{
    if ((unsigned)(ch - 1) < (UNICODE_SELF - 1)) {
	return 1;
    }
    if (ch <= 0x7FF) {
	return 2;
    }

    if (((unsigned)(ch - 0x10000) <= 0xFFFFF)) {
	return 4;
    }

    return 3;
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UniCharToUtf --

int
Tcl_UniCharToUtf(
    int ch,			/* The Tcl_UniChar to be stored in the
				 * buffer. */
    char *buf)			/* Buffer in which the UTF-8 representation of
				 * the Tcl_UniChar is stored. Buffer must be
				 * large enough to hold the UTF-8 character
				 * (at most 4 bytes). */
{
    if ((unsigned)(ch - 1) < (UNICODE_SELF - 1)) {
	buf[0] = (char) ch;
	return 1;
    }
    if (ch >= 0) {
	if (ch <= 0x7FF) {
	    buf[1] = (char) ((ch | 0x80) & 0xBF);
	    buf[0] = (char) ((ch >> 6) | 0xC0);
	    return 2;
	}
	if (ch <= 0xFFFF) {

	    if ((ch & 0xF800) == 0xD800) {
		if (ch & 0x0400) {
		    /* Low surrogate */
		    if (((buf[0] & 0xF8) == 0xF0) && ((buf[1] & 0xC0) == 0x80)
			    && ((buf[2] & 0xCF) == 0)) {
			/* Previous Tcl_UniChar was a High surrogate, so combine */
			buf[3] = (char) ((ch & 0x3F) | 0x80);
			buf[2] |= (char) (((ch >> 6) & 0x0F) | 0x80);
			return 4;
		    }
		    /* Previous Tcl_UniChar was not a High surrogate, so just output */
		} else {
		    /* High surrogate */
		    ch += 0x40;
		    /* Fill buffer with specific 3-byte (invalid) byte combination,
		       so following Low surrogate can recognize it and combine */
		    buf[2] = (char) ((ch << 4) & 0x30);
		    buf[1] = (char) (((ch >> 2) & 0x3F) | 0x80);
		    buf[0] = (char) (((ch >> 8) & 0x07) | 0xF0);
		    return 0;
		}
	    }

	    goto three;
	}


	if (ch <= 0x10FFFF) {
	    buf[3] = (char) ((ch | 0x80) & 0xBF);
	    buf[2] = (char) (((ch >> 6) | 0x80) & 0xBF);
	    buf[1] = (char) (((ch >> 12) | 0x80) & 0xBF);
	    buf[0] = (char) ((ch >> 18) | 0xF0);
	    return 4;
	}

    }

    ch = 0xFFFD;
three:
    buf[2] = (char) ((ch | 0x80) & 0xBF);
    buf[1] = (char) (((ch >> 6) | 0x80) & 0xBF);
    buf[0] = (char) ((ch >> 12) | 0xE0);

 *	number of bytes from the UTF-8 string that were consumed.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */

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
};

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. */
{
    register int byte;

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

    byte = *((unsigned char *) src);
    if (byte < 0xC0) {
	/*
	 * Handles properly formed UTF-8 characters between 0x01 and 0x7F.
	 * Treats naked trail bytes 0x80 to 0x9F as valid characters from
	 * the cp1252 table. See: <https://en.wikipedia.org/wiki/UTF-8>
	 * Also treats \0 and other naked trail bytes 0xA0 to 0xBF as valid
	 * characters representing themselves.
	 */

	if ((unsigned)(byte-0x80) < (unsigned) 0x20) {
	    *chPtr = (Tcl_UniChar) cp1252[byte-0x80];
	} else {
	    *chPtr = (Tcl_UniChar) byte;
	}
	return 1;
    } else if (byte < 0xE0) {
	if ((src[1] & 0xC0) == 0x80) {
	    /*
	     * Two-byte-character lead-byte followed by a trail-byte.
	     */

	 */
    }
    else if (byte < 0xF8) {
	if (((src[1] & 0xC0) == 0x80) && ((src[2] & 0xC0) == 0x80) && ((src[3] & 0xC0) == 0x80)) {
	    /*
	     * Four-byte-character lead byte followed by three trail bytes.
	     */
#if TCL_UTF_MAX <= 4










	    Tcl_UniChar surrogate;

	    byte = (((byte & 0x07) << 18) | ((src[1] & 0x3F) << 12)
		    | ((src[2] & 0x3F) << 6) | (src[3] & 0x3F)) - 0x10000;
	    surrogate = (Tcl_UniChar) (0xD800 + (byte >> 10));
	    if (byte & 0x100000) {
		/* out of range, < 0x10000 or > 0x10ffff */

{
    int len, fullchar;
    Tcl_UniChar find = 0;

    while (1) {
	len = TclUtfToUniChar(src, &find);
	fullchar = find;
#if TCL_UTF_MAX <= 4
	if (!len) {
	    len += TclUtfToUniChar(src, &find);
	    fullchar = (((fullchar & 0x3ff) << 10) | (find & 0x3ff)) + 0x10000;
	}
#endif
	if (fullchar == ch) {
	    return src;

    Tcl_UniChar find = 0;
    const char *last;

    last = NULL;
    while (1) {
	len = TclUtfToUniChar(src, &find);
	fullchar = find;
#if TCL_UTF_MAX <= 4
	if (!len) {
	    len += TclUtfToUniChar(src, &find);
	    fullchar = (((fullchar & 0x3ff) << 10) | (find & 0x3ff)) + 0x10000;
	}
#endif
	if (fullchar == ch) {
	    last = src;

const char *
Tcl_UtfNext(
    const char *src)		/* The current location in the string. */
{
    Tcl_UniChar ch = 0;
    int len = TclUtfToUniChar(src, &ch);

#if TCL_UTF_MAX <= 4
    if (len == 0) {
      len = TclUtfToUniChar(src, &ch);
    }
#endif
    return src + len;
}


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

int
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;
    int fullchar = 0;
#if TCL_UTF_MAX <= 4
	int len = 1;
#endif

    while (index-- >= 0) {
#if TCL_UTF_MAX <= 4
	src += (len = TclUtfToUniChar(src, &ch));
#else
	src += TclUtfToUniChar(src, &ch);
#endif
    }
    fullchar = ch;
#if TCL_UTF_MAX <= 4
     if (!len) {
	/* If last Tcl_UniChar was an upper surrogate, combine with lower surrogate */
	(void)TclUtfToUniChar(src, &ch);
	fullchar = (((fullchar & 0x3ff) << 10) | (ch & 0x3ff)) + 0x10000;
    }
#endif
    return fullchar;
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UtfAtIndex --
 *

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

int
Tcl_UtfToUpper(
    char *str)			/* String to convert in place. */
{
    Tcl_UniChar ch = 0;
    int upChar;
    char *src, *dst;
    int bytes;

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

    src = dst = str;
    while (*src) {
	bytes = TclUtfToUniChar(src, &ch);
	upChar = ch;
	if (!bytes) {
	    /* TclUtfToUniChar only returns 0 for chars > 0xffff ! */
	    bytes = TclUtfToUniChar(src, &ch);
	    /* Combine surrogates */
	    upChar = (((upChar & 0x3ff) << 10) | (ch & 0x3ff)) + 0x10000;
	}
	upChar = Tcl_UniCharToUpper(upChar);

	/*
	 * 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.
	 */

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

int
Tcl_UtfToLower(
    char *str)			/* String to convert in place. */
{
    Tcl_UniChar ch = 0;
    int lowChar;
    char *src, *dst;
    int bytes;

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

    src = dst = str;
    while (*src) {
	bytes = TclUtfToUniChar(src, &ch);
	lowChar = ch;
	if (!bytes) {
	    /* TclUtfToUniChar only returns 0 for chars > 0xffff ! */
	    bytes = TclUtfToUniChar(src, &ch);
	    /* Combine surrogates */
	    lowChar = (((lowChar & 0x3ff) << 10) | (ch & 0x3ff)) + 0x10000;
	}
	lowChar = Tcl_UniCharToLower(lowChar);

	/*
	 * 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.
	 */

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

int
Tcl_UtfToTitle(
    char *str)			/* String to convert in place. */
{
    Tcl_UniChar ch = 0;
    int titleChar, lowChar;
    char *src, *dst;
    int bytes;

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

    src = dst = str;

    if (*src) {
	bytes = TclUtfToUniChar(src, &ch);
	titleChar = ch;
	if (!bytes) {
	    /* TclUtfToUniChar only returns 0 for chars > 0xffff ! */
	    bytes = TclUtfToUniChar(src, &ch);
	    /* Combine surrogates */
	    titleChar = (((titleChar & 0x3ff) << 10) | (ch & 0x3ff)) + 0x10000;
	}
	titleChar = Tcl_UniCharToTitle(titleChar);

	if (bytes < TclUtfCount(titleChar)) {
	    memcpy(dst, src, (size_t) bytes);
	    dst += bytes;
	} else {
	    dst += Tcl_UniCharToUtf(titleChar, dst);
	}
	src += bytes;
    }
    while (*src) {
	bytes = TclUtfToUniChar(src, &ch);
	lowChar = ch;
	if (!bytes) {
	    /* TclUtfToUniChar only returns 0 for chars > 0xffff ! */
	    bytes = TclUtfToUniChar(src, &ch);
	    /* Combine surrogates */
	    lowChar = (((lowChar & 0x3ff) << 10) | (ch & 0x3ff)) + 0x10000;
	}
	lowChar = Tcl_UniCharToLower(lowChar);

	if (bytes < TclUtfCount(lowChar)) {
	    memcpy(dst, src, (size_t) bytes);
	    dst += bytes;
	} else {
	    dst += Tcl_UniCharToUtf(lowChar, dst);
	}

	 * 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;

	 * 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;

{
    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;

{
    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;

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

int
Tcl_UniCharToUpper(
    int ch)			/* Unicode character to convert. */
{
    if (!UNICODE_OUT_OF_RANGE(ch)) {
	int info = GetUniCharInfo(ch);

	if (GetCaseType(info) & 0x04) {
	    ch -= GetDelta(info);
	}
    }
    return ch & 0x1FFFFF;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharToLower --
 *
 *	Compute the lowercase equivalent of the given Unicode character.
 *
 * Results:
 *	Returns the lowercase Unicode character.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UniCharToLower(
    int ch)			/* Unicode character to convert. */
{
    if (!UNICODE_OUT_OF_RANGE(ch)) {
	int info = GetUniCharInfo(ch);

	if (GetCaseType(info) & 0x02) {
	    ch += GetDelta(info);
	}
    }
    return ch & 0x1FFFFF;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharToTitle --
 *
 *	Compute the titlecase equivalent of the given Unicode character.
 *
 * Results:
 *	Returns the titlecase Unicode character.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UniCharToTitle(
    int ch)			/* Unicode character to convert. */
{
    if (!UNICODE_OUT_OF_RANGE(ch)) {
	int info = GetUniCharInfo(ch);
	int mode = GetCaseType(info);

	if (mode & 0x1) {
	    /*
	     * Subtract or add one depending on the original case.
	     */

	    ch += ((mode & 0x4) ? -1 : 1);
	} else if (mode == 0x4) {
	    ch -= GetDelta(info);
	}
    }
    return ch & 0x1FFFFF;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharLen --
 *

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

int
Tcl_UniCharIsAlnum(
    int ch)			/* Unicode character to test. */
{

    if (UNICODE_OUT_OF_RANGE(ch)) {
	return 0;
    }

    return (((ALPHA_BITS | DIGIT_BITS) >> GetCategory(ch)) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsAlpha --

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

int
Tcl_UniCharIsAlpha(
    int ch)			/* Unicode character to test. */
{

    if (UNICODE_OUT_OF_RANGE(ch)) {
	return 0;
    }

    return ((ALPHA_BITS >> GetCategory(ch)) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsControl --

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

int
Tcl_UniCharIsControl(
    int ch)			/* Unicode character to test. */
{

    if (UNICODE_OUT_OF_RANGE(ch)) {
	ch &= 0x1FFFFF;
	if ((ch == 0xE0001) || ((ch >= 0xE0020) && (ch <= 0xE007f))) {
	    return 1;
	}
	if ((ch >= 0xF0000) && ((ch & 0xFFFF) <= 0xFFFD)) {
	    return 1;
	}
	return 0;
    }

    return ((CONTROL_BITS >> GetCategory(ch)) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsDigit --

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

int
Tcl_UniCharIsDigit(
    int ch)			/* Unicode character to test. */
{

    if (UNICODE_OUT_OF_RANGE(ch)) {
	return 0;
    }

    return (GetCategory(ch) == DECIMAL_DIGIT_NUMBER);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsGraph --

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

int
Tcl_UniCharIsGraph(
    int ch)			/* Unicode character to test. */
{

    if (UNICODE_OUT_OF_RANGE(ch)) {
	ch &= 0x1FFFFF;
	return (ch >= 0xE0100) && (ch <= 0xE01EF);
    }

    return ((GRAPH_BITS >> GetCategory(ch)) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsLower --

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

int
Tcl_UniCharIsLower(
    int ch)			/* Unicode character to test. */
{

    if (UNICODE_OUT_OF_RANGE(ch)) {
	return 0;
    }

    return (GetCategory(ch) == LOWERCASE_LETTER);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsPrint --

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

int
Tcl_UniCharIsPrint(
    int ch)			/* Unicode character to test. */
{

    if (UNICODE_OUT_OF_RANGE(ch)) {
	ch &= 0x1FFFFF;
	return (ch >= 0xE0100) && (ch <= 0xE01EF);
    }

    return (((GRAPH_BITS|SPACE_BITS) >> GetCategory(ch)) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsPunct --

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

int
Tcl_UniCharIsPunct(
    int ch)			/* Unicode character to test. */
{

    if (UNICODE_OUT_OF_RANGE(ch)) {
	return 0;
    }

    return ((PUNCT_BITS >> GetCategory(ch)) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsSpace --

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

int
Tcl_UniCharIsSpace(
    int ch)			/* Unicode character to test. */
{

    /* Ignore upper 11 bits. */
    ch &= 0x1FFFFF;





    /*
     * If the character is within the first 127 characters, just use the
     * standard C function, otherwise consult the Unicode table.
     */

    if (ch < 0x80) {
	return TclIsSpaceProc((char) ch);

    } else if (UNICODE_OUT_OF_RANGE(ch)) {
	return 0;

    } else if (ch == 0x0085 || ch == 0x180E || ch == 0x200B
	    || ch == 0x202F || ch == 0x2060 || ch == 0xFEFF) {
	return 1;
    } else {
	return ((SPACE_BITS >> GetCategory(ch)) & 1);
    }
}

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

int
Tcl_UniCharIsUpper(
    int ch)			/* Unicode character to test. */
{

    if (UNICODE_OUT_OF_RANGE(ch)) {
	return 0;
    }

    return (GetCategory(ch) == UPPERCASE_LETTER);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsWordChar --

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

int
Tcl_UniCharIsWordChar(
    int ch)			/* Unicode character to test. */
{

    if (UNICODE_OUT_OF_RANGE(ch)) {
	return 0;
    }

    return ((WORD_BITS >> GetCategory(ch)) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharCaseMatch --