]> git.saurik.com Git - wxWidgets.git/blobdiff - src/common/strconv.cpp
Committing in .
[wxWidgets.git] / src / common / strconv.cpp
index 997899e769822839b58763a49ff0bdb9ed55bad8..82656dd1d26c83f019a210ce196a76a2ee74f2a8 100644 (file)
@@ -1,11 +1,11 @@
 /////////////////////////////////////////////////////////////////////////////
 // Name:        strconv.cpp
 // Purpose:     Unicode conversion classes
 /////////////////////////////////////////////////////////////////////////////
 // Name:        strconv.cpp
 // Purpose:     Unicode conversion classes
-// Author:      Ove Kaaven, Robert Roebling, Vadim Zeitlin
+// Author:      Ove Kaaven, Robert Roebling, Vadim Zeitlin, Vaclav Slavik
 // Modified by:
 // Created:     29/01/98
 // RCS-ID:      $Id$
 // Modified by:
 // Created:     29/01/98
 // RCS-ID:      $Id$
-// Copyright:   (c) 1999 Ove Kaaven, Robert Roebling, Vadim Zeitlin
+// Copyright:   (c) 1999 Ove Kaaven, Robert Roebling, Vadim Zeitlin, Vaclav Slavik
 // Licence:     wxWindows license
 /////////////////////////////////////////////////////////////////////////////
 
 // Licence:     wxWindows license
 /////////////////////////////////////////////////////////////////////////////
 
   #pragma hdrstop
 #endif
 
   #pragma hdrstop
 #endif
 
+#ifdef __WXMSW__
+  #include "wx/msw/private.h"
+#endif
+
+#include <errno.h>
 #include <ctype.h>
 #include <string.h>
 #include <stdlib.h>
 
 #include <ctype.h>
 #include <string.h>
 #include <stdlib.h>
 
-#ifdef __SALFORDC__
-  #include <clib.h>
-#endif
 
 #include "wx/debug.h"
 #include "wx/strconv.h"
 
 #include "wx/debug.h"
 #include "wx/strconv.h"
+#include "wx/intl.h"
+#include "wx/log.h"
 
 // ----------------------------------------------------------------------------
 // globals
 
 // ----------------------------------------------------------------------------
 // globals
 
 WXDLLEXPORT_DATA(wxMBConv *) wxConvCurrent = &wxConvLibc;
 
 
 WXDLLEXPORT_DATA(wxMBConv *) wxConvCurrent = &wxConvLibc;
 
+
 // ============================================================================
 // implementation
 // ============================================================================
 
 #if wxUSE_WCHAR_T
 
 // ============================================================================
 // implementation
 // ============================================================================
 
 #if wxUSE_WCHAR_T
 
+#ifdef __SALFORDC__
+  #include <clib.h>
+#endif
+
+#ifdef HAVE_ICONV_H
+  #include <iconv.h>
+#endif
+
+#ifdef __WXMSW__
+  #include <windows.h>
+#endif
+
+#define BSWAP_UCS4(str, len) { unsigned _c; for (_c=0; _c<len; _c++) str[_c]=wxUINT32_SWAP_ALWAYS(str[_c]); }
+#define BSWAP_UTF16(str, len) { unsigned _c; for (_c=0; _c<len; _c++) str[_c]=wxUINT16_SWAP_ALWAYS(str[_c]); }
+
+// under Unix SIZEOF_WCHAR_T is defined by configure, but under other platforms
+// it might be not defined - assume the most common value
+#ifndef SIZEOF_WCHAR_T
+    #define SIZEOF_WCHAR_T 2
+#endif // !defined(SIZEOF_WCHAR_T)
+
+#if SIZEOF_WCHAR_T == 4
+    #define WC_NAME         "UCS4"
+    #define WC_BSWAP         BSWAP_UCS4
+    #ifdef WORDS_BIGENDIAN
+      #define WC_NAME_BEST  "UCS-4BE"
+    #else
+      #define WC_NAME_BEST  "UCS-4LE"
+    #endif
+#elif SIZEOF_WCHAR_T == 2
+    #define WC_NAME         "UTF16"
+    #define WC_BSWAP         BSWAP_UTF16
+    #define WC_UTF16
+    #ifdef WORDS_BIGENDIAN
+      #define WC_NAME_BEST  "UTF-16BE"
+    #else
+      #define WC_NAME_BEST  "UTF-16LE"
+    #endif
+#else // sizeof(wchar_t) != 2 nor 4
+    // I don't know what to do about this
+    #error "Weird sizeof(wchar_t): please report your platform details to wx-users mailing list"
+#endif
+
+
+#ifdef WC_UTF16
+
+static size_t encode_utf16(wxUint32 input, wchar_t *output)
+{
+    if (input<=0xffff)
+    {
+        if (output) *output++ = input;
+        return 1;
+    }
+    else if (input>=0x110000)
+    {
+        return (size_t)-1;
+    }
+    else
+    {
+        if (output)
+        {
+            *output++ = (input >> 10)+0xd7c0;
+            *output++ = (input&0x3ff)+0xdc00;
+        }
+        return 2;
+    }
+}
+
+static size_t decode_utf16(const wchar_t* 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;
+    }
+}
+
+#endif // WC_UTF16
+
 // ----------------------------------------------------------------------------
 // wxMBConv
 // ----------------------------------------------------------------------------
 // ----------------------------------------------------------------------------
 // wxMBConv
 // ----------------------------------------------------------------------------
@@ -126,26 +220,32 @@ WXDLLEXPORT_DATA(wxMBConvGdk) wxConvGdk;
 
 size_t wxMBConvGdk::MB2WC(wchar_t *buf, const char *psz, size_t n) const
 {
 
 size_t wxMBConvGdk::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;
-    return len;
-  }
+    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;
+        return len;
+    }
 }
 
 size_t wxMBConvGdk::WC2MB(char *buf, const wchar_t *psz, size_t n) const
 {
 }
 
 size_t wxMBConvGdk::WC2MB(char *buf, const wchar_t *psz, size_t n) const
 {
-  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;
+    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
 }
 
 #endif // GTK > 1.0
@@ -187,72 +287,110 @@ size_t wxMBConvUTF7::WC2MB(char * WXUNUSED(buf),
 
 WXDLLEXPORT_DATA(wxMBConvUTF8) wxConvUTF8;
 
 
 WXDLLEXPORT_DATA(wxMBConvUTF8) wxConvUTF8;
 
-static unsigned long 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 wxMBConvUTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const
 {
-  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;
-        unsigned long 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);
+    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++;
         }
         }
-        if (res<=utf8_max[ocnt]) {
-          // illegal UTF-8 encoding
-          return (size_t)-1;
+        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
+            }
         }
         }
-        if (buf) *buf++=res;
-        len++;
-      }
     }
     }
-  }
-  if (buf && (len<n)) *buf = 0;
-  return len;
+    if (buf && (len < n))
+        *buf = 0;
+    return len;
 }
 
 size_t wxMBConvUTF8::WC2MB(char *buf, const wchar_t *psz, size_t n) const
 {
 }
 
 size_t wxMBConvUTF8::WC2MB(char *buf, const wchar_t *psz, size_t n) const
 {
-  size_t len = 0;
-
-  while (*psz && ((!buf) || (len<n))) {
-    unsigned long cc=(*psz++)&0x7fffffff;
-    unsigned cnt;
-    for (cnt=0; cc>utf8_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<n)) *buf = 0;
-  return len;
+    size_t len = 0;
+
+    while (*psz && ((!buf) || (len < n)))
+    {
+        wxUint32 cc;
+#ifdef WC_UTF16
+        size_t pa = decode_utf16(psz, cc);
+        psz += (pa == (size_t)-1) ? 1 : pa;
+#else
+        cc=(*psz++) & 0x7fffffff;
+#endif
+        unsigned cnt;
+        for (cnt = 0; cc > utf8_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<n)) *buf = 0;
+    return len;
 }
 
 // ----------------------------------------------------------------------------
 }
 
 // ----------------------------------------------------------------------------
@@ -264,104 +402,570 @@ WXDLLEXPORT_DATA(wxCSConv) wxConvLocal((const wxChar *)NULL);
 #include "wx/encconv.h"
 #include "wx/fontmap.h"
 
 #include "wx/encconv.h"
 #include "wx/fontmap.h"
 
+// TODO: add some tables here
+// - perhaps common encodings to common codepages (for Win32)
+// - perhaps common encodings to objects ("UTF8" -> wxConvUTF8)
+// - move wxEncodingConverter meat in here
+
+#if defined(__WIN32__) && !defined(__WXMICROWIN__)
+#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);
+
+        if (!key.Exists()) break;
+
+        // 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.HasValue(wxT("InternetEncoding")) &&
+            key.QueryValue(wxT("InternetEncoding"), &CP)) break;
+
+        // no encoding, see if it's an alias
+        if (!key.HasValue(wxT("AliasForCharset")) ||
+            !key.QueryValue(wxT("AliasForCharset"), cn)) break;
+    } while (1);
+
+    return CP;
+}
+#endif
+
 class wxCharacterSet
 {
 public:
 class wxCharacterSet
 {
 public:
-  // temporarily just use wxEncodingConverter stuff,
-  // so that it works while a better implementation is built
-  wxFontEncoding enc;
-  wxEncodingConverter m2w, w2m;
-  wxCharacterSet(wxFontEncoding e) : enc(e)
-  {
-    m2w.Init(enc, wxFONTENCODING_UNICODE);
-    w2m.Init(wxFONTENCODING_UNICODE, enc);
-  }
+    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 HAVE_ICONV_H
+
+bool g_wcNeedsSwap = FALSE;
+static const char *g_wcCharset = NULL;
+
+// VS: glibc 2.1.3 is broken in that iconv() conversion to/from UCS4 fails with E2BIG
+//     if output buffer is _exactly_ as big as needed. Such case is (unless there's
+//     yet another bug in glibc) the only case when iconv() returns with (size_t)-1
+//     (which means error) and says there are 0 bytes left in the input buffer --
+//     when _real_ error occurs, bytes-left-in-input buffer is non-zero. Hence,
+//     this alternative test for iconv() failure.
+//     [This bug does not appear in glibc 2.2.]
+#if defined(__GLIBC__) && __GLIBC__ == 2 && __GLIBC_MINOR__ <= 1
+#define ICONV_FAILED(cres, bufLeft) ((cres == (size_t)-1) && \
+                                     (errno != E2BIG || bufLeft != 0))
+#else
+#define ICONV_FAILED(cres, bufLeft)  (cres == (size_t)-1)
+#endif
+
+class IC_CharSet : public wxCharacterSet
+{
+public:
+    IC_CharSet(const wxChar *name)
+        : wxCharacterSet(name)
+    {
+        // check for charset that represents wchar_t:
+        if (g_wcCharset == NULL)
+        {
+            g_wcNeedsSwap = FALSE;
+
+            // try charset with explicit bytesex info (e.g. "UCS-4LE"):
+            g_wcCharset = WC_NAME_BEST;
+            m2w = iconv_open(g_wcCharset, wxConvLibc.cWX2MB(name));
+
+            if (m2w == (iconv_t)-1)
+            {
+                // try charset w/o bytesex info (e.g. "UCS4")
+                // and check for bytesex ourselves: 
+                g_wcCharset = WC_NAME;
+                m2w = iconv_open(g_wcCharset, wxConvLibc.cWX2MB(name));
+
+                // last bet, try if it knows WCHAR_T pseudo-charset
+                if (m2w == (iconv_t)-1)
+                {
+                    g_wcCharset = "WCHAR_T";
+                    m2w = iconv_open(g_wcCharset, wxConvLibc.cWX2MB(name));
+                }
+                
+                if (m2w != (iconv_t)-1)
+                {
+                    char    buf[2], *bufPtr;
+                    wchar_t wbuf[2], *wbufPtr;
+                    size_t  insz, outsz;
+                    size_t  res;
+                    
+                    buf[0] = 'A';
+                    buf[1] = 0;
+                    wbuf[0] = 0;
+                    insz = 2;
+                    outsz = SIZEOF_WCHAR_T * 2;
+                    wbufPtr = wbuf;
+                    bufPtr = buf;
+
+                    #ifdef WX_ICONV_TAKES_CHAR
+                    res = iconv(m2w, (char**)&bufPtr, &insz, (char**)&wbufPtr, &outsz);
+                    #else
+                    res = iconv(m2w, (const char**)&bufPtr, &insz, (char**)&wbufPtr, &outsz);
+                    #endif
+                    if (ICONV_FAILED(res, insz))
+                    {
+                        g_wcCharset = NULL;
+                        wxLogLastError(wxT("iconv"));
+                        wxLogError(_("Convertion to charset '%s' doesn't work."), name);
+                    }
+                    else
+                    {
+                        g_wcNeedsSwap = (wbuf[0] != (wchar_t)buf[0]);
+                    }
+                }
+                else
+                {
+                    g_wcCharset = NULL;
+                    wxLogError(_("Don't know how to convert to/from charset '%s'."), name);
+                }
+            }
+            wxLogTrace(wxT("strconv"), wxT("wchar_t charset is '%s', needs swap: %i"), g_wcCharset, g_wcNeedsSwap);
+        }
+        else
+            m2w = iconv_open(g_wcCharset, wxConvLibc.cWX2MB(name));
+    
+        w2m = iconv_open(wxConvLibc.cWX2MB(name), g_wcCharset);
+    }
+    
+    ~IC_CharSet() 
+    {
+        if ( m2w != (iconv_t)-1 )
+            iconv_close(m2w);
+        if ( w2m != (iconv_t)-1 )
+            iconv_close(w2m);
+    }
+
+    size_t MB2WC(wchar_t *buf, const char *psz, size_t n)
+    {
+        size_t inbuf = strlen(psz);
+        size_t outbuf = n * SIZEOF_WCHAR_T;
+        size_t res, cres;
+        // VS: Use these instead of psz, buf because iconv() modifies its arguments:
+        wchar_t *bufPtr = buf;
+        const char *pszPtr = psz;
+
+        if (buf)
+        {
+            // have destination buffer, convert there
+#ifdef WX_ICONV_TAKES_CHAR
+            cres = iconv(m2w, (char**)&pszPtr, &inbuf, (char**)&bufPtr, &outbuf);
+#else
+            cres = iconv(m2w, &pszPtr, &inbuf, (char**)&bufPtr, &outbuf);
+#endif
+            res = n - (outbuf / SIZEOF_WCHAR_T);
+
+            if (g_wcNeedsSwap)
+            {
+                // convert to native endianness
+                WC_BSWAP(buf /* _not_ bufPtr */, res)
+            }
+        }
+        else
+        {
+            // no destination buffer... convert using temp buffer
+            // to calculate destination buffer requirement
+            wchar_t tbuf[8];
+            res = 0;
+            do {
+                bufPtr = tbuf; outbuf = 8*SIZEOF_WCHAR_T;
+#ifdef WX_ICONV_TAKES_CHAR
+                cres = iconv( m2w, (char**)&pszPtr, &inbuf, (char**)&bufPtr, &outbuf );
+#else
+                cres = iconv( m2w, &pszPtr, &inbuf, (char**)&bufPtr, &outbuf );
+#endif
+                res += 8-(outbuf/SIZEOF_WCHAR_T);
+            } while ((cres==(size_t)-1) && (errno==E2BIG));
+        }
+
+        if (ICONV_FAILED(cres, inbuf))
+        {
+            //VS: it is ok if iconv fails, hence trace only
+            wxLogTrace(wxT("strconv"), wxT("iconv failed: %s"), wxSysErrorMsg(wxSysErrorCode()));
+            return (size_t)-1;
+        }
+
+        return res;
+    }
+
+    size_t WC2MB(char *buf, const wchar_t *psz, size_t n)
+    {
+#if defined(__BORLANDC__) && (__BORLANDC__ > 0x530)
+        size_t inbuf = std::wcslen(psz) * SIZEOF_WCHAR_T;
+#else
+        size_t inbuf = ::wcslen(psz) * SIZEOF_WCHAR_T;
+#endif
+        size_t outbuf = n;
+        size_t res, cres;
+        
+        wchar_t *tmpbuf;
+
+        if (g_wcNeedsSwap)
+        {
+            // 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)
+            tmpbuf=(wchar_t*)malloc((inbuf+1)*SIZEOF_WCHAR_T);
+            memcpy(tmpbuf,psz,(inbuf+1)*SIZEOF_WCHAR_T);
+            WC_BSWAP(tmpbuf, inbuf)
+            psz=tmpbuf;
+        }
+
+        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));
+        }
+
+        if (g_wcNeedsSwap)
+        {
+            free(tmpbuf);
+        }
+
+        if (ICONV_FAILED(cres, inbuf))
+        {
+            //VS: it is ok if iconv fails, hence trace only
+            wxLogTrace(wxT("strconv"), wxT("iconv failed: %s"), wxSysErrorMsg(wxSysErrorCode()));
+            return (size_t)-1;
+        }
+
+        return res;
+    }
+
+    bool usable()
+        { return (m2w != (iconv_t)-1) && (w2m != (iconv_t)-1); }
+
+protected:
+    iconv_t m2w, w2m;
+};
+#endif
+
+#if defined(__WIN32__) && !defined(__WXMICROWIN__)
+class CP_CharSet : public wxCharacterSet
+{
+public:
+    CP_CharSet(const wxChar*name)
+        : wxCharacterSet(name), CodePage(CharsetToCodepage(name)) {}
+
+    size_t MB2WC(wchar_t *buf, const char *psz, size_t n)
+    {
+        size_t len =
+            MultiByteToWideChar(CodePage, 0, psz, -1, buf, buf ? n : 0);
+        //VS: returns # of written chars for buf!=NULL and *size*
+        //    needed buffer for buf==NULL
+        return len ? (buf ? len : len-1) : (size_t)-1;
+    }
+
+    size_t WC2MB(char *buf, const wchar_t *psz, size_t n)
+    {
+        size_t len = WideCharToMultiByte(CodePage, 0, psz, -1, buf,
+                                         buf ? n : 0, NULL, NULL);
+        //VS: returns # of written chars for buf!=NULL and *size*
+        //    needed buffer for buf==NULL
+        return len ? (buf ? len : len-1) : (size_t)-1;
+    }
+
+    bool usable()
+        { return CodePage != -1; }
+
+public:
+    long CodePage;
+};
+#endif // __WIN32__
+
+#if wxUSE_FONTMAP
+
+class EC_CharSet : public wxCharacterSet
+{
+public:
+    // temporarily just use wxEncodingConverter stuff,
+    // so that it works while a better implementation is built
+    EC_CharSet(const wxChar*name) : wxCharacterSet(name),
+                                    enc(wxFONTENCODING_SYSTEM)
+    {
+        if (name)
+            enc = wxTheFontMapper->CharsetToEncoding(name, FALSE);
+        m2w.Init(enc, wxFONTENCODING_UNICODE);
+        w2m.Init(wxFONTENCODING_UNICODE, enc);
+    }
+
+    size_t MB2WC(wchar_t *buf, const char *psz, size_t n)
+    {
+        size_t inbuf = strlen(psz);
+        if (buf)
+            m2w.Convert(psz,buf);
+        return inbuf;
+    }
+
+    size_t WC2MB(char *buf, const wchar_t *psz, size_t n)
+    {
+#if defined(__BORLANDC__) && (__BORLANDC__ > 0x530)
+        size_t inbuf = std::wcslen(psz);
+#else
+        size_t inbuf = ::wcslen(psz);
+#endif
+        if (buf)
+            w2m.Convert(psz,buf);
+
+        return inbuf;
+    }
+
+    bool usable()
+        { return (enc!=wxFONTENCODING_SYSTEM) && (enc!=wxFONTENCODING_DEFAULT); }
+
+public:
+    wxFontEncoding enc;
+    wxEncodingConverter m2w, w2m;
+};
+
+#endif // wxUSE_FONTMAP
+
 static wxCharacterSet *wxGetCharacterSet(const wxChar *name)
 {
 static wxCharacterSet *wxGetCharacterSet(const wxChar *name)
 {
-  wxFontEncoding enc = name ? wxTheFontMapper->CharsetToEncoding(name, FALSE)
-                            : wxFONTENCODING_SYSTEM;
-  wxCharacterSet *cset = (enc != wxFONTENCODING_SYSTEM) ? new wxCharacterSet(enc)
-                                                        : (wxCharacterSet *)NULL;
-  return cset;
+    wxCharacterSet *cset = NULL;
+    if (name)
+    {
+        if (wxStricmp(name, wxT("UTF8")) == 0 || wxStricmp(name, wxT("UTF-8")) == 0)
+        {
+            cset = new ID_CharSet(name, &wxConvUTF8);
+        }
+        else
+        {
+#ifdef HAVE_ICONV_H
+            cset = new IC_CharSet(name); // may not take NULL
+#endif
+        }
+    }
+
+    if (cset && cset->usable())
+        return cset;
+
+    if (cset)
+    {
+        delete cset;
+        cset = NULL;
+    }
+
+#if defined(__WIN32__) && !defined(__WXMICROWIN__)
+    cset = new CP_CharSet(name); // may take NULL
+    if (cset->usable())
+        return cset;
+
+    delete cset;
+#endif // __WIN32__
+
+#if wxUSE_FONTMAP
+    cset = new EC_CharSet(name);
+    if (cset->usable())
+        return cset;
+#endif // wxUSE_FONTMAP
+
+    delete cset;
+    wxLogError(_("Unknown encoding '%s'!"), name);
+    return NULL;
 }
 
 wxCSConv::wxCSConv(const wxChar *charset)
 {
 }
 
 wxCSConv::wxCSConv(const wxChar *charset)
 {
-  m_name = (wxChar *) NULL;
-  m_cset = (wxCharacterSet *) NULL;
-  m_deferred = TRUE;
-  SetName(charset);
+    m_name = (wxChar *)NULL;
+    m_cset = (wxCharacterSet *) NULL;
+    m_deferred = TRUE;
+
+    SetName(charset);
 }
 
 wxCSConv::~wxCSConv()
 {
 }
 
 wxCSConv::~wxCSConv()
 {
-  if (m_name) free(m_name);
-  if (m_cset) delete m_cset;
+    free(m_name);
+    delete m_cset;
 }
 
 void wxCSConv::SetName(const wxChar *charset)
 {
 }
 
 void wxCSConv::SetName(const wxChar *charset)
 {
-  if (charset) {
-    m_name = wxStrdup(charset);
-    m_deferred = TRUE;
-  }
+    if (charset)
+    {
+        m_name = wxStrdup(charset);
+        m_deferred = TRUE;
+    }
 }
 
 void wxCSConv::LoadNow()
 {
 }
 
 void wxCSConv::LoadNow()
 {
-//  wxPrintf(wxT("Conversion request\n"));
-  if (m_deferred) {
-    if (!m_name) {
-#ifdef __UNIX__
-      wxChar *lang = wxGetenv(wxT("LC_ALL"));
-      if (!lang) lang = wxGetenv(wxT("LANG"));
-      wxChar *dot = lang ? wxStrchr(lang, wxT('.')) : (wxChar *)NULL;
-      if (dot) SetName(dot+1);
-#endif
+    if (m_deferred)
+    {
+        if ( !m_name )
+        {
+            wxString name = wxLocale::GetSystemEncodingName();
+            if ( !name.empty() )
+                SetName(name);
+        }
+
+        // wxGetCharacterSet() complains about NULL name
+        m_cset = m_name ? wxGetCharacterSet(m_name) : NULL;
+        m_deferred = FALSE;
     }
     }
-    m_cset = wxGetCharacterSet(m_name);
-    m_deferred = FALSE;
-  }
 }
 
 size_t wxCSConv::MB2WC(wchar_t *buf, const char *psz, size_t n) const
 {
 }
 
 size_t wxCSConv::MB2WC(wchar_t *buf, const char *psz, size_t n) const
 {
-  ((wxCSConv *)this)->LoadNow(); // discard constness
-  if (buf) {
-    if (m_cset) {
-      m_cset->m2w.Convert(psz, buf);
-    } else {
-      // latin-1 (direct)
-      for (size_t c=0; c<n; c++)
-        buf[c] = (unsigned char)(psz[c]);
-    }
-    return n;
-  }
-  return strlen(psz);
+    ((wxCSConv *)this)->LoadNow(); // discard constness
+
+    if (m_cset)
+        return m_cset->MB2WC(buf, psz, n);
+
+    // latin-1 (direct)
+    size_t len = strlen(psz);
+
+    if (buf)
+    {
+        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
 {
 }
 
 size_t wxCSConv::WC2MB(char *buf, const wchar_t *psz, size_t n) const
 {
-  ((wxCSConv *)this)->LoadNow(); // discard constness
-  if (buf) {
-    if (m_cset) {
-      m_cset->w2m.Convert(psz, buf);
-    } else {
-      // latin-1 (direct)
-      for (size_t c=0; c<n; c++)
-        buf[c] = (psz[c]>0xff) ? '?' : psz[c];
-    }
-    return n;
-  }
+    ((wxCSConv *)this)->LoadNow(); // discard constness
+
+    if (m_cset)
+        return m_cset->WC2MB(buf, psz, n);
+
+    // latin-1 (direct)
 #if defined(__BORLANDC__) && (__BORLANDC__ > 0x530)
 #if defined(__BORLANDC__) && (__BORLANDC__ > 0x530)
-  return std::wcslen(psz);
+    size_t len=std::wcslen(psz);
 #else
 #else
-  return ::wcslen(psz);
+    size_t len=::wcslen(psz);
 #endif
 #endif
+    if (buf)
+    {
+        for (size_t c = 0; c <= len; c++)
+            buf[c] = (psz[c] > 0xff) ? '?' : psz[c];
+    }
+
+    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 );
+#else
+        size_t res = iconv( cnv, &psz, &inbuf, &buf, &outbuf );
+#endif
+        if (res == (size_t)-1)
+            return (size_t)-1;
+        return (n - outbuf);
+    }
+
+public:
+    iconv_t cnv;
+};
+
+#endif // HAVE_ICONV_H
+
+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;
+};
+
 #else // !wxUSE_WCHAR_T
 
 // ----------------------------------------------------------------------------
 #else // !wxUSE_WCHAR_T
 
 // ----------------------------------------------------------------------------