*
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */

#undef Tcl_UniCharToUtf
Tcl_Size
Tcl_UniCharToUtf(
    int ch,	/* The Tcl_UniChar to be stored in the
		 * buffer. Can be or'ed with flag TCL_COMBINE
		 */
    char *buf)	/* Buffer in which the UTF-8 representation of
		 * ch is stored. Must be large enough to hold the UTF-8
		 * character (at most 4 bytes).
		 */
{
    int flags = ch;

    if (ch >= 0) {
	if (ch <= 0x7FF) {
	    buf[1] = (char) (0x80 | (0x3F & ch));
	    buf[0] = (char) (0xC0 | (ch >> 6));
	    return 2;
	}
	if (ch <= 0xFFFF) {
	    if (
		    (flags & TCL_COMBINE) &&
		    ((ch & 0xF800) == 0xD800)) {
		if (ch & 0x0400) {
		    /* Low surrogate */
		    if (   (0x80 == (0xC0 & buf[0]))
			&& (0    == (0xCF & buf[1]))) {
			/* Previous Tcl_UniChar was a high surrogate, so combine */
			buf[2]  = (char) (0x80 | (0x3F & ch));

 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */

#undef Tcl_UniCharToUtfDString
char *
Tcl_UniCharToUtfDString(
    const int *uniStr,	/* Unicode string to convert to UTF-8. */
    Tcl_Size uniLength,		/* Length of Unicode string. Negative for nul
    				 * nul terminated string */
    Tcl_DString *dsPtr)		/* UTF-8 representation of string is appended
				 * to this previously initialized DString. */
{
    const int *w, *wEnd;
    char *p, *string;
    Tcl_Size oldLength;

    const unsigned short *uniStr,/* Utf-16 string to convert to UTF-8. */
    Tcl_Size uniLength,		/* Length of Utf-16 string. */
    Tcl_DString *dsPtr)		/* UTF-8 representation of string is appended
				 * to this previously initialized DString. */
{
    const unsigned short *w, *wEnd;
    char *p, *string;
    size_t oldLength;
    int len = 1;

    /*
     * UTF-8 string length in bytes will be <= Utf16 string length * 3.
     */

    if (uniStr == NULL) {

static const unsigned short cp1252[32] = {
  0x20AC,   0x81, 0x201A, 0x0192, 0x201E, 0x2026, 0x2020, 0x2021,
  0x02C6, 0x2030, 0x0160, 0x2039, 0x0152,   0x8D, 0x017D,   0x8F,
    0x90, 0x2018, 0x2019, 0x201C, 0x201D, 0x2022, 0x2013, 0x2014,
   0x2DC, 0x2122, 0x0161, 0x203A, 0x0153,   0x9D, 0x017E, 0x0178
};

#undef Tcl_UtfToUniChar
Tcl_Size
Tcl_UtfToUniChar(
    const char *src,	/* The UTF-8 string. */
    int *chPtr)/* Filled with the Unicode character represented by
				 * the UTF-8 string. */
{
    int byte;

 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */

#undef Tcl_UtfToUniCharDString
int *
Tcl_UtfToUniCharDString(
    const char *src,		/* UTF-8 string to convert to Unicode. */
    Tcl_Size length,		/* Length of UTF-8 string in bytes, or -1 for
				 * strlen(). */
    Tcl_DString *dsPtr)		/* Unicode representation of string is
				 * appended to this previously initialized

 *---------------------------------------------------------------------------
 */

const char *
Tcl_UtfNext(
    const char *src)		/* The current location in the string. */
{
    size_t left;
    const char *next;

    if (((*src) & 0xC0) == 0x80) {
	/* Continuation byte, so we start 'inside' a (possible valid) UTF-8
	 * sequence. Since we are not allowed to access src[-1], we cannot
	 * check if the sequence is actually valid, the best we can do is
	 * just assume it is valid and locate the end. */

 */

const char *
Tcl_UtfAtIndex(
    const char *src,	/* The UTF-8 string. */
    Tcl_Size index)	/* The position of the desired character. */
{
    int ch = 0;

    if (index > 0) {
	while (index--) {
	    /* Make use of the #undef Tcl_UtfToUniChar above, which already handles UCS4. */
	    src += Tcl_UtfToUniChar(src, &ch);
	}
    }
    return src;
}

const char *
TclUtfAtIndex(
    const char *src,	/* The UTF-8 string. */

 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

#undef Tcl_UniCharLen
Tcl_Size
Tcl_UniCharLen(
    const int *uniStr)	/* Unicode string to find length of. */
{
    Tcl_Size len = 0;

    while (*uniStr != '\0') {