Tcl Source Code

				 * into UTF-8. */
    Tcl_EncodingConvertProc *fromUtfProc;
				/* Function to convert from UTF-8 into
				 * external encoding. */
    Tcl_EncodingFreeProc *freeProc;
				/* If non-NULL, function to call when this
				 * encoding is deleted. */
    void *clientData;	/* Arbitrary value associated with encoding
				 * type. Passed to conversion functions. */
    Tcl_Size nullSize;	/* Number of 0x00 bytes that signify
				 * end-of-string in this encoding. This number
				 * is used to determine the source string
				 * length when the srcLen argument is
				 * negative. This number can be 1, 2, or 4. */
    LengthProc *lengthProc;	/* Function to compute length of
				 * null-terminated strings in this encoding.
				 * If nullSize is 1, this is strlen; if

				 * conversion. */
    char prefixBytes[256];	/* If a byte in the input stream is the first
				 * character of one of the escape sequences in
				 * the following array, the corresponding
				 * entry in this array is 1, otherwise it is
				 * 0. */
    int numSubTables;		/* Length of following array. */
    EscapeSubTable subTables[TCLFLEXARRAY];/* Information about each EscapeSubTable used

				 * by this encoding type. The actual size is
				 * as large as necessary to hold all
				 * EscapeSubTables. */
} EscapeEncodingData;

/*
 * Constants used when loading an encoding file to identify the type of the

    int value;
} encodingProfiles[] = {
    {"replace", TCL_ENCODING_PROFILE_REPLACE},
    {"strict", TCL_ENCODING_PROFILE_STRICT},
    {"tcl8", TCL_ENCODING_PROFILE_TCL8},
};

#define PROFILE_TCL8(flags_)                                           \
    (ENCODING_PROFILE_GET(flags_) == TCL_ENCODING_PROFILE_TCL8)

#define PROFILE_REPLACE(flags_)                                        \
    (ENCODING_PROFILE_GET(flags_) == TCL_ENCODING_PROFILE_REPLACE)

#define PROFILE_STRICT(flags_)                                         \
    (!PROFILE_TCL8(flags_) && !PROFILE_REPLACE(flags_))

#define UNICODE_REPLACE_CHAR 0xFFFD
#define SURROGATE(c_)      (((c_) & ~0x7FF) == 0xD800)
#define HIGH_SURROGATE(c_) (((c_) & ~0x3FF) == 0xD800)
#define LOW_SURROGATE(c_)  (((c_) & ~0x3FF) == 0xDC00)

/*
 * The following variable is used in the sparse matrix code for a
 * TableEncoding to represent a page in the table that has no entries.
 */

static unsigned short emptyPage[256];

static Tcl_EncodingConvertProc	UtfToUtf16Proc;
static Tcl_EncodingConvertProc	UtfToUcs2Proc;
static Tcl_EncodingConvertProc	UtfToUtfProc;
static Tcl_EncodingConvertProc	Iso88591FromUtfProc;
static Tcl_EncodingConvertProc	Iso88591ToUtfProc;

/*
 * A Tcl_ObjType for holding a cached Tcl_Encoding in the twoPtrValue.ptr1 field
 * of the internalrep. This should help the lifetime of encodings be more useful.
 * See concerns raised in [Bug 1077262].
 */

static const Tcl_ObjType encodingType = {
    "encoding",
    FreeEncodingInternalRep,
    DupEncodingInternalRep,
    NULL,

/*
 * NOTE: THESE BIT DEFINITIONS SHOULD NOT OVERLAP WITH INTERNAL USE BITS
 * DEFINED IN tcl.h (TCL_ENCODING_* et al). Be cognizant of this
 * when adding bits. TODO - should really be defined in a single file.
 *
 * To prevent conflicting bits, only define bits within 0xff00 mask here.
 */

#define TCL_ENCODING_LE	0x100   /* Used to distinguish LE/BE variants */
#define ENCODING_UTF	0x200	/* For UTF-8 encoding, allow 4-byte output sequences */

#define ENCODING_INPUT	0x400   /* For UTF-8/CESU-8 encoding, means external -> internal */



void
TclInitEncodingSubsystem(void)
{
    Tcl_EncodingType type;
    TableEncodingData *dataPtr;
    unsigned size;

    type.clientData	= INT2PTR(ENCODING_UTF);
    tclUtf8Encoding = Tcl_CreateEncoding(&type);
    type.clientData	= NULL;
    type.encodingName	= "cesu-8";
    Tcl_CreateEncoding(&type);

    type.toUtfProc	= Utf16ToUtfProc;
    type.fromUtfProc    = UtfToUcs2Proc;
    type.freeProc	= NULL;
    type.nullSize	= 2;
    type.encodingName   = "ucs-2le";
    type.clientData	= INT2PTR(TCL_ENCODING_LE);
    Tcl_CreateEncoding(&type);
    type.encodingName   = "ucs-2be";
    type.clientData	= NULL;
    Tcl_CreateEncoding(&type);
    type.encodingName   = "ucs-2";
    type.clientData	= INT2PTR(leFlags);
    Tcl_CreateEncoding(&type);

    type.toUtfProc	= Utf32ToUtfProc;
    type.fromUtfProc    = UtfToUtf32Proc;
    type.freeProc	= NULL;
    type.nullSize	= 4;
    type.encodingName   = "utf-32le";
    type.clientData	= INT2PTR(TCL_ENCODING_LE);
    Tcl_CreateEncoding(&type);
    type.encodingName   = "utf-32be";
    type.clientData	= NULL;
    Tcl_CreateEncoding(&type);
    type.encodingName   = "utf-32";
    type.clientData	= INT2PTR(leFlags);
    Tcl_CreateEncoding(&type);

    type.toUtfProc	= Utf16ToUtfProc;
    type.fromUtfProc    = UtfToUtf16Proc;
    type.freeProc	= NULL;
    type.nullSize	= 2;
    type.encodingName   = "utf-16le";
    type.clientData	= INT2PTR(TCL_ENCODING_LE);
    Tcl_CreateEncoding(&type);
    type.encodingName   = "utf-16be";
    type.clientData	= NULL;
    Tcl_CreateEncoding(&type);
    type.encodingName   = "utf-16";
    type.clientData	= INT2PTR(leFlags);
    Tcl_CreateEncoding(&type);

#ifndef TCL_NO_DEPRECATED
    type.encodingName   = "unicode";
    Tcl_CreateEncoding(&type);
#endif

    /*
     * Need the iso8859-1 encoding in order to process binary data, so force
     * it to always be embedded. Note that this encoding *must* be a proper
     * table encoding or some of the escape encodings crash! Hence the ugly


/*
 *-------------------------------------------------------------------------
 *
 * Tcl_GetEncodingNulLength --
 *
 *	Given an encoding, return the number of nul bytes used for the
 *      string termination.
 *
 * Results:
 *	The number of nul bytes used for the string termination.
 *
 * Side effects:
 *	None.
 *

 *	"flags" controls the behavior if any of the bytes in
 *	the source buffer are invalid or cannot be represented in utf-8.
 *	Possible flags values:
 *	target encoding. It should be composed by OR-ing the following:
 *	- *At most one* of TCL_ENCODING_PROFILE{DEFAULT,TCL8,STRICT}
 *
 * Results:
 *      The return value is one of
 *        TCL_OK: success. Converted string in *dstPtr
 *        TCL_ERROR: error in passed parameters. Error message in interp
 *        TCL_CONVERT_MULTIBYTE: source ends in truncated multibyte sequence
 *        TCL_CONVERT_SYNTAX: source is not conformant to encoding definition
 *        TCL_CONVERT_UNKNOWN: source contained a character that could not
 *            be represented in target encoding.
 *
 * Side effects:
 *
 *      TCL_OK: The converted bytes are stored in the DString and NUL
 *          terminated in an encoding-specific manner.
 *      TCL_ERROR: an error, message is stored in the interp if not NULL.
 *      TCL_CONVERT_*: if errorLocPtr is NULL, an error message is stored
 *          in the interpreter (if not NULL). If errorLocPtr is not NULL,
 *          no error message is stored as it is expected the caller is
 *          interested in whatever is decoded so far and not treating this
 *          as an error condition.
 *
 *      In addition, *dstPtr is always initialized and must be cleared
 *      by the caller irrespective of the return code.
 *
 *-------------------------------------------------------------------------
 */

int
Tcl_ExternalToUtfDStringEx(
    Tcl_Interp *interp,         /* For error messages. May be NULL. */
    Tcl_Encoding encoding,	/* The encoding for the source string, or NULL
				 * for the default system encoding. */
    const char *src,		/* Source string in specified encoding. */
    Tcl_Size srcLen,		/* Source string length in bytes, or < 0 for
				 * encoding-specific string length. */
    int flags,			/* Conversion control flags. */
    Tcl_DString *dstPtr,	/* Uninitialized or free DString in which the
				 * converted string is stored. */
    Tcl_Size *errorLocPtr)      /* Where to store the error location
                                 * (or TCL_INDEX_NONE if no error). May
				 * be NULL. */
{
    char *dst;
    Tcl_EncodingState state;
    const Encoding *encodingPtr;
    int result;
    Tcl_Size dstLen, soFar;

	    Tcl_DStringSetLength(dstPtr, soFar);
	    if (errorLocPtr) {
		/*
		 * Do not write error message into interpreter if caller
		 * wants to know error location.
		 */
		*errorLocPtr = result == TCL_OK ? TCL_INDEX_NONE : nBytesProcessed;

	    } else {
		/* Caller wants error message on failure */
		if (result != TCL_OK && interp != NULL) {
		    char buf[TCL_INTEGER_SPACE];
		    snprintf(buf, sizeof(buf), "%" TCL_SIZE_MODIFIER "d", nBytesProcessed);

		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
			    "unexpected byte sequence starting at index %"
			    TCL_SIZE_MODIFIER "d: '\\x%02X'",
			    nBytesProcessed, UCHAR(srcStart[nBytesProcessed])));
		    Tcl_SetErrorCode(
			    interp, "TCL", "ENCODING", "ILLEGALSEQUENCE", buf, (void *)NULL);

		}
	    }
	    if (result != TCL_OK) {
		errno = (result == TCL_CONVERT_NOSPACE) ? ENOMEM : EILSEQ;
	    }
	    return result;
	}

int
Tcl_ExternalToUtf(
    TCL_UNUSED(Tcl_Interp *),	/* TODO: Re-examine this. */
    Tcl_Encoding encoding,	/* The encoding for the source string, or NULL
				 * for the default system encoding. */
    const char *src,		/* Source string in specified encoding. */
    Tcl_Size srcLen,		/* Source string length in bytes, or TCL_INDEX_NONE for
				 * encoding-specific string length. */

    int flags,			/* Conversion control flags. */
    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
				 * information used during a piecewise
				 * conversion. Contents of statePtr are
				 * initialized and/or reset by conversion
				 * routine under control of flags argument. */
    char *dst,			/* Output buffer in which converted string is

 *	Convert a source buffer from UTF-8 to the specified encoding.
 *	The parameter flags controls the behavior, if any of the bytes in
 *	the source buffer are invalid or cannot be represented in the
 *	target encoding. It should be composed by OR-ing the following:
 *	- *At most one* of TCL_ENCODING_PROFILE_*
 *
 * Results:
 *      The return value is one of
 *        TCL_OK: success. Converted string in *dstPtr
 *        TCL_ERROR: error in passed parameters. Error message in interp
 *        TCL_CONVERT_MULTIBYTE: source ends in truncated multibyte sequence
 *        TCL_CONVERT_SYNTAX: source is not conformant to encoding definition
 *        TCL_CONVERT_UNKNOWN: source contained a character that could not
 *            be represented in target encoding.
 *
 * Side effects:
 *
 *      TCL_OK: The converted bytes are stored in the DString and NUL
 *          terminated in an encoding-specific manner
 *      TCL_ERROR: an error, message is stored in the interp if not NULL.
 *      TCL_CONVERT_*: if errorLocPtr is NULL, an error message is stored
 *          in the interpreter (if not NULL). If errorLocPtr is not NULL,
 *          no error message is stored as it is expected the caller is
 *          interested in whatever is decoded so far and not treating this
 *          as an error condition.
 *
 *      In addition, *dstPtr is always initialized and must be cleared
 *      by the caller irrespective of the return code.
 *
 *-------------------------------------------------------------------------
 */

int
Tcl_UtfToExternalDStringEx(
    Tcl_Interp *interp,         /* For error messages. May be NULL. */
    Tcl_Encoding encoding,	/* The encoding for the converted string, or
				 * NULL for the default system encoding. */
    const char *src,		/* Source string in UTF-8. */
    Tcl_Size srcLen,		/* Source string length in bytes, or < 0 for
				 * strlen(). */
    int flags,			/* Conversion control flags. */
    Tcl_DString *dstPtr,	/* Uninitialized or free DString in which the
				 * converted string is stored. */
    Tcl_Size *errorLocPtr)      /* Where to store the error location
                                 * (or TCL_INDEX_NONE if no error). May
				 * be NULL. */
{
    char *dst;
    Tcl_EncodingState state;
    const Encoding *encodingPtr;
    int result;
    const char *srcStart = src;

		Tcl_DStringSetLength(dstPtr, i--);
	    }
	    if (errorLocPtr) {
		/*
		 * Do not write error message into interpreter if caller
		 * wants to know error location.
		 */
		*errorLocPtr = result == TCL_OK ? TCL_INDEX_NONE : nBytesProcessed;

	    } else {
		/* Caller wants error message on failure */
		if (result != TCL_OK && interp != NULL) {
		    Tcl_Size pos = Tcl_NumUtfChars(srcStart, nBytesProcessed);
		    int ucs4;
		    char buf[TCL_INTEGER_SPACE];

		    TclUtfToUniChar(&srcStart[nBytesProcessed], &ucs4);
		    snprintf(buf, sizeof(buf), "%" TCL_SIZE_MODIFIER "d", nBytesProcessed);

		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
			    "unexpected character at index %" TCL_SIZE_MODIFIER
			    "u: 'U+%06X'",
			    pos, ucs4));
		    Tcl_SetErrorCode(interp, "TCL", "ENCODING", "ILLEGALSEQUENCE",
			    buf, (void *)NULL);
		}

int
Tcl_UtfToExternal(
    TCL_UNUSED(Tcl_Interp *),	/* TODO: Re-examine this. */
    Tcl_Encoding encoding,	/* The encoding for the converted string, or
				 * NULL for the default system encoding. */
    const char *src,		/* Source string in UTF-8. */
    Tcl_Size srcLen,		/* Source string length in bytes, or TCL_INDEX_NONE for
				 * strlen(). */
    int flags,			/* Conversion control flags. */
    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
				 * information used during a piecewise
				 * conversion. Contents of statePtr are
				 * initialized and/or reset by conversion
				 * routine under control of flags argument. */
    char *dst,			/* Output buffer in which converted string

	    TclSetProcessGlobalValue(&encodingFileMap, map);
	}
    }

    if ((NULL == chan) && (interp != NULL)) {
	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
		"unknown encoding \"%s\"", name));
	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ENCODING", name, (void *)NULL);

    }
    Tcl_DecrRefCount(fileNameObj);
    Tcl_DecrRefCount(searchPath);

    return chan;
}


    case 'E':
	encoding = LoadEscapeEncoding(name, chan);
	break;
    }
    if ((encoding == NULL) && (interp != NULL)) {
	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
		"invalid encoding file \"%s\"", name));
	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ENCODING", name, (void *)NULL);

    }
    Tcl_CloseEx(NULL, chan, 0);

    return encoding;
}

/*

		/*
		 * To avoid infinite recursion in [encoding system iso2022-*]
		 */

		e = (Encoding *) Tcl_GetEncoding(NULL, est.name);
		if ((e != NULL) && (e->toUtfProc != TableToUtfProc)
			&& (e->toUtfProc != Iso88591ToUtfProc)) {
		   Tcl_FreeEncoding((Tcl_Encoding) e);
		   e = NULL;
		}
		est.encodingPtr = e;
		Tcl_DStringAppend(&escapeData, (char *) &est, sizeof(est));
	    }
	}
	Tcl_Free(argv);
	Tcl_DStringFree(&lineString);

	    result = TCL_CONVERT_MULTIBYTE;
	    break;
	}
	if (dst > dstEnd) {
	    result = TCL_CONVERT_NOSPACE;
	    break;
	}

	if (UCHAR(*src) < 0x80 && !((UCHAR(*src) == 0) && (flags & ENCODING_INPUT))) {
	    /*
	     * Copy 7bit characters, but skip null-bytes when we are in input
	     * mode, so that they get converted to \xC0\x80.
	     */
	    *dst++ = *src++;
	} else if ((UCHAR(*src) == 0xC0) && (src + 1 < srcEnd) &&
		 (UCHAR(src[1]) == 0x80) &&
		 (!(flags & ENCODING_INPUT) || !PROFILE_TCL8(profile))) {
	    /* Special sequence \xC0\x80 */
	    if (!PROFILE_TCL8(profile) && (flags & ENCODING_INPUT)) {
		if (PROFILE_REPLACE(profile)) {
		   dst += Tcl_UniCharToUtf(UNICODE_REPLACE_CHAR, dst);
		   src += 2;
		} else {
		   /* PROFILE_STRICT */
		   result = TCL_CONVERT_SYNTAX;
		   break;
		}
	    } else {
		/*
		 * Convert 0xC080 to real nulls when we are in output mode,
		 * irrespective of the profile.
		 */
		*dst++ = 0;

	     * the user has explicitly asked to be told.
	     */

	    if (flags & ENCODING_INPUT) {
		/* Incomplete bytes for modified UTF-8 target */
		if (PROFILE_STRICT(profile)) {
		    result = (flags & TCL_ENCODING_CHAR_LIMIT)
			       ? TCL_CONVERT_MULTIBYTE
			       : TCL_CONVERT_SYNTAX;
		    break;
		}
	    }
	    if (PROFILE_REPLACE(profile)) {
		ch = UNICODE_REPLACE_CHAR;
		++src;
	    } else {
		/* TCL_ENCODING_PROFILE_TCL8 */
		char chbuf[2];
		chbuf[0] = UCHAR(*src++); chbuf[1] = 0;
		TclUtfToUniChar(chbuf, &ch);
	    }
	    dst += Tcl_UniCharToUtf(ch, dst);
	} else {
	    size_t len = TclUtfToUniChar(src, &ch);
	    if (flags & ENCODING_INPUT) {
		if (((len < 2) && (ch != 0)) || ((ch > 0xFFFF) && !(flags & ENCODING_UTF))) {

		    if (PROFILE_STRICT(profile)) {
			result = TCL_CONVERT_SYNTAX;
			break;
		    } else if (PROFILE_REPLACE(profile)) {
			ch = UNICODE_REPLACE_CHAR;
		    }
		}
	    }

	    const char *saveSrc = src;
	    src += len;
	    if (!(flags & ENCODING_UTF) && !(flags & ENCODING_INPUT) && (ch > 0x3FF)) {

		if (ch > 0xFFFF) {
		    /* CESU-8 6-byte sequence for chars > U+FFFF */
		    ch -= 0x10000;
		    *dst++ = 0xED;
		    *dst++ = (char) (((ch >> 16) & 0x0F) | 0xA0);
		    *dst++ = (char) (((ch >> 10) & 0x3F) | 0x80);
		    ch = (ch & 0x0CFF) | 0xDC00;
		}
		*dst++ = (char)(((ch >> 12) | 0xE0) & 0xEF);
		*dst++ = (char)(((ch >> 6) | 0x80) & 0xBF);
		*dst++ = (char)((ch | 0x80) & 0xBF);
		continue;
	    } else if (SURROGATE(ch)) {
		if (PROFILE_STRICT(profile)) {
		    result = (flags & ENCODING_INPUT) ? TCL_CONVERT_SYNTAX : TCL_CONVERT_UNKNOWN;

		    src = saveSrc;
		    break;
		} else if (PROFILE_REPLACE(profile)) {
		    ch = UNICODE_REPLACE_CHAR;
		}
	    }
	    dst += Tcl_UniCharToUtf(ch, dst);

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

static int
Utf32ToUtfProc(
    void *clientData,	/* additional flags, e.g. TCL_ENCODING_LE */
    const char *src,		/* Source string in Unicode. */
    int srcLen,			/* Source string length in bytes. */
    int flags,			/* Conversion control flags. */
    TCL_UNUSED(Tcl_EncodingState *),
    char *dst,			/* Output buffer in which converted string is
				 * stored. */
    int dstLen,			/* The maximum length of output buffer in

    for (numChars = 0; src < srcEnd && numChars <= charLimit; numChars++) {
	if (dst > dstEnd) {
	    result = TCL_CONVERT_NOSPACE;
	    break;
	}

	if (flags & TCL_ENCODING_LE) {
	    ch = (unsigned int)(src[3] & 0xFF) << 24 | (src[2] & 0xFF) << 16 | (src[1] & 0xFF) << 8 | (src[0] & 0xFF);

	} else {
	    ch = (unsigned int)(src[0] & 0xFF) << 24 | (src[1] & 0xFF) << 16 | (src[2] & 0xFF) << 8 | (src[3] & 0xFF);

	}
	if ((unsigned)ch > 0x10FFFF) {
	    if (PROFILE_STRICT(flags)) {
		result = TCL_CONVERT_SYNTAX;
		break;
	    }
	    ch = UNICODE_REPLACE_CHAR;

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

static int
UtfToUtf32Proc(
    void *clientData,	/* additional flags, e.g. TCL_ENCODING_LE */
    const char *src,		/* Source string in UTF-8. */
    int srcLen,			/* Source string length in bytes. */
    int flags,			/* Conversion control flags. */
    TCL_UNUSED(Tcl_EncodingState *),
    char *dst,			/* Output buffer in which converted string is
				 * stored. */
    int dstLen,			/* The maximum length of output buffer in

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

static int
Utf16ToUtfProc(
    void *clientData,	/* additional flags, e.g. TCL_ENCODING_LE */
    const char *src,		/* Source string in Unicode. */
    int srcLen,			/* Source string length in bytes. */
    int flags,			/* Conversion control flags. */
    TCL_UNUSED(Tcl_EncodingState *),
    char *dst,			/* Output buffer in which converted string is
				 * stored. */
    int dstLen,			/* The maximum length of output buffer in

    srcStart = src;
    srcEnd = src + srcLen;

    dstStart = dst;
    dstEnd = dst + dstLen - TCL_UTF_MAX;

    for (numChars = 0; src < srcEnd && numChars <= charLimit; src += 2, numChars++) {

	if (dst > dstEnd) {
	    result = TCL_CONVERT_NOSPACE;
	    break;
	}

	unsigned short prev = ch;
	if (flags & TCL_ENCODING_LE) {
	    ch = (src[1] & 0xFF) << 8 | (src[0] & 0xFF);
	} else {
	    ch = (src[0] & 0xFF) << 8 | (src[1] & 0xFF);
	}
	if (HIGH_SURROGATE(prev) && !LOW_SURROGATE(ch)) {
	    if (PROFILE_STRICT(flags)) {
		result = TCL_CONVERT_SYNTAX;
		src -= 2; /* Go back to beginning of high surrogate */
		dst--; /* Also undo writing a single byte too much */
		numChars--;
		break;
	    } else if (PROFILE_REPLACE(flags)) {
		/*
		 * Previous loop wrote a single byte to mark the high surrogate.
		 * Replace it with the replacement character. Further, restart
		 * current loop iteration since need to recheck destination space
		 * and reset processing of current character.
		 */
		ch = UNICODE_REPLACE_CHAR;
		dst--;
		dst += Tcl_UniCharToUtf(ch, dst);
		src -= 2;
		numChars--;
		continue;
	    } else {

	    /* Bug [10c2c17c32]. If Hi surrogate not followed by Lo surrogate, finish 3-byte UTF-8 */


		dst += Tcl_UniCharToUtf(-1, dst);
	    }
	}

	/*
	 * Special case for 1-byte utf chars for speed. Make sure we work with
	 * unsigned short-size data.

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

static int
UtfToUtf16Proc(
    void *clientData,	/* additional flags, e.g. TCL_ENCODING_LE */
    const char *src,		/* Source string in UTF-8. */
    int srcLen,			/* Source string length in bytes. */
    int flags,			/* Conversion control flags. */
    TCL_UNUSED(Tcl_EncodingState *),
    char *dst,			/* Output buffer in which converted string is
				 * stored. */
    int dstLen,			/* The maximum length of output buffer in

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

static int
UtfToUcs2Proc(
    void *clientData,	/* additional flags, e.g. TCL_ENCODING_LE */
    const char *src,		/* Source string in UTF-8. */
    int srcLen,			/* Source string length in bytes. */
    int flags,			/* Conversion control flags. */
    TCL_UNUSED(Tcl_EncodingState *),
    char *dst,			/* Output buffer in which converted string is
				 * stored. */
    int dstLen,			/* The maximum length of output buffer in

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

static int
TableToUtfProc(
    void *clientData,	/* TableEncodingData that specifies
				 * encoding. */
    const char *src,		/* Source string in specified encoding. */
    int srcLen,			/* Source string length in bytes. */
    int flags,			/* Conversion control flags. */
    TCL_UNUSED(Tcl_EncodingState *),
    char *dst,			/* Output buffer in which converted string is
				 * stored. */

		    break;
		} else if (PROFILE_STRICT(flags)) {
		    result = TCL_CONVERT_SYNTAX;
		    break;
		} else if (PROFILE_REPLACE(flags)) {
		    ch = UNICODE_REPLACE_CHAR;
		} else {
		    /* For prefix bytes, we don't fallback to cp1252, see [1355b9a874] */

		    ch = byte;
		}
	    } else {
		ch = toUnicode[byte][*((unsigned char *)++src)];
	    }
	} else {
	    ch = pageZero[byte];

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

static int
TableFromUtfProc(
    void *clientData,	/* TableEncodingData that specifies
				 * encoding. */
    const char *src,		/* Source string in UTF-8. */
    int srcLen,			/* Source string length in bytes. */
    int flags,			/* Conversion control flags. */
    TCL_UNUSED(Tcl_EncodingState *),
    char *dst,			/* Output buffer in which converted string is
				 * stored. */

 *	Memory freed.
 *
 *---------------------------------------------------------------------------
 */

static void
TableFreeProc(
    void *clientData)	/* TableEncodingData that specifies
				 * encoding. */
{
    TableEncodingData *dataPtr = (TableEncodingData *)clientData;

    /*
     * Make sure we aren't freeing twice on shutdown. [Bug 219314]
     */

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

static int
EscapeToUtfProc(
    void *clientData,	/* EscapeEncodingData that specifies
				 * encoding. */
    const char *src,		/* Source string in specified encoding. */
    int srcLen,			/* Source string length in bytes. */
    int flags,			/* Conversion control flags. */
    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
				 * information used during a piecewise
				 * conversion. Contents of statePtr are

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

static int
EscapeFromUtfProc(
    void *clientData,	/* EscapeEncodingData that specifies
				 * encoding. */
    const char *src,		/* Source string in UTF-8. */
    int srcLen,			/* Source string length in bytes. */
    int flags,			/* Conversion control flags. */
    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
				 * information used during a piecewise
				 * conversion. Contents of statePtr are

	memcpy(dst, dataPtr->init, dataPtr->initLen);
	dst += dataPtr->initLen;
    } else {
	state = PTR2INT(*statePtr);
    }

    encodingPtr = GetTableEncoding(dataPtr, state);
    tableDataPtr = (const TableEncodingData *)encodingPtr->clientData;
    tablePrefixBytes = tableDataPtr->prefixBytes;
    tableFromUnicode = (const unsigned short *const *)
	    tableDataPtr->fromUnicode;

    for (numChars = 0; src < srcEnd; numChars++) {
	unsigned len;
	int word;

	if ((word == 0) && (ch != 0)) {
	    int oldState;
	    const EscapeSubTable *subTablePtr;

	    oldState = state;
	    for (state = 0; state < dataPtr->numSubTables; state++) {
		encodingPtr = GetTableEncoding(dataPtr, state);
		tableDataPtr = (const TableEncodingData *)encodingPtr->clientData;
		word = tableDataPtr->fromUnicode[(ch >> 8)][ch & 0xFF];
		if (word != 0) {
		    break;
		}
	    }

	    if (word == 0) {
		state = oldState;
		if (PROFILE_STRICT(flags)) {
		    result = TCL_CONVERT_UNKNOWN;
		    break;
		}
		encodingPtr = GetTableEncoding(dataPtr, state);
		tableDataPtr = (const TableEncodingData *)encodingPtr->clientData;
		word = tableDataPtr->fallback;
	    }

	    tablePrefixBytes = (const char *) tableDataPtr->prefixBytes;
	    tableFromUnicode = (const unsigned short *const *)
		    tableDataPtr->fromUnicode;

 *	Memory is freed.
 *
 *---------------------------------------------------------------------------
 */

static void
EscapeFreeProc(
    void *clientData)	/* EscapeEncodingData that specifies
				 * encoding. */
{
    EscapeEncodingData *dataPtr = (EscapeEncodingData *)clientData;
    EscapeSubTable *subTablePtr;
    int i;

    if (dataPtr == NULL) {

    }
    if (interp) {
	/* This code assumes at least two profiles :-) */
	Tcl_Obj *errorObj = Tcl_ObjPrintf("bad profile name \"%s\": must be",
		profileName);
	for (i = 0; i < (numProfiles - 1); ++i) {
	    Tcl_AppendStringsToObj(
		errorObj, " ", encodingProfiles[i].name, ",", (void *)NULL);

	}
	Tcl_AppendStringsToObj(
	    errorObj, " or ", encodingProfiles[numProfiles-1].name, (void *)NULL);


	Tcl_SetObjResult(interp, errorObj);
	Tcl_SetErrorCode(
	    interp, "TCL", "ENCODING", "PROFILE", profileName, (void *)NULL);

    }
    return TCL_ERROR;
}

/*
 *------------------------------------------------------------------------
 *

const char *
TclEncodingProfileIdToName(
    Tcl_Interp *interp,		/* For error messages. May be NULL */
    int profileValue)		/* Profile #define value */
{
    size_t i;

    for (i = 0; i < sizeof(encodingProfiles) / sizeof(encodingProfiles[0]); ++i) {

	if (profileValue == encodingProfiles[i].value) {
	    return encodingProfiles[i].name;
	}
    }
    if (interp) {
	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
		"Internal error. Bad profile id \"%d\".",
		profileValue));
	Tcl_SetErrorCode(
	    interp, "TCL", "ENCODING", "PROFILEID", (void *)NULL);
    }
    return NULL;
}

/*
 *------------------------------------------------------------------------
 *

    Tcl_Interp *interp)
{
    size_t i, n;
    Tcl_Obj *objPtr;
    n = sizeof(encodingProfiles) / sizeof(encodingProfiles[0]);
    objPtr = Tcl_NewListObj(n, NULL);
    for (i = 0; i < n; ++i) {
	Tcl_ListObjAppendElement(
	    interp, objPtr, Tcl_NewStringObj(encodingProfiles[i].name, TCL_INDEX_NONE));
    }
    Tcl_SetObjResult(interp, objPtr);
}

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