};

/*
 * Functions used only in this module.
 */

static int		Invalid(unsigned char *src);

#define UCS4ToUpper Tcl_UniCharToUpper
#define UCS4ToLower Tcl_UniCharToLower
#define UCS4ToTitle Tcl_UniCharToTitle


/*
 *---------------------------------------------------------------------------
 *
 * TclUtfCount --
 *
 *	Find the number of bytes in the Utf character "ch".

static int
Invalid(
    unsigned char *src)	/* Points to lead byte of a UTF-8 byte sequence */
{
    unsigned char byte = *src;
    int index;

    if (byte % 0x04) {
    if ((byte & 0xC3) != 0xC0) {
	/* Only lead bytes 0xC0, 0xE0, 0xF0, 0xF4 need examination */
	return 0;
    }
    index = (byte - 0xC0) >> 1;
    if (src[1] < bounds[index] || src[1] > bounds[index+1]) {
	/* Out of bounds - report invalid. */
	return 1;

   0x2DC, 0x2122, 0x0161, 0x203A, 0x0153,   0x9D, 0x017E, 0x0178
};

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

    /*
     * Unroll 1 to 4 byte UTF-8 sequences.
     */

	}

	/*
	 * A three-byte-character lead-byte not followed by two trail-bytes
	 * represents itself.
	 */
    }
    else if (byte < 0xF8) {
    else if (byte < 0xF5) {
	if (((src[1] & 0xC0) == 0x80) && ((src[2] & 0xC0) == 0x80) && ((src[3] & 0xC0) == 0x80)) {
	    /*
	     * Four-byte-character lead byte followed by three trail bytes.
	     */
	    *chPtr = (((byte & 0x07) << 18) | ((src[1] & 0x3F) << 12)
		    | ((src[2] & 0x3F) << 6) | (src[3] & 0x3F));
	    if ((unsigned)(*chPtr - 0x10000) <= 0xFFFFF) {

    *chPtr = byte;
    return 1;
}

int
Tcl_UtfToChar16(
    const char *src,	/* The UTF-8 string. */
    unsigned short *chPtr)/* Filled with the unsigned short represented by
				 * the UTF-8 string. */
    unsigned short *chPtr)/* Filled with the Tcl_UniChar represented by
				 * the UTF-8 string. This could be a surrogate too. */
{
    unsigned short byte;

    /*
     * Unroll 1 to 4 byte UTF-8 sequences.
     */

	}

	/*
	 * A three-byte-character lead-byte not followed by two trail-bytes
	 * represents itself.
	 */
    }
    else if (byte < 0xF8) {
	if (((src[1] & 0xC0) == 0x80) && ((src[2] & 0xC0) == 0x80) && ((src[3] & 0xC0) == 0x80)) {
    else if (byte < 0xF5) {
	if (((src[1] & 0xC0) == 0x80) && ((src[2] & 0xC0) == 0x80)) {
	    /*
	     * Four-byte-character lead byte followed by three trail bytes.
	     * Four-byte-character lead byte followed by at least two trail bytes.
	     * We don't test the validity of 3th trail byte, see [ed29806ba]
	     */
	    unsigned short high = (((byte & 0x07) << 8) | ((src[1] & 0x3F) << 2)
	    Tcl_UniChar high = (((byte & 0x07) << 8) | ((src[1] & 0x3F) << 2)
		    | ((src[2] & 0x3F) >> 4)) - 0x40;
	    if (high >= 0x400) {
	    if (high < 0x400) {
		/* out of range, < 0x10000 or > 0x10FFFF */
	    } else {
		/* produce high surrogate, advance source pointer */
		*chPtr = 0xD800 + high;
		return 1;
	    }
	    /* out of range, < 0x10000 or > 0x10FFFF */
	}

	/*
	 * A four-byte-character lead-byte not followed by three trail-bytes
	 * represents itself.
	 */
    }

int
Tcl_UtfCharComplete(
    const char *src,		/* String to check if first few bytes contain
				 * a complete UTF-8 character. */
    int length)			/* Length of above string in bytes. */
{
    return length >= complete[(unsigned char)*src];
    return length >= complete[UCHAR(*src)];
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_NumUtfChars --
 *

    if (length < 0) {
	while (*src != '\0') {
	    src += TclUtfToUniChar(src, &ch);
	    i++;
	}
	if (i < 0) i = INT_MAX; /* Bug [2738427] */
    } else {
	const char *endPtr = src + length - 4;
	const char *endPtr = src + length - TCL_UTF_MAX;

	while (src < endPtr) {
	    src += TclUtfToUniChar(src, &ch);
	    i++;
	}
	endPtr += 4;
	endPtr += TCL_UTF_MAX;
	while ((src < endPtr) && Tcl_UtfCharComplete(src, endPtr - src)) {
	    src += TclUtfToUniChar(src, &ch);
	    i++;
	}
	if (src < endPtr) {
	    i += endPtr - src;
	}

	     * before the needs of the lead byte were satisfied.
	     * Let the (malformed) lead byte alone be a character
	     */
	    return src + 1;
	}
	next++;
    }
    if ((next == src + 1) || Invalid((unsigned char *)src)) {
    if (Invalid((unsigned char *)src)) {
	return src + 1;
    }
    return next;
}

/*
 *---------------------------------------------------------------------------

	     * and exit this loop before we run past the start of the string.
	     */
	    return fallback;
	}

	/* Continue the search backwards... */
	look--;
    } while (trailBytesSeen < 4);
    } while (trailBytesSeen < TCL_UTF_MAX);

    /*
     * We've seen TCL_UTF_MAX trail bytes, so we know there will not be a
     * properly formed byte sequence to find, and we can stop looking,
     * accepting the fallback (for TCL_UTF_MAX > 3) or just go back as
     * far as we can.
     */
#if TCL_UTF_MAX > 3
    return fallback;
#else
    return src - TCL_UTF_MAX;
#endif
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UniCharAtIndex --
 *

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

    src = dst = str;
    while (*src) {
	len = TclUtfToUCS4(src, &ch);
	upChar = UCS4ToUpper(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.
	 */

	if ((len < TclUtfCount(upChar)) || ((upChar & 0xF800) == 0xD800)) {
	if ((len < TclUtfCount(upChar)) || ((upChar & ~0x7FF) == 0xD800)) {
	    memmove(dst, src, len);
	    dst += len;
	} else {
	    dst += Tcl_UniCharToUtf(upChar, dst);
	}
	src += len;
    }

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

    src = dst = str;
    while (*src) {
	len = TclUtfToUCS4(src, &ch);
	lowChar = UCS4ToLower(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.
	 */

	if ((len < TclUtfCount(lowChar)) || ((lowChar & 0xF800) == 0xD800)) {
	if ((len < TclUtfCount(lowChar)) || ((lowChar & ~0x7FF) == 0xD800)) {
	    memmove(dst, src, len);
	    dst += len;
	} else {
	    dst += Tcl_UniCharToUtf(lowChar, dst);
	}
	src += len;
    }

     * characters until we get to a null.
     */

    src = dst = str;

    if (*src) {
	len = TclUtfToUCS4(src, &ch);
	titleChar = UCS4ToTitle(ch);
	titleChar = Tcl_UniCharToTitle(ch);

	if ((len < TclUtfCount(titleChar)) || ((titleChar & 0xF800) == 0xD800)) {
	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);
	lowChar = ch;
	/* Special exception for Georgian Asomtavruli chars, no titlecase. */
	if ((unsigned)(lowChar - 0x1C90) >= 0x30) {
	    lowChar = UCS4ToLower(lowChar);
	    lowChar = Tcl_UniCharToLower(lowChar);
	}

	if ((len < TclUtfCount(lowChar)) || ((lowChar & 0xF800) == 0xD800)) {
	if ((len < TclUtfCount(lowChar)) || ((lowChar & ~0x7FF) == 0xD800)) {
	    memmove(dst, src, len);
	    dst += len;
	} else {
	    dst += Tcl_UniCharToUtf(lowChar, dst);
	}
	src += len;
    }

    if (!UNICODE_OUT_OF_RANGE(ch)) {
	int info = GetUniCharInfo(ch);

	if (GetCaseType(info) & 0x04) {
	    ch -= GetDelta(info);
	}
    }
    /* Clear away extension bits, if any */
    return ch & 0x1FFFFF;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharToLower --

	int info = GetUniCharInfo(ch);
	int mode = GetCaseType(info);

	if ((mode & 0x02) && (mode != 0x7)) {
	    ch += GetDelta(info);
	}
    }
    /* Clear away extension bits, if any */
    return ch & 0x1FFFFF;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharToTitle --

	    if (mode != 0x7) {
		ch += ((mode & 0x4) ? -1 : 1);
	    }
	} else if (mode == 0x4) {
	    ch -= GetDelta(info);
	}
    }
    /* Clear away extension bits, if any */
    return ch & 0x1FFFFF;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharLen --

 */

int
Tcl_UniCharIsControl(
    int ch)			/* Unicode character to test. */
{
    if (UNICODE_OUT_OF_RANGE(ch)) {
	/* Clear away extension bits, if any */
	ch &= 0x1FFFFF;
	if ((ch == 0xE0001) || ((ch >= 0xE0020) && (ch <= 0xE007F))) {
	    return 1;
	}
	if ((ch >= 0xF0000) && ((ch & 0xFFFF) <= 0xFFFD)) {
	    return 1;
	}