*	None.
 *
 *---------------------------------------------------------------------------
 */

int
Tcl_UtfToUniChar(
    register const char *src,	/* The UTF-8 string. */
    register Tcl_UniChar *chPtr)/* Filled with the Tcl_UniChar represented by
				 * the UTF-8 string. */
{
    Tcl_UniChar byte;

    /*
     * Unroll 1 to 3 (or 4) byte UTF-8 sequences.
     */

 *	None.
 *
 *---------------------------------------------------------------------------
 */

int
Tcl_NumUtfChars(
    register const char *src,	/* The UTF-8 string to measure. */
    int length)			/* The length of the string in bytes, or -1
				 * for strlen(string). */
{
    Tcl_UniChar ch = 0;
    register int i = 0;

    /*
     * The separate implementations are faster.
     *
     * Since this is a time-sensitive function, we also do the check for the
     * single-byte char case specially.
     */

    if (length < 0) {
	while (*src != '\0') {
	    src += TclUtfToUniChar(src, &ch);
	    i++;
	}
	if (i < 0) i = INT_MAX; /* Bug [2738427] */
    } else {
	register const char *endPtr = src + length - TCL_UTF_MAX;

	while (src < endPtr) {

	    if (((unsigned)UCHAR(*src) - 0xF0) < 5) {
		/* treat F0 - F4 as single character */
		ch = 0;
		src++;
	    } else {

		src += TclUtfToUniChar(src, &ch);
	    }
	    i++;
	}
	endPtr += TCL_UTF_MAX;
	while ((src < endPtr) && Tcl_UtfCharComplete(src, endPtr - src)) {

	    if (((unsigned)UCHAR(*src) - 0xF0) < 5) {
		/* treat F0 - F4 as single character */
		ch = 0;
		src++;
	    } else {

		src += TclUtfToUniChar(src, &ch);
	    }
	    i++;
	}
	if (src < endPtr) {
	    i += endPtr - src;
	}
    }
    return i;

 *	None.
 *
 *---------------------------------------------------------------------------
 */

Tcl_UniChar
Tcl_UniCharAtIndex(
    register const char *src,	/* The UTF-8 string to dereference. */
    register int index)		/* The position of the desired character. */
{
    Tcl_UniChar ch = 0;

    TclUtfToUniChar(Tcl_UtfAtIndex(src, index), &ch);
    return ch;
}


 *	None.
 *
 *---------------------------------------------------------------------------
 */

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

    while (index-- > 0) {
	len = TclUtfToUniChar(src, &ch);
	src += len;

{
    /*
     * We can't simply call 'memcmp(cs, ct, numBytes);' because we need to
     * check for Tcl's \xC0\x80 non-utf-8 null encoding. Otherwise utf-8 lexes
     * fine in the strcmp manner.
     */

    register int result = 0;

    for ( ; numBytes != 0; numBytes--, cs++, ct++) {
	if (*cs != *ct) {
	    result = UCHAR(*cs) - UCHAR(*ct);
	    break;
	}
    }