Tcl Source Code

2011-05-07  Miguel Sofer  <[email protected]>

	* generic/tclInt.h: fix USE_TCLALLOC so that it can be enabled
	* unix/Makefile.in: without editing the Makefile

2011-05-05  Don Porter  <[email protected]>

	* generic/tclListObj.c:	Stop generating string rep of dict when
	converting to list.  Tolerate NULL interps more completely.

2011-05-03  Don Porter  <[email protected]>

	* generic/tclUtil.c:	Tighten Tcl_SplitList().
	* generic/tclListObj.c:	Tighten SetListFromAny().
	* generic/tclDictObj.c:	Tighten SetDictFromAny().
	* tests/join.test:
	* tests/mathop.test:

2011-05-02  Don Porter  <[email protected]>

	* generic/tclCmdMZ.c:	Revised TclFindElement() interface.  The
	* generic/tclDictObj.c:	final argument had been bracePtr, the address
	* generic/tclListObj.c:	of a boolean var, where the caller can be told
	* generic/tclParse.c:	whether or not the parsed list element was
	* generic/tclUtil.c:	enclosed in braces.  In practice, no callers
	really care about that.  What the callers really want to know is 
	whether the list element value exists as a literal substring of the
	string being parsed, or whether a call to TclCopyAndCollpase() is
	needed to produce the list element value.  Now the final argument
	is changed to do what callers actually need.  This is a better fit
	for the calls in tclParse.c, where now a good deal of post-processing
	checking for "naked backslashes" is no longer necessary.
	***POTENTIAL INCOMPATIBILITY***
	For any callers calling in via the internal stubs table who really
	do use the final argument explicitly to check for the enclosing brace
	scenario.  Simply looking for the braces where they must be is the
	revision available to those callers, and it will backport cleanly.

	* tests/parse.test:	Tests for expanded literals quoting detection.

	* generic/tclCompCmdsSZ.c:	New TclFindElement() is also a better
	fit for the [switch] compiler.

	* generic/tclInt.h:	Replace TclCountSpaceRuns() with
	* generic/tclListObj.c:	TclMaxListLength() which is the function we
	* generic/tclUtil.c:	actually want.
	* generic/tclCompCmdsSZ.c:

	* generic/tclCompCmdsSZ.c: Rewrite of parts of the switch compiler to
	better use the powers of TclFindElement() and do less parsing on
	its own.

2011-04-28  Don Porter  <[email protected]>

	* generic/tclInt.h:	New utility routines:
	* generic/tclParse.c:	TclIsSpaceProc() and
	* generic/tclUtil.c:	TclCountSpaceRuns()

	* generic/tclCmdMZ.c:	Use new routines to replace calls to
	* generic/tclListObj.c:	isspace() and their /* INTL */ risk.
	* generic/tclStrToD.c:
	* generic/tclUtf.c:
	* unix/tclUnixFile.c:
	
	* generic/tclStringObj.c:	Improved reaction to out of memory.

2011-04-27  Don Porter  <[email protected]>

	* generic/tclCmdMZ.c:	TclFreeIntRep() correction & cleanup.
	* generic/tclExecute.c:
	* generic/tclIndexObj.c:
	* generic/tclInt.h:
	* generic/tclListObj.c:
	* generic/tclNamesp.c:
	* generic/tclResult.c:
	* generic/tclStringObj.c:
	* generic/tclVar.c:

	* generic/tclListObj.c:	FreeListInternalRep() cleanup.

2011-04-21  Don Porter  <[email protected]>

	* generic/tclInt.h:	Use macro to set List intreps.
	* generic/tclListObj.c:

	* generic/tclCmdIL.c:	Limits on list length were too strict.
	* generic/tclInt.h:	Revised panics to errors where possible.
	* generic/tclListObj.c:
	* tests/lrepeat.test:

	* generic/tclCompile.c:	Make sure SetFooFromAny routines react
	* generic/tclIO.c:	reasonably when passed a NULL interp.
	* generic/tclIndexObj.c:
	* generic/tclListObj.c:
	* generic/tclNamesp.c:
	* generic/tclObj.c:
	* generic/tclProc.c:
	* macosx/tclMacOSXFCmd.c:

2011-04-21  Jan Nijtmans  <[email protected]>

	* generic/tcl.h:       fix for [Bug 3288345]: Wrong Tcl_StatBuf
	* generic/tclInt.h:    used on MinGW. Make sure that all _WIN32
	* win/tclWinFile.c:    compilers use exactly the same layout
	* win/configure.in:    for Tcl_StatBuf - the one used by MSVC6 -
	* win/configure:       in all situations.

2011-04-19  Don Porter  <[email protected]>

	* generic/tclConfig.c:	Reduce internals access in the implementation
	of [<foo>::pkgconfig list].

2011-04-18  Don Porter  <[email protected]>

	* generic/tclCmdIL.c:	Use ListRepPtr(.) and other cleanup.
	* generic/tclConfig.c:
	* generic/tclListObj.c:

	* generic/tclInt.h:	Define and use macros that test whether
	* generic/tclBasic.c:	a Tcl list value is canonical.
	* generic/tclUtil.c:

2011-04-18  Donal K. Fellows  <[email protected]>

	* doc/dict.n: [Bug 3288696]: Command summary was confusingly wrong
	when it came to [dict filter] with a 'value' filter.





2011-04-16  Donal K. Fellows  <[email protected]>

	* generic/tclFCmd.c (TclFileAttrsCmd): Add comments to make this code
	easier to understand. Added a panic to handle the case where the VFS
	layer does something odd.

2011-04-13  Don Porter  <[email protected]>

of arguments.  \fBTcl_VarEval\fR is now deprecated.
.PP
\fBTcl_VarEvalVA\fR is the same as \fBTcl_VarEval\fR except that
instead of taking a variable number of arguments it takes an argument
list. Like \fBTcl_VarEval\fR, \fBTcl_VarEvalVA\fR is deprecated.

.SH "FLAG BITS"
.PP
Any ORed combination of the following values may be used for the
\fIflags\fR argument to procedures such as \fBTcl_EvalObjEx\fR:
.TP 23
\fBTCL_EVAL_DIRECT\fR
.
This flag is only used by \fBTcl_EvalObjEx\fR; it is ignored by
other procedures.  If this flag bit is set, the script is not
compiled to bytecodes; instead it is executed directly
as is done by \fBTcl_EvalEx\fR.  The
\fBTCL_EVAL_DIRECT\fR flag is useful in situations where the
contents of an object are going to change immediately, so the
bytecodes will not be reused in a future execution.  In this case,
it is faster to execute the script directly.
.TP 23
\fBTCL_EVAL_GLOBAL\fR
.
If this flag is set, the script is processed at global level.  This
means that it is evaluated in the global namespace and its variable
context consists of global variables only (it ignores any Tcl
procedures that are active).

.SH "MISCELLANEOUS DETAILS"
.PP
During the processing of a Tcl command it is legal to make nested
calls to evaluate other commands (this is how procedures and
some control structures are implemented).
If a code other than \fBTCL_OK\fR is returned

 *
 * Note on converting between Tcl_WideInt and strings. This implementation (in
 * tclObj.c) depends on the function
 * sprintf(...,"%" TCL_LL_MODIFIER "d",...).
 */

#if !defined(TCL_WIDE_INT_TYPE)&&!defined(TCL_WIDE_INT_IS_LONG)
#   if defined(__WIN32__) && !defined(__CYGWIN__)
#      define TCL_WIDE_INT_TYPE __int64
#      ifdef __BORLANDC__
typedef struct stati64 Tcl_StatBuf;
#         define TCL_LL_MODIFIER	"L"
#      else /* __BORLANDC__ */
#         if defined(_WIN64)
typedef struct __stat64 Tcl_StatBuf;
#         elif (defined(_MSC_VER) && (_MSC_VER < 1400))
typedef struct _stati64	Tcl_StatBuf;
#         else
typedef struct _stat32i64 Tcl_StatBuf;
#         endif /* _MSC_VER < 1400 */
#         define TCL_LL_MODIFIER	"I64"
#      endif /* __BORLANDC__ */
#   elif defined(__GNUC__)
#      define TCL_WIDE_INT_TYPE long long
#      define TCL_LL_MODIFIER	"ll"
#      if defined(__WIN32__)
typedef struct _stat32i64 Tcl_StatBuf;
#      else
typedef struct stat	Tcl_StatBuf;
#      endif
#   else /* ! __WIN32__ && ! __GNUC__ */
/*
 * Don't know what platform it is and configure hasn't discovered what is
 * going on for us. Try to guess...
 */
#      ifdef NO_LIMITS_H
#	  error please define either TCL_WIDE_INT_TYPE or TCL_WIDE_INT_IS_LONG
#      else /* !NO_LIMITS_H */

    /*
     * Skip forward to the interesting arguments now we've finished parsing.
     */

    objc -= 2;
    objv += 2;


    /* Final sanity check. Do not exceed limits on max list length. */






    if (elementCount && objc > LIST_MAX/elementCount) {
	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
		"max length of a Tcl list (%d elements) exceeded", LIST_MAX));
        Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
	return TCL_ERROR;
    }
    totalElems = objc * elementCount;





    /*
     * Get an empty list object that is allocated large enough to hold each
     * init value elementCount times.
     */

    listPtr = Tcl_NewListObj(totalElems, NULL);

	    result = 0;
	    failat = 0;
	} else {
	    failat = stop - string1;
	    if (stop < end) {
		result = 0;
		TclFreeIntRep(objPtr);

	    }
	}
	break;
    }
    case STR_IS_GRAPH:
	chcomp = Tcl_UniCharIsGraph;
	break;

		 * so return failure index as the point where parsing stopped.
		 * Clear out the internal rep, since keeping it would leave
		 * *objPtr in an inconsistent state.
		 */

		failat = stop - string1;
		TclFreeIntRep(objPtr);

	    }
	} else {
	    /*
	     * No prefix is a valid integer. Fail at beginning.
	     */

	    failat = 0;

	    /*
	     * Need to figure out where the list parsing failed, which is
	     * fairly expensive. This is adapted from the core of
	     * SetListFromAny().
	     */

	    const char *elemStart, *nextElem;
	    int lenRemain, elemSize;
	    register const char *p;

	    string1 = TclGetStringFromObj(objPtr, &length1);
	    end = string1 + length1;
	    failat = -1;
	    for (p=string1, lenRemain=length1; lenRemain > 0;
		    p=nextElem, lenRemain=end-nextElem) {
		if (TCL_ERROR == TclFindElement(NULL, p, lenRemain,
			&elemStart, &nextElem, &elemSize, NULL)) {
		    Tcl_Obj *tmpStr;

		    /*
		     * This is the simplest way of getting the number of
		     * characters parsed. Note that this is not the same as
		     * the number of bytes when parsing strings with non-ASCII
		     * characters in them.
		     *
		     * Skip leading spaces first. This is only really an issue
		     * if it is the first "element" that has the failure.
		     */

		    while (TclIsSpaceProc(*p)) {
			p++;
		    }
		    TclNewStringObj(tmpStr, string1, p-string1);
		    failat = Tcl_GetCharLength(tmpStr);
		    TclDecrRefCount(tmpStr);
		    break;
		}

				 * created by Tcl_ParseCommand. */
    Command *cmdPtr,		/* Points to defintion of command being
				 * compiled. */
    CompileEnv *envPtr)		/* Holds resulting instructions. */
{
    Tcl_Token *tokenPtr;	/* Pointer to tokens in command. */
    int numWords;		/* Number of words in command. */

    Tcl_Token *valueTokenPtr;	/* Token for the value to switch on. */
    enum {Switch_Exact, Switch_Glob, Switch_Regexp} mode;
				/* What kind of switch are we doing? */

    Tcl_Token *bodyTokenArray;	/* Array of real pattern list items. */
    Tcl_Token **bodyToken;	/* Array of pointers to pattern list items. */
    int *bodyLines;		/* Array of line numbers for body list
				 * items. */
    int **bodyContLines;	/* Array of continuation line info. */
    int noCase;			/* Has the -nocase flag been given? */
    int foundMode = 0;		/* Have we seen a mode flag yet? */

    int i, valueIndex;
    int result = TCL_ERROR;
    DefineLineInformation;	/* TIP #280 */
    int *clNext = envPtr->clNext;

    /*
     * Only handle the following versions:

     * that in the case of the quoted bodies, this is tricky as we cannot use
     * copies of the string from the input token for the generated tokens (it
     * causes a crash during exception handling). When multiple tokens are
     * available at this point, this is pretty easy.
     */

    if (numWords == 1) {

	const char *bytes;
	int maxLen, numBytes;
	int bline;		/* TIP #280: line of the pattern/action list,
				 * and start of list for when tracking the
				 * location. This list comes immediately after
				 * the value we switch on. */








	if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
	    return TCL_ERROR;
	}
	bytes = tokenPtr[1].start;

	numBytes = tokenPtr[1].size;






	/* Allocate enough space to work in. */





	maxLen = TclMaxListLength(bytes, numBytes, NULL);
	if (maxLen < 2)  {

	    return TCL_ERROR;
	}


	bodyTokenArray = ckalloc(sizeof(Tcl_Token) * maxLen);
	bodyToken = ckalloc(sizeof(Tcl_Token *) * maxLen);
	bodyLines = ckalloc(sizeof(int) * maxLen);
	bodyContLines = ckalloc(sizeof(int*) * maxLen);





	bline = mapPtr->loc[eclIndex].line[valueIndex+1];








	numWords = 0;

	while (numBytes > 0) {

	    const char *prevBytes = bytes;
	    int literal;



	    if (TCL_OK != TclFindElement(NULL, bytes, numBytes,
		    &(bodyTokenArray[numWords].start), &bytes,
		    &(bodyTokenArray[numWords].size), &literal) || !literal) {
	    abort:
		ckfree((char *) bodyToken);
		ckfree((char *) bodyTokenArray);
		ckfree((char *) bodyLines);
		ckfree((char *) bodyContLines);
		return TCL_ERROR;






	    }






	    bodyTokenArray[numWords].type = TCL_TOKEN_TEXT;
	    bodyTokenArray[numWords].numComponents = 0;
	    bodyToken[numWords] = bodyTokenArray + numWords;

	    /*
	     * TIP #280: Now determine the line the list element starts on
	     * (there is no need to do it earlier, due to the possibility of
	     * aborting, see above).
	     */

	    TclAdvanceLines(&bline, prevBytes, bodyTokenArray[numWords].start);
	    TclAdvanceContinuations(&bline, &clNext,
		    bodyTokenArray[numWords].start - envPtr->source);
	    bodyLines[numWords] = bline;
	    bodyContLines[numWords] = clNext;
	    TclAdvanceLines(&bline, bodyTokenArray[numWords].start, bytes);
	    TclAdvanceContinuations(&bline, &clNext, bytes - envPtr->source);







	    numBytes -= (bytes - prevBytes);
	    numWords++;



	}


	if (numWords % 2) {





	    goto abort;


	}

    } else if (numWords % 2 || numWords == 0) {
	/*
	 * Odd number of words (>1) available, or no words at all available.
	 * Both are error cases, so punt and let the interpreted-version
	 * generate the error message. Note that the second case probably
	 * should get caught earlier, but it's easy to check here again anyway
	 * because it'd cause a nasty crash otherwise.

	for (i=0 ; i<numWords ; i++) {
	    /*
	     * We only handle the very simplest case. Anything more complex is
	     * a good reason to go to the interpreted case anyway due to
	     * traces, etc.
	     */

	    if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {

		goto freeTemporaries;
	    }
	    bodyToken[i] = tokenPtr+1;

	    /*
	     * TIP #280: Copy line information from regular cmd info.
	     */

     * Now we commit to generating code; the parsing stage per se is done.
     * Check if we can generate a jump table, since if so that's faster than
     * doing an explicit compare with each body. Note that we're definitely
     * over-conservative with determining whether we can do the jump table,
     * but it handles the most common case well enough.
     */

    if (mode == Switch_Exact) {
	IssueSwitchJumpTable(interp, envPtr, mapPtr, eclIndex, valueIndex,
		valueTokenPtr, numWords, bodyToken, bodyLines, bodyContLines);
    } else {
	IssueSwitchChainedTests(interp, envPtr, mapPtr, eclIndex, mode,noCase,
		valueIndex, valueTokenPtr, numWords, bodyToken, bodyLines,
		bodyContLines);
    }

 *
 * TclSetByteCodeFromAny --
 *
 *	Part of the bytecode Tcl object type implementation. Attempts to
 *	generate an byte code internal form for the Tcl object "objPtr" by
 *	compiling its string representation. This function also takes a hook
 *	procedure that will be invoked to perform any needed post processing
 *	on the compilation results before generating byte codes. interp is
 *	compilation context and may not be NULL.
 *
 * Results:
 *	The return value is a standard Tcl object result. If an error occurs
 *	during compilation, an error message is left in the interpreter's
 *	result.
 *
 * Side effects:
 *	Frees the old internal representation. If no error occurs, then the
 *	compiled code is stored as "objPtr"s bytecode representation. Also, if
 *	debugging, initializes the "tcl_traceCompile" Tcl variable used to
 *	trace compilations.
 *

static int
SetByteCodeFromAny(
    Tcl_Interp *interp,		/* The interpreter for which the code is being
				 * compiled. Must not be NULL. */
    Tcl_Obj *objPtr)		/* The object to make a ByteCode object. */
{
    if (interp == NULL) {
	return TCL_ERROR;
    }
    TclSetByteCodeFromAny(interp, objPtr, NULL, NULL);
    return TCL_OK;
}

/*
 *----------------------------------------------------------------------
 *

	    Tcl_SetResult(interp, "insufficient memory to create list",
		    TCL_STATIC);
	    Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
	    return TCL_ERROR;
	}

	if (n) {

	    Tcl_DictSearch s;
	    Tcl_Obj *key;
	    int done;




	    for (Tcl_DictObjFirst(interp, pkgDict, &s, &key, NULL, &done);
		    !done; Tcl_DictObjNext(&s, &key, NULL, &done)) {
		Tcl_ListObjAppendElement(NULL, listPtr, key);

	    }
	}

	Tcl_SetObjResult(interp, listPtr);
	return TCL_OK;

    default:

 */

static int
SetDictFromAny(
    Tcl_Interp *interp,
    Tcl_Obj *objPtr)
{

    Tcl_HashEntry *hPtr;

    int isNew, result;



    Dict *dict = ckalloc(sizeof(Dict));

    InitChainTable(dict);

    /*
     * Since lists and dictionaries have very closely-related string
     * representations (i.e. the same parsing code) we can safely special-case
     * the conversion from lists to dictionaries.
     */

    if (objPtr->typePtr == &tclListType) {
	int objc, i;
	Tcl_Obj **objv;

	/* Cannot fail, we already know the Tcl_ObjType is "list". */
	TclListObjGetElements(NULL, objPtr, &objc, &objv);


	if (objc & 1) {

	    goto missingValue;


	}


	for (i=0 ; i<objc ; i+=2) {



	




	    /* Store key and value in the hash table we're building. */


	    hPtr = CreateChainEntry(dict, objv[i], &isNew);
	    if (!isNew) {
		Tcl_Obj *discardedValue = Tcl_GetHashValue(hPtr);

		/*
		 * Not really a well-formed dictionary as there are duplicate
		 * keys, so better get the string rep here so that we can
		 * convert back.
		 */

		(void) Tcl_GetString(objPtr);

		TclDecrRefCount(discardedValue);
	    }
	    Tcl_SetHashValue(hPtr, objv[i+1]);
	    Tcl_IncrRefCount(objv[i+1]); /* Since hash now holds ref to it */
	}
    } else {



	int length;







	const char *nextElem = TclGetStringFromObj(objPtr, &length);
	const char *limit = (nextElem + length);



	while (nextElem < limit) {
	    Tcl_Obj *keyPtr, *valuePtr;

	    const char *elemStart;

	    int elemSize, literal;



	    result = TclFindElement(interp, nextElem, (limit - nextElem),
		    &elemStart, &nextElem, &elemSize, &literal);
	    if (result != TCL_OK) {



		goto errorExit;
	    }
	    if (elemStart == limit) {
		break;
	    }
	    if (nextElem == limit) {
		goto missingValue;
	    }






	    if (literal) {
		TclNewStringObj(keyPtr, elemStart, elemSize);

	    } else {
		/* Avoid double copy */
		TclNewObj(keyPtr);
		keyPtr->bytes = ckalloc((unsigned) elemSize + 1);
		keyPtr->length = TclCopyAndCollapse(elemSize, elemStart,



			keyPtr->bytes);

	    }






	    result = TclFindElement(interp, nextElem, (limit - nextElem),
		    &elemStart, &nextElem, &elemSize, &literal);
	    if (result != TCL_OK) {



		TclDecrRefCount(keyPtr);
		goto errorExit;
	    }



	    if (literal) {




		TclNewStringObj(valuePtr, elemStart, elemSize);




	    } else {


		/* Avoid double copy */
		TclNewObj(valuePtr);
		valuePtr->bytes = ckalloc((unsigned) elemSize + 1);
		valuePtr->length = TclCopyAndCollapse(elemSize, elemStart,
			valuePtr->bytes);
	    }

	    /* Store key and value in the hash table we're building. */


	    hPtr = CreateChainEntry(dict, keyPtr, &isNew);
	    if (!isNew) {
		Tcl_Obj *discardedValue = Tcl_GetHashValue(hPtr);

		TclDecrRefCount(keyPtr);
		TclDecrRefCount(discardedValue);
	    }
	    Tcl_SetHashValue(hPtr, valuePtr);
	    Tcl_IncrRefCount(valuePtr); /* since hash now holds ref to it */
	}
    }

    /*
     * Free the old internalRep before setting the new one. We do this as late
     * as possible to allow the conversion code, in particular
     * Tcl_GetStringFromObj, to use that old internalRep.
     */

    TclFreeIntRep(objPtr);
    dict->epoch = 0;
    dict->chain = NULL;
    dict->refcount = 1;
    objPtr->internalRep.otherValuePtr = dict;
    objPtr->typePtr = &tclDictType;
    return TCL_OK;

  missingValue:
    if (interp != NULL) {
	Tcl_SetResult(interp, "missing value to go with key", TCL_STATIC);
	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DICTIONARY", NULL);
    }

    result = TCL_ERROR;

  errorExit:
    if (interp != NULL) {
	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DICTIONARY", NULL);
    }
    DeleteChainTable(dict);
    ckfree(dict);
    return result;
}

/*
 *----------------------------------------------------------------------

		/* TODO: convert panic to error ? */
		Tcl_Panic("max size for a Tcl value (%d bytes) exceeded",
			INT_MAX);
	    }
#if !TCL_COMPILE_DEBUG
	    if (bytes != tclEmptyStringRep && !Tcl_IsShared(objResultPtr)) {
		TclFreeIntRep(objResultPtr);

		objResultPtr->bytes = ckrealloc(bytes, length+appendLen+1);
		objResultPtr->length = length + appendLen;
		p = TclGetString(objResultPtr) + length;
		currPtr = &OBJ_AT_DEPTH(opnd - 2);
	    } else
#endif
	    {

SetChannelFromAny(
    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
    register Tcl_Obj *objPtr)	/* The object to convert. */
{
    ChannelState *statePtr;
    Interp *interpPtr;

    if (interp == NULL) {
	return TCL_ERROR;
    }
    if (objPtr->typePtr == &tclChannelType) {
	/*
	 * The channel is valid until any call to DetachChannel occurs.
	 * Ensure consistency checks are done.
	 */

	statePtr = GET_CHANNELSTATE(objPtr);

     * Most commands are plugged directly together, but some are done via
     * alias-like rewriting; [chan configure] is this way for security reasons
     * (want overwriting of [fconfigure] to control that nicely), and [chan
     * names] because the functionality isn't available as a separate command
     * function at the moment.
     */
    static const EnsembleImplMap initMap[] = {
	{"blocked",	Tcl_FblockedObjCmd, NULL, NULL, NULL, 0},
	{"close",	Tcl_CloseObjCmd, NULL, NULL, NULL, 0},
	{"copy",	Tcl_FcopyObjCmd, NULL, NULL, NULL, 0},
	{"create",	TclChanCreateObjCmd, NULL, NULL, NULL, 0},		/* TIP #219 */
	{"eof",		Tcl_EofObjCmd, NULL, NULL, NULL, 0},
	{"event",	Tcl_FileEventObjCmd, NULL, NULL, NULL, 0},
	{"flush",	Tcl_FlushObjCmd, NULL, NULL, NULL, 0},
	{"gets",	Tcl_GetsObjCmd, NULL, NULL, NULL, 0},
	{"names",	TclChannelNamesCmd, NULL, NULL, NULL, 0},
	{"pending",	ChanPendingObjCmd, NULL, NULL, NULL, 0},		/* TIP #287 */
	{"pop",		TclChanPopObjCmd, NULL, NULL, NULL, 0},		/* TIP #230 */
	{"postevent",	TclChanPostEventObjCmd, NULL, NULL, NULL, 0},	/* TIP #219 */
	{"push",	TclChanPushObjCmd, NULL, NULL, NULL, 0},		/* TIP #230 */
	{"puts",	Tcl_PutsObjCmd, NULL, NULL, NULL, 0},
	{"read",	Tcl_ReadObjCmd, NULL, NULL, NULL, 0},
	{"seek",	Tcl_SeekObjCmd, NULL, NULL, NULL, 0},
	{"pipe",	ChanPipeObjCmd, NULL, NULL, NULL, 0},		/* TIP #304 */
	{"tell",	Tcl_TellObjCmd, NULL, NULL, NULL, 0},
	{"truncate",	ChanTruncateObjCmd, NULL, NULL, NULL, 0},		/* TIP #208 */
	{NULL, NULL, NULL, NULL, NULL, 0}
    };
    static const char *const extras[] = {
	"configure",	"::fconfigure",
	NULL
    };
    Tcl_Command ensemble;

	    sizeof(char *), msg, flags, indexPtr);

    /*
     * The internal rep must be cleared since tablePtr will go away.
     */

    TclFreeIntRep(objPtr);

    ckfree(tablePtr);

    return result;
}

/*
 *----------------------------------------------------------------------

 */

static int
SetIndexFromAny(
    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
    register Tcl_Obj *objPtr)	/* The object to convert. */
{
    if (interp) {
    Tcl_SetObjResult(interp, Tcl_NewStringObj(
	    "can't convert value to index except via Tcl_GetIndexFromObj API",
	    -1));
    }
    return TCL_ERROR;
}

/*
 *----------------------------------------------------------------------
 *
 * UpdateStringOfIndex --

				 * derived from the list representation. May
				 * be ignored if there is no string rep at
				 * all.*/
    Tcl_Obj *elements;		/* First list element; the struct is grown to
				 * accomodate all elements. */
} List;

#define LIST_MAX \
	(1 + (int)(((size_t)UINT_MAX - sizeof(List))/sizeof(Tcl_Obj *)))

/*
 * Macro used to get the elements of a list object.
 */

#define ListRepPtr(listPtr) \
    ((List *) (listPtr)->internalRep.twoPtrValue.ptr1)

#define ListSetIntRep(objPtr, listRepPtr) \
    (objPtr)->internalRep.twoPtrValue.ptr1 = (void *)(listRepPtr), \
    (objPtr)->internalRep.twoPtrValue.ptr2 = NULL, \
    (listRepPtr)->refCount++, \
    (objPtr)->typePtr = &tclListType

#define ListObjGetElements(listPtr, objc, objv) \
    ((objv) = &(ListRepPtr(listPtr)->elements), \
     (objc) = ListRepPtr(listPtr)->elemCount)

#define ListObjLength(listPtr, len) \
    ((len) = ListRepPtr(listPtr)->elemCount)

MODULE_SCOPE void	TclInitLimitSupport(Tcl_Interp *interp);
MODULE_SCOPE void	TclInitNamespaceSubsystem(void);
MODULE_SCOPE void	TclInitNotifier(void);
MODULE_SCOPE void	TclInitObjSubsystem(void);
MODULE_SCOPE void	TclInitSubsystems(void);
MODULE_SCOPE int	TclInterpReady(Tcl_Interp *interp);
MODULE_SCOPE int	TclIsLocalScalar(const char *src, int len);
MODULE_SCOPE int	TclIsSpaceProc(char byte);
MODULE_SCOPE int	TclJoinThread(Tcl_ThreadId id, int *result);
MODULE_SCOPE void	TclLimitRemoveAllHandlers(Tcl_Interp *interp);
MODULE_SCOPE Tcl_Obj *	TclLindexList(Tcl_Interp *interp,
			    Tcl_Obj *listPtr, Tcl_Obj *argPtr);
MODULE_SCOPE Tcl_Obj *	TclLindexFlat(Tcl_Interp *interp, Tcl_Obj *listPtr,
			    int indexCount, Tcl_Obj *const indexArray[]);
/* TIP #280 */
MODULE_SCOPE void	TclListLines(Tcl_Obj *listObj, int line, int n,
			    int *lines, Tcl_Obj *const *elems);
MODULE_SCOPE Tcl_Obj *	TclListObjCopy(Tcl_Interp *interp, Tcl_Obj *listPtr);
MODULE_SCOPE Tcl_Obj *	TclLsetList(Tcl_Interp *interp, Tcl_Obj *listPtr,
			    Tcl_Obj *indexPtr, Tcl_Obj *valuePtr);
MODULE_SCOPE Tcl_Obj *	TclLsetFlat(Tcl_Interp *interp, Tcl_Obj *listPtr,
			    int indexCount, Tcl_Obj *const indexArray[],
			    Tcl_Obj *valuePtr);
MODULE_SCOPE Tcl_Command TclMakeEnsemble(Tcl_Interp *interp, const char *name,
			    const EnsembleImplMap map[]);
MODULE_SCOPE int	TclMaxListLength(CONST char *bytes, int numBytes,
			    CONST char **endPtr);
MODULE_SCOPE int	TclMergeReturnOptions(Tcl_Interp *interp, int objc,
			    Tcl_Obj *const objv[], Tcl_Obj **optionsPtrPtr,
			    int *codePtr, int *levelPtr);
MODULE_SCOPE int	TclNokia770Doubles(void);
MODULE_SCOPE void	TclNsDecrRefCount(Namespace *nsPtr);
MODULE_SCOPE void	TclObjVarErrMsg(Tcl_Interp *interp, Tcl_Obj *part1Ptr,
			    Tcl_Obj *part2Ptr, const char *operation,

 * "prototype" for this macro is:
 *
 * MODULE_SCOPE void	TclFreeIntRep(Tcl_Obj *objPtr);
 *----------------------------------------------------------------
 */

#define TclFreeIntRep(objPtr) \
    if ((objPtr)->typePtr != NULL) { \
	if ((objPtr)->typePtr->freeIntRepProc != NULL) { \
	    (objPtr)->typePtr->freeIntRepProc(objPtr); \
	} \
	(objPtr)->typePtr = NULL; \
    }

/*
 *----------------------------------------------------------------
 * Macro used by the Tcl core to clean out an object's string representation.
 * The ANSI C "prototype" for this macro is:

#include "tclInt.h"

/*
 * Prototypes for functions defined later in this file:
 */

static List *		AttemptNewList(Tcl_Interp *interp, int objc,
			    Tcl_Obj *const objv[]);
static List *		NewListIntRep(int objc, Tcl_Obj *const objv[], int p);
static void		DupListInternalRep(Tcl_Obj *srcPtr, Tcl_Obj *copyPtr);
static void		FreeListInternalRep(Tcl_Obj *listPtr);
static int		SetListFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
static void		UpdateStringOfList(Tcl_Obj *listPtr);

/*
 * The structure below defines the list Tcl object type by means of functions

};

/*
 *----------------------------------------------------------------------
 *
 * NewListIntRep --
 *
 *	Creates a list internal rep with space for objc elements.  objc
 *	must be > 0.  If objv!=NULL, initializes with the first objc values
 *	in that array.  If objv==NULL, initalize list internal rep to have
 *	0 elements, with space to add objc more.  Flag value "p" indicates
 *	how to behave on failure.
 *
 * Results:
 *	A new List struct with refCount 0 is returned. If some failure
 *	prevents this then if p=0, NULL is returned and otherwise the
 *	routine panics.
 *
 * Side effects:
 *	The ref counts of the elements in objv are incremented since the
 *	resulting list now refers to them.
 *
 *----------------------------------------------------------------------
 */

#define Elems2Size(n)				\
    (sizeof(List) - sizeof(Tcl_Obj *) + n*sizeof(Tcl_Obj *))

#define Size2Elems(s)							\
    (s - (sizeof(List) - sizeof(Tcl_Obj *)))/sizeof(Tcl_Obj *)

static List *
NewListIntRep(
    int objc,
    Tcl_Obj *const objv[],
    int p)
{
    List *listRepPtr;

    if (objc <= 0) {
	Tcl_Panic("NewListIntRep: expects postive element count");
    }

    /*
     * First check to see if we'd overflow and try to allocate an object
     * larger than our memory allocator allows. Note that this is actually a
     * fairly small value when you're on a serious 64-bit machine, but that
     * requires API changes to fix. See [Bug 219196] for a discussion.
     */

    if ((size_t)objc > LIST_MAX) {
	if (p) {
	    Tcl_Panic("max length of a Tcl list (%d elements) exceeded",
		    LIST_MAX);
	}
	return NULL;
    }

    listRepPtr = attemptckalloc(Elems2Size(objc));
    if (listRepPtr == NULL) {
	if (p) {
	    Tcl_Panic("list creation failed: unable to alloc %u bytes",
		    (unsigned)(sizeof(List) + ((objc-1) * sizeof(Tcl_Obj *))));
	}
	return NULL;
    }

    listRepPtr->canonicalFlag = 0;
    listRepPtr->refCount = 0;
    listRepPtr->maxElemCount = objc;

	    Tcl_IncrRefCount(elemPtrs[i]);
	}
    } else {
	listRepPtr->elemCount = 0;
    }
    return listRepPtr;
}

/*
 *----------------------------------------------------------------------
 *
 * AttemptNewList --
 *
 *	Creates a list internal rep with space for objc elements.  objc
 *	must be > 0.  If objv!=NULL, initializes with the first objc values
 *	in that array.  If objv==NULL, initalize list internal rep to have
 *	0 elements, with space to add objc more.  
 *
 * Results:
 *	A new List struct with refCount 0 is returned. If some failure
 *	prevents this then NULL is returned, and an error message is left
 *	in the interp result, unless interp is NULL.
 *
 * Side effects:
 *	The ref counts of the elements in objv are incremented since the
 *	resulting list now refers to them.
 *
 *----------------------------------------------------------------------
 */

static List *
AttemptNewList(
    Tcl_Interp *interp,
    int objc,
    Tcl_Obj *CONST objv[])
{
    List *listRepPtr = NewListIntRep(objc, objv, 0);

    if (interp != NULL && listRepPtr == NULL) {
	if (objc > LIST_MAX) {
	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
		    "max length of a Tcl list (%d elements) exceeded",
		    LIST_MAX));
	} else {
	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
		    "list creation failed: unable to alloc %u bytes",
		    (unsigned)(sizeof(List) + ((objc-1) * sizeof(Tcl_Obj *)))));
	}
	Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
    }
    return listRepPtr;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_NewListObj --
 *
 *	This function is normally called when not debugging: i.e., when

	return listPtr;
    }

    /*
     * Create the internal rep.
     */

    listRepPtr = NewListIntRep(objc, objv, 1);




    /*
     * Now create the object.
     */

    Tcl_InvalidateStringRep(listPtr);



    ListSetIntRep(listPtr, listRepPtr);

    return listPtr;
}
#endif /* if TCL_MEM_DEBUG */

/*
 *----------------------------------------------------------------------
 *

	return listPtr;
    }

    /*
     * Create the internal rep.
     */

    listRepPtr = NewListIntRep(objc, objv, 1);




    /*
     * Now create the object.
     */

    Tcl_InvalidateStringRep(listPtr);



    ListSetIntRep(listPtr, listRepPtr);

    return listPtr;
}

#else /* if not TCL_MEM_DEBUG */

Tcl_Obj *

    }

    /*
     * Free any old string rep and any internal rep for the old type.
     */

    TclFreeIntRep(objPtr);

    Tcl_InvalidateStringRep(objPtr);

    /*
     * Set the object's type to "list" and initialize the internal rep.
     * However, if there are no elements to put in the list, just give the
     * object an empty string rep and a NULL type.
     */

    if (objc > 0) {
	listRepPtr = NewListIntRep(objc, objv, 1);
	ListSetIntRep(objPtr, listRepPtr);






    } else {
	objPtr->bytes = tclEmptyStringRep;
	objPtr->length = 0;
    }
}

/*

    if (Tcl_IsShared(listPtr)) {
	Tcl_Panic("%s called with shared object", "Tcl_ListObjAppendElement");
    }
    if (listPtr->typePtr != &tclListType) {
	int result, length;

	if (listPtr->typePtr == &tclDictType) {
	    (void) Tcl_DictObjSize(NULL, listPtr, &length);
	} else {
	    (void) TclGetStringFromObj(listPtr, &length);
	}
	if (!length) {
	    Tcl_SetListObj(listPtr, 1, &objPtr);
	    return TCL_OK;
	}

	result = SetListFromAny(interp, listPtr);
	if (result != TCL_OK) {

	newSize = 0;
    }

    if (listRepPtr->refCount > 1) {
	List *oldListRepPtr = listRepPtr;
	Tcl_Obj **oldElems;

	listRepPtr = AttemptNewList(interp, newMax, NULL);
	if (listRepPtr == NULL) {
	    return TCL_ERROR;
	}
	oldElems = &oldListRepPtr->elements;
	elemPtrs = &listRepPtr->elements;
	for (i=0; i<numElems; i++) {
	    elemPtrs[i] = oldElems[i];
	    Tcl_IncrRefCount(elemPtrs[i]);
	}

    Tcl_Obj **objPtrPtr)	/* The resulting Tcl_Obj* is stored here. */
{
    register List *listRepPtr;

    if (listPtr->typePtr != &tclListType) {
	int result, length;

	if (listPtr->typePtr == &tclDictType) {
	    (void) Tcl_DictObjSize(NULL, listPtr, &length);
	} else {
	    (void) TclGetStringFromObj(listPtr, &length);
	}
	if (!length) {
	    *objPtrPtr = NULL;
	    return TCL_OK;
	}

	result = SetListFromAny(interp, listPtr);
	if (result != TCL_OK) {

    register int *intPtr)	/* The resulting int is stored here. */
{
    register List *listRepPtr;

    if (listPtr->typePtr != &tclListType) {
	int result, length;

	if (listPtr->typePtr == &tclDictType) {
	    (void) Tcl_DictObjSize(NULL, listPtr, &length);
	    /*
	     * It's tempting to just report 2*length as the list length
	     * of this dict, but arguably that's false since the max sizes
	     * for dicts and lists are not the same, so some dicts don't
	     * actually convert to lists, and it's good to get that error
	     * back from the SetListFromAny() call below instead of a false
	     * indication we can treat the value as a list.  ([llength $val]
	     * often used as a "listiness" test)
	     */
	} else {
	    (void) TclGetStringFromObj(listPtr, &length);
	}
	if (!length) {
	    *intPtr = 0;
	    return TCL_OK;
	}

	result = SetListFromAny(interp, listPtr);
	if (result != TCL_OK) {

    if (Tcl_IsShared(listPtr)) {
	Tcl_Panic("%s called with shared object", "Tcl_ListObjReplace");
    }
    if (listPtr->typePtr != &tclListType) {
	int length;

	if (listPtr->typePtr == &tclDictType) {
	    (void) Tcl_DictObjSize(NULL, listPtr, &length);
	} else {
	    (void) TclGetStringFromObj(listPtr, &length);
	}
	if (!length) {
	    if (objc) {
		Tcl_SetListObj(listPtr, objc, NULL);
	    } else {
		return TCL_OK;
	    }
	} else {

	if (numRequired > listRepPtr->maxElemCount){
	    newMax = 2 * numRequired;
	} else {
	    newMax = listRepPtr->maxElemCount;
	}

	listRepPtr = AttemptNewList(interp, newMax, NULL);
	if (listRepPtr == NULL) {
	    return TCL_ERROR;
	}

	listPtr->internalRep.twoPtrValue.ptr1 = listRepPtr;
	listRepPtr->refCount++;

	elemPtrs = &listRepPtr->elements;

	    indexArray++;
	    break;
	}
	indexArray++;

	if (index < 0 || index > elemCount) {
	    /* ...the index points outside the sublist. */
	    if (interp != NULL) {
		Tcl_SetObjResult(interp,
			Tcl_NewStringObj("list index out of range", -1));
	    }
	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LSET", "BADINDEX",
		    NULL);
	    break;
	}

	/*
	 * No error conditions.  As long as we're not yet on the last index,

    if (Tcl_IsShared(listPtr)) {
	Tcl_Panic("%s called with shared object", "TclListObjSetElement");
    }
    if (listPtr->typePtr != &tclListType) {
	int length, result;

	if (listPtr->typePtr == &tclDictType) {
	    (void) Tcl_DictObjSize(NULL, listPtr, &length);
	} else {
	    (void) TclGetStringFromObj(listPtr, &length);
	}
	if (!length) {
	    if (interp != NULL) {
		Tcl_SetObjResult(interp,
			Tcl_NewStringObj("list index out of range", -1));
	    }
	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LSET", "BADINDEX",
		    NULL);
	    return TCL_ERROR;
	}
	result = SetListFromAny(interp, listPtr);
	if (result != TCL_OK) {
	    return result;

     */

    if (listRepPtr->refCount > 1) {
	List *oldListRepPtr = listRepPtr;
	Tcl_Obj **oldElemPtrs = elemPtrs;
	int i;

	listRepPtr = AttemptNewList(interp, listRepPtr->maxElemCount, NULL);
	if (listRepPtr == NULL) {
	    return TCL_ERROR;
	}
	listRepPtr->canonicalFlag = oldListRepPtr->canonicalFlag;
	elemPtrs = &listRepPtr->elements;
	for (i=0; i < elemCount; i++) {
	    elemPtrs[i] = oldElemPtrs[i];
	    Tcl_IncrRefCount(elemPtrs[i]);
	}

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

static void
FreeListInternalRep(
    Tcl_Obj *listPtr)		/* List object with internal rep to free. */
{
    List *listRepPtr = ListRepPtr(listPtr);

    if (--listRepPtr->refCount <= 0) {
	Tcl_Obj **elemPtrs = &listRepPtr->elements;

	int i, numElems = listRepPtr->elemCount;



	for (i = 0;  i < numElems;  i++) {

	    Tcl_DecrRefCount(elemPtrs[i]);
	}
	ckfree(listRepPtr);
    }

    listPtr->internalRep.twoPtrValue.ptr1 = NULL;
    listPtr->internalRep.twoPtrValue.ptr2 = NULL;
    listPtr->typePtr = NULL;

static void
DupListInternalRep(
    Tcl_Obj *srcPtr,		/* Object with internal rep to copy. */
    Tcl_Obj *copyPtr)		/* Object with internal rep to set. */
{
    List *listRepPtr = ListRepPtr(srcPtr);

    ListSetIntRep(copyPtr, listRepPtr);



}

/*
 *----------------------------------------------------------------------
 *
 * SetListFromAny --
 *

 */

static int
SetListFromAny(
    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
    Tcl_Obj *objPtr)		/* The object to convert. */
{








    List *listRepPtr;
    Tcl_Obj **elemPtrs;

    /*
     * Dictionaries are a special case; they have a string representation such
     * that *all* valid dictionaries are valid lists. Hence we can convert
     * more directly. Only do this when there's no existing string rep; if
     * there is, it is the string rep that's authoritative (because it could
     * describe duplicate keys).

	 * the reverse back to a dictionary) are both order-preserving. Also
	 * note that since we know we've got a valid dictionary (by
	 * representation) we also know that fetching the size of the
	 * dictionary or iterating over it will not fail.
	 */

	Tcl_DictObjSize(NULL, objPtr, &size);
	listRepPtr = AttemptNewList(interp, size > 0 ? 2*size : 1, NULL);
	if (!listRepPtr) {




	    return TCL_ERROR;
	}
	listRepPtr->elemCount = 2 * size;

	/*
	 * Populate the list representation.
	 */

	elemPtrs = &listRepPtr->elements;
	Tcl_DictObjFirst(NULL, objPtr, &search, &keyPtr, &valuePtr, &done);

	while (!done) {
	    *elemPtrs++ = keyPtr;
	    *elemPtrs++ = valuePtr;
	    Tcl_IncrRefCount(keyPtr);
	    Tcl_IncrRefCount(valuePtr);
	    Tcl_DictObjNext(&search, &keyPtr, &valuePtr, &done);
	}
    } else {



	int estCount, length;







	const char *limit, *nextElem = TclGetStringFromObj(objPtr, &length);

	/*



	 * Allocate enough space to hold a (Tcl_Obj *) for each
	 * (possible) list element.

	 */


	estCount = TclMaxListLength(nextElem, length, &limit);


	estCount += (estCount == 0); /* Smallest List struct holds 1 element. */
	listRepPtr = AttemptNewList(interp, estCount, NULL);
	if (listRepPtr == NULL) {
	    return TCL_ERROR;
	}
	elemPtrs = &listRepPtr->elements;

	/* Each iteration, parse and store a list element */




	while (nextElem < limit) {
	    const char *elemStart;

	    int elemSize, literal;







	    if (TCL_OK != TclFindElement(interp, nextElem, (limit - nextElem),
		    &elemStart, &nextElem, &elemSize, &literal)) {
		while (--elemPtrs >= &listRepPtr->elements) {
		    Tcl_DecrRefCount(*elemPtrs);
		}










		ckfree((char *) listRepPtr);



		return TCL_ERROR;
	    }
	    if (elemStart == limit) {
		break;
	    }



	    /* TODO: replace panic with error on alloc failure? */




	    if (literal) {


		TclNewStringObj(*elemPtrs, elemStart, elemSize);

	    } else {
		TclNewObj(*elemPtrs);
		(*elemPtrs)->bytes = ckalloc((unsigned) elemSize + 1);
		(*elemPtrs)->length = TclCopyAndCollapse(elemSize, elemStart,
			(*elemPtrs)->bytes);
	    }





	    Tcl_IncrRefCount(*elemPtrs++);/* Since list now holds ref to it. */
	}

 	listRepPtr->elemCount = elemPtrs - &listRepPtr->elements;
    }

    /*
     * Free the old internalRep before setting the new one. We do this as late
     * as possible to allow the conversion code, in particular
     * Tcl_GetStringFromObj, to use that old internalRep.
     */



    TclFreeIntRep(objPtr);
    ListSetIntRep(objPtr, listRepPtr);


    return TCL_OK;
}

/*
 *----------------------------------------------------------------------
 *
 * UpdateStringOfList --

/*
 * Array of values describing how to implement each standard subcommand of the
 * "namespace" command.
 */

static const EnsembleImplMap defaultNamespaceMap[] = {
    {"children",        NamespaceChildrenCmd, NULL, NULL, NULL, 0},
    {"code",            NamespaceCodeCmd, NULL, NULL, NULL, 0},
    {"current",         NamespaceCurrentCmd, NULL, NULL, NULL, 0},
    {"delete",          NamespaceDeleteCmd, NULL, NULL, NULL, 0},
    {"ensemble",        TclNamespaceEnsembleCmd, NULL, NULL, NULL, 0},
    {"eval",            NamespaceEvalCmd,       NULL, NRNamespaceEvalCmd, NULL, 0},
    {"exists",          NamespaceExistsCmd, NULL, NULL, NULL, 0},
    {"export",          NamespaceExportCmd, NULL, NULL, NULL, 0},
    {"forget",          NamespaceForgetCmd, NULL, NULL, NULL, 0},
    {"import",          NamespaceImportCmd, NULL, NULL, NULL, 0},
    {"inscope",         NamespaceInscopeCmd,    NULL, NULL, NRNamespaceInscopeCmd, 0},
    {"origin",          NamespaceOriginCmd, NULL, NULL, NULL, 0},
    {"parent",          NamespaceParentCmd, NULL, NULL, NULL, 0},
    {"path",            NamespacePathCmd, NULL, NULL, NULL, 0},
    {"qualifiers",      NamespaceQualifiersCmd, NULL, NULL, NULL, 0},
    {"tail",            NamespaceTailCmd, NULL, NULL, NULL, 0},
    {"unknown",         NamespaceUnknownCmd, NULL, NULL, NULL, 0},
    {"upvar",           NamespaceUpvarCmd,      TclCompileNamespaceUpvarCmd, NULL, NULL, 0},
    {"which",           NamespaceWhichCmd, NULL, NULL, NULL, 0},
    {NULL, NULL, NULL, NULL, NULL, 0}
};

/*
 *----------------------------------------------------------------------
 *
 * TclInitNamespaceSubsystem --

				 * name. Also used for error reporting if not
				 * NULL. */
    register Tcl_Obj *objPtr)	/* The object to convert. */
{
    const char *dummy;
    Namespace *nsPtr, *dummy1Ptr, *dummy2Ptr;
    register ResolvedNsName *resNamePtr;
    const char *name;

    if (interp == NULL) {
	return TCL_ERROR;
    }

    name = TclGetString(objPtr);
    TclGetNamespaceForQualName(interp, name, NULL, TCL_FIND_ONLY_NS,
	     &nsPtr, &dummy1Ptr, &dummy2Ptr, &dummy);

    /*
     * If we found a namespace, then create a new ResolvedNsName structure
     * that holds a reference to it.
     */

    if ((nsPtr == NULL) || (nsPtr->flags & NS_DYING)) {
	/*
	 * Our failed lookup proves any previously cached nsName intrep is no
	 * longer valid. Get rid of it so we no longer waste memory storing
	 * it, nor time determining its invalidity again and again.
	 */

	if (objPtr->typePtr == &nsNameType) {
	    TclFreeIntRep(objPtr);

	}
	return TCL_ERROR;
    }

    nsPtr->refCount++;
    resNamePtr = ckalloc(sizeof(ResolvedNsName));
    resNamePtr->nsPtr = nsPtr;

    register Tcl_Obj *objPtr)	/* The object to convert. */
{
    Interp *iPtr = (Interp *) interp;
    const char *name;
    register Command *cmdPtr;
    Namespace *currNsPtr;
    register ResolvedCmdName *resPtr;

    if (interp == NULL) {
	return TCL_ERROR;
    }

    /*
     * Find the Command structure, if any, that describes the command called
     * "name". Build a ResolvedCmdName that holds a cached pointer to this
     * Command, and bump the reference count in the referenced Command
     * structure. A Command structure will not be deleted as long as it is
     * referenced from a CmdName object.

		if (tokenPtr[i].type != TCL_TOKEN_TEXT) {
		    isLiteral = 0;
		    break;
		}
	    }

	    if (isLiteral) {
		int elemCount = 0, code = TCL_OK, literal = 1;
		const char *nextElem, *listEnd, *elemStart;

		/*
		 * The word to be expanded is a literal, so determine the
		 * boundaries of the literal string to be treated as a list
		 * and expanded. That literal string starts at
		 * tokenPtr[1].start, and includes all bytes up to, but not

		/*
		 * Step through the literal string, parsing and counting list
		 * elements.
		 */

		while (nextElem < listEnd) {
		    int size;

		    code = TclFindElement(NULL, nextElem, listEnd - nextElem,
			    &elemStart, &nextElem, &size, &literal);
		    if ((code != TCL_OK) || !literal) {
			break;










		    }
		    if (elemStart < listEnd) {
			elemCount++;
		    }
		}

		if ((code != TCL_OK) || !literal) {
		    /*
		     * Some list element could not be parsed, or is not
		     * present as a literal substring of the script.  The
		     * compiler cannot handle list elements that get generated
		     * by a call to TclCopyAndCollapse(). Defer  the
		     * handling of  this to  compile/eval time, where  code is
		     * already  in place to  report the  "attempt to  expand a
		     * non-list" error or expand lists that require
		     * substitution.
		     */

		    tokenPtr->type = TCL_TOKEN_EXPAND_WORD;
		} else if (elemCount == 0) {
		    /*

		    parsePtr->numTokens = wordIndex;
		} else {
		    /*
		     * Recalculate the number of Tcl_Tokens needed to store
		     * tokens representing the expanded list.
		     */

		    CONST char *listStart;
		    int growthNeeded = wordIndex + 2*elemCount
			    - parsePtr->numTokens;

		    parsePtr->numWords += elemCount - 1;
		    if (growthNeeded > 0) {
			TclGrowParseTokenArray(parsePtr, growthNeeded);
			tokenPtr = &parsePtr->tokenPtr[wordIndex];
		    }
		    parsePtr->numTokens = wordIndex + 2*elemCount;

		    /*
		     * Generate a TCL_TOKEN_SIMPLE_WORD token sequence for
		     * each element of the literal list we are expanding in
		     * place. Take care with the start and size fields of each
		     * token so they point to the right literal characters in
		     * the original script to represent the right expanded
		     * word value.
		     */

		    listStart = nextElem = tokenPtr[1].start;
		    while (nextElem < listEnd) {
			int quoted;
	
			tokenPtr->type = TCL_TOKEN_SIMPLE_WORD;
			tokenPtr->numComponents = 1;

			tokenPtr++;
			tokenPtr->type = TCL_TOKEN_TEXT;
			tokenPtr->numComponents = 0;
			TclFindElement(NULL, nextElem, listEnd - nextElem,
				&(tokenPtr->start), &nextElem,
				&(tokenPtr->size), NULL);

			quoted = (tokenPtr->start[-1] == '{'
				|| tokenPtr->start[-1] == '"')
				&& tokenPtr->start > listStart;
			tokenPtr[-1].start = tokenPtr->start - quoted;
			tokenPtr[-1].size = tokenPtr->start + tokenPtr->size
				- tokenPtr[-1].start + quoted;

			tokenPtr++;
		    }
		}

    return TCL_OK;

  error:
    Tcl_FreeParse(parsePtr);
    parsePtr->commandSize = parsePtr->end - parsePtr->commandStart;
    return TCL_ERROR;
}

/*
 *----------------------------------------------------------------------
 *
 * TclIsSpaceProc --
 *
 *	Report whether byte is in the set of whitespace characters used by
 *	Tcl to separate words in scripts or elements in lists.
 *
 * Results:
 *	Returns 1, if byte is in the set, 0 otherwise.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
TclIsSpaceProc(
    char byte)
{
    return CHAR_TYPE(byte) & (TYPE_SPACE) || byte == '\n';
}

/*
 *----------------------------------------------------------------------
 *
 * ParseWhiteSpace --
 *
 *	Scans up to numBytes bytes starting at src, consuming white space

	    case '{':
		openBrace = 1;
		break;
	    case '\n':
		openBrace = 0;
		break;
	    case '#' :
		if (openBrace && TclIsSpaceProc(src[-1])) {
		    Tcl_AppendResult(parsePtr->interp,
			    ": possible unbalanced brace in comment", NULL);
		    goto error;
		}
		break;
	    }
	}

	    return TCL_OK;
	}

	if (pathPtr->bytes == NULL) {
	    UpdateStringOfFsPath(pathPtr);
	}
	FreeFsPathInternalRep(pathPtr);

    }

    return Tcl_ConvertToType(interp, pathPtr, &tclFsPathType);

    /*
     * We used to have more complex code here:
     *
     * FsPath *fsPathPtr = PATHOBJ(pathPtr);
     * if (fsPathPtr->cwdPtr == NULL || PATHFLAGS(pathPtr) != 0) {
     *     return TCL_OK;
     * } else {
     *     if (TclFSCwdPointerEquals(&fsPathPtr->cwdPtr)) {
     *         return TCL_OK;
     *     } else {
     *         if (pathPtr->bytes == NULL) {
     *             UpdateStringOfFsPath(pathPtr);
     *         }
     *         FreeFsPathInternalRep(pathPtr);

     *         return Tcl_ConvertToType(interp, pathPtr, &tclFsPathType);
     *     }
     * }
     *
     * But we no longer believe this is necessary.
     */
}

    if (fsPathPtr->cwdPtr != NULL) {
	if (!TclFSCwdPointerEquals(&fsPathPtr->cwdPtr)) {
	    if (pathPtr->bytes == NULL) {
		UpdateStringOfFsPath(pathPtr);
	    }
	    FreeFsPathInternalRep(pathPtr);

	    if (Tcl_ConvertToType(interp, pathPtr, &tclFsPathType) != TCL_OK) {
		return NULL;
	    }
	    fsPathPtr = PATHOBJ(pathPtr);
	} else if (fsPathPtr->normPathPtr == NULL) {
	    int cwdLen;
	    Tcl_Obj *copy;

	 * We have to discard the stale representation and recalculate it.
	 */

	if (pathPtr->bytes == NULL) {
	    UpdateStringOfFsPath(pathPtr);
	}
	FreeFsPathInternalRep(pathPtr);

	if (SetFsPathFromAny(NULL, pathPtr) != TCL_OK) {
	    return TCL_ERROR;
	}
	srcFsPathPtr = PATHOBJ(pathPtr);
    }

    /*

		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "PROC",
			"CROSSINTERPBYTECODE", NULL);
		return TCL_ERROR;
	    }
	    codePtr->compileEpoch = iPtr->compileEpoch;
	    codePtr->nsPtr = nsPtr;
	} else {

	    TclFreeIntRep(bodyPtr);
	}
    }

    if (bodyPtr->typePtr != &tclByteCodeType) {
	Tcl_HashEntry *hePtr;

#ifdef TCL_COMPILE_DEBUG

    register Tcl_Obj *objPtr)	/* The object to convert. */
{
    Interp *iPtr = (Interp *) interp;
    const char *name;
    Tcl_Obj *argsPtr, *bodyPtr, *nsObjPtr, **objv, *errPtr;
    int objc, result;
    Proc *procPtr;

    if (interp == NULL) {
	return TCL_ERROR;
    }

    /*
     * Convert objPtr to list type first; if it cannot be converted, or if its
     * length is not 2, then it cannot be converted to lambdaType.
     */

    result = TclListObjGetElements(NULL, objPtr, &objc, &objv);
    if ((result != TCL_OK) || ((objc != 2) && (objc != 3))) {
	TclNewLiteralStringObj(errPtr, "can't interpret \"");
	Tcl_AppendObjToObj(errPtr, objPtr);
	Tcl_AppendToObj(errPtr, "\" as a lambda expression", -1);
	Tcl_SetObjResult(interp, errPtr);
	Tcl_SetErrorCode(interp, "TCL", "VALUE", "LAMBDA", NULL);
	return TCL_ERROR;

    /*
     * Free the list internalrep of objPtr - this will free argsPtr, but
     * bodyPtr retains a reference from the Proc structure. Then finish the
     * conversion to lambdaType.
     */

    TclFreeIntRep(objPtr);

    objPtr->internalRep.twoPtrValue.ptr1 = procPtr;
    objPtr->internalRep.twoPtrValue.ptr2 = nsObjPtr;
    objPtr->typePtr = &lambdaType;
    return TCL_OK;
}


	    if (objResultPtr->bytes) {
		ckfree(objResultPtr->bytes);
	    }
	    objResultPtr->bytes = tclEmptyStringRep;
	    objResultPtr->length = 0;
	}
	TclFreeIntRep(objResultPtr);

    }
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_SetErrorCodeVA --

	case INITIAL:
	    /*
	     * Initial state. Acceptable characters are +, -, digits, period,
	     * I, N, and whitespace.
	     */

	    if (TclIsSpaceProc(c)) {
		if (flags & TCL_PARSE_NO_WHITESPACE) {
		    goto endgame;
		}
		break;
	    } else if (c == '+') {
		state = SIGNUM;
		break;

	case sNANHEX:
	    if (c == ')') {
		state = sNANFINISH;
		break;
	    }
	    /* FALLTHROUGH */
	case sNANPAREN:
	    if (TclIsSpaceProc(c)) {
		break;
	    }
	    if (numSigDigs < 13) {
		if (c >= '0' && c <= '9') {
		    d = c - '0';
		} else if (c >= 'a' && c <= 'f') {
		    d = 10 + c - 'a';

	p = acceptPoint;
	len = acceptLen;
	if (!(flags & TCL_PARSE_NO_WHITESPACE)) {
	    /*
	     * Accept trailing whitespace.
	     */

	    while (len != 0 && TclIsSpaceProc(*p)) {
		p++;
		len--;
	    }
	}
	if (endPtrPtr == NULL) {
	    if ((len != 0) && ((numBytes > 0) || (*p != '\0'))) {
		status = TCL_ERROR;

#define STRING_SIZE(numChars) \
	(sizeof(String) + ((numChars) * sizeof(Tcl_UniChar)))
#define stringCheckLimits(numChars) \
    if ((numChars) < 0 || (numChars) > STRING_MAXCHARS) { \
	Tcl_Panic("max length for a Tcl unicode value (%d chars) exceeded", \
		STRING_MAXCHARS); \
    }
#define stringAttemptAlloc(numChars) \
	(String *) attemptckalloc((unsigned) STRING_SIZE(numChars) )
#define stringAlloc(numChars) \
	(String *) ckalloc((unsigned) STRING_SIZE(numChars) )
#define stringRealloc(ptr, numChars) \
    (String *) ckrealloc((ptr), (unsigned) STRING_SIZE(numChars) )
#define stringAttemptRealloc(ptr, numChars) \
    (String *) attemptckrealloc((ptr), (unsigned) STRING_SIZE(numChars) )
#define GET_STRING(objPtr) \

    }

    /*
     * Set the type to NULL and free any internal rep for the old type.
     */

    TclFreeIntRep(objPtr);


    /*
     * Free any old string rep, then set the string rep to a copy of the
     * length bytes starting at "bytes".
     */

    TclInvalidateStringRep(objPtr);

	int copyMaxChars;

	if (srcStringPtr->maxChars / 2 >= srcStringPtr->numChars) {
	    copyMaxChars = 2 * srcStringPtr->numChars;
	} else {
	    copyMaxChars = srcStringPtr->maxChars;
	}
	copyStringPtr = stringAttemptAlloc(copyMaxChars);
	if (copyStringPtr == NULL) {
	    copyMaxChars = srcStringPtr->numChars;
	    copyStringPtr = stringAlloc(copyMaxChars);
	}
	copyStringPtr->maxChars = copyMaxChars;
	memcpy(copyStringPtr->unicode, srcStringPtr->unicode,
		srcStringPtr->numChars * sizeof(Tcl_UniChar));
	copyStringPtr->unicode[srcStringPtr->numChars] = 0;
    } else {
	copyStringPtr = stringAlloc(0);
	copyStringPtr->maxChars = 0;

     * object, clear out the object's cached state.
     */

    if (objv[3]->typePtr != NULL
	    && !strcmp("index", objv[3]->typePtr->name)) {
	indexRep = objv[3]->internalRep.otherValuePtr;
	if (indexRep->tablePtr == (void *) argv) {

	    TclFreeIntRep(objv[3]);
	}
    }

    result = Tcl_GetIndexFromObj((setError? interp : NULL), objv[3],
	    argv, "token", (allowAbbrev? 0 : TCL_EXACT), &index);
    ckfree(argv);
    if (result == TCL_OK) {

    /*
     * 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 {
	category = (GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK);
	return ((SPACE_BITS >> category) & 1);
    }
}

/*

    UpdateStringOfEndOffset,		/* updateStringProc */
    SetEndOffsetFromAny
};

/*
 *----------------------------------------------------------------------
 *
 * TclMaxListLength --
 *
 *	Given 'bytes' pointing to 'numBytes' bytes, scan through them and
 *	count the number of whitespace runs that could be list element
 *	separators.  If 'numBytes' is -1, scan to the terminating '\0'.
 *	Not a full list parser.  Typically used to get a quick and dirty
 *	overestimate of length size in order to allocate space for an
 *	actual list parser to operate with.
 *
 * Results:
 *	Returns the largest number of list elements that could possibly
 *	be in this string, interpreted as a Tcl list.  If 'endPtr' is not
 *	NULL, writes a pointer to the end of the string scanned there.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
TclMaxListLength(
    CONST char *bytes,
    int numBytes,
    CONST char **endPtr)
{
    int count = 0;

    if ((numBytes == 0) || ((numBytes == -1) && (*bytes == '\0'))) {
	/* Empty string case - quick exit */
	goto done;
    }

    /* No list element before leading white space */
    count += 1 - TclIsSpaceProc(*bytes); 

    /* Count white space runs as potential element separators */
    while (numBytes) {
	if ((numBytes == -1) && (*bytes == '\0')) {
	    break;
	}
	if (TclIsSpaceProc(*bytes)) {
	    /* Space run started; bump count */
	    count++;
	    do {
		bytes++;
		numBytes -= (numBytes != -1);
	    } while (numBytes && TclIsSpaceProc(*bytes));
	    if (numBytes == 0) {
		break;
	    }
	    /* (*bytes) is non-space; return to counting state */
	}
	bytes++;
	numBytes -= (numBytes != -1);
    }

    /* No list element following trailing white space */
    count -= TclIsSpaceProc(bytes[-1]); 

    done:
    if (endPtr) {
	*endPtr = bytes;
    }
    return count;
}

/*
 *----------------------------------------------------------------------
 *
 * TclFindElement --
 *
 *	Given a pointer into a Tcl list, locate the first (or next) element in
 *	the list.
 *
 * Results:
 *	The return value is normally TCL_OK, which means that the element was

 *	that's part of the element. If this is the last argument in the list,
 *	then *nextPtr will point just after the last character in the list
 *	(i.e., at the character at list+listLength). If sizePtr is non-NULL,
 *	*sizePtr is filled in with the number of characters in the element. If
 *	the element is in braces, then *elementPtr will point to the character
 *	after the opening brace and *sizePtr will not include either of the
 *	braces. If there isn't an element in the list, *sizePtr will be zero,
 *	and both *elementPtr and *nextPtr will point just after the last
 *	character in the list. If literalPtr is non-NULL, *literalPtr is set
 *	to a boolean value indicating whether the substring returned as
 *	the values of **elementPtr and *sizePtr is the literal value of
 *	a list element.  If not, a call to TclCopyAndCollapse() is needed
 *	to produce the actual value of the list element.  Note: this function
 *	does NOT collapse backslash sequences, but uses *literalPtr to tell
 * 	callers when it is required for them to do so.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

    const char **elementPtr,	/* Where to put address of first significant
				 * character in first element of list. */
    const char **nextPtr,	/* Fill in with location of character just
				 * after all white space following end of
				 * argument (next arg or end of list). */
    int *sizePtr,		/* If non-zero, fill in with size of
				 * element. */
    int *literalPtr)		/* If non-zero, fill in with non-zero/zero to
				 * indicate that the substring of *sizePtr
				 * bytes starting at **elementPtr is/is not
				 * the literal list element and therefore
				 * does not/does require a call to 
				 * TclCopyAndCollapse() by the caller. */
{
    const char *p = list;
    const char *elemStart;	/* Points to first byte of first element. */
    const char *limit;		/* Points just after list's last byte. */
    int openBraces = 0;		/* Brace nesting level during parse. */
    int inQuotes = 0;
    int size = 0;		/* lint. */
    int numChars;
    int literal = 1;
    const char *p2;

    /*
     * Skim off leading white space and check for an opening brace or quote.
     * We treat embedded NULLs in the list as bytes belonging to a list
     * element.
     */

    limit = (list + listLength);
    while ((p < limit) && (TclIsSpaceProc(*p))) {
	p++;
    }
    if (p == limit) {		/* no element found */
	elemStart = limit;
	goto done;
    }

    if (*p == '{') {
	openBraces = 1;
	p++;
    } else if (*p == '"') {
	inQuotes = 1;
	p++;
    }
    elemStart = p;




    /*
     * Find element's end (a space, close brace, or the end of the string).
     */

    while (p < limit) {
	switch (*p) {

	case '}':
	    if (openBraces > 1) {
		openBraces--;
	    } else if (openBraces == 1) {
		size = (p - elemStart);
		p++;
		if ((p >= limit) || TclIsSpaceProc(*p)) {

		    goto done;
		}

		/*
		 * Garbage after the closing brace; return an error.
		 */

		if (interp != NULL) {
		    p2 = p;
		    while ((p2 < limit) && (!TclIsSpaceProc(*p2))

			    && (p2 < p+20)) {
			p2++;
		    }
		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
			    "list element in braces followed by \"%.*s\" "
			    "instead of space", (int) (p2-p), p));
		    Tcl_SetErrorCode(interp, "TCL", "VALUE", "LIST", "JUNK",
			    NULL);
		}
		return TCL_ERROR;
	    }
	    break;

	    /*
	     * Backslash: skip over everything up to the end of the backslash
	     * sequence.
	     */

	case '\\':
	    if (openBraces == 0) {
		/*
		 * A backslash sequence not within a brace quoted element
		 * means the value of the element is different from the
		 * substring we are parsing.  A call to TclCopyAndCollapse()
		 * is needed to produce the element value.  Inform the caller.
		 */
		literal = 0;
	    }
	    TclParseBackslash(p, limit - p, &numChars, NULL);
	    p += (numChars - 1);
	    break;

	    /*
	     * Space: ignore if element is in braces or quotes; otherwise
	     * terminate element.

	     * Double-quote: if element is in quotes then terminate it.
	     */

	case '"':
	    if (inQuotes) {
		size = (p - elemStart);
		p++;
		if ((p >= limit) || TclIsSpaceProc(*p)) {

		    goto done;
		}

		/*
		 * Garbage after the closing quote; return an error.
		 */

		if (interp != NULL) {
		    p2 = p;
		    while ((p2 < limit) && (!TclIsSpaceProc(*p2))

			    && (p2 < p+20)) {
			p2++;
		    }
		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
			    "list element in quotes followed by \"%.*s\" "
			    "instead of space", (int) (p2-p), p));
		    Tcl_SetErrorCode(interp, "TCL", "VALUE", "LIST", "JUNK",

	    }
	    return TCL_ERROR;
	}
	size = (p - elemStart);
    }

  done:
    while ((p < limit) && (TclIsSpaceProc(*p))) {
	p++;
    }
    *elementPtr = elemStart;
    *nextPtr = p;
    if (sizePtr != 0) {
	*sizePtr = size;
    }
    if (literalPtr != 0) {
	*literalPtr = literal;
    }
    return TCL_OK;
}

/*
 *----------------------------------------------------------------------
 *

				 * NULL, no error message is left. */
    const char *list,		/* Pointer to string with list structure. */
    int *argcPtr,		/* Pointer to location to fill in with the
				 * number of elements in the list. */
    const char ***argvPtr)	/* Pointer to place to store pointer to array
				 * of pointers to list elements. */
{
    const char **argv, *end, *element;
    char *p;
    int length, size, i, result, elSize;

    /*
     * Allocate enough space to work in. A (CONST char *) for each
     * (possible) list element plus one more for terminating NULL,



     * plus as many bytes as in the original string value, plus one



     * more for a terminating '\0'.  Space used to hold element separating

     * white space in the original string gets re-purposed to hold '\0'
     * characters in the argv array.
     */



    size = TclMaxListLength(list, -1, &end) + 1;











    length = end - list;
    argv = ckalloc((size * sizeof(char *)) + length + 1);

    for (i = 0, p = ((char *) argv) + size*sizeof(char *);
	    *list != 0;  i++) {
	const char *prevList = list;
	int literal;

	result = TclFindElement(interp, list, length, &element, &list,
		&elSize, &literal);
	length -= (list - prevList);
	if (result != TCL_OK) {
	    ckfree(argv);
	    return result;
	}
	if (*element == 0) {
	    break;
	}
	if (i >= size) {
	    ckfree(argv);
	    if (interp != NULL) {
		Tcl_SetResult(interp, "internal error in Tcl_SplitList",
			TCL_STATIC);
		Tcl_SetErrorCode(interp, "TCL", "INTERNAL", "Tcl_SplitList",
			NULL);
	    }
	    return TCL_ERROR;
	}
	argv[i] = p;
	if (literal) {
	    memcpy(p, element, (size_t) elSize);
	    p += elSize;
	    *p = 0;
	    p++;
	} else {
	    p += 1 + TclCopyAndCollapse(elSize, element, p);

	}
    }

    argv[i] = NULL;
    *argvPtr = argv;
    *argcPtr = i;
    return TCL_OK;

    bytes = TclGetStringFromObj(objPtr, &length);

    /*
     * Leading whitespace is acceptable in an index.
     */

    while (length && TclIsSpaceProc(*bytes)) {
	bytes++;
	length--;
    }

    if (TclParseNumber(NULL, NULL, NULL, bytes, length, (const char **)&opPtr,
	    TCL_PARSE_INTEGER_ONLY | TCL_PARSE_NO_WHITESPACE) == TCL_OK) {
	int code, first, second;
	char savedOp = *opPtr;

	if ((savedOp != '+') && (savedOp != '-')) {
	    goto parseError;
	}
	if (TclIsSpaceProc(opPtr[1])) {
	    goto parseError;
	}
	*opPtr = '\0';
	code = Tcl_GetInt(interp, bytes, &first);
	*opPtr = savedOp;
	if (code == TCL_ERROR) {
	    goto parseError;

	offset = 0;
    } else if ((length > 4) && ((bytes[3] == '-') || (bytes[3] == '+'))) {
	/*
	 * This is our limited string expression evaluator. Pass everything
	 * after "end-" to Tcl_GetInt, then reverse for offset.
	 */

	if (TclIsSpaceProc(bytes[4])) {
	    goto badIndexFormat;
	}
	if (Tcl_GetInt(interp, bytes+4, &offset) != TCL_OK) {
	    return TCL_ERROR;
	}
	if (bytes[3] == '-') {
	    offset = -offset;

    register const char *p = value;

    /*
     * A frequent mistake is invalid octal values due to an unwanted leading
     * zero. Try to generate a meaningful error message.
     */

    while (TclIsSpaceProc(*p)) {
	p++;
    }
    if (*p == '+' || *p == '-') {
	p++;
    }
    if (*p == '0') {
	if ((p[1] == 'o') || p[1] == 'O') {
	    p += 2;
	}
	while (isdigit(UCHAR(*p))) {	/* INTL: digit. */
	    p++;
	}
	while (TclIsSpaceProc(*p)) {
	    p++;
	}
	if (*p == '\0') {
	    /*
	     * Reached end of string.
	     */

  doneParsing:
    /*
     * part1Ptr is not an array element; look it up, and convert it to one of
     * the cached types if possible.
     */

    TclFreeIntRep(part1Ptr);


    varPtr = TclLookupSimpleVar(interp, part1Ptr, flags, createPart1,
	    &errMsg, &index);
    if (varPtr == NULL) {
	if ((errMsg != NULL) && (flags & TCL_LEAVE_ERR_MSG)) {
	    TclObjVarErrMsg(interp, part1Ptr, part2Ptr, msg, errMsg, -1);
	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "VARNAME",

     * Try to avoid keeping the Var struct allocated due to a tclNsVarNameType
     * keeping a reference. This removes some additional exteriorisations of
     * [Bug 736729], but may be a good thing independently of the bug.
     */

    if (part1Ptr->typePtr == &tclNsVarNameType) {
	TclFreeIntRep(part1Ptr);

    }
#endif

    /*
     * Finally, if the variable is truly not in use then free up its Var
     * structure and remove it from its hash table, if any. The ref count of
     * its value object, if any, was decremented above.

    Tcl_DString ds;
    Tcl_Encoding encoding = Tcl_GetEncoding(NULL, "macRoman");

    string = Tcl_GetStringFromObj(objPtr, &length);
    Tcl_UtfToExternalDString(encoding, string, length, &ds);

    if (Tcl_DStringLength(&ds) > 4) {
	if (interp) {
	    Tcl_AppendResult(interp, "expected Macintosh OS type but got \"",
		    string, "\": ", NULL);
	    Tcl_SetErrorCode(interp, "TCL", "VALUE", "MAC_OSTYPE", NULL);
	}
	result = TCL_ERROR;
    } else {
	OSType osType;
	char bytes[4] = {'\0','\0','\0','\0'};

	memcpy(bytes, Tcl_DStringValue(&ds), (size_t)Tcl_DStringLength(&ds));
	osType = (OSType) bytes[0] << 24 |

    list [catch join msg] $msg $errorCode
} {1 {wrong # args: should be "join list ?joinString?"} {TCL WRONGARGS}}
test join-2.2 {join errors} {
    list [catch {join a b c} msg] $msg $errorCode
} {1 {wrong # args: should be "join list ?joinString?"} {TCL WRONGARGS}}
test join-2.3 {join errors} {
    list [catch {join "a \{ c" 111} msg] $msg $errorCode
} {1 {unmatched open brace in list} {TCL VALUE LIST BRACE}}

test join-3.1 {joinString is binary ok} {
  string length [join {a b c} a\0b]
} 9
test join-3.2 {join is binary ok} {
  string length [join "a\0b a\0b a\0b"]
} 11

    -body {
	lrepeat 0 a b c
    }
    -result {}
}
test lrepeat-1.8 {Do not build enormous lists - Bug 2130992} -body {
     lrepeat 0x10000000 a b c d e f g h
} -returnCodes error -match glob -result *

## Okay
test lrepeat-2.1 {normal cases} {
    lrepeat 10 a
} {a a a a a a a a a a}
test lrepeat-2.2 {normal cases} {
    lrepeat 3 [lrepeat 3 0]

        lappend res [TestOp $op 5.0 1]
        lappend exp "can't use floating-point value as operand of \"$op\" ARITH DOMAIN {floating-point value}"
        lappend res [TestOp $op 1 5.0]
        lappend exp "can't use floating-point value as operand of \"$op\" ARITH DOMAIN {floating-point value}"
    }
    foreach op {in ni} {
        lappend res [TestOp $op 5 "a b \{ c"]
        lappend exp "unmatched open brace in list TCL VALUE LIST BRACE"
    }
    lappend res [TestOp % 5 0]
    lappend exp "divide by zero ARITH DIVZERO {divide by zero}"
    lappend res [TestOp % 9838923468297346238478737647637375 0]
    lappend exp "divide by zero ARITH DIVZERO {divide by zero}"
    lappend res [TestOp / 5 0]
    lappend exp "divide by zero ARITH DIVZERO {divide by zero}"

} {- \{*\}\\\n\ foo\ bar 3 simple {{*}} 1 text * 0 simple foo 1 text foo 0 simple bar 1 text bar 0 {}}
test parse-5.28 {Tcl_ParseCommand: {*} parsing, expanded literals} testparser {
    testparser {{*}{a b}} 0
} {- {{*}{a b}} 2 simple a 1 text a 0 simple b 1 text b 0 {}}
test parse-5.29 {Tcl_ParseCommand: {*} parsing, expanded literals, naked backslashes} testparser {
    testparser {{*}{a \n b}} 0
} {- {{*}{a \n b}} 1 expand {{*}{a \n b}} 1 text {a \n b} 0 {}}
test parse-5.30 {Tcl_ParseCommand: {*} parsing, expanded literals} testparser {
    testparser {{*}"a b"} 0
} {- {{*}"a b"} 2 simple a 1 text a 0 simple b 1 text b 0 {}}
test parse-5.31 {Tcl_ParseCommand: {*} parsing, expanded literals, naked backslashes} testparser {
    testparser {{*}"a \n b"} 0
} {- {{*}"a \n b"} 1 expand {{*}"a \n b"} 3 text {a } 0 backslash {\n} 0 text { b} 0 {}}

test parse-6.1 {ParseTokens procedure, empty word} testparser {
    testparser {""} 0
} {- {""} 1 simple {""} 1 text {} 0 {}}
test parse-6.2 {ParseTokens procedure, simple range} testparser {
    testparser {"abc$x.e"} 0
} {- {"abc$x.e"} 1 word {"abc$x.e"} 4 text abc 0 variable {$x} 1 text x 0 text .e 0 {}}

    set result ""
    set numbers [list 1.0 +1.0 ++1.0 +-1.0 -+1.0 -1.0 --1.0 "- +1.0"]
    foreach num $numbers {
	lappend result [string is double -strict $num]
    }
    set result
} {1 1 0 0 0 1 0 0}
test string-6.92 {string is integer, 32-bit overflow} {
    # Bug 718878
    set x 0x100000000
    list [string is integer -failindex var $x] $var
} {0 -1}
test string-6.93 {string is integer, 32-bit overflow} {
    # Bug 718878
    set x 0x100000000
    append x ""
    list [string is integer -failindex var $x] $var
} {0 -1}
test string-6.94 {string is integer, 32-bit overflow} {
    # Bug 718878
    set x 0x100000000
    list [string is integer -failindex var [expr {$x}]] $var
} {0 -1}
test string-6.95 {string is wideinteger, true} {
    string is wideinteger +1234567890
} 1

#!/bin/sh
# install - install a program, script, or datafile

scriptversion=2011-04-20.01; # UTC

# This originates from X11R5 (mit/util/scripts/install.sh), which was
# later released in X11R6 (xc/config/util/install.sh) with the
# following copyright and license.
#
# Copyright (C) 1994 X Consortium
#

  -c            (ignored)
  -C            install only if different (preserve the last data modification time)
  -d            create directories instead of installing files.
  -g GROUP      $chgrpprog installed files to GROUP.
  -m MODE       $chmodprog installed files to MODE.
  -o USER       $chownprog installed files to USER.
  -s            $stripprog installed files.
  -S            $stripprog installed files.
  -t DIRECTORY  install into DIRECTORY.
  -T            report an error if DSTFILE is a directory.

Environment variables override the default commands:
  CHGRPPROG CHMODPROG CHOWNPROG CMPPROG CPPROG MKDIRPROG MVPROG
  RMPROG STRIPPROG
"

	esac
	shift;;

    -o) chowncmd="$chownprog $2"
	shift;;

    -s) stripcmd=$stripprog;;

      -S) stripcmd="$stripprog $2" 
         shift;; 

    -t) dst_arg=$2
	shift;;

    -T) no_target_directory=true;;

    --version) echo "$0 $scriptversion"; exit $?;;

 * a #define of TCL_LOCAL_APPINIT instead of rewriting this entire file. The
 * #if checks for that #define and uses Tcl_AppInit if it doesn't exist.
 */

#ifndef TCL_LOCAL_APPINIT
#define TCL_LOCAL_APPINIT Tcl_AppInit
#endif
#ifndef MODULE_SCOPE
#   define MODULE_SCOPE extern
#endif
MODULE_SCOPE int TCL_LOCAL_APPINIT(Tcl_Interp *);
MODULE_SCOPE int main(int, char **);

/*
 * The following #if block allows you to change how Tcl finds the startup
 * script, prime the library or encoding paths, fiddle with the argv, etc.,
 * without needing to rewrite Tcl_Main()

    /*
     * Search through all the directories named in the PATH variable to see if
     * argv[0] is in one of them. If so, use that file name.
     */

    while (1) {
	while (TclIsSpaceProc(*p)) {
	    p++;
	}
	name = p;
	while ((*p != ':') && (*p != 0)) {
	    p++;
	}
	Tcl_DStringSetLength(&buffer, 0);

    }
}
#endif /* TCL_THREADS */

/*
 *----------------------------------------------------------------------
 *
 * TclpReaddir, TclpInetNtoa --
 *
 *	These procedures replace core C versions to be used in a threaded
 *	environment.
 *
 * Results:
 *	See documentation of C functions.
 *

echo "${ECHO_T}$tcl_cv_cast_to_union" >&6
if test "$tcl_cv_cast_to_union" = "yes"; then

cat >>confdefs.h <<\_ACEOF
#define HAVE_CAST_TO_UNION 1
_ACEOF

fi

# Check to see if struct _stat32i64 exists in mingw's sys/stat.h

echo "$as_me:$LINENO: checking struct _stat32i64" >&5
echo $ECHO_N "checking struct _stat32i64... $ECHO_C" >&6
if test "${tcl_struct_stat32i64+set}" = set; then
  echo $ECHO_N "(cached) $ECHO_C" >&6
else
  cat >conftest.$ac_ext <<_ACEOF
/* confdefs.h.  */
_ACEOF
cat confdefs.h >>conftest.$ac_ext
cat >>conftest.$ac_ext <<_ACEOF
/* end confdefs.h.  */

#include <sys/types.h>
#include <sys/stat.h>

int
main ()
{

  struct _stat32i64 foo;

  ;
  return 0;
}
_ACEOF
rm -f conftest.$ac_objext
if { (eval echo "$as_me:$LINENO: \"$ac_compile\"") >&5
  (eval $ac_compile) 2>conftest.er1
  ac_status=$?
  grep -v '^ *+' conftest.er1 >conftest.err
  rm -f conftest.er1
  cat conftest.err >&5
  echo "$as_me:$LINENO: \$? = $ac_status" >&5
  (exit $ac_status); } &&
	 { ac_try='test -z "$ac_c_werror_flag"
			 || test ! -s conftest.err'
  { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
  (eval $ac_try) 2>&5
  ac_status=$?
  echo "$as_me:$LINENO: \$? = $ac_status" >&5
  (exit $ac_status); }; } &&
	 { ac_try='test -s conftest.$ac_objext'
  { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
  (eval $ac_try) 2>&5
  ac_status=$?
  echo "$as_me:$LINENO: \$? = $ac_status" >&5
  (exit $ac_status); }; }; then
  tcl_struct_stat32i64=yes
else
  echo "$as_me: failed program was:" >&5
sed 's/^/| /' conftest.$ac_ext >&5

tcl_struct_stat32i64=no
fi
rm -f conftest.err conftest.$ac_objext conftest.$ac_ext

fi
echo "$as_me:$LINENO: result: $tcl_struct_stat32i64" >&5
echo "${ECHO_T}$tcl_struct_stat32i64" >&6
if test "$tcl_struct_stat32i64" = "yes" ; then

cat >>confdefs.h <<\_ACEOF
#define HAVE_STRUCT_STAT32I64 1
_ACEOF

fi


# See if declarations like FINDEX_INFO_LEVELS are
# missing from winbase.h. This is known to be
# a problem with VC++ 5.2.

        tcl_cv_cast_to_union=yes,
        tcl_cv_cast_to_union=no)
)
if test "$tcl_cv_cast_to_union" = "yes"; then
    AC_DEFINE(HAVE_CAST_TO_UNION, 1,
            [Defined when compiler supports casting to union type.])
fi

# Check to see if struct _stat32i64 exists in mingw's sys/stat.h

AC_CACHE_CHECK(struct _stat32i64,
    tcl_struct_stat32i64,
AC_TRY_COMPILE([
#include <sys/types.h>
#include <sys/stat.h>
],
[
  struct _stat32i64 foo;
],
        tcl_struct_stat32i64=yes,
        tcl_struct_stat32i64=no)
)
if test "$tcl_struct_stat32i64" = "yes" ; then
    AC_DEFINE(HAVE_STRUCT_STAT32I64, 1,
            [Defined when sys/stat.h has struct_stat32i64])
fi


# See if declarations like FINDEX_INFO_LEVELS are
# missing from winbase.h. This is known to be
# a problem with VC++ 5.2.

AC_CACHE_CHECK(for FINDEX_INFO_LEVELS in winbase.h,

TCLREGLIB	= $(OUT_DIR)\$(TCLREGLIBNAME)

TCLDDELIBNAME	= $(PROJECT)dde$(DDEVERSION)$(SUFX:t=).$(EXT)
TCLDDELIB	= $(OUT_DIR)\$(TCLDDELIBNAME)

TCLTEST		= $(OUT_DIR)\$(PROJECT)test.exe
CAT32		= $(OUT_DIR)\cat32.exe

# Can we run what we build? IX86 runs on all architectures.
!ifndef TCLSH_NATIVE
!if "$(MACHINE)" == "IX86" || "$(MACHINE)" == "$(NATIVE_ARCH)"
TCLSH_NATIVE	= $(TCLSH)
!else
!error You must explicitly set TCLSH_NATIVE for cross-compilation
!endif
!endif

### Make sure we use backslash only.
LIB_INSTALL_DIR		= $(_INSTALLDIR)\lib
BIN_INSTALL_DIR		= $(_INSTALLDIR)\bin
DOC_INSTALL_DIR		= $(_INSTALLDIR)\doc
SCRIPT_INSTALL_DIR	= $(_INSTALLDIR)\lib\tcl$(DOTVERSION)
INCLUDE_INSTALL_DIR	= $(_INSTALLDIR)\include

		"$(SCRIPT_INSTALL_DIR)\encoding\"

#" emacs fix

install-tzdata:
	@echo Installing time zone data
	@set TCL_LIBRARY=$(ROOT)/library
	@$(TCLSH_NATIVE) "$(ROOT)/tools/installData.tcl" \
	    "$(ROOT)/library/tzdata" "$(SCRIPT_INSTALL_DIR)/tzdata"

install-msgs:
	@echo Installing message catalogs
	@set TCL_LIBRARY=$(ROOT)/library
	@$(TCLSH_NATIVE) "$(ROOT)/tools/installData.tcl" \
	    "$(ROOT)/library/msgs" "$(SCRIPT_INSTALL_DIR)/msgs"

#---------------------------------------------------------------------
# Clean up
#---------------------------------------------------------------------

tidy:

### Assume the normal default.
_INSTALLDIR	= C:\Program Files\Tcl
!else
### Fix the path separators.
_INSTALLDIR	= $(INSTALLDIR:/=\)
!endif













#----------------------------------------------------------
# Set the proper copy method to avoid overwrite questions
# to the user when copying files and selecting the right
# "delete all" method.
#----------------------------------------------------------

!if "$(OS)" == "Windows_NT"

CPY	= xcopy /i >_JUNK.OUT # On Win98 NUL does not work here.
COPY	= copy >_JUNK.OUT # On Win98 NUL does not work here.
RMDIR	= deltree /Y
NULL    = \NUL # Used in testing directory existence
ERRNULL = >NUL # Win9x shell cannot redirect stderr
!endif
MKDIR   = mkdir

#------------------------------------------------------------------------------
# Determine the host and target architectures and compiler version.
#------------------------------------------------------------------------------

_HASH=^#
_VC_MANIFEST_EMBED_EXE=
_VC_MANIFEST_EMBED_DLL=
VCVER=0
!if ![echo VCVERSION=_MSC_VER > vercl.x] \
    && ![echo $(_HASH)if defined(_M_IX86) >> vercl.x] \
    && ![echo ARCH=IX86 >> vercl.x] \
    && ![echo $(_HASH)elif defined(_M_AMD64) >> vercl.x] \
    && ![echo ARCH=AMD64 >> vercl.x] \
    && ![echo $(_HASH)endif >> vercl.x] \
    && ![cl -nologo -TC -P vercl.x $(ERRNULL)]
!include vercl.i
!if ![echo VCVER= ^\> vercl.vc] \
    && ![set /a $(VCVERSION) / 100 - 6 >> vercl.vc]
!include vercl.vc
!endif
!endif
!if ![del $(ERRNUL) /q/f vercl.x vercl.i vercl.vc]
!endif

!if ![reg query HKLM\Hardware\Description\System\CentralProcessor\0 /v Identifier | findstr /i x86]
NATIVE_ARCH=IX86
!else
NATIVE_ARCH=AMD64
!endif

# Since MSVC8 we must deal with manifest resources.
!if $(VCVERSION) >= 1400
_VC_MANIFEST_EMBED_EXE=if exist [email protected] mt -nologo -manifest [email protected] -outputresource:$@;1
_VC_MANIFEST_EMBED_DLL=if exist [email protected] mt -nologo -manifest [email protected] -outputresource:$@;2
!endif

!ifndef MACHINE
MACHINE=$(ARCH)
!endif

!ifndef CFG_ENCODING
CFG_ENCODING	= \"cp1252\"
!endif

!message ===============================================================================

#----------------------------------------------------------
# build the helper app we need to overcome nmake's limiting
# environment.
#----------------------------------------------------------

LINKERFLAGS     =

!if [nmakehlp -l -ltcg]
LINKERFLAGS     =-ltcg
!endif































#----------------------------------------------------------
# Decode the options requested.
#----------------------------------------------------------

!if "$(OPTS)" == "" || [nmakehlp -f "$(OPTS)" "none"]
STATIC_BUILD	= 0
TCL_THREADS	= 1

#----------------------------------------------------------

!message *** Intermediate directory will be '$(TMP_DIR)'
!message *** Output directory will be '$(OUT_DIR)'
!message *** Suffix for binaries will be '$(SUFX)'
!message *** Optional defines are '$(OPTDEFINES)'
!message *** Compiler version $(VCVER). Target machine is $(MACHINE)
!message *** Host architecture is $(NATIVE_ARCH)
!message *** Compiler options '$(COMPILERFLAGS) $(OPTIMIZATIONS) $(DEBUGFLAGS) $(WARNINGS)'
!message *** Link options '$(LINKERFLAGS)'

!endif

 *
 * Copyright (c) 1995-1998 Sun Microsystems, Inc.
 *
 * See the file "license.terms" for information on usage and redistribution of
 * this file, and for a DISCLAIMER OF ALL WARRANTIES.
 */

#ifndef _WIN64
#   define _USE_32BIT_TIME_T
#endif
#include "tclWinInt.h"
#include "tclFileSystem.h"
#include <winioctl.h>
#include <sys/stat.h>
#include <shlobj.h>
#include <lm.h>		/* For TclpGetUserHome(). */

#include <malloc.h>
#include <process.h>
#include <signal.h>
#include <limits.h>

#ifdef __CYGWIN__
#   include <unistd.h>
#   ifndef _vsnprintf
#	define _vsnprintf vsnprintf
#   endif
#   ifndef _wcsicmp
#	define _wcsicmp wcscasecmp
#   endif
#else
#   ifndef strncasecmp
#	define strncasecmp strnicmp
#   endif

#   else
#	include <sys/utime.h>
#   endif /* __BORLANDC__ */
#endif /* __MWERKS__ */

#include <time.h>

/*
 * Not all mingw32 versions have this struct.
 */
#if !defined(__BORLANDC__) && !defined(_MSC_VER) && !defined(_WIN64) && !defined(HAVE_STRUCT_STAT32I64)
  struct _stat32i64 {
    dev_t st_dev;
    ino_t st_ino;
    unsigned short st_mode;
    short st_nlink;
    short st_uid;
    short st_gid;
    dev_t st_rdev;
    __int64 st_size;
#ifdef __CYGWIN__
    struct {long tv_sec;} st_atim;
    struct {long tv_sec;} st_mtim;
    struct {long tv_sec;} st_ctim;
#else
    long st_atime;
    long st_mtime;
    long st_ctime;
#endif
  };
#endif

/*
 * The following defines redefine the Windows Socket errors as
 * BSD errors so Tcl_PosixError can do the right thing.
 */

#ifndef ENOTEMPTY
#   define ENOTEMPTY 	41	/* Directory not empty */