*	The caller must ensure that the source buffer is long enough that this
 *	routine does not run off the end and dereference non-existent memory
 *	looking for trail bytes. If the source buffer is known to be '\0'
 *	terminated, this cannot happen. Otherwise, the caller should call
 *	Tcl_UtfCharComplete() before calling this routine to ensure that
 *	enough bytes remain in the string.
 *
 *	If TCL_UTF_MAX <= 4, special handling of Surrogate pairs is done:
 *	Special handling of Surrogate pairs is done:
 *	For any UTF-8 string containing a character outside of the BMP, the
 *	first call to this function will fill *chPtr with the high surrogate
 *	and generate a return value of 1. Calling Tcl_UtfToUniChar again
 *	will produce the low surrogate and a return value of 3. Because *chPtr
 *	is used to remember whether the high surrogate is already produced, it
 *	is recommended to initialize the variable it points to as 0 before
 *	the first call to Tcl_UtfToUniChar is done.

    wString = (int *) (Tcl_DStringValue(dsPtr) + oldLength);

    w = wString;
    p = src;
    endPtr = src + length;
    optPtr = endPtr - 4;
    while (p <= optPtr) {
	p += TclUtfToUCS4(p, &ch);
	p += Tcl_UtfToUniChar(p, &ch);
	*w++ = ch;
    }
    while ((p < endPtr) && Tcl_UtfCharComplete(p, endPtr-p)) {
	p += TclUtfToUCS4(p, &ch);
	p += Tcl_UtfToUniChar(p, &ch);
	*w++ = ch;
    }
    while (p < endPtr) {
	*w++ = UCHAR(*p++);
    }
    *w = '\0';
    Tcl_DStringSetLength(dsPtr,

	    }
	    i++;
	}
    }
    return i;
}

#if (TCL_UTF_MAX > 3) && !defined(TCL_NO_DEPRECATED)
#if !defined(TCL_NO_DEPRECATED)
#undef Tcl_NumUtfChars
int
Tcl_NumUtfChars(
    const char *src,	/* The UTF-8 string to measure. */
    int length)		/* The length of the string in bytes, or -1
			 * for strlen(string). */
{

const char *
Tcl_UtfFindFirst(
    const char *src,		/* The UTF-8 string to be searched. */
    int ch)			/* The Unicode character to search for. */
{
    while (1) {
	int find, len = TclUtfToUCS4(src, &find);
	int find, len = Tcl_UtfToUniChar(src, &find);

	if (find == ch) {
	    return src;
	}
	if (*src == '\0') {
	    return NULL;
	}

Tcl_UtfFindLast(
    const char *src,		/* The UTF-8 string to be searched. */
    int ch)			/* The Unicode character to search for. */
{
    const char *last = NULL;

    while (1) {
	int find, len = TclUtfToUCS4(src, &find);
	int find, len = Tcl_UtfToUniChar(src, &find);

	if (find == ch) {
	    last = src;
	}
	if (*src == '\0') {
	    break;
	}

	i = Tcl_UtfToChar16(src, &ch);
	src += i;
    }
    if ((ch >= 0xD800) && (i < 3)) {
	/* Index points at character following high Surrogate */
	return -1;
    }
    TclUtfToUCS4(src, &i);
    Tcl_UtfToUniChar(src, &i);
    return i;
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UtfAtIndex --
 *
 *	Returns a pointer to the specified character (not byte) position in
 *	the UTF-8 string. If TCL_UTF_MAX < 4, characters > U+FFFF count as
 *	the UTF-8 string.
 *	2 positions, but then the pointer should never be placed between
 *	the two positions.
 *
 * Results:
 *	As above.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */

#if TCL_UTF_MAX < 4
#   undef Tcl_UtfToUniChar
#   define Tcl_UtfToUniChar Tcl_UtfToChar16
#endif

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

    while (index-- > 0) {
	len = (Tcl_UtfToUniChar)(src, &ch);
	src += len;
    }
#if TCL_UTF_MAX < 4
    if ((ch >= 0xD800) && (len < 3)) {
	/* Index points at character following high Surrogate */
	src += (Tcl_UtfToUniChar)(src, &ch);
    }
#endif
    return src;
}

#if (TCL_UTF_MAX > 3) && !defined(TCL_NO_DEPRECATED)
#if !defined(TCL_NO_DEPRECATED)
#undef Tcl_UtfAtIndex
const char *
Tcl_UtfAtIndex(
    const char *src,	/* The UTF-8 string. */
    int index)		/* The position of the desired character. */
{
    unsigned short ch = 0;

    /*
     * Iterate over the string until we hit the terminating null.
     */

    src = dst = str;
    while (*src) {
	len = TclUtfToUCS4(src, &ch);
	len = Tcl_UtfToUniChar(src, &ch);
	upChar = Tcl_UniCharToUpper(ch);

	/*
	 * To keep badly formed Utf strings from getting inflated by the
	 * conversion (thereby causing a segfault), only copy the upper case
	 * char to dst if its size is <= the original char.
	 */

    /*
     * Iterate over the string until we hit the terminating null.
     */

    src = dst = str;
    while (*src) {
	len = TclUtfToUCS4(src, &ch);
	len = Tcl_UtfToUniChar(src, &ch);
	lowChar = Tcl_UniCharToLower(ch);

	/*
	 * To keep badly formed Utf strings from getting inflated by the
	 * conversion (thereby causing a segfault), only copy the lower case
	 * char to dst if its size is <= the original char.
	 */

     * Capitalize the first character and then lowercase the rest of the
     * characters until we get to a null.
     */

    src = dst = str;

    if (*src) {
	len = TclUtfToUCS4(src, &ch);
	len = Tcl_UtfToUniChar(src, &ch);
	titleChar = Tcl_UniCharToTitle(ch);

	if ((len < TclUtfCount(titleChar)) || ((titleChar & ~0x7FF) == 0xD800)) {
	    memmove(dst, src, len);
	    dst += len;
	} else {
	    dst += Tcl_UniCharToUtf(titleChar, dst);
	}
	src += len;
    }
    while (*src) {
	len = TclUtfToUCS4(src, &ch);
	len = Tcl_UtfToUniChar(src, &ch);
	lowChar = ch;
	/* Special exception for Georgian Asomtavruli chars, no titlecase. */
	if ((unsigned)(lowChar - 0x1C90) >= 0x30) {
	    lowChar = Tcl_UniCharToLower(lowChar);
	}

	if ((len < TclUtfCount(lowChar)) || ((lowChar & ~0x7FF) == 0xD800)) {

	 * only when both strings are of at least n chars long (no need for \0
	 * check)
	 */

	cs += TclUtfToUniChar(cs, &ch1);
	ct += TclUtfToUniChar(ct, &ch2);
	if (ch1 != ch2) {
#if TCL_UTF_MAX < 4
	    /* Surrogates always report higher than non-surrogates */
	    if (((ch1 & 0xFC00) == 0xD800)) {
	    if ((ch2 & 0xFC00) != 0xD800) {
		return ch1;
	    }
	    } else if ((ch2 & 0xFC00) == 0xD800) {
		return -ch2;
	    }
#endif
	    return (ch1 - ch2);
	}
    }
    return 0;
}

/*

	 * n must be interpreted as chars, not bytes.
	 * This should be called only when both strings are of
	 * at least n chars long (no need for \0 check)
	 */
	cs += TclUtfToUniChar(cs, &ch1);
	ct += TclUtfToUniChar(ct, &ch2);
	if (ch1 != ch2) {
#if TCL_UTF_MAX < 4
	    /* Surrogates always report higher than non-surrogates */
	    if (((ch1 & 0xFC00) == 0xD800)) {
	    if ((ch2 & 0xFC00) != 0xD800) {
		return ch1;
	    }
	    } else if ((ch2 & 0xFC00) == 0xD800) {
		return -ch2;
	    }
#endif
	    ch1 = Tcl_UniCharToLower(ch1);
	    ch2 = Tcl_UniCharToLower(ch2);
	    if (ch1 != ch2) {
		return (ch1 - ch2);
	    }
	}
    }

{
    Tcl_UniChar ch1 = 0, ch2 = 0;

    while (*cs && *ct) {
	cs += TclUtfToUniChar(cs, &ch1);
	ct += TclUtfToUniChar(ct, &ch2);
	if (ch1 != ch2) {
#if TCL_UTF_MAX < 4
	    /* Surrogates always report higher than non-surrogates */
	    if (((ch1 & 0xFC00) == 0xD800)) {
	    if ((ch2 & 0xFC00) != 0xD800) {
		return ch1;
	    }
	    } else if ((ch2 & 0xFC00) == 0xD800) {
		return -ch2;
	    }
#endif
	    return ch1 - ch2;
	}
    }
    return UCHAR(*cs) - UCHAR(*ct);
}



{
    Tcl_UniChar ch1 = 0, ch2 = 0;

    while (*cs && *ct) {
	cs += TclUtfToUniChar(cs, &ch1);
	ct += TclUtfToUniChar(ct, &ch2);
	if (ch1 != ch2) {
#if TCL_UTF_MAX < 4
	    /* Surrogates always report higher than non-surrogates */
	    if (((ch1 & 0xFC00) == 0xD800)) {
	    if ((ch2 & 0xFC00) != 0xD800) {
		return ch1;
	    }
	    } else if ((ch2 & 0xFC00) == 0xD800) {
		return -ch2;
	    }
#endif
	    ch1 = Tcl_UniCharToLower(ch1);
	    ch2 = Tcl_UniCharToLower(ch2);
	    if (ch1 != ch2) {
		return ch1 - ch2;
	    }
	}
    }

int
TclUniCharNcmp(
    const Tcl_UniChar *ucs,	/* Unicode string to compare to uct. */
    const Tcl_UniChar *uct,	/* Unicode string ucs is compared to. */
    unsigned long numChars)	/* Number of unichars to compare. */
{
#if defined(WORDS_BIGENDIAN) && (TCL_UTF_MAX > 3)
#if defined(WORDS_BIGENDIAN)
    /*
     * We are definitely on a big-endian machine; memcmp() is safe
     */

    return memcmp(ucs, uct, numChars*sizeof(Tcl_UniChar));

#else /* !WORDS_BIGENDIAN */
    /*
     * We can't simply call memcmp() because that is not lexically correct.
     */

    for ( ; numChars != 0; ucs++, uct++, numChars--) {
	if (*ucs != *uct) {
	    return (*ucs - *uct);
	}
    }
    return 0;
#endif /* WORDS_BIGENDIAN */
}

#if (TCL_UTF_MAX > 3) && !defined(TCL_NO_DEPRECATED)
#if !defined(TCL_NO_DEPRECATED)
int
Tcl_UniCharNcmp(
    const unsigned short *ucs,	/* Unicode string to compare to uct. */
    const unsigned short *uct,	/* Unicode string ucs is compared to. */
    unsigned long numChars)	/* Number of unichars to compare. */
{
#if defined(WORDS_BIGENDIAN) && (TCL_UTF_MAX > 3)
#if defined(WORDS_BIGENDIAN)
    /*
     * We are definitely on a big-endian machine; memcmp() is safe
     */

    return memcmp(ucs, uct, numChars*sizeof(Tcl_UniChar));

#else /* !WORDS_BIGENDIAN */

		return (lcs - lct);
	    }
	}
    }
    return 0;
}

#if (TCL_UTF_MAX > 3) && !defined(TCL_NO_DEPRECATED)
#if !defined(TCL_NO_DEPRECATED)
int
Tcl_UniCharNcasecmp(
    const unsigned short *ucs,	/* Unicode string to compare to uct. */
    const unsigned short *uct,	/* Unicode string ucs is compared to. */
    unsigned long numChars)	/* Number of unichars to compare. */
{
    for ( ; numChars != 0; numChars--, ucs++, uct++) {

	    return 0;
	}
	uniStr++;
	uniPattern++;
    }
}

#if (TCL_UTF_MAX > 3) && !defined(TCL_NO_DEPRECATED)
#if !defined(TCL_NO_DEPRECATED)
int
Tcl_UniCharCaseMatch(
    const unsigned short *uniStr,	/* Unicode String. */
    const unsigned short *uniPattern,
				/* Pattern, which may contain special
				 * characters. */
    int nocase)			/* 0 for case sensitive, 1 for insensitive */

	    return 0;
	}
	string++;
	pattern++;
    }
}

/*
 *---------------------------------------------------------------------------
 *
 * TclUtfToUCS4 --
 *
 *	Extracts the 4-byte codepoint from the leading bytes of the
 *	Modified UTF-8 string "src".  This is a utility routine to
 *	contain the surrogate gymnastics in one place.
 *
 *	The caller must ensure that the source buffer is long enough that this
 *	routine does not run off the end and dereference non-existent memory
 *	looking for trail bytes. If the source buffer is known to be '\0'
 *	terminated, this cannot happen. Otherwise, the caller should call
 *	Tcl_UtfCharComplete() before calling this routine to ensure that
 *	enough bytes remain in the string.
 *
 * Results:
 *	Fills *usc4Ptr with the UCS4 code point and returns the number of bytes
 *	consumed from the source string.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */

#if TCL_UTF_MAX < 4
int
TclUtfToUCS4(
    const char *src,	/* The UTF-8 string. */
    int *ucs4Ptr)	/* Filled with the UCS4 codepoint represented
			 * by the UTF-8 string. */
{
#   undef Tcl_UtfToUniChar
    return Tcl_UtfToUniChar(src, ucs4Ptr);
}

int
TclUniCharToUCS4(
    const Tcl_UniChar *src,	/* The Tcl_UniChar string. */
    int *ucs4Ptr)	/* Filled with the UCS4 codepoint represented
			 * by the Tcl_UniChar string. */
{
    if (((src[0] & 0xFC00) == 0xD800) && ((src[1] & 0xFC00) == 0xDC00)) {
	*ucs4Ptr = (((src[0] & 0x3FF) << 10) | (src[1] & 0x3FF)) + 0x10000;
	return 2;
    }
    *ucs4Ptr = src[0];
    return 1;
}

const Tcl_UniChar *TclUCS4Prev(const Tcl_UniChar *src, const Tcl_UniChar *ptr) {
    if (src <= ptr + 1) {
    	return ptr;
    }
    if (((src[-1] & 0xFC00) == 0xDC00) && ((src[-2] & 0xFC00) == 0xD800)) {
	return src - 2;
    }
    return src - 1;
}



#endif

/*
 * Local Variables:
 * mode: c
 * c-basic-offset: 4
 * fill-column: 78
 * End:
 */