Tcl Source Code

 badIndex:
    errorString = "not enough arguments for all format specifiers";
    goto error;

 badField:
    {
	Tcl_UniChar ch = 0;
	char buf[TCL_UTF_MAX + 1] = "";

	TclUtfToUniChar(errorString, &ch);
	buf[Tcl_UniCharToUtf(ch, buf)] = '\0';
	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
		"bad field specifier \"%s\"", buf));
	return TCL_ERROR;
    }

 badIndex:
    errorString = "not enough arguments for all format specifiers";
    goto error;

 badField:
    {
	Tcl_UniChar ch = 0;
	char buf[TCL_UTF_MAX + 1] = "";

	TclUtfToUniChar(errorString, &ch);
	buf[Tcl_UniCharToUtf(ch, buf)] = '\0';
	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
		"bad field specifier \"%s\"", buf));
	return TCL_ERROR;
    }

	for ( ; stringPtr < end; stringPtr += len) {
	    int fullchar;
	    len = TclUtfToUniChar(stringPtr, &ch);
	    fullchar = ch;

#if TCL_UTF_MAX == 4
	    if ((ch >= 0xD800) && (len < 3)) {
		len += TclUtfToUniChar(stringPtr + len, &ch);
		fullchar = (((fullchar & 0x3ff) << 10) | (ch & 0x3ff)) + 0x10000;
	    }
#endif

	    /*
	     * Assume Tcl_UniChar is an integral type...
	     */

	 */

	if (TclIsPureByteArray(objv[1])) {
	    unsigned char uch = (unsigned char) ch;

	    Tcl_SetObjResult(interp, Tcl_NewByteArrayObj(&uch, 1));
	} else {
	    char buf[TCL_UTF_MAX] = "";

	    length = Tcl_UniCharToUtf(ch, buf);
#if TCL_UTF_MAX > 3
	    if ((ch >= 0xD800) && (length < 3)) {
		length += Tcl_UniCharToUtf(-1, buf + length);
	    }
#endif
	    Tcl_SetObjResult(interp, Tcl_NewStringObj(buf, length));
	}
    }
    return TCL_OK;
}

/*

	}
	end = string1 + length1;
	for (; string1 < end; string1 += length2, failat++) {
	    int fullchar;
	    length2 = TclUtfToUniChar(string1, &ch);
	    fullchar = ch;
#if TCL_UTF_MAX == 4
	    if ((ch >= 0xD800) && (length2 < 3)) {
	    	length2 += TclUtfToUniChar(string1 + length2, &ch);
	    	fullchar = (((fullchar & 0x3ff) << 10) | (ch & 0x3ff)) + 0x10000;
	    }
#endif
	    if (!chcomp(fullchar)) {
		result = 0;
		break;
	    }

	return TCL_ERROR;
    }

    if (objc == 4) {
	const char *string = TclGetStringFromObj(objv[1], &length2);

	if ((length2 > 1) &&
		strncmp(string, "-nocase", length2) == 0) {
	    nocase = 1;
	} else {
	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
		    "bad option \"%s\": must be -nocase", string));
	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "INDEX", "option",
		    string, NULL);
	    return TCL_ERROR;

    }

    if (objc == 4) {
	int length;
	const char *string = TclGetStringFromObj(objv[1], &length);

	if ((length > 1) &&
	    strncmp(string, "-nocase", length) == 0) {
	    nocase = TCL_MATCH_NOCASE;
	} else {
	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
		    "bad option \"%s\": must be -nocase", string));
	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "INDEX", "option",
		    string, NULL);
	    return TCL_ERROR;

	Tcl_WrongNumArgs(interp, 1, objv,
		"?-nocase? ?-length int? string1 string2");
	return TCL_ERROR;
    }

    for (i = 1; i < objc-2; i++) {
	string2 = TclGetStringFromObj(objv[i], &length2);
	if ((length2 > 1) && !strncmp(string2, "-nocase", length2)) {
	    nocase = 1;
	} else if ((length2 > 1)
		&& !strncmp(string2, "-length", length2)) {
	    if (i+1 >= objc-2) {
		goto str_cmp_args;
	    }
	    i++;
	    if (TclGetIntFromObj(interp, objv[i], &reqlength) != TCL_OK) {
		return TCL_ERROR;
	    }

	Tcl_WrongNumArgs(interp, 1, objv,
		"?-nocase? ?-length int? string1 string2");
	return TCL_ERROR;
    }

    for (i = 1; i < objc-2; i++) {
	string = TclGetStringFromObj(objv[i], &length);
	if ((length > 1) && !strncmp(string, "-nocase", length)) {
	    *nocase = 1;
	} else if ((length > 1)
		&& !strncmp(string, "-length", length)) {
	    if (i+1 >= objc-2) {
		goto str_cmp_args;
	    }
	    i++;
	    if (TclGetIntFromObj(interp, objv[i], reqlength) != TCL_OK) {
		return TCL_ERROR;
	    }

	PUSH("");
	count++;
    }

    for (endTokenPtr = tokenPtr + parse.numTokens;
	    tokenPtr < endTokenPtr; tokenPtr = TokenAfter(tokenPtr)) {
	int length, literal, catchRange, breakJump;
	char buf[TCL_UTF_MAX] = "";
	JumpFixup startFixup, okFixup, returnFixup, breakFixup;
	JumpFixup continueFixup, otherFixup, endFixup;

	switch (tokenPtr->type) {
	case TCL_TOKEN_TEXT:
	    literal = TclRegisterNewLiteral(envPtr,
		    tokenPtr->start, tokenPtr->size);

	    if (tempPtr != NULL) {
		Tcl_AppendToObj(tempPtr, tokenPtr->start, tokenPtr->size);
	    }
	    break;

	case TCL_TOKEN_BS:
	    if (tempPtr != NULL) {
		char utfBuf[TCL_UTF_MAX] = "";
		int length = TclParseBackslash(tokenPtr->start,
			tokenPtr->size, NULL, utfBuf);

		Tcl_AppendToObj(tempPtr, utfBuf, length);
	    }
	    break;

				 * compile. */
    int count,			/* Number of tokens to consider at tokenPtr.
				 * Must be at least 1. */
    CompileEnv *envPtr)		/* Holds the resulting instructions. */
{
    Tcl_DString textBuffer;	/* Holds concatenated chars from adjacent
				 * TCL_TOKEN_TEXT, TCL_TOKEN_BS tokens. */
    char buffer[TCL_UTF_MAX] = "";
    int i, numObjsToConcat, length, adjust;
    unsigned char *entryCodeNext = envPtr->codeNext;
#define NUM_STATIC_POS 20
    int isLiteral, maxNumCL, numCL;
    int *clPosition = NULL;
    int depth = TclGetStackDepth(envPtr);

	    src += 1;
	    dst += Tcl_UniCharToUtf(*chPtr, dst);
	} else {
	    int len = TclUtfToUniChar(src, chPtr);
	    src += len;
	    dst += Tcl_UniCharToUtf(*chPtr, dst);
#if TCL_UTF_MAX == 4
	    if ((*chPtr >= 0xD800) && (len < 3)) {
		src += TclUtfToUniChar(src + len, chPtr);
		dst += Tcl_UniCharToUtf(*chPtr, dst);
	    }
#endif
	}
    }

    *srcReadPtr = src - srcStart;

	/*
	 * Check for illegal characters.
	 */

	if (ch > 0xff
#if TCL_UTF_MAX == 4
		|| ((ch >= 0xD800) && (len < 3))
#endif
		) {
	    if (flags & TCL_ENCODING_STOPONERROR) {
		result = TCL_CONVERT_UNKNOWN;
		break;
	    }
#if TCL_UTF_MAX == 4
	    if ((ch >= 0xD800) && (len < 3)) len = 4;
#endif

	    /*
	     * Plunge on, using '?' as a fallback character.
	     */

	    ch = (Tcl_UniChar) '?';

		     */

		    state = oldState;
		    result = TCL_CONVERT_NOSPACE;
		    break;
		}
		memcpy(dst, subTablePtr->sequence,
			subTablePtr->sequenceLen);
		dst += subTablePtr->sequenceLen;
	    }
	}

	if (tablePrefixBytes[(word >> 8)] != 0) {
	    if (dst + 1 > dstEnd) {
		result = TCL_CONVERT_NOSPACE;

	} else if (TclIsPureByteArray(valuePtr)) {
	    objResultPtr = Tcl_NewByteArrayObj(
		    Tcl_GetByteArrayFromObj(valuePtr, NULL)+index, 1);
	} else if (valuePtr->bytes && length == valuePtr->length) {
	    objResultPtr = Tcl_NewStringObj((const char *)
		    valuePtr->bytes+index, 1);
	} else {
	    char buf[TCL_UTF_MAX] = "";
	    Tcl_UniChar ch = Tcl_GetUniChar(valuePtr, index);

	    /*
	     * This could be: Tcl_NewUnicodeObj((const Tcl_UniChar *)&ch, 1)
	     * but creating the object as a string seems to be faster in
	     * practical use.
	     */

				 * written. At most TCL_UTF_MAX bytes will be
				 * written there. */
{
    register const char *p = src+1;
    Tcl_UniChar unichar = 0;
    int result;
    int count;
    char buf[TCL_UTF_MAX] = "";

    if (numBytes == 0) {
	if (readPtr != NULL) {
	    *readPtr = 0;
	}
	return 0;
    }

  done:
    if (readPtr != NULL) {
	*readPtr = count;
    }
    count = Tcl_UniCharToUtf(result, dst);
#if TCL_UTF_MAX > 3
     if ((result >= 0xD800) && (count < 3)) {
	count += Tcl_UniCharToUtf(-1, dst + count);
    }
#endif
    return count;
}

/*
 *----------------------------------------------------------------------

    adjust = 0;
    result = NULL;
    for (; count>0 && code==TCL_OK ; count--, tokenPtr++) {
	Tcl_Obj *appendObj = NULL;
	const char *append = NULL;
	int appendByteLength = 0;
	char utfCharBytes[TCL_UTF_MAX] = "";

	switch (tokenPtr->type) {
	case TCL_TOKEN_TEXT:
	    append = tokenPtr->start;
	    appendByteLength = tokenPtr->size;
	    break;

				 * required. */
{
    int gotXpg, gotSequential, value, i, flags;
    char *end;
    Tcl_UniChar ch = 0;
    int objIndex, xpgSize, nspace = numVars;
    int *nassign = TclStackAlloc(interp, nspace * sizeof(int));
    char buf[TCL_UTF_MAX+1] = "";
    Tcl_Obj *errorMsg;		/* Place to build an error messages. Note that
				 * these are messy operations because we do
				 * not want to use the formatting engine;
				 * we're inside there! */

    /*
     * Initialize an array that records the number of times a variable is

	     * Scan a single Unicode character.
	     */

	    offset = TclUtfToUniChar(string, &sch);
	    i = (int)sch;
#if TCL_UTF_MAX == 4
	    if (!offset) {
		offset = TclUtfToUniChar(string, &sch);
		i = (((i<<10) & 0x0FFC00) + 0x10000) + (sch & 0x3FF);
	    }
#endif
	    string += offset;
	    if (!(flags & SCAN_SUPPRESS)) {
		objPtr = Tcl_NewIntObj(i);
		Tcl_IncrRefCount(objPtr);

	    int code, length;

	    if (TclGetIntFromObj(interp, segment, &code) != TCL_OK) {
		goto error;
	    }
	    length = Tcl_UniCharToUtf(code, buf);
#if TCL_UTF_MAX > 3
	    if ((code >= 0xD800) && (length < 3)) {
		/* Special case for handling high surrogates. */
		length += Tcl_UniCharToUtf(-1, buf + length);
	    }
#endif
	    segment = Tcl_NewStringObj(buf, length);
	    Tcl_IncrRefCount(segment);
	    allocSegment = 1;
	    break;
	}

    if (size > stringPtr->allocated) {
	GrowStringBuffer(objPtr, size, 1);
    }

  copyBytes:
    dst = objPtr->bytes + origLength;
    for (i = 0; i < numChars; i++) {
	dst += Tcl_UniCharToUtf(unicode[i], dst);
    }
    *dst = '\0';
    objPtr->length = dst - objPtr->bytes;
    return numChars;
}

/*

    for (w = (wchar_t *)string; w < wEnd; ) {
	if (!blen && ((*w & 0xFC00) != 0xDC00)) {
	    /* Special case for handling high surrogates. */
	    p += Tcl_UniCharToUtf(-1, p);
	}
	blen = Tcl_UniCharToUtf(*w, p);
	p += blen;
	if ((*w >= 0xD800) && (blen < 3)) {
	    /* Indication that high surrogate is handled */
	    blen = 0;
	}
	w++;
    }
    if (!blen) {
	/* Special case for handling high surrogates. */
	p += Tcl_UniCharToUtf(-1, p);
    }
    Tcl_DStringSetLength(dsPtr, oldLength + (p - result));

	    return 2;
	}
	if (ch <= 0xFFFF) {
#if TCL_UTF_MAX > 3
	    if ((ch & 0xF800) == 0xD800) {
		if (ch & 0x0400) {
		    /* Low surrogate */
		    if (((buf[0] & 0xC0) == 0x80) && ((buf[1] & 0xCF) == 0)) {

			/* Previous Tcl_UniChar was a high surrogate, so combine */
			buf[2] = (char) ((ch & 0x3F) | 0x80);
			buf[1] |= (char) (((ch >> 6) & 0x0F) | 0x80);
			return 3;
		    }
		    /* 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 1;
		}
	    }
#endif
	    goto three;
	}

#if TCL_UTF_MAX > 3
	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;
	}
    } else if (ch == -1) {
	if (((buf[0] & 0xC0) == 0x80) && ((buf[1] & 0xCF) == 0)
		&& ((buf[-1] & 0xF8) == 0xF0)) {
	    ch = 0xD7C0 + ((buf[-1] & 0x07) << 8) + ((buf[0] & 0x3F) << 2)
		    + ((buf[1] & 0x30) >> 4);
	    buf[1] = (char) ((ch | 0x80) & 0xBF);
	    buf[0] = (char) (((ch >> 6) | 0x80) & 0xBF);
	    buf[-1] = (char) ((ch >> 12) | 0xE0);
	    return 2;
	}
#endif
    }

    ch = 0xFFFD;
three:
    buf[2] = (char) ((ch | 0x80) & 0xBF);

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.
     */

    byte = *((unsigned char *) src);
    if (byte < 0xC0) {
	/*
	 * Handles properly formed UTF-8 characters between 0x01 and 0x7F.
	 * Also treats \0 and naked trail bytes 0x80 to 0xBF as valid
	 * characters representing themselves.
	 */

#if TCL_UTF_MAX == 4
	/* If *chPtr contains a high surrogate (produced by a previous
	 * Tcl_UtfToUniChar() call) and the next 3 bytes are UTF-8 continuation
	 * bytes, then we must produce a follow-up low surrogate. We only
	 * do that if the high surrogate matches the bits we encounter.
	 */
	if ((byte >= 0x80)
		&& (((((byte - 0x10) << 2) & 0xFC) | 0xD800) == (*chPtr & 0xFCFC))
		&& ((src[1] & 0xF0) == (((*chPtr << 4) & 0x30) | 0x80))
		&& ((src[2] & 0xC0) == 0x80)) {
	    *chPtr = ((src[1] & 0x0F) << 6) + (src[2] & 0x3F) + 0xDC00;
	    return 3;
	}
#endif
	*chPtr = byte;
	return 1;
    } else if (byte < 0xE0) {
	if ((src[1] & 0xC0) == 0x80) {
	    /*
	     * Two-byte-character lead-byte followed by a trail-byte.
	     */

	    *chPtr = (((byte & 0x1F) << 6) | (src[1] & 0x3F));
	    if ((unsigned)(*chPtr - 1) >= (UNICODE_SELF - 1)) {
		return 2;
	    }
	}

	/*
	 * A two-byte-character lead-byte not followed by trail-byte
	 * represents itself.
	 */
    } else if (byte < 0xF0) {
	if (((src[1] & 0xC0) == 0x80) && ((src[2] & 0xC0) == 0x80)) {
	    /*
	     * Three-byte-character lead byte followed by two trail bytes.
	     */

	    *chPtr = (((byte & 0x0F) << 12)
		    | ((src[1] & 0x3F) << 6) | (src[2] & 0x3F));
	    if (*chPtr > 0x7FF) {
		return 3;
	    }
	}

	/*
	 * A three-byte-character lead-byte not followed by two trail-bytes
	 * represents itself.
	 */
    }
#if TCL_UTF_MAX > 3
    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 high = (((byte & 0x07) << 8) | ((src[1] & 0x3F) << 2)


		    | ((src[2] & 0x3F) >> 4)) - 0x40;

	    if (high >= 0x400) {
		/* out of range, < 0x10000 or > 0x10ffff */




	    } else {
		/* produce high surrogate, advance source pointer */
		*chPtr = 0xD800 + high;
		return 1;
	    }
#else
	    *chPtr = (((byte & 0x07) << 18) | ((src[1] & 0x3F) << 12)
		    | ((src[2] & 0x3F) << 6) | (src[3] & 0x3F));
	    if ((*chPtr - 0x10000) <= 0xFFFFF) {
		return 4;
	    }
#endif
	}

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

    *chPtr = byte;
    return 1;
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UtfToUniCharDString --

    int len, fullchar;
    Tcl_UniChar find = 0;

    while (1) {
	len = TclUtfToUniChar(src, &find);
	fullchar = find;
#if TCL_UTF_MAX == 4
	if ((ch >= 0xD800) && (len < 3)) {
	    len += TclUtfToUniChar(src + len, &find);
	    fullchar = (((fullchar & 0x3ff) << 10) | (find & 0x3ff)) + 0x10000;
	}
#endif
	if (fullchar == ch) {
	    return src;
	}
	if (*src == '\0') {

    const char *last;

    last = NULL;
    while (1) {
	len = TclUtfToUniChar(src, &find);
	fullchar = find;
#if TCL_UTF_MAX == 4
	if ((ch >= 0xD800) && (len < 3)) {
	    len += TclUtfToUniChar(src + len, &find);
	    fullchar = (((fullchar & 0x3ff) << 10) | (find & 0x3ff)) + 0x10000;
	}
#endif
	if (fullchar == ch) {
	    last = src;
	}
	if (*src == '\0') {

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 ((ch >= 0xD800) && (len < 3)) {
	len += TclUtfToUniChar(src + len, &ch);
    }
#endif
    return src + len;
}

/*
 *---------------------------------------------------------------------------

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;
    }
#if TCL_UTF_MAX == 4
    if ((ch >= 0xD800) && (len < 3)) {
	/* Index points at character following high Surrogate */
	src += TclUtfToUniChar(src, &ch);
    }
#endif
    return src;
}

/*

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

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

    src = dst = str;
    while (*src) {
	len = TclUtfToUniChar(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.
	 */

	if (len < UtfCount(upChar)) {
	    memcpy(dst, src, len);
	    dst += len;
	} else {
	    dst += Tcl_UniCharToUtf(upChar, dst);
	}
	src += len;
    }
    *dst = '\0';
    return (dst - str);
}

/*
 *----------------------------------------------------------------------

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

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

    src = dst = str;
    while (*src) {
	len = TclUtfToUniChar(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.
	 */

	if (len < UtfCount(lowChar)) {
	    memcpy(dst, src, len);
	    dst += len;
	} else {
	    dst += Tcl_UniCharToUtf(lowChar, dst);
	}
	src += len;
    }
    *dst = '\0';
    return (dst - str);
}

/*
 *----------------------------------------------------------------------

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

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

    src = dst = str;

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

	if (len < UtfCount(titleChar)) {
	    memcpy(dst, src, len);
	    dst += len;
	} else {
	    dst += Tcl_UniCharToUtf(titleChar, dst);
	}
	src += len;
    }
    while (*src) {
	len = TclUtfToUniChar(src, &ch);
	lowChar = ch;
	/* Special exception for Georgian Asomtavruli chars, no titlecase. */
	if ((unsigned)(lowChar - 0x1C90) >= 0x30) {
	    lowChar = Tcl_UniCharToLower(lowChar);
	}

	if (len < UtfCount(lowChar)) {
	    memcpy(dst, src, len);
	    dst += len;
	} else {
	    dst += Tcl_UniCharToUtf(lowChar, dst);
	}
	src += len;
    }
    *dst = '\0';
    return (dst - str);
}

/*
 *----------------------------------------------------------------------

char
Tcl_Backslash(
    const char *src,		/* Points to the backslash character of a
				 * backslash sequence. */
    int *readPtr)		/* Fill in with number of characters read from
				 * src, unless NULL. */
{
    char buf[TCL_UTF_MAX] = "";
    Tcl_UniChar ch = 0;

    Tcl_UtfBackslash(src, readPtr, buf);
    TclUtfToUniChar(buf, &ch);
    return (char) ch;
}


    for (w = (TCHAR *)string; w < wEnd; ) {
	if (!blen && ((*w & 0xFC00) != 0xDC00)) {
	    /* Special case for handling high surrogates. */
	    p += Tcl_UniCharToUtf(-1, p);
	}
	blen = Tcl_UniCharToUtf(*w, p);
	p += blen;
	if ((*w >= 0xD800) && (blen < 3)) {
	    /* Indication that high surrogate is handled */
	    blen = 0;
	}
	w++;
    }
    if (!blen) {
	/* Special case for handling high surrogates. */
	p += Tcl_UniCharToUtf(-1, p);
    }
    Tcl_DStringSetLength(dsPtr, oldLength + (p - result));