To: vim-dev@vim.org
Subject: Patch 7.2.049 (extra)
Fcc: outbox
From: Bram Moolenaar <Bram@moolenaar.net>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
------------

Patch 7.2.049 (extra)
Problem:    Win32: the clipboard doesn't support UTF-16.
Solution:   Change UCS-2 support to UTF-16 support. (Jia Yanwei)
Files:	    src/gui_w32.c, src/gui_w48.c, src/mbyte.c, src/misc1.c,
	    src/os_mswin.c, src/os_win32.c, src/proto/os_mswin.pro


*** ../vim-7.2.048/src/gui_w32.c	Tue Jun 24 23:46:14 2008
--- src/gui_w32.c	Thu Nov 20 16:47:10 2008
***************
*** 992,998 ****
  			    SendMessage(lpdi->hdr.hwndFrom, TTM_SETMAXTIPWIDTH,
  								      0, 500);
  
! 			    tt_text = enc_to_ucs2(str, NULL);
  			    lpdi->lpszText = tt_text;
  			    /* can't show tooltip if failed */
  			}
--- 992,998 ----
  			    SendMessage(lpdi->hdr.hwndFrom, TTM_SETMAXTIPWIDTH,
  								      0, 500);
  
! 			    tt_text = enc_to_utf16(str, NULL);
  			    lpdi->lpszText = tt_text;
  			    /* can't show tooltip if failed */
  			}
***************
*** 1935,1941 ****
      if (buf == NULL)
  	return NULL;
  
!     convbuf = ucs2_to_enc(buf, lenp);
      pImmReleaseContext(hwnd, hIMC);
      vim_free(buf);
      return convbuf;
--- 1935,1941 ----
      if (buf == NULL)
  	return NULL;
  
!     convbuf = utf16_to_enc(buf, lenp);
      pImmReleaseContext(hwnd, hIMC);
      vim_free(buf);
      return convbuf;
***************
*** 2566,2572 ****
  	    {
  		/* 'encoding' differs from active codepage: convert menu name
  		 * and use wide function */
! 		wn = enc_to_ucs2(menu->name, NULL);
  		if (wn != NULL)
  		{
  		    MENUITEMINFOW	infow;
--- 2566,2572 ----
  	    {
  		/* 'encoding' differs from active codepage: convert menu name
  		 * and use wide function */
! 		wn = enc_to_utf16(menu->name, NULL);
  		if (wn != NULL)
  		{
  		    MENUITEMINFOW	infow;
***************
*** 2728,2734 ****
  	{
  	    /* 'encoding' differs from active codepage: convert menu item name
  	     * and use wide function */
! 	    wn = enc_to_ucs2(menu->name, NULL);
  	    if (wn != NULL)
  	    {
  		n = InsertMenuW(parent->submenu_id, (UINT)idx,
--- 2728,2734 ----
  	{
  	    /* 'encoding' differs from active codepage: convert menu item name
  	     * and use wide function */
! 	    wn = enc_to_utf16(menu->name, NULL);
  	    if (wn != NULL)
  	    {
  		n = InsertMenuW(parent->submenu_id, (UINT)idx,
***************
*** 3570,3576 ****
      if (enc_codepage == 0 && (int)GetACP() != enc_codepage)
      {
  	/* Not a codepage, use our own conversion function. */
! 	wn = enc_to_ucs2(lpAnsiIn, NULL);
  	if (wn != NULL)
  	{
  	    wcscpy(lpWCStr, wn);
--- 3570,3576 ----
      if (enc_codepage == 0 && (int)GetACP() != enc_codepage)
      {
  	/* Not a codepage, use our own conversion function. */
! 	wn = enc_to_utf16(lpAnsiIn, NULL);
  	if (wn != NULL)
  	{
  	    wcscpy(lpWCStr, wn);
*** ../vim-7.2.048/src/gui_w48.c	Wed Nov 12 13:35:31 2008
--- src/gui_w48.c	Thu Nov 20 16:42:32 2008
***************
*** 547,553 ****
  	else
  	{
  	    len = 1;
! 	    ws = ucs2_to_enc(wstring, &len);
  	    if (ws == NULL)
  		len = 0;
  	    else
--- 547,553 ----
  	else
  	{
  	    len = 1;
! 	    ws = utf16_to_enc(wstring, &len);
  	    if (ws == NULL)
  		len = 0;
  	    else
***************
*** 2128,2134 ****
      {
  	/* 'encoding' differs from active codepage: convert text and use wide
  	 * function */
! 	wstr = enc_to_ucs2(str, &wlen);
  	if (wstr != NULL)
  	{
  	    n = GetTextExtentPointW(hdc, wstr, wlen, &size);
--- 2128,2134 ----
      {
  	/* 'encoding' differs from active codepage: convert text and use wide
  	 * function */
! 	wstr = enc_to_utf16(str, &wlen);
  	if (wstr != NULL)
  	{
  	    n = GetTextExtentPointW(hdc, wstr, wlen, &size);
***************
*** 2252,2258 ****
      {
  	/* 'encoding' differs from active codepage: convert menu name
  	 * and use wide function */
! 	wn = enc_to_ucs2(item_text, NULL);
  	if (wn != NULL)
  	{
  	    MENUITEMINFOW	infow;
--- 2252,2258 ----
      {
  	/* 'encoding' differs from active codepage: convert menu name
  	 * and use wide function */
! 	wn = enc_to_utf16(item_text, NULL);
  	if (wn != NULL)
  	{
  	    MENUITEMINFOW	infow;
***************
*** 2422,2428 ****
  	if (use_unicode)
  	{
  	    /* Need to go through Unicode. */
! 	    wstr = enc_to_ucs2(NameBuff, NULL);
  	    if (wstr != NULL)
  	    {
  		TCITEMW		tiw;
--- 2422,2428 ----
  	if (use_unicode)
  	{
  	    /* Need to go through Unicode. */
! 	    wstr = enc_to_utf16(NameBuff, NULL);
  	    if (wstr != NULL)
  	    {
  		TCITEMW		tiw;
***************
*** 2521,2528 ****
  	WCHAR	*wbuf;
  	int	n;
  
! 	/* Convert the title from 'encoding' to ucs2. */
! 	wbuf = (WCHAR *)enc_to_ucs2((char_u *)title, NULL);
  	if (wbuf != NULL)
  	{
  	    n = SetWindowTextW(hwnd, wbuf);
--- 2521,2528 ----
  	WCHAR	*wbuf;
  	int	n;
  
! 	/* Convert the title from 'encoding' to UTF-16. */
! 	wbuf = (WCHAR *)enc_to_utf16((char_u *)title, NULL);
  	if (wbuf != NULL)
  	{
  	    n = SetWindowTextW(hwnd, wbuf);
***************
*** 3222,3228 ****
  	char_u *initdir,
  	char_u *filter)
  {
!     /* We always use the wide function.  This means enc_to_ucs2() must work,
       * otherwise it fails miserably! */
      OPENFILENAMEW	fileStruct;
      WCHAR		fileBuf[MAXPATHL];
--- 3222,3228 ----
  	char_u *initdir,
  	char_u *filter)
  {
!     /* We always use the wide function.  This means enc_to_utf16() must work,
       * otherwise it fails miserably! */
      OPENFILENAMEW	fileStruct;
      WCHAR		fileBuf[MAXPATHL];
***************
*** 3238,3244 ****
  	fileBuf[0] = NUL;
      else
      {
! 	wp = enc_to_ucs2(dflt, NULL);
  	if (wp == NULL)
  	    fileBuf[0] = NUL;
  	else
--- 3238,3244 ----
  	fileBuf[0] = NUL;
      else
      {
! 	wp = enc_to_utf16(dflt, NULL);
  	if (wp == NULL)
  	    fileBuf[0] = NUL;
  	else
***************
*** 3263,3273 ****
  #endif
  
      if (title != NULL)
! 	titlep = enc_to_ucs2(title, NULL);
      fileStruct.lpstrTitle = titlep;
  
      if (ext != NULL)
! 	extp = enc_to_ucs2(ext, NULL);
      fileStruct.lpstrDefExt = extp;
  
      fileStruct.lpstrFile = fileBuf;
--- 3263,3273 ----
  #endif
  
      if (title != NULL)
! 	titlep = enc_to_utf16(title, NULL);
      fileStruct.lpstrTitle = titlep;
  
      if (ext != NULL)
! 	extp = enc_to_utf16(ext, NULL);
      fileStruct.lpstrDefExt = extp;
  
      fileStruct.lpstrFile = fileBuf;
***************
*** 3278,3284 ****
      if (initdir != NULL && *initdir != NUL)
      {
  	/* Must have backslashes here, no matter what 'shellslash' says */
! 	initdirp = enc_to_ucs2(initdir, NULL);
  	if (initdirp != NULL)
  	{
  	    for (wp = initdirp; *wp != NUL; ++wp)
--- 3278,3284 ----
      if (initdir != NULL && *initdir != NUL)
      {
  	/* Must have backslashes here, no matter what 'shellslash' says */
! 	initdirp = enc_to_utf16(initdir, NULL);
  	if (initdirp != NULL)
  	{
  	    for (wp = initdirp; *wp != NUL; ++wp)
***************
*** 3318,3324 ****
      vim_free(extp);
  
      /* Convert from UCS2 to 'encoding'. */
!     p = ucs2_to_enc(fileBuf, NULL);
      if (p != NULL)
  	/* when out of memory we get garbage for non-ASCII chars */
  	STRCPY(fileBuf, p);
--- 3318,3324 ----
      vim_free(extp);
  
      /* Convert from UCS2 to 'encoding'. */
!     p = utf16_to_enc(fileBuf, NULL);
      if (p != NULL)
  	/* when out of memory we get garbage for non-ASCII chars */
  	STRCPY(fileBuf, p);
***************
*** 3518,3524 ****
  	{
  #ifdef FEAT_MBYTE
  	    if (DragQueryFileW(hDrop, i, wszFile, BUFPATHLEN) > 0)
! 		fnames[i] = ucs2_to_enc(wszFile, NULL);
  	    else
  #endif
  	    {
--- 3518,3524 ----
  	{
  #ifdef FEAT_MBYTE
  	    if (DragQueryFileW(hDrop, i, wszFile, BUFPATHLEN) > 0)
! 		fnames[i] = utf16_to_enc(wszFile, NULL);
  	    else
  #endif
  	    {
*** ../vim-7.2.048/src/mbyte.c	Thu Nov 20 11:55:53 2008
--- src/mbyte.c	Thu Nov 20 16:47:41 2008
***************
*** 6101,6107 ****
  
  	    /* 1. codepage/UTF-8  ->  ucs-2. */
  	    if (vcp->vc_cpfrom == 0)
! 		tmp_len = utf8_to_ucs2(ptr, len, NULL, NULL);
  	    else
  		tmp_len = MultiByteToWideChar(vcp->vc_cpfrom, 0,
  							      ptr, len, 0, 0);
--- 6101,6107 ----
  
  	    /* 1. codepage/UTF-8  ->  ucs-2. */
  	    if (vcp->vc_cpfrom == 0)
! 		tmp_len = utf8_to_utf16(ptr, len, NULL, NULL);
  	    else
  		tmp_len = MultiByteToWideChar(vcp->vc_cpfrom, 0,
  							      ptr, len, 0, 0);
***************
*** 6109,6121 ****
  	    if (tmp == NULL)
  		break;
  	    if (vcp->vc_cpfrom == 0)
! 		utf8_to_ucs2(ptr, len, tmp, unconvlenp);
  	    else
  		MultiByteToWideChar(vcp->vc_cpfrom, 0, ptr, len, tmp, tmp_len);
  
  	    /* 2. ucs-2  ->  codepage/UTF-8. */
  	    if (vcp->vc_cpto == 0)
! 		retlen = ucs2_to_utf8(tmp, tmp_len, NULL);
  	    else
  		retlen = WideCharToMultiByte(vcp->vc_cpto, 0,
  						    tmp, tmp_len, 0, 0, 0, 0);
--- 6109,6121 ----
  	    if (tmp == NULL)
  		break;
  	    if (vcp->vc_cpfrom == 0)
! 		utf8_to_utf16(ptr, len, tmp, unconvlenp);
  	    else
  		MultiByteToWideChar(vcp->vc_cpfrom, 0, ptr, len, tmp, tmp_len);
  
  	    /* 2. ucs-2  ->  codepage/UTF-8. */
  	    if (vcp->vc_cpto == 0)
! 		retlen = utf16_to_utf8(tmp, tmp_len, NULL);
  	    else
  		retlen = WideCharToMultiByte(vcp->vc_cpto, 0,
  						    tmp, tmp_len, 0, 0, 0, 0);
***************
*** 6123,6129 ****
  	    if (retval != NULL)
  	    {
  		if (vcp->vc_cpto == 0)
! 		    ucs2_to_utf8(tmp, tmp_len, retval);
  		else
  		    WideCharToMultiByte(vcp->vc_cpto, 0,
  					  tmp, tmp_len, retval, retlen, 0, 0);
--- 6123,6129 ----
  	    if (retval != NULL)
  	    {
  		if (vcp->vc_cpto == 0)
! 		    utf16_to_utf8(tmp, tmp_len, retval);
  		else
  		    WideCharToMultiByte(vcp->vc_cpto, 0,
  					  tmp, tmp_len, retval, retlen, 0, 0);
*** ../vim-7.2.048/src/misc1.c	Sun Nov  9 13:43:25 2008
--- src/misc1.c	Thu Nov 20 16:42:59 2008
***************
*** 8728,8734 ****
  	/* The active codepage differs from 'encoding'.  Attempt using the
  	 * wide function.  If it fails because it is not implemented fall back
  	 * to the non-wide version (for Windows 98) */
! 	wn = enc_to_ucs2(buf, NULL);
  	if (wn != NULL)
  	{
  	    hFind = FindFirstFileW(wn, &wfb);
--- 8728,8734 ----
  	/* The active codepage differs from 'encoding'.  Attempt using the
  	 * wide function.  If it fails because it is not implemented fall back
  	 * to the non-wide version (for Windows 98) */
! 	wn = enc_to_utf16(buf, NULL);
  	if (wn != NULL)
  	{
  	    hFind = FindFirstFileW(wn, &wfb);
***************
*** 8756,8762 ****
  #ifdef WIN3264
  # ifdef FEAT_MBYTE
  	if (wn != NULL)
! 	    p = ucs2_to_enc(wfb.cFileName, NULL);   /* p is allocated here */
  	else
  # endif
  	    p = (char_u *)fb.cFileName;
--- 8756,8762 ----
  #ifdef WIN3264
  # ifdef FEAT_MBYTE
  	if (wn != NULL)
! 	    p = utf16_to_enc(wfb.cFileName, NULL);   /* p is allocated here */
  	else
  # endif
  	    p = (char_u *)fb.cFileName;
***************
*** 8830,8836 ****
  	    if (wn != NULL)
  	    {
  		vim_free(wn);
! 		wn = enc_to_ucs2(buf, NULL);
  		if (wn != NULL)
  		    hFind = FindFirstFileW(wn, &wfb);
  	    }
--- 8830,8836 ----
  	    if (wn != NULL)
  	    {
  		vim_free(wn);
! 		wn = enc_to_utf16(buf, NULL);
  		if (wn != NULL)
  		    hFind = FindFirstFileW(wn, &wfb);
  	    }
*** ../vim-7.2.048/src/os_mswin.c	Sat Aug  9 19:37:30 2008
--- src/os_mswin.c	Thu Nov 20 17:05:10 2008
***************
*** 309,315 ****
  	if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
  	{
  	    /* Convert the title from 'encoding' to the active codepage. */
! 	    WCHAR	*wp = enc_to_ucs2(title, NULL);
  	    int	n;
  
  	    if (wp != NULL)
--- 309,315 ----
  	if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
  	{
  	    /* Convert the title from 'encoding' to the active codepage. */
! 	    WCHAR	*wp = enc_to_utf16(title, NULL);
  	    int	n;
  
  	    if (wp != NULL)
***************
*** 406,415 ****
  	     * - invoke _wfullpath()
  	     * - convert the result from UCS2 to 'encoding'.
  	     */
! 	    wname = enc_to_ucs2(fname, NULL);
  	    if (wname != NULL && _wfullpath(wbuf, wname, MAX_PATH - 1) != NULL)
  	    {
! 		cname = ucs2_to_enc((short_u *)wbuf, NULL);
  		if (cname != NULL)
  		{
  		    vim_strncpy(buf, cname, len - 1);
--- 406,415 ----
  	     * - invoke _wfullpath()
  	     * - convert the result from UCS2 to 'encoding'.
  	     */
! 	    wname = enc_to_utf16(fname, NULL);
  	    if (wname != NULL && _wfullpath(wbuf, wname, MAX_PATH - 1) != NULL)
  	    {
! 		cname = utf16_to_enc((short_u *)wbuf, NULL);
  		if (cname != NULL)
  		{
  		    vim_strncpy(buf, cname, len - 1);
***************
*** 507,513 ****
  # endif
         )
      {
! 	WCHAR	*wp = enc_to_ucs2(buf, NULL);
  	int	n;
  
  	if (wp != NULL)
--- 507,513 ----
  # endif
         )
      {
! 	WCHAR	*wp = enc_to_utf16(buf, NULL);
  	int	n;
  
  	if (wp != NULL)
***************
*** 668,674 ****
  #ifdef FEAT_MBYTE
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	WCHAR	*p = enc_to_ucs2(path, NULL);
  	int	n;
  
  	if (p != NULL)
--- 668,674 ----
  #ifdef FEAT_MBYTE
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	WCHAR	*p = enc_to_utf16(path, NULL);
  	int	n;
  
  	if (p != NULL)
***************
*** 891,909 ****
  
  #if defined(FEAT_MBYTE) || defined(PROTO)
  /*
!  * Convert an UTF-8 string to UCS-2.
   * "instr[inlen]" is the input.  "inlen" is in bytes.
!  * When "outstr" is NULL only return the number of UCS-2 words produced.
   * Otherwise "outstr" must be a buffer of sufficient size.
!  * Returns the number of UCS-2 words produced.
   */
      int
! utf8_to_ucs2(char_u *instr, int inlen, short_u *outstr, int *unconvlenp)
  {
      int		outlen = 0;
      char_u	*p = instr;
      int		todo = inlen;
      int		l;
  
      while (todo > 0)
      {
--- 891,910 ----
  
  #if defined(FEAT_MBYTE) || defined(PROTO)
  /*
!  * Convert an UTF-8 string to UTF-16.
   * "instr[inlen]" is the input.  "inlen" is in bytes.
!  * When "outstr" is NULL only return the number of UTF-16 words produced.
   * Otherwise "outstr" must be a buffer of sufficient size.
!  * Returns the number of UTF-16 words produced.
   */
      int
! utf8_to_utf16(char_u *instr, int inlen, short_u *outstr, int *unconvlenp)
  {
      int		outlen = 0;
      char_u	*p = instr;
      int		todo = inlen;
      int		l;
+     int		ch;
  
      while (todo > 0)
      {
***************
*** 917,924 ****
  	    break;
  	}
  
! 	if (outstr != NULL)
! 	    *outstr++ = utf_ptr2char(p);
  	++outlen;
  	p += l;
  	todo -= l;
--- 918,936 ----
  	    break;
  	}
  
! 	ch = utf_ptr2char(p);
! 	if (ch >= 0x10000)
! 	{
! 	    /* non-BMP character, encoding with surrogate pairs */
! 	    ++outlen;
! 	    if (outstr != NULL)
! 	    {
! 		*outstr++ = (0xD800 - (0x10000 >> 10)) + (ch >> 10);
! 		*outstr++ = 0xDC00 | (ch & 0x3FF);
! 	    }
! 	}
! 	else if (outstr != NULL)
! 	    *outstr++ = ch;
  	++outlen;
  	p += l;
  	todo -= l;
***************
*** 928,956 ****
  }
  
  /*
!  * Convert an UCS-2 string to UTF-8.
!  * The input is "instr[inlen]" with "inlen" in number of ucs-2 words.
   * When "outstr" is NULL only return the required number of bytes.
   * Otherwise "outstr" must be a buffer of sufficient size.
   * Return the number of bytes produced.
   */
      int
! ucs2_to_utf8(short_u *instr, int inlen, char_u *outstr)
  {
      int		outlen = 0;
      int		todo = inlen;
      short_u	*p = instr;
      int		l;
  
      while (todo > 0)
      {
  	if (outstr != NULL)
  	{
! 	    l = utf_char2bytes(*p, outstr);
  	    outstr += l;
  	}
  	else
! 	    l = utf_char2len(*p);
  	++p;
  	outlen += l;
  	--todo;
--- 940,981 ----
  }
  
  /*
!  * Convert an UTF-16 string to UTF-8.
!  * The input is "instr[inlen]" with "inlen" in number of UTF-16 words.
   * When "outstr" is NULL only return the required number of bytes.
   * Otherwise "outstr" must be a buffer of sufficient size.
   * Return the number of bytes produced.
   */
      int
! utf16_to_utf8(short_u *instr, int inlen, char_u *outstr)
  {
      int		outlen = 0;
      int		todo = inlen;
      short_u	*p = instr;
      int		l;
+     int		ch, ch2;
  
      while (todo > 0)
      {
+ 	ch = *p;
+ 	if (ch >= 0xD800 && ch <= 0xDBFF && todo > 1)
+ 	{
+ 	    /* surrogate pairs handling */
+ 	    ch2 = p[1];
+ 	    if (ch2 >= 0xDC00 && ch2 <= 0xDFFF)
+ 	    {
+ 		ch = ((ch - 0xD800) << 10) + (ch2 & 0x3FF) + 0x10000;
+ 		++p;
+ 		--todo;
+ 	    }
+ 	}
  	if (outstr != NULL)
  	{
! 	    l = utf_char2bytes(ch, outstr);
  	    outstr += l;
  	}
  	else
! 	    l = utf_char2len(ch);
  	++p;
  	outlen += l;
  	--todo;
***************
*** 1079,1092 ****
   */
  
  /*
!  * Convert "str" from 'encoding' to UCS-2.
   * Input in "str" with length "*lenp".  When "lenp" is NULL, use strlen().
   * Output is returned as an allocated string.  "*lenp" is set to the length of
   * the result.  A trailing NUL is always added.
   * Returns NULL when out of memory.
   */
      short_u *
! enc_to_ucs2(char_u *str, int *lenp)
  {
      vimconv_T	conv;
      WCHAR	*ret;
--- 1104,1117 ----
   */
  
  /*
!  * Convert "str" from 'encoding' to UTF-16.
   * Input in "str" with length "*lenp".  When "lenp" is NULL, use strlen().
   * Output is returned as an allocated string.  "*lenp" is set to the length of
   * the result.  A trailing NUL is always added.
   * Returns NULL when out of memory.
   */
      short_u *
! enc_to_utf16(char_u *str, int *lenp)
  {
      vimconv_T	conv;
      WCHAR	*ret;
***************
*** 1102,1108 ****
  
      if (enc_codepage > 0)
      {
! 	/* We can do any CP### -> UCS-2 in one pass, and we can do it
  	 * without iconv() (convert_* may need iconv). */
  	MultiByteToWideChar_alloc(enc_codepage, 0, str, *lenp, &ret, &length);
      }
--- 1127,1133 ----
  
      if (enc_codepage > 0)
      {
! 	/* We can do any CP### -> UTF-16 in one pass, and we can do it
  	 * without iconv() (convert_* may need iconv). */
  	MultiByteToWideChar_alloc(enc_codepage, 0, str, *lenp, &ret, &length);
      }
***************
*** 1123,1133 ****
  	}
  	convert_setup(&conv, NULL, NULL);
  
! 	length = utf8_to_ucs2(str, *lenp, NULL, NULL);
  	ret = (WCHAR *)alloc((unsigned)((length + 1) * sizeof(WCHAR)));
  	if (ret != NULL)
  	{
! 	    utf8_to_ucs2(str, *lenp, (short_u *)ret, NULL);
  	    ret[length] = 0;
  	}
  
--- 1148,1158 ----
  	}
  	convert_setup(&conv, NULL, NULL);
  
! 	length = utf8_to_utf16(str, *lenp, NULL, NULL);
  	ret = (WCHAR *)alloc((unsigned)((length + 1) * sizeof(WCHAR)));
  	if (ret != NULL)
  	{
! 	    utf8_to_utf16(str, *lenp, (short_u *)ret, NULL);
  	    ret[length] = 0;
  	}
  
***************
*** 1139,1145 ****
  }
  
  /*
!  * Convert an UCS-2 string to 'encoding'.
   * Input in "str" with length (counted in wide characters) "*lenp".  When
   * "lenp" is NULL, use wcslen().
   * Output is returned as an allocated string.  If "*lenp" is not NULL it is
--- 1164,1170 ----
  }
  
  /*
!  * Convert an UTF-16 string to 'encoding'.
   * Input in "str" with length (counted in wide characters) "*lenp".  When
   * "lenp" is NULL, use wcslen().
   * Output is returned as an allocated string.  If "*lenp" is not NULL it is
***************
*** 1147,1153 ****
   * Returns NULL when out of memory.
   */
      char_u *
! ucs2_to_enc(short_u *str, int *lenp)
  {
      vimconv_T	conv;
      char_u	*utf8_str = NULL, *enc_str = NULL;
--- 1172,1178 ----
   * Returns NULL when out of memory.
   */
      char_u *
! utf16_to_enc(short_u *str, int *lenp)
  {
      vimconv_T	conv;
      char_u	*utf8_str = NULL, *enc_str = NULL;
***************
*** 1161,1167 ****
  
      if (enc_codepage > 0)
      {
! 	/* We can do any UCS-2 -> CP### in one pass. */
  	int length;
  
  	WideCharToMultiByte_alloc(enc_codepage, 0, str, *lenp,
--- 1186,1192 ----
  
      if (enc_codepage > 0)
      {
! 	/* We can do any UTF-16 -> CP### in one pass. */
  	int length;
  
  	WideCharToMultiByte_alloc(enc_codepage, 0, str, *lenp,
***************
*** 1171,1180 ****
      }
  
      /* Avoid allocating zero bytes, it generates an error message. */
!     utf8_str = alloc(ucs2_to_utf8(str, *lenp == 0 ? 1 : *lenp, NULL));
      if (utf8_str != NULL)
      {
! 	*lenp = ucs2_to_utf8(str, *lenp, utf8_str);
  
  	/* We might be called before we have p_enc set up. */
  	conv.vc_type = CONV_NONE;
--- 1196,1205 ----
      }
  
      /* Avoid allocating zero bytes, it generates an error message. */
!     utf8_str = alloc(utf16_to_utf8(str, *lenp == 0 ? 1 : *lenp, NULL));
      if (utf8_str != NULL)
      {
! 	*lenp = utf16_to_utf8(str, *lenp, utf8_str);
  
  	/* We might be called before we have p_enc set up. */
  	conv.vc_type = CONV_NONE;
***************
*** 1308,1314 ****
  		    if (hMemWstr[str_size] == NUL)
  			break;
  	    }
! 	    to_free = str = ucs2_to_enc((short_u *)hMemWstr, &str_size);
  	    GlobalUnlock(hMemW);
  	}
      }
--- 1333,1339 ----
  		    if (hMemWstr[str_size] == NUL)
  			break;
  	    }
! 	    to_free = str = utf16_to_enc((short_u *)hMemWstr, &str_size);
  	    GlobalUnlock(hMemW);
  	}
      }
***************
*** 1340,1346 ****
  
  # if defined(FEAT_MBYTE) && defined(WIN3264)
  	    /* The text is in the active codepage.  Convert to 'encoding',
! 	     * going through UCS-2. */
  	    acp_to_enc(str, str_size, &to_free, &maxlen);
  	    if (to_free != NULL)
  	    {
--- 1365,1371 ----
  
  # if defined(FEAT_MBYTE) && defined(WIN3264)
  	    /* The text is in the active codepage.  Convert to 'encoding',
! 	     * going through UTF-16. */
  	    acp_to_enc(str, str_size, &to_free, &maxlen);
  	    if (to_free != NULL)
  	    {
***************
*** 1404,1410 ****
      if (widestr != NULL)
      {
  	++*outlen;	/* Include the 0 after the string */
! 	*out = ucs2_to_enc((short_u *)widestr, outlen);
  	vim_free(widestr);
      }
  }
--- 1429,1435 ----
      if (widestr != NULL)
      {
  	++*outlen;	/* Include the 0 after the string */
! 	*out = utf16_to_enc((short_u *)widestr, outlen);
  	vim_free(widestr);
      }
  }
***************
*** 1466,1474 ****
  	WCHAR		*out;
  	int		len = metadata.txtlen;
  
! 	/* Convert the text to UCS-2. This is put on the clipboard as
  	 * CF_UNICODETEXT. */
! 	out = (WCHAR *)enc_to_ucs2(str, &len);
  	if (out != NULL)
  	{
  	    WCHAR *lpszMemW;
--- 1491,1499 ----
  	WCHAR		*out;
  	int		len = metadata.txtlen;
  
! 	/* Convert the text to UTF-16. This is put on the clipboard as
  	 * CF_UNICODETEXT. */
! 	out = (WCHAR *)enc_to_utf16(str, &len);
  	if (out != NULL)
  	{
  	    WCHAR *lpszMemW;
***************
*** 1488,1494 ****
  	    WideCharToMultiByte(GetACP(), 0, out, len,
  						  str, metadata.txtlen, 0, 0);
  
! 	    /* Allocate memory for the UCS-2 text, add one NUL word to
  	     * terminate the string. */
  	    hMemW = (LPSTR)GlobalAlloc(GMEM_MOVEABLE | GMEM_DDESHARE,
  						   (len + 1) * sizeof(WCHAR));
--- 1513,1519 ----
  	    WideCharToMultiByte(GetACP(), 0, out, len,
  						  str, metadata.txtlen, 0, 0);
  
! 	    /* Allocate memory for the UTF-16 text, add one NUL word to
  	     * terminate the string. */
  	    hMemW = (LPSTR)GlobalAlloc(GMEM_MOVEABLE | GMEM_DDESHARE,
  						   (len + 1) * sizeof(WCHAR));
*** ../vim-7.2.048/src/os_win32.c	Thu Jul 24 20:50:30 2008
--- src/os_win32.c	Thu Nov 20 16:44:33 2008
***************
*** 1587,1593 ****
  #ifdef FEAT_MBYTE
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	WCHAR	*p = enc_to_ucs2(name, NULL);
  	WCHAR	fnamew[_MAX_PATH];
  	WCHAR	*dumw;
  	long	n;
--- 1587,1593 ----
  #ifdef FEAT_MBYTE
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	WCHAR	*p = enc_to_utf16(name, NULL);
  	WCHAR	fnamew[_MAX_PATH];
  	WCHAR	*dumw;
  	long	n;
***************
*** 2440,2446 ****
  
  	if (GetCurrentDirectoryW(_MAX_PATH, wbuf) != 0)
  	{
! 	    char_u  *p = ucs2_to_enc(wbuf, NULL);
  
  	    if (p != NULL)
  	    {
--- 2440,2446 ----
  
  	if (GetCurrentDirectoryW(_MAX_PATH, wbuf) != 0)
  	{
! 	    char_u  *p = utf16_to_enc(wbuf, NULL);
  
  	    if (p != NULL)
  	    {
***************
*** 2466,2472 ****
  #ifdef FEAT_MBYTE
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	WCHAR	*p = enc_to_ucs2(name, NULL);
  	long	n;
  
  	if (p != NULL)
--- 2466,2472 ----
  #ifdef FEAT_MBYTE
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	WCHAR	*p = enc_to_utf16(name, NULL);
  	long	n;
  
  	if (p != NULL)
***************
*** 2495,2501 ****
  #ifdef FEAT_MBYTE
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	WCHAR	*p = enc_to_ucs2(name, NULL);
  	long	n;
  
  	if (p != NULL)
--- 2495,2501 ----
  #ifdef FEAT_MBYTE
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	WCHAR	*p = enc_to_utf16(name, NULL);
  	long	n;
  
  	if (p != NULL)
***************
*** 2522,2528 ****
      WCHAR	*p = NULL;
  
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
! 	p = enc_to_ucs2(name, NULL);
  #endif
  
  #ifdef FEAT_MBYTE
--- 2522,2528 ----
      WCHAR	*p = NULL;
  
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
! 	p = enc_to_utf16(name, NULL);
  #endif
  
  #ifdef FEAT_MBYTE
***************
*** 2590,2596 ****
      WCHAR	*wn = NULL;
  
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
! 	wn = enc_to_ucs2(fname, NULL);
      if (wn != NULL)
      {
  	hFile = CreateFileW(wn,		/* file name */
--- 2590,2596 ----
      WCHAR	*wn = NULL;
  
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
! 	wn = enc_to_utf16(fname, NULL);
      if (wn != NULL)
      {
  	hFile = CreateFileW(wn,		/* file name */
***************
*** 4239,4245 ****
  
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	wn = enc_to_ucs2(name, NULL);
  	if (wn != NULL)
  	{
  	    SetFileAttributesW(wn, FILE_ATTRIBUTE_NORMAL);
--- 4239,4245 ----
  
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	wn = enc_to_utf16(name, NULL);
  	if (wn != NULL)
  	{
  	    SetFileAttributesW(wn, FILE_ATTRIBUTE_NORMAL);
***************
*** 4382,4389 ****
  
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	wold = enc_to_ucs2((char_u *)pszOldFile, NULL);
! 	wnew = enc_to_ucs2((char_u *)pszNewFile, NULL);
  	if (wold != NULL && wnew != NULL)
  	    retval = mch_wrename(wold, wnew);
  	vim_free(wold);
--- 4382,4389 ----
  
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	wold = enc_to_utf16((char_u *)pszOldFile, NULL);
! 	wnew = enc_to_utf16((char_u *)pszNewFile, NULL);
  	if (wold != NULL && wnew != NULL)
  	    retval = mch_wrename(wold, wnew);
  	vim_free(wold);
***************
*** 4492,4498 ****
      WCHAR	*wn = NULL;
  
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
! 	wn = enc_to_ucs2(n, NULL);
  #endif
  
      if (mch_isdir(n))
--- 4492,4498 ----
      WCHAR	*wn = NULL;
  
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
! 	wn = enc_to_utf16(n, NULL);
  #endif
  
      if (mch_isdir(n))
***************
*** 4618,4624 ****
  
  #if defined(FEAT_MBYTE) || defined(PROTO)
  /*
!  * Version of open() that may use ucs2 file name.
   */
      int
  mch_open(char *name, int flags, int mode)
--- 4618,4624 ----
  
  #if defined(FEAT_MBYTE) || defined(PROTO)
  /*
!  * Version of open() that may use UTF-16 file name.
   */
      int
  mch_open(char *name, int flags, int mode)
***************
*** 4630,4636 ****
  
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	wn = enc_to_ucs2(name, NULL);
  	if (wn != NULL)
  	{
  	    f = _wopen(wn, flags, mode);
--- 4630,4636 ----
  
      if (enc_codepage >= 0 && (int)GetACP() != enc_codepage)
      {
! 	wn = enc_to_utf16(name, NULL);
  	if (wn != NULL)
  	{
  	    f = _wopen(wn, flags, mode);
***************
*** 4648,4654 ****
  }
  
  /*
!  * Version of fopen() that may use ucs2 file name.
   */
      FILE *
  mch_fopen(char *name, char *mode)
--- 4648,4654 ----
  }
  
  /*
!  * Version of fopen() that may use UTF-16 file name.
   */
      FILE *
  mch_fopen(char *name, char *mode)
***************
*** 4675,4682 ****
  	else if (newMode == 'b')
  	    _set_fmode(_O_BINARY);
  # endif
! 	wn = enc_to_ucs2(name, NULL);
! 	wm = enc_to_ucs2(mode, NULL);
  	if (wn != NULL && wm != NULL)
  	    f = _wfopen(wn, wm);
  	vim_free(wn);
--- 4675,4682 ----
  	else if (newMode == 'b')
  	    _set_fmode(_O_BINARY);
  # endif
! 	wn = enc_to_utf16(name, NULL);
! 	wm = enc_to_utf16(mode, NULL);
  	if (wn != NULL && wm != NULL)
  	    f = _wfopen(wn, wm);
  	vim_free(wn);
***************
*** 4776,4783 ****
      int			len;
  
      /* Convert the file names to wide characters. */
!     fromw = enc_to_ucs2(from, NULL);
!     tow = enc_to_ucs2(to, NULL);
      if (fromw != NULL && tow != NULL)
      {
  	/* Open the file for reading. */
--- 4776,4783 ----
      int			len;
  
      /* Convert the file names to wide characters. */
!     fromw = enc_to_utf16(from, NULL);
!     tow = enc_to_utf16(to, NULL);
      if (fromw != NULL && tow != NULL)
      {
  	/* Open the file for reading. */
***************
*** 5122,5128 ****
      for (i = 0; i < used_file_count; ++i)
      {
  	idx = used_file_indexes[i];
! 	str = ucs2_to_enc(ArglistW[idx], NULL);
  	if (str != NULL)
  	{
  #ifdef FEAT_DIFF
--- 5122,5128 ----
      for (i = 0; i < used_file_count; ++i)
      {
  	idx = used_file_indexes[i];
! 	str = utf16_to_enc(ArglistW[idx], NULL);
  	if (str != NULL)
  	{
  #ifdef FEAT_DIFF
*** ../vim-7.2.048/src/proto/os_mswin.pro	Sat May  5 19:07:50 2007
--- src/proto/os_mswin.pro	Thu Nov 20 16:41:00 2008
***************
*** 22,35 ****
  int can_end_termcap_mode __ARGS((int give_msg));
  int mch_screenmode __ARGS((char_u *arg));
  int mch_libcall __ARGS((char_u *libname, char_u *funcname, char_u *argstring, int argint, char_u **string_result, int *number_result));
! int utf8_to_ucs2 __ARGS((char_u *instr, int inlen, short_u *outstr, int *unconvlenp));
! int ucs2_to_utf8 __ARGS((short_u *instr, int inlen, char_u *outstr));
  void MultiByteToWideChar_alloc __ARGS((UINT cp, DWORD flags, LPCSTR in, int inlen, LPWSTR *out, int *outlen));
  void WideCharToMultiByte_alloc __ARGS((UINT cp, DWORD flags, LPCWSTR in, int inlen, LPSTR *out, int *outlen, LPCSTR def, LPBOOL useddef));
  int clip_mch_own_selection __ARGS((VimClipboard *cbd));
  void clip_mch_lose_selection __ARGS((VimClipboard *cbd));
! short_u *enc_to_ucs2 __ARGS((char_u *str, int *lenp));
! char_u *ucs2_to_enc __ARGS((short_u *str, int *lenp));
  void clip_mch_request_selection __ARGS((VimClipboard *cbd));
  void acp_to_enc __ARGS((char_u *str, int str_size, char_u **out, int *outlen));
  void clip_mch_set_selection __ARGS((VimClipboard *cbd));
--- 22,35 ----
  int can_end_termcap_mode __ARGS((int give_msg));
  int mch_screenmode __ARGS((char_u *arg));
  int mch_libcall __ARGS((char_u *libname, char_u *funcname, char_u *argstring, int argint, char_u **string_result, int *number_result));
! int utf8_to_utf16 __ARGS((char_u *instr, int inlen, short_u *outstr, int *unconvlenp));
! int utf16_to_utf8 __ARGS((short_u *instr, int inlen, char_u *outstr));
  void MultiByteToWideChar_alloc __ARGS((UINT cp, DWORD flags, LPCSTR in, int inlen, LPWSTR *out, int *outlen));
  void WideCharToMultiByte_alloc __ARGS((UINT cp, DWORD flags, LPCWSTR in, int inlen, LPSTR *out, int *outlen, LPCSTR def, LPBOOL useddef));
  int clip_mch_own_selection __ARGS((VimClipboard *cbd));
  void clip_mch_lose_selection __ARGS((VimClipboard *cbd));
! short_u *enc_to_utf16 __ARGS((char_u *str, int *lenp));
! char_u *utf16_to_enc __ARGS((short_u *str, int *lenp));
  void clip_mch_request_selection __ARGS((VimClipboard *cbd));
  void acp_to_enc __ARGS((char_u *str, int str_size, char_u **out, int *outlen));
  void clip_mch_set_selection __ARGS((VimClipboard *cbd));
*** ../vim-7.2.048/src/version.c	Thu Nov 20 16:11:03 2008
--- src/version.c	Thu Nov 20 17:08:07 2008
***************
*** 678,679 ****
--- 678,681 ----
  {   /* Add new patch number below this line */
+ /**/
+     49,
  /**/

-- 
FIRST HEAD: All right! All right! We'll kill him first and then have tea and
            biscuits.
                 "Monty Python and the Holy Grail" PYTHON (MONTY) PICTURES LTD

 /// Bram Moolenaar -- Bram@Moolenaar.net -- http://www.Moolenaar.net   \\\
///        sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\
\\\        download, build and distribute -- http://www.A-A-P.org        ///
 \\\            help me help AIDS victims -- http://ICCF-Holland.org    ///