) - { return work ? work->MB2WC(buf,psz,n) : (size_t)-1; } - - size_t WC2MB(char*buf, const wchar_t*psz, size_t n) - { return work ? work->WC2MB(buf,psz,n) : (size_t)-1; } - - bool usable() - { return work!=NULL; } -public: - wxMBConv*work; -}; +#ifdef WC_UTF16 -#ifdef HAVE_ICONV_H -class IC_CharSet : public wxCharacterSet +// copy 16bit MB to 16bit String +size_t wxMBConvUTF16straight::MB2WC(wchar_t *buf, const char *psz, size_t n) const { -public: - IC_CharSet(const wxChar*name) - : wxCharacterSet(name), m2w((iconv_t)-1), w2m((iconv_t)-1) {} - ~IC_CharSet() + size_t len=0; + + while (*(wxUint16*)psz && (!buf || len < n)) { - if (m2w!=(iconv_t)-1) iconv_close(m2w); - if (w2m!=(iconv_t)-1) iconv_close(w2m); - } - - void LoadM2W() - { - if (m2w==(iconv_t)-1) - m2w=iconv_open(WC_NAME,wxConvLibc.cWX2MB(cname)); - } - - void LoadW2M() - { - if (w2m==(iconv_t)-1) - w2m=iconv_open(wxConvLibc.cWX2MB(cname),WC_NAME); + if (buf) + *buf++ = *(wxUint16*)psz; + len++; + + psz += sizeof(wxUint16); } - - size_t MB2WC(wchar_t*buf, const char*psz, size_t n) + if (buf && len 0x530) - size_t inbuf = std::wcslen(psz); -#else - size_t inbuf = ::wcslen(psz); -#endif - size_t outbuf = n; - size_t res, cres; - fprintf(stderr,"IC Convert from WC using %s\n",(const char*)wxConvLibc.cWX2MB(cname)); -#ifdef WC_NEED_BSWAP - // need to copy to temp buffer to switch endianness - // this absolutely doesn't rock! - // (no, doing WC_BSWAP twice on the original buffer won't help, as it - // could be in read-only memory, or be accessed in some other thread) - wchar_t*tmpbuf=(wchar_t*)malloc((inbuf+1)*SIZEOF_WCHAR_T); - memcpy(tmpbuf,psz,(inbuf+1)*SIZEOF_WCHAR_T); - WC_BSWAP(tmpbuf, inbuf) - psz=tmpbuf; -#endif if (buf) { - // have destination buffer, convert there -#ifdef WX_ICONV_TAKES_CHAR - cres = iconv( w2m, (char**)&psz, &inbuf, &buf, &outbuf ); -#else - cres = iconv( w2m, (const char**)&psz, &inbuf, &buf, &outbuf ); -#endif - res = n-outbuf; - } - else - { - // no destination buffer... convert using temp buffer - // to calculate destination buffer requirement - char tbuf[16]; - res = 0; - do { - buf = tbuf; outbuf = 16; -#ifdef WX_ICONV_TAKES_CHAR - cres = iconv( w2m, (char**)&psz, &inbuf, &buf, &outbuf ); -#else - cres = iconv( w2m, (const char**)&psz, &inbuf, &buf, &outbuf ); -#endif - res += 16 - outbuf; - } while ((cres==(size_t)-1) && (errno==E2BIG)); + ((char *)buf)[0] = psz[1]; + ((char *)buf)[1] = psz[0]; + buf++; } -#ifdef WC_NEED_BSWAP - free(tmpbuf); -#endif - if (cres==(size_t)-1) - return (size_t)-1; - - return res; + len++; + psz += sizeof(wxUint16); } - - bool usable() - { return TRUE; } - -public: - iconv_t m2w, w2m; -}; -#endif + if (buf && len 1) + { + *(wxUint16*)buf = cc[1]; + buf += sizeof(wxUint16); + } + } + + len += pa*sizeof(wxUint16); + psz++; + } + if (buf && len<=n-sizeof(wxUint16)) *(wxUint16*)buf=0; + + return len; +} + + +// swap 16bit MB to 32bit String +size_t wxMBConvUTF16swap::MB2WC(wchar_t *buf, const char *psz, size_t n) const +{ + size_t len=0; + + while (*(wxUint16*)psz && (!buf || len < n)) + { + wxUint32 cc; + char tmp[4]; + tmp[0]=psz[1]; tmp[1]=psz[0]; + tmp[2]=psz[3]; tmp[3]=psz[2]; + + size_t pa=decode_utf16((wxUint16*)tmp, cc); + if (pa == (size_t)-1) + return pa; + + if (buf) + *buf++ = cc; + + len++; + psz += pa * sizeof(wxUint16); + } + if (buf && len 1) + { + *buf++ = ((char*)cc)[3]; + *buf++ = ((char*)cc)[2]; + } + } + + len += pa*sizeof(wxUint16); + psz++; + } + if (buf && len<=n-sizeof(wxUint16)) *(wxUint16*)buf=0; + + return len; +} + +#endif // WC_UTF16 + + +// ---------------------------------------------------------------------------- +// UTF-32 +// ---------------------------------------------------------------------------- + +#ifdef WORDS_BIGENDIAN +#define wxMBConvUTF32straight wxMBConvUTF32BE +#define wxMBConvUTF32swap wxMBConvUTF32LE +#else +#define wxMBConvUTF32swap wxMBConvUTF32BE +#define wxMBConvUTF32straight wxMBConvUTF32LE +#endif + + +WXDLLIMPEXP_DATA_BASE(wxMBConvUTF32LE) wxConvUTF32LE; +WXDLLIMPEXP_DATA_BASE(wxMBConvUTF32BE) wxConvUTF32BE; + + +#ifdef WC_UTF16 + +// copy 32bit MB to 16bit String +size_t wxMBConvUTF32straight::MB2WC(wchar_t *buf, const char *psz, size_t n) const +{ + size_t len=0; + + while (*(wxUint32*)psz && (!buf || len < n)) + { + wxUint16 cc[2]; + + size_t pa=encode_utf16(*(wxUint32*)psz, cc); + if (pa == (size_t)-1) + return pa; + + if (buf) + { + *buf++ = cc[0]; + if (pa > 1) + *buf++ = cc[1]; + } + len += pa; + psz += sizeof(wxUint32); + } + if (buf && len 1) + *buf++ = cc[1]; + } + len += pa; + psz += sizeof(wxUint32); + } + + if (buf && len= 4 && verMin >= 10; + break; + + case wxWINDOWS_NT: + s_isWin98Or2k = verMaj >= 5; + break; + + default: + // unknown, be conseravtive by default + s_isWin98Or2k = 0; + } + + wxASSERT_MSG( s_isWin98Or2k != -1, _T("should be set above") ); + } + + return s_isWin98Or2k == 1; + } + + long m_CodePage; +}; + +#endif // wxHAVE_WIN32_MB2WC + +// ============================================================================ +// Cocoa conversion classes +// ============================================================================ + +#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). + +#include +#include + +CFStringEncoding wxCFStringEncFromFontEnc(wxFontEncoding encoding) +{ + CFStringEncoding enc = 0 ; + if ( encoding == wxFONTENCODING_DEFAULT ) + { +#if wxUSE_GUI + encoding = wxFont::GetDefaultEncoding() ; +#else + encoding = wxLocale::GetSystemEncoding() ; +#endif + } + else switch( encoding) + { + case wxFONTENCODING_ISO8859_1 : + enc = kCFStringEncodingISOLatin1 ; + break ; + case wxFONTENCODING_ISO8859_2 : + enc = kCFStringEncodingISOLatin2; + break ; + case wxFONTENCODING_ISO8859_3 : + enc = kCFStringEncodingISOLatin3 ; + break ; + case wxFONTENCODING_ISO8859_4 : + enc = kCFStringEncodingISOLatin4; + break ; + case wxFONTENCODING_ISO8859_5 : + enc = kCFStringEncodingISOLatinCyrillic; + break ; + case wxFONTENCODING_ISO8859_6 : + enc = kCFStringEncodingISOLatinArabic; + break ; + case wxFONTENCODING_ISO8859_7 : + enc = kCFStringEncodingISOLatinGreek; + break ; + case wxFONTENCODING_ISO8859_8 : + enc = kCFStringEncodingISOLatinHebrew; + break ; + case wxFONTENCODING_ISO8859_9 : + enc = kCFStringEncodingISOLatin5; + break ; + case wxFONTENCODING_ISO8859_10 : + enc = kCFStringEncodingISOLatin6; + break ; + case wxFONTENCODING_ISO8859_11 : + enc = kCFStringEncodingISOLatinThai; + break ; + case wxFONTENCODING_ISO8859_13 : + enc = kCFStringEncodingISOLatin7; + break ; + case wxFONTENCODING_ISO8859_14 : + enc = kCFStringEncodingISOLatin8; + break ; + case wxFONTENCODING_ISO8859_15 : + enc = kCFStringEncodingISOLatin9; + break ; + + case wxFONTENCODING_KOI8 : + enc = kCFStringEncodingKOI8_R; + break ; + case wxFONTENCODING_ALTERNATIVE : // MS-DOS CP866 + enc = kCFStringEncodingDOSRussian; + break ; + +// case wxFONTENCODING_BULGARIAN : +// enc = ; +// break ; + + case wxFONTENCODING_CP437 : + enc =kCFStringEncodingDOSLatinUS ; + break ; + case wxFONTENCODING_CP850 : + enc = kCFStringEncodingDOSLatin1; + break ; + case wxFONTENCODING_CP852 : + enc = kCFStringEncodingDOSLatin2; + break ; + case wxFONTENCODING_CP855 : + enc = kCFStringEncodingDOSCyrillic; + break ; + case wxFONTENCODING_CP866 : + enc =kCFStringEncodingDOSRussian ; + break ; + case wxFONTENCODING_CP874 : + enc = kCFStringEncodingDOSThai; + break ; + case wxFONTENCODING_CP932 : + enc = kCFStringEncodingDOSJapanese; + break ; + case wxFONTENCODING_CP936 : + enc =kCFStringEncodingDOSChineseSimplif ; + break ; + case wxFONTENCODING_CP949 : + enc = kCFStringEncodingDOSKorean; + break ; + case wxFONTENCODING_CP950 : + enc = kCFStringEncodingDOSChineseTrad; + break ; + + case wxFONTENCODING_CP1250 : + enc = kCFStringEncodingWindowsLatin2; + break ; + case wxFONTENCODING_CP1251 : + enc =kCFStringEncodingWindowsCyrillic ; + break ; + case wxFONTENCODING_CP1252 : + enc =kCFStringEncodingWindowsLatin1 ; + break ; + case wxFONTENCODING_CP1253 : + enc = kCFStringEncodingWindowsGreek; + break ; + case wxFONTENCODING_CP1254 : + enc = kCFStringEncodingWindowsLatin5; + break ; + case wxFONTENCODING_CP1255 : + enc =kCFStringEncodingWindowsHebrew ; + break ; + case wxFONTENCODING_CP1256 : + enc =kCFStringEncodingWindowsArabic ; + break ; + case wxFONTENCODING_CP1257 : + enc = kCFStringEncodingWindowsBalticRim; + break ; + case wxFONTENCODING_UTF7 : + enc = kCFStringEncodingNonLossyASCII ; + break ; + case wxFONTENCODING_UTF8 : + enc = kCFStringEncodingUTF8 ; + break ; + case wxFONTENCODING_EUC_JP : + enc = kCFStringEncodingEUC_JP; + break ; + case wxFONTENCODING_UTF16 : + enc = kCFStringEncodingUnicode ; + break ; + case wxFONTENCODING_MACROMAN : + enc = kCFStringEncodingMacRoman ; + break ; + case wxFONTENCODING_MACJAPANESE : + enc = kCFStringEncodingMacJapanese ; + break ; + case wxFONTENCODING_MACCHINESETRAD : + enc = kCFStringEncodingMacChineseTrad ; + break ; + case wxFONTENCODING_MACKOREAN : + enc = kCFStringEncodingMacKorean ; + break ; + case wxFONTENCODING_MACARABIC : + enc = kCFStringEncodingMacArabic ; + break ; + case wxFONTENCODING_MACHEBREW : + enc = kCFStringEncodingMacHebrew ; + break ; + case wxFONTENCODING_MACGREEK : + enc = kCFStringEncodingMacGreek ; + break ; + case wxFONTENCODING_MACCYRILLIC : + enc = kCFStringEncodingMacCyrillic ; + break ; + case wxFONTENCODING_MACDEVANAGARI : + enc = kCFStringEncodingMacDevanagari ; + break ; + case wxFONTENCODING_MACGURMUKHI : + enc = kCFStringEncodingMacGurmukhi ; + break ; + case wxFONTENCODING_MACGUJARATI : + enc = kCFStringEncodingMacGujarati ; + break ; + case wxFONTENCODING_MACORIYA : + enc = kCFStringEncodingMacOriya ; + break ; + case wxFONTENCODING_MACBENGALI : + enc = kCFStringEncodingMacBengali ; + break ; + case wxFONTENCODING_MACTAMIL : + enc = kCFStringEncodingMacTamil ; + break ; + case wxFONTENCODING_MACTELUGU : + enc = kCFStringEncodingMacTelugu ; + break ; + case wxFONTENCODING_MACKANNADA : + enc = kCFStringEncodingMacKannada ; + break ; + case wxFONTENCODING_MACMALAJALAM : + enc = kCFStringEncodingMacMalayalam ; + break ; + case wxFONTENCODING_MACSINHALESE : + enc = kCFStringEncodingMacSinhalese ; + break ; + case wxFONTENCODING_MACBURMESE : + enc = kCFStringEncodingMacBurmese ; + break ; + case wxFONTENCODING_MACKHMER : + enc = kCFStringEncodingMacKhmer ; + break ; + case wxFONTENCODING_MACTHAI : + enc = kCFStringEncodingMacThai ; + break ; + case wxFONTENCODING_MACLAOTIAN : + enc = kCFStringEncodingMacLaotian ; + break ; + case wxFONTENCODING_MACGEORGIAN : + enc = kCFStringEncodingMacGeorgian ; + break ; + case wxFONTENCODING_MACARMENIAN : + enc = kCFStringEncodingMacArmenian ; + break ; + case wxFONTENCODING_MACCHINESESIMP : + enc = kCFStringEncodingMacChineseSimp ; + break ; + case wxFONTENCODING_MACTIBETAN : + enc = kCFStringEncodingMacTibetan ; + break ; + case wxFONTENCODING_MACMONGOLIAN : + enc = kCFStringEncodingMacMongolian ; + break ; + case wxFONTENCODING_MACETHIOPIC : + enc = kCFStringEncodingMacEthiopic ; + break ; + case wxFONTENCODING_MACCENTRALEUR : + enc = kCFStringEncodingMacCentralEurRoman ; + break ; + case wxFONTENCODING_MACVIATNAMESE : + enc = kCFStringEncodingMacVietnamese ; + break ; + case wxFONTENCODING_MACARABICEXT : + enc = kCFStringEncodingMacExtArabic ; + break ; + case wxFONTENCODING_MACSYMBOL : + enc = kCFStringEncodingMacSymbol ; + break ; + case wxFONTENCODING_MACDINGBATS : + enc = kCFStringEncodingMacDingbats ; + break ; + case wxFONTENCODING_MACTURKISH : + enc = kCFStringEncodingMacTurkish ; + break ; + case wxFONTENCODING_MACCROATIAN : + enc = kCFStringEncodingMacCroatian ; + break ; + case wxFONTENCODING_MACICELANDIC : + enc = kCFStringEncodingMacIcelandic ; + break ; + case wxFONTENCODING_MACROMANIAN : + enc = kCFStringEncodingMacRomanian ; + break ; + case wxFONTENCODING_MACCELTIC : + enc = kCFStringEncodingMacCeltic ; + break ; + case wxFONTENCODING_MACGAELIC : + enc = kCFStringEncodingMacGaelic ; + break ; +// case wxFONTENCODING_MACKEYBOARD : +// enc = kCFStringEncodingMacKeyboardGlyphs ; +// break ; + default : + // because gcc is picky + break ; + } ; + return enc ; +} + +wxFontEncoding wxFontEncFromCFStringEnc(CFStringEncoding encoding) +{ + wxFontEncoding enc = wxFONTENCODING_DEFAULT ; + + switch( encoding) + { + case kCFStringEncodingISOLatin1 : + enc = wxFONTENCODING_ISO8859_1 ; + break ; + case kCFStringEncodingISOLatin2 : + enc = wxFONTENCODING_ISO8859_2; + break ; + case kCFStringEncodingISOLatin3 : + enc = wxFONTENCODING_ISO8859_3 ; + break ; + case kCFStringEncodingISOLatin4 : + enc = wxFONTENCODING_ISO8859_4; + break ; + case kCFStringEncodingISOLatinCyrillic : + enc = wxFONTENCODING_ISO8859_5; + break ; + case kCFStringEncodingISOLatinArabic : + enc = wxFONTENCODING_ISO8859_6; + break ; + case kCFStringEncodingISOLatinGreek : + enc = wxFONTENCODING_ISO8859_7; + break ; + case kCFStringEncodingISOLatinHebrew : + enc = wxFONTENCODING_ISO8859_8; + break ; + case kCFStringEncodingISOLatin5 : + enc = wxFONTENCODING_ISO8859_9; + break ; + case kCFStringEncodingISOLatin6 : + enc = wxFONTENCODING_ISO8859_10; + break ; + case kCFStringEncodingISOLatin7 : + enc = wxFONTENCODING_ISO8859_13; + break ; + case kCFStringEncodingISOLatin8 : + enc = wxFONTENCODING_ISO8859_14; + break ; + case kCFStringEncodingISOLatin9 : + enc =wxFONTENCODING_ISO8859_15 ; + break ; + + case kCFStringEncodingKOI8_R : + enc = wxFONTENCODING_KOI8; + break ; + +// case : +// enc = wxFONTENCODING_BULGARIAN; +// break ; + + case kCFStringEncodingDOSLatinUS : + enc = wxFONTENCODING_CP437; + break ; + case kCFStringEncodingDOSLatin1 : + enc = wxFONTENCODING_CP850; + break ; + case kCFStringEncodingDOSLatin2 : + enc =wxFONTENCODING_CP852 ; + break ; + case kCFStringEncodingDOSCyrillic : + enc = wxFONTENCODING_CP855; + break ; + case kCFStringEncodingDOSRussian : + enc = wxFONTENCODING_CP866; + break ; + case kCFStringEncodingDOSThai : + enc =wxFONTENCODING_CP874 ; + break ; + case kCFStringEncodingDOSJapanese : + enc = wxFONTENCODING_CP932; + break ; + case kCFStringEncodingDOSChineseSimplif : + enc = wxFONTENCODING_CP936; + break ; + case kCFStringEncodingDOSKorean : + enc = wxFONTENCODING_CP949; + break ; + case kCFStringEncodingDOSChineseTrad : + enc = wxFONTENCODING_CP950; + break ; + + case kCFStringEncodingWindowsLatin2 : + enc = wxFONTENCODING_CP1250; + break ; + case kCFStringEncodingWindowsCyrillic : + enc = wxFONTENCODING_CP1251; + break ; + case kCFStringEncodingWindowsLatin1 : + enc = wxFONTENCODING_CP1252; + break ; + case kCFStringEncodingWindowsGreek : + enc = wxFONTENCODING_CP1253; + break ; + case kCFStringEncodingWindowsLatin5 : + enc = wxFONTENCODING_CP1254; + break ; + case kCFStringEncodingWindowsHebrew : + enc = wxFONTENCODING_CP1255; + break ; + case kCFStringEncodingWindowsArabic : + enc = wxFONTENCODING_CP1256; + break ; + case kCFStringEncodingWindowsBalticRim : + enc =wxFONTENCODING_CP1257 ; + break ; + case kCFStringEncodingEUC_JP : + enc = wxFONTENCODING_EUC_JP; + break ; + case kCFStringEncodingUnicode : + enc = wxFONTENCODING_UTF16; + break; + case kCFStringEncodingMacRoman : + enc = wxFONTENCODING_MACROMAN ; + break ; + case kCFStringEncodingMacJapanese : + enc = wxFONTENCODING_MACJAPANESE ; + break ; + case kCFStringEncodingMacChineseTrad : + enc = wxFONTENCODING_MACCHINESETRAD ; + break ; + case kCFStringEncodingMacKorean : + enc = wxFONTENCODING_MACKOREAN ; + break ; + case kCFStringEncodingMacArabic : + enc =wxFONTENCODING_MACARABIC ; + break ; + case kCFStringEncodingMacHebrew : + enc = wxFONTENCODING_MACHEBREW ; + break ; + case kCFStringEncodingMacGreek : + enc = wxFONTENCODING_MACGREEK ; + break ; + case kCFStringEncodingMacCyrillic : + enc = wxFONTENCODING_MACCYRILLIC ; + break ; + case kCFStringEncodingMacDevanagari : + enc = wxFONTENCODING_MACDEVANAGARI ; + break ; + case kCFStringEncodingMacGurmukhi : + enc = wxFONTENCODING_MACGURMUKHI ; + break ; + case kCFStringEncodingMacGujarati : + enc = wxFONTENCODING_MACGUJARATI ; + break ; + case kCFStringEncodingMacOriya : + enc =wxFONTENCODING_MACORIYA ; + break ; + case kCFStringEncodingMacBengali : + enc =wxFONTENCODING_MACBENGALI ; + break ; + case kCFStringEncodingMacTamil : + enc = wxFONTENCODING_MACTAMIL ; + break ; + case kCFStringEncodingMacTelugu : + enc = wxFONTENCODING_MACTELUGU ; + break ; + case kCFStringEncodingMacKannada : + enc = wxFONTENCODING_MACKANNADA ; + break ; + case kCFStringEncodingMacMalayalam : + enc = wxFONTENCODING_MACMALAJALAM ; + break ; + case kCFStringEncodingMacSinhalese : + enc = wxFONTENCODING_MACSINHALESE ; + break ; + case kCFStringEncodingMacBurmese : + enc = wxFONTENCODING_MACBURMESE ; + break ; + case kCFStringEncodingMacKhmer : + enc = wxFONTENCODING_MACKHMER ; + break ; + case kCFStringEncodingMacThai : + enc = wxFONTENCODING_MACTHAI ; + break ; + case kCFStringEncodingMacLaotian : + enc = wxFONTENCODING_MACLAOTIAN ; + break ; + case kCFStringEncodingMacGeorgian : + enc = wxFONTENCODING_MACGEORGIAN ; + break ; + case kCFStringEncodingMacArmenian : + enc = wxFONTENCODING_MACARMENIAN ; + break ; + case kCFStringEncodingMacChineseSimp : + enc = wxFONTENCODING_MACCHINESESIMP ; + break ; + case kCFStringEncodingMacTibetan : + enc = wxFONTENCODING_MACTIBETAN ; + break ; + case kCFStringEncodingMacMongolian : + enc = wxFONTENCODING_MACMONGOLIAN ; + break ; + case kCFStringEncodingMacEthiopic : + enc = wxFONTENCODING_MACETHIOPIC ; + break ; + case kCFStringEncodingMacCentralEurRoman: + enc = wxFONTENCODING_MACCENTRALEUR ; + break ; + case kCFStringEncodingMacVietnamese: + enc = wxFONTENCODING_MACVIATNAMESE ; + break ; + case kCFStringEncodingMacExtArabic : + enc = wxFONTENCODING_MACARABICEXT ; + break ; + case kCFStringEncodingMacSymbol : + enc = wxFONTENCODING_MACSYMBOL ; + break ; + case kCFStringEncodingMacDingbats : + enc = wxFONTENCODING_MACDINGBATS ; + break ; + case kCFStringEncodingMacTurkish : + enc = wxFONTENCODING_MACTURKISH ; + break ; + case kCFStringEncodingMacCroatian : + enc = wxFONTENCODING_MACCROATIAN ; + break ; + case kCFStringEncodingMacIcelandic : + enc = wxFONTENCODING_MACICELANDIC ; + break ; + case kCFStringEncodingMacRomanian : + enc = wxFONTENCODING_MACROMANIAN ; + break ; + case kCFStringEncodingMacCeltic : + enc = wxFONTENCODING_MACCELTIC ; + break ; + case kCFStringEncodingMacGaelic : + enc = wxFONTENCODING_MACGAELIC ; + break ; +// case kCFStringEncodingMacKeyboardGlyphs : +// enc = wxFONTENCODING_MACKEYBOARD ; +// break ; + } ; + return enc ; +} + +class wxMBConv_cocoa : public wxMBConv +{ +public: + wxMBConv_cocoa() + { + Init(CFStringGetSystemEncoding()) ; + } + + wxMBConv_cocoa(const wxChar* name) + { + Init( wxCFStringEncFromFontEnc(wxFontMapper::Get()->CharsetToEncoding(name, false) ) ) ; + } + + wxMBConv_cocoa(wxFontEncoding encoding) + { + Init( wxCFStringEncFromFontEnc(encoding) ); + } + + ~wxMBConv_cocoa() + { + } + + void Init( CFStringEncoding encoding) + { + m_char_encoding = encoding ; + m_unicode_encoding = kCFStringEncodingUnicode; + } + + size_t MB2WC(wchar_t * szOut, const char * szUnConv, size_t nOutSize) const + { + wxASSERT(szUnConv); + + size_t nBufSize = strlen(szUnConv) + 1; + size_t nRealOutSize; + + UniChar* szUniCharBuffer = (UniChar*) szOut; + wchar_t* szConvBuffer = szOut; + + if (szConvBuffer == NULL && nOutSize != 0) + { + szConvBuffer = new wchar_t[nOutSize] ; + } + +#if SIZEOF_WCHAR_T == 4 + szUniCharBuffer = new UniChar[nOutSize]; +#endif + + CFDataRef theData = CFDataCreateWithBytesNoCopy ( + NULL, //allocator + (const UInt8*)szUnConv, + nBufSize - 1, + NULL //deallocator + ); + + wxASSERT(theData); + + CFStringRef theString = CFStringCreateFromExternalRepresentation ( + NULL, + theData, + m_char_encoding + ); + + wxASSERT(theString); + + if (nOutSize == 0) + { + nRealOutSize = CFStringGetLength(theString) + 1; + CFRelease(theString); + return nRealOutSize - 1; + } + + CFRange theRange = { 0, CFStringGetLength(theString) }; + + CFStringGetCharacters(theString, theRange, szUniCharBuffer); + + + nRealOutSize = (CFStringGetLength(theString) + 1); + + CFRelease(theString); + + szUniCharBuffer[nRealOutSize-1] = '\0' ; + +#if SIZEOF_WCHAR_T == 4 + wxMBConvUTF16 converter ; + converter.MB2WC(szConvBuffer , (const char*)szUniCharBuffer , nRealOutSize ) ; + delete[] szUniCharBuffer; +#endif + if ( szOut == NULL ) + delete [] szConvBuffer; + + return nRealOutSize ; + } + + size_t WC2MB(char *szOut, const wchar_t *szUnConv, size_t nOutSize) const + { + size_t nBufSize = wxWcslen(szUnConv) + 1; + size_t nRealOutSize; + char* szBuffer = szOut; + UniChar* szUniBuffer = (UniChar*) szUnConv; + + if (szOut == NULL) + { + // worst case + nRealOutSize = wxString::WorstEncodingCase(nBufSize - 1, *this)+1 ; + szBuffer = new char[ nRealOutSize ] ; + } + else + nRealOutSize = nOutSize; + +#if SIZEOF_WCHAR_T == 4 + wxMBConvUTF16BE converter ; + nBufSize = converter.WC2MB( NULL , szUnConv , 0 ); + szUniBuffer = new UniChar[ (nBufSize / sizeof(UniChar)) + 1] ; + converter.WC2MB( (char*) szUniBuffer , szUnConv, nBufSize + sizeof(UniChar)) ; + nBufSize /= sizeof(UniChar); + ++nBufSize; +#endif + + CFStringRef theString = CFStringCreateWithCharactersNoCopy( + NULL, //allocator + szUniBuffer, + nBufSize, + NULL //deallocator + ); + + wxASSERT(theString); + + //Note that CER puts a BOM when converting to unicode + //so we may want to check and use getchars instead in that case + CFDataRef theData = CFStringCreateExternalRepresentation( + NULL, //allocator + theString, + m_char_encoding, + 0 //what to put in characters that can't be converted - + //0 tells CFString to return NULL if it meets such a character + ); + + if(!theData) + return (size_t)-1; + + CFRelease(theString); + + nRealOutSize = CFDataGetLength(theData); + + if ( szOut == NULL ) + delete[] szBuffer; + + if(nOutSize == 0) + { +//TODO: This gets flagged as a non-malloced address by the debugger... +//#if SIZEOF_WCHAR_T == 4 +// delete[] szUniBuffer; +//#endif + CFRelease(theData); + return nRealOutSize - 1; + } + + CFRange theRange = {0, CFDataGetLength(theData) }; + CFDataGetBytes(theData, theRange, (UInt8*) szBuffer); + + CFRelease(theData); + +//TODO: This gets flagged as a non-malloced address by the debugger... +//#if SIZEOF_WCHAR_T == 4 +// delete[] szUniBuffer; +//#endif + return nRealOutSize - 1; + } + + bool IsOk() const + { + //TODO: check for invalid en/de/coding + return true; + } + +private: + CFStringEncoding m_char_encoding ; + CFStringEncoding m_unicode_encoding ; +}; + +#endif // defined(__WXCOCOA__) + +// ============================================================================ +// Mac conversion classes +// ============================================================================ + +#if defined(__WXMAC__) && defined(TARGET_CARBON) + +class wxMBConv_mac : public wxMBConv +{ +public: + wxMBConv_mac() + { + Init(CFStringGetSystemEncoding()) ; + } + + wxMBConv_mac(const wxChar* name) + { + Init( wxMacGetSystemEncFromFontEnc(wxFontMapper::Get()->CharsetToEncoding(name, false) ) ) ; + } + + wxMBConv_mac(wxFontEncoding encoding) + { + Init( wxMacGetSystemEncFromFontEnc(encoding) ); + } + + ~wxMBConv_mac() + { + OSStatus status = noErr ; + status = TECDisposeConverter(m_MB2WC_converter); + status = TECDisposeConverter(m_WC2MB_converter); + } + + + void Init( TextEncodingBase encoding) + { + OSStatus status = noErr ; + m_char_encoding = encoding ; + m_unicode_encoding = CreateTextEncoding(kTextEncodingUnicodeDefault,0,kUnicode16BitFormat) ; + + status = TECCreateConverter(&m_MB2WC_converter, + m_char_encoding, + m_unicode_encoding); + status = TECCreateConverter(&m_WC2MB_converter, + m_unicode_encoding, + m_char_encoding); + } + + size_t MB2WC(wchar_t *buf, const char *psz, size_t n) const + { + OSStatus status = noErr ; + ByteCount byteOutLen ; + ByteCount byteInLen = strlen(psz) ; + wchar_t *tbuf = NULL ; + UniChar* ubuf = NULL ; + size_t res = 0 ; + + if (buf == NULL) + { + n = 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); +#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 ; + wxMBConvUTF16BE converter ; + res = converter.MB2WC( (buf ? buf : tbuf) , (const char*)ubuf , n ) ; + free( ubuf ) ; +#else + res = byteOutLen / sizeof( UniChar ) ; +#endif + if ( buf == NULL ) + free(tbuf) ; + + if ( buf && res < n) + buf[res] = 0; + + return res ; + } + + size_t WC2MB(char *buf, const wchar_t *psz, size_t n) const + { + OSStatus status = noErr ; + ByteCount byteOutLen ; + ByteCount byteInLen = wxWcslen(psz) * SIZEOF_WCHAR_T ; + + char *tbuf = NULL ; + + if (buf == NULL) + { + // worst case + n = wxString::WorstEncodingCase(byteInLen / SIZEOF_WCHAR_T, *this) + SIZEOF_WCHAR_T; + tbuf = (char*) malloc( n ) ; + } + + ByteCount byteBufferLen = n ; + UniChar* ubuf = NULL ; +#if SIZEOF_WCHAR_T == 4 + wxMBConvUTF16BE converter ; + size_t unicharlen = converter.WC2MB( NULL , psz , 0 ) ; + byteInLen = unicharlen ; + ubuf = (UniChar*) malloc( byteInLen + 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); +#if SIZEOF_WCHAR_T == 4 + free( ubuf ) ; +#endif + if ( buf == NULL ) + free(tbuf) ; + + size_t res = byteOutLen ; + if ( buf && res < n) + buf[res] = 0; + + return res ; + } + + bool IsOk() const + { return m_MB2WC_converter != NULL && m_WC2MB_converter != NULL ; } + +private: + TECObjectRef m_MB2WC_converter ; + TECObjectRef m_WC2MB_converter ; + + TextEncodingBase m_char_encoding ; + TextEncodingBase m_unicode_encoding ; +}; + +#endif // defined(__WXMAC__) && defined(TARGET_CARBON) + +// ============================================================================ +// wxEncodingConverter based conversion classes +// ============================================================================ + +#if wxUSE_FONTMAP + +class wxMBConv_wxwin : public wxMBConv +{ +private: + void Init() + { + m_ok = m2w.Init(m_enc, wxFONTENCODING_UNICODE) && + w2m.Init(wxFONTENCODING_UNICODE, m_enc); + } + +public: + // temporarily just use wxEncodingConverter stuff, + // so that it works while a better implementation is built + wxMBConv_wxwin(const wxChar* name) { if (name) - enc = wxTheFontMapper->CharsetToEncoding(name, FALSE); - m2w.Init(enc, wxFONTENCODING_UNICODE); - w2m.Init(wxFONTENCODING_UNICODE, enc); + m_enc = wxFontMapper::Get()->CharsetToEncoding(name, false); + else + m_enc = wxFONTENCODING_SYSTEM; + + Init(); + } + + wxMBConv_wxwin(wxFontEncoding enc) + { + m_enc = enc; + + Init(); } - - size_t MB2WC(wchar_t*buf, const char*psz, size_t n) + + size_t MB2WC(wchar_t *buf, const char *psz, size_t WXUNUSED(n)) const { size_t inbuf = strlen(psz); - fprintf(stderr,"EC Convert to WC using %d\n",enc); - if (buf) m2w.Convert(psz,buf); + if (buf) + m2w.Convert(psz,buf); return inbuf; } - - size_t WC2MB(char*buf, const wchar_t*psz, size_t n) + + size_t WC2MB(char *buf, const wchar_t *psz, size_t WXUNUSED(n)) const { -#if defined(__BORLANDC__) && (__BORLANDC__ > 0x530) - size_t inbuf = std::wcslen(psz); -#else - size_t inbuf = ::wcslen(psz); -#endif - fprintf(stderr,"EC Convert from WC using %d\n",enc); + const size_t inbuf = wxWcslen(psz); if (buf) w2m.Convert(psz,buf); - + return inbuf; } - - bool usable() - { return (enc!=wxFONTENCODING_SYSTEM) && (enc!=wxFONTENCODING_DEFAULT); } + + bool IsOk() const { return m_ok; } public: - wxFontEncoding enc; + wxFontEncoding m_enc; wxEncodingConverter m2w, w2m; + + // were we initialized successfully? + bool m_ok; + + DECLARE_NO_COPY_CLASS(wxMBConv_wxwin) }; -static wxCharacterSet *wxGetCharacterSet(const wxChar *name) +#endif // wxUSE_FONTMAP + +// ============================================================================ +// wxCSConv implementation +// ============================================================================ + +void wxCSConv::Init() +{ + m_name = NULL; + m_convReal = NULL; + m_deferred = true; +} + +wxCSConv::wxCSConv(const wxChar *charset) { - wxCharacterSet *cset = NULL; - if (name) + Init(); + + if ( charset ) { - if (!wxStricmp(name, wxT("UTF8")) || !wxStricmp(name, wxT("UTF-8"))) - { - cset = new ID_CharSet(name, &wxConvUTF8); - } - else - { -#ifdef HAVE_ICONV_H - cset = new IC_CharSet(name); // may not take NULL -#endif - } + SetName(charset); } - - if (cset && cset->usable()) return cset; - if (cset) delete cset; -#ifdef __WIN32__ - cset = new CP_CharSet(name); // may take NULL - if (cset->usable()) return cset; -#endif - if (cset) delete cset; - cset = new EC_CharSet(name); - if (cset->usable()) return cset; - delete cset; - return NULL; + + m_encoding = wxFONTENCODING_SYSTEM; } -wxCSConv::wxCSConv(const wxChar *charset) +wxCSConv::wxCSConv(wxFontEncoding encoding) { - m_name = (wxChar *) NULL; - m_cset = (wxCharacterSet *) NULL; - m_deferred = TRUE; - SetName(charset); + if ( encoding == wxFONTENCODING_MAX || encoding == wxFONTENCODING_DEFAULT ) + { + wxFAIL_MSG( _T("invalid encoding value in wxCSConv ctor") ); + + encoding = wxFONTENCODING_SYSTEM; + } + + Init(); + + m_encoding = encoding; } wxCSConv::~wxCSConv() { - if (m_name) free(m_name); - if (m_cset) delete m_cset; + Clear(); +} + +wxCSConv::wxCSConv(const wxCSConv& conv) + : wxMBConv() +{ + Init(); + + SetName(conv.m_name); + m_encoding = conv.m_encoding; +} + +wxCSConv& wxCSConv::operator=(const wxCSConv& conv) +{ + Clear(); + + SetName(conv.m_name); + m_encoding = conv.m_encoding; + + return *this; +} + +void wxCSConv::Clear() +{ + free(m_name); + delete m_convReal; + + m_name = NULL; + m_convReal = NULL; } void wxCSConv::SetName(const wxChar *charset) @@ -671,138 +2441,269 @@ void wxCSConv::SetName(const wxChar *charset) if (charset) { m_name = wxStrdup(charset); - m_deferred = TRUE; + m_deferred = true; } } -void wxCSConv::LoadNow() +wxMBConv *wxCSConv::DoCreate() const { -// wxPrintf(wxT("Conversion request\n")); - if (m_deferred) + // check for the special case of ASCII or ISO8859-1 charset: as we have + // special knowledge of it anyhow, we don't need to create a special + // conversion object + if ( m_encoding == wxFONTENCODING_ISO8859_1 ) + { + // don't convert at all + return NULL; + } + + // we trust OS to do conversion better than we can so try external + // conversion methods first + // + // the full order is: + // 1. OS conversion (iconv() under Unix or Win32 API) + // 2. hard coded conversions for UTF + // 3. wxEncodingConverter as fall back + + // step (1) +#ifdef HAVE_ICONV +#if !wxUSE_FONTMAP + if ( m_name ) +#endif // !wxUSE_FONTMAP + { + wxString name(m_name); + +#if wxUSE_FONTMAP + if ( name.empty() ) + name = wxFontMapper::Get()->GetEncodingName(m_encoding); +#endif // wxUSE_FONTMAP + + wxMBConv_iconv *conv = new wxMBConv_iconv(name); + if ( conv->IsOk() ) + return conv; + + delete conv; + } +#endif // HAVE_ICONV + +#ifdef wxHAVE_WIN32_MB2WC + { +#if wxUSE_FONTMAP + wxMBConv_win32 *conv = m_name ? new wxMBConv_win32(m_name) + : new wxMBConv_win32(m_encoding); + if ( conv->IsOk() ) + return conv; + + delete conv; +#else + return NULL; +#endif + } +#endif // wxHAVE_WIN32_MB2WC +#if defined(__WXMAC__) { - if (!m_name) + if ( m_name || ( m_encoding < wxFONTENCODING_UTF16BE ) ) { -#ifdef __UNIX__ -#if defined(HAVE_LANGINFO_H) && defined(CODESET) - // GNU libc provides current character set this way - char*alang = nl_langinfo(CODESET); - if (alang) - { - SetName(wxConvLibc.cMB2WX(alang)); - } - else + + wxMBConv_mac *conv = m_name ? new wxMBConv_mac(m_name) + : new wxMBConv_mac(m_encoding); + if ( conv->IsOk() ) + return conv; + + delete conv; + } + } #endif - { - // if we can't get at the character set directly, - // try to see if it's in the environment variables - // (in most cases this won't work, but I was out of ideas) - wxChar *lang = wxGetenv(wxT("LC_ALL")); - if (!lang) lang = wxGetenv(wxT("LC_CTYPE")); - if (!lang) lang = wxGetenv(wxT("LANG")); - wxChar *dot = lang ? wxStrchr(lang, wxT('.')) : (wxChar *)NULL; - if (dot) SetName(dot+1); - } +#if defined(__WXCOCOA__) + { + if ( m_name || ( m_encoding <= wxFONTENCODING_UTF16 ) ) + { + + wxMBConv_cocoa *conv = m_name ? new wxMBConv_cocoa(m_name) + : new wxMBConv_cocoa(m_encoding); + if ( conv->IsOk() ) + return conv; + + delete conv; + } + } #endif + // step (2) + wxFontEncoding enc = m_encoding; +#if wxUSE_FONTMAP + if ( enc == wxFONTENCODING_SYSTEM && m_name ) + { + // use "false" to suppress interactive dialogs -- we can be called from + // anywhere and popping up a dialog from here is the last thing we want to + // do + enc = wxFontMapper::Get()->CharsetToEncoding(m_name, false); + } +#endif // wxUSE_FONTMAP + + switch ( enc ) + { + case wxFONTENCODING_UTF7: + return new wxMBConvUTF7; + + case wxFONTENCODING_UTF8: + return new wxMBConvUTF8; + + case wxFONTENCODING_UTF16BE: + return new wxMBConvUTF16BE; + + case wxFONTENCODING_UTF16LE: + return new wxMBConvUTF16LE; + + case wxFONTENCODING_UTF32BE: + return new wxMBConvUTF32BE; + + case wxFONTENCODING_UTF32LE: + return new wxMBConvUTF32LE; + + default: + // nothing to do but put here to suppress gcc warnings + ; + } + + // step (3) +#if wxUSE_FONTMAP + { + wxMBConv_wxwin *conv = m_name ? new wxMBConv_wxwin(m_name) + : new wxMBConv_wxwin(m_encoding); + if ( conv->IsOk() ) + return conv; + + delete conv; + } +#endif // wxUSE_FONTMAP + + // NB: This is a hack to prevent deadlock. What could otherwise happen + // in Unicode build: wxConvLocal creation ends up being here + // because of some failure and logs the error. But wxLog will try to + // attach timestamp, for which it will need wxConvLocal (to convert + // time to char* and then wchar_t*), but that fails, tries to log + // error, but wxLog has a (already locked) critical section that + // guards static buffer. + static bool alreadyLoggingError = false; + if (!alreadyLoggingError) + { + alreadyLoggingError = true; + wxLogError(_("Cannot convert from the charset '%s'!"), + m_name ? m_name + : +#if wxUSE_FONTMAP + wxFontMapper::GetEncodingDescription(m_encoding).c_str() +#else // !wxUSE_FONTMAP + wxString::Format(_("encoding %s"), m_encoding).c_str() +#endif // wxUSE_FONTMAP/!wxUSE_FONTMAP + ); + alreadyLoggingError = false; + } + + return NULL; +} + +void wxCSConv::CreateConvIfNeeded() const +{ + if ( m_deferred ) + { + wxCSConv *self = (wxCSConv *)this; // const_cast + +#if wxUSE_INTL + // if we don't have neither the name nor the encoding, use the default + // encoding for this system + if ( !m_name && m_encoding == wxFONTENCODING_SYSTEM ) + { + self->m_name = wxStrdup(wxLocale::GetSystemEncodingName()); } - m_cset = wxGetCharacterSet(m_name); - m_deferred = FALSE; +#endif // wxUSE_INTL + + self->m_convReal = DoCreate(); + self->m_deferred = false; } } size_t wxCSConv::MB2WC(wchar_t *buf, const char *psz, size_t n) const { - ((wxCSConv *)this)->LoadNow(); // discard constness - - if (m_cset) - return m_cset->MB2WC(buf, psz, n); + CreateConvIfNeeded(); + + if (m_convReal) + return m_convReal->MB2WC(buf, psz, n); // latin-1 (direct) - size_t len=strlen(psz); - + size_t len = strlen(psz); + if (buf) { - for (size_t c=0; c<=len; c++) + for (size_t c = 0; c <= len; c++) buf[c] = (unsigned char)(psz[c]); } - + return len; } size_t wxCSConv::WC2MB(char *buf, const wchar_t *psz, size_t n) const { - ((wxCSConv *)this)->LoadNow(); // discard constness - - if (m_cset) - return m_cset->WC2MB(buf, psz, n); + CreateConvIfNeeded(); + + if (m_convReal) + return m_convReal->WC2MB(buf, psz, n); // latin-1 (direct) -#if defined(__BORLANDC__) && (__BORLANDC__ > 0x530) - size_t len=std::wcslen(psz); -#else - size_t len=::wcslen(psz); -#endif + const size_t len = wxWcslen(psz); if (buf) { - for (size_t c=0; c<=len; c++) - buf[c] = (psz[c]>0xff) ? '?' : psz[c]; + for (size_t c = 0; c <= len; c++) + { + if (psz[c] > 0xFF) + return (size_t)-1; + buf[c] = (char)psz[c]; + } + } + else + { + for (size_t c = 0; c <= len; c++) + { + if (psz[c] > 0xFF) + return (size_t)-1; + } } - + return len; } -#ifdef HAVE_ICONV_H -class IC_CharSetConverter -{ -public: - IC_CharSetConverter(IC_CharSet*from,IC_CharSet*to) - { cnv = iconv_open(wxConvLibc.cWX2MB(to->cname),wxConvLibc.cWX2MB(from->cname)); } - - ~IC_CharSetConverter() - { if (cnv!=(iconv_t)-1) iconv_close(cnv); } - - size_t Convert(char*buf, const char*psz, size_t n) - { - size_t inbuf = strlen(psz); - size_t outbuf = n; -#ifdef WX_ICONV_TAKES_CHAR - size_t res = iconv( cnv, (char**)&psz, &inbuf, &buf, &outbuf ); +// ---------------------------------------------------------------------------- +// globals +// ---------------------------------------------------------------------------- + +#ifdef __WINDOWS__ + static wxMBConv_win32 wxConvLibcObj; +#elif defined(__WXMAC__) && !defined(__MACH__) + static wxMBConv_mac wxConvLibcObj ; #else - size_t res = iconv( cnv, &psz, &inbuf, &buf, &outbuf ); + static wxMBConvLibc wxConvLibcObj; #endif - if (res==(size_t)-1) return (size_t)-1; - return n-outbuf; - } -public: - iconv_t cnv; -}; -#endif +static wxCSConv wxConvLocalObj(wxFONTENCODING_SYSTEM); +static wxCSConv wxConvISO8859_1Obj(wxFONTENCODING_ISO8859_1); +static wxMBConvUTF7 wxConvUTF7Obj; +static wxMBConvUTF8 wxConvUTF8Obj; -class EC_CharSetConverter -{ -public: - EC_CharSetConverter(EC_CharSet*from,EC_CharSet*to) - { cnv.Init(from->enc,to->enc); } - - size_t Convert(char*buf, const char*psz, size_t n) - { - size_t inbuf = strlen(psz); - if (buf) cnv.Convert(psz,buf); - return inbuf; - } - -public: - wxEncodingConverter cnv; -}; + +WXDLLIMPEXP_DATA_BASE(wxMBConv&) wxConvLibc = wxConvLibcObj; +WXDLLIMPEXP_DATA_BASE(wxCSConv&) wxConvLocal = wxConvLocalObj; +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; #else // !wxUSE_WCHAR_T -// ---------------------------------------------------------------------------- // stand-ins in absence of wchar_t -// ---------------------------------------------------------------------------- - -WXDLLEXPORT_DATA(wxMBConv) wxConvLibc, wxConvFile; +WXDLLIMPEXP_DATA_BASE(wxMBConv) wxConvLibc, + wxConvISO8859_1, + wxConvLocal, + wxConvUTF8; -#endif // wxUSE_WCHAR_T +#endif // wxUSE_WCHAR_T/!wxUSE_WCHAR_T