os/persistentdata/persistentstorage/sqlite3api/TEST/TCL/tcldistribution/generic/tclScan.c
changeset 0 bde4ae8d615e
     1.1 --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
     1.2 +++ b/os/persistentdata/persistentstorage/sqlite3api/TEST/TCL/tcldistribution/generic/tclScan.c	Fri Jun 15 03:10:57 2012 +0200
     1.3 @@ -0,0 +1,1222 @@
     1.4 +/* 
     1.5 + * tclScan.c --
     1.6 + *
     1.7 + *	This file contains the implementation of the "scan" command.
     1.8 + *
     1.9 + * Copyright (c) 1998 by Scriptics Corporation.
    1.10 + *
    1.11 + * See the file "license.terms" for information on usage and redistribution
    1.12 + * of this file, and for a DISCLAIMER OF ALL WARRANTIES.
    1.13 + *
    1.14 + * RCS: @(#) $Id: tclScan.c,v 1.12.2.2 2005/10/23 22:01:30 msofer Exp $
    1.15 + */
    1.16 +
    1.17 +#include "tclInt.h"
    1.18 +/*
    1.19 + * For strtoll() and strtoull() declarations on some platforms...
    1.20 + */
    1.21 +#include "tclPort.h"
    1.22 +
    1.23 +/*
    1.24 + * Flag values used by Tcl_ScanObjCmd.
    1.25 + */
    1.26 +
    1.27 +#define SCAN_NOSKIP	0x1		  /* Don't skip blanks. */
    1.28 +#define SCAN_SUPPRESS	0x2		  /* Suppress assignment. */
    1.29 +#define SCAN_UNSIGNED	0x4		  /* Read an unsigned value. */
    1.30 +#define SCAN_WIDTH	0x8		  /* A width value was supplied. */
    1.31 +
    1.32 +#define SCAN_SIGNOK	0x10		  /* A +/- character is allowed. */
    1.33 +#define SCAN_NODIGITS	0x20		  /* No digits have been scanned. */
    1.34 +#define SCAN_NOZERO	0x40		  /* No zero digits have been scanned. */
    1.35 +#define SCAN_XOK	0x80		  /* An 'x' is allowed. */
    1.36 +#define SCAN_PTOK	0x100		  /* Decimal point is allowed. */
    1.37 +#define SCAN_EXPOK	0x200		  /* An exponent is allowed. */
    1.38 +
    1.39 +#define SCAN_LONGER	0x400		  /* Asked for a wide value. */
    1.40 +
    1.41 +/*
    1.42 + * The following structure contains the information associated with
    1.43 + * a character set.
    1.44 + */
    1.45 +
    1.46 +typedef struct CharSet {
    1.47 +    int exclude;		/* 1 if this is an exclusion set. */
    1.48 +    int nchars;
    1.49 +    Tcl_UniChar *chars;
    1.50 +    int nranges;
    1.51 +    struct Range {
    1.52 +	Tcl_UniChar start;
    1.53 +	Tcl_UniChar end;
    1.54 +    } *ranges;
    1.55 +} CharSet;
    1.56 +
    1.57 +/*
    1.58 + * Declarations for functions used only in this file.
    1.59 + */
    1.60 +
    1.61 +static char *	BuildCharSet _ANSI_ARGS_((CharSet *cset, char *format));
    1.62 +static int	CharInSet _ANSI_ARGS_((CharSet *cset, int ch));
    1.63 +static void	ReleaseCharSet _ANSI_ARGS_((CharSet *cset));
    1.64 +static int	ValidateFormat _ANSI_ARGS_((Tcl_Interp *interp, char *format,
    1.65 +		    int numVars, int *totalVars));
    1.66 +
    1.67 +/*
    1.68 + *----------------------------------------------------------------------
    1.69 + *
    1.70 + * BuildCharSet --
    1.71 + *
    1.72 + *	This function examines a character set format specification
    1.73 + *	and builds a CharSet containing the individual characters and
    1.74 + *	character ranges specified.
    1.75 + *
    1.76 + * Results:
    1.77 + *	Returns the next format position.
    1.78 + *
    1.79 + * Side effects:
    1.80 + *	Initializes the charset.
    1.81 + *
    1.82 + *----------------------------------------------------------------------
    1.83 + */
    1.84 +
    1.85 +static char *
    1.86 +BuildCharSet(cset, format)
    1.87 +    CharSet *cset;
    1.88 +    char *format;		/* Points to first char of set. */
    1.89 +{
    1.90 +    Tcl_UniChar ch, start;
    1.91 +    int offset, nranges;
    1.92 +    char *end;
    1.93 +
    1.94 +    memset(cset, 0, sizeof(CharSet));
    1.95 +    
    1.96 +    offset = Tcl_UtfToUniChar(format, &ch);
    1.97 +    if (ch == '^') {
    1.98 +	cset->exclude = 1;
    1.99 +	format += offset;
   1.100 +	offset = Tcl_UtfToUniChar(format, &ch);
   1.101 +    }
   1.102 +    end = format + offset;
   1.103 +
   1.104 +    /*
   1.105 +     * Find the close bracket so we can overallocate the set.
   1.106 +     */
   1.107 +
   1.108 +    if (ch == ']') {
   1.109 +	end += Tcl_UtfToUniChar(end, &ch);
   1.110 +    }
   1.111 +    nranges = 0;
   1.112 +    while (ch != ']') {
   1.113 +	if (ch == '-') {
   1.114 +	    nranges++;
   1.115 +	}
   1.116 +	end += Tcl_UtfToUniChar(end, &ch);
   1.117 +    }
   1.118 +
   1.119 +    cset->chars = (Tcl_UniChar *) ckalloc(sizeof(Tcl_UniChar)
   1.120 +	    * (end - format - 1));
   1.121 +    if (nranges > 0) {
   1.122 +	cset->ranges = (struct Range *) ckalloc(sizeof(struct Range)*nranges);
   1.123 +    } else {
   1.124 +	cset->ranges = NULL;
   1.125 +    }
   1.126 +
   1.127 +    /*
   1.128 +     * Now build the character set.
   1.129 +     */
   1.130 +
   1.131 +    cset->nchars = cset->nranges = 0;
   1.132 +    format += Tcl_UtfToUniChar(format, &ch);
   1.133 +    start = ch;
   1.134 +    if (ch == ']' || ch == '-') {
   1.135 +	cset->chars[cset->nchars++] = ch;
   1.136 +	format += Tcl_UtfToUniChar(format, &ch);
   1.137 +    }
   1.138 +    while (ch != ']') {
   1.139 +	if (*format == '-') {
   1.140 +	    /*
   1.141 +	     * This may be the first character of a range, so don't add
   1.142 +	     * it yet.
   1.143 +	     */
   1.144 +
   1.145 +	    start = ch;
   1.146 +	} else if (ch == '-') {
   1.147 +	    /*
   1.148 +	     * Check to see if this is the last character in the set, in which
   1.149 +	     * case it is not a range and we should add the previous character
   1.150 +	     * as well as the dash.
   1.151 +	     */
   1.152 +
   1.153 +	    if (*format == ']') {
   1.154 +		cset->chars[cset->nchars++] = start;
   1.155 +		cset->chars[cset->nchars++] = ch;
   1.156 +	    } else {
   1.157 +		format += Tcl_UtfToUniChar(format, &ch);
   1.158 +
   1.159 +		/*
   1.160 +		 * Check to see if the range is in reverse order.
   1.161 +		 */
   1.162 +
   1.163 +		if (start < ch) {
   1.164 +		    cset->ranges[cset->nranges].start = start;
   1.165 +		    cset->ranges[cset->nranges].end = ch;
   1.166 +		} else {
   1.167 +		    cset->ranges[cset->nranges].start = ch;
   1.168 +		    cset->ranges[cset->nranges].end = start;
   1.169 +		}		    
   1.170 +		cset->nranges++;
   1.171 +	    }
   1.172 +	} else {
   1.173 +	    cset->chars[cset->nchars++] = ch;
   1.174 +	}
   1.175 +	format += Tcl_UtfToUniChar(format, &ch);
   1.176 +    }
   1.177 +    return format;
   1.178 +}
   1.179 +
   1.180 +/*
   1.181 + *----------------------------------------------------------------------
   1.182 + *
   1.183 + * CharInSet --
   1.184 + *
   1.185 + *	Check to see if a character matches the given set.
   1.186 + *
   1.187 + * Results:
   1.188 + *	Returns non-zero if the character matches the given set.
   1.189 + *
   1.190 + * Side effects:
   1.191 + *	None.
   1.192 + *
   1.193 + *----------------------------------------------------------------------
   1.194 + */
   1.195 +
   1.196 +static int
   1.197 +CharInSet(cset, c)
   1.198 +    CharSet *cset;
   1.199 +    int c;			/* Character to test, passed as int because
   1.200 +				 * of non-ANSI prototypes. */
   1.201 +{
   1.202 +    Tcl_UniChar ch = (Tcl_UniChar) c;
   1.203 +    int i, match = 0;
   1.204 +    for (i = 0; i < cset->nchars; i++) {
   1.205 +	if (cset->chars[i] == ch) {
   1.206 +	    match = 1;
   1.207 +	    break;
   1.208 +	}
   1.209 +    }
   1.210 +    if (!match) {
   1.211 +	for (i = 0; i < cset->nranges; i++) {
   1.212 +	    if ((cset->ranges[i].start <= ch)
   1.213 +		    && (ch <= cset->ranges[i].end)) {
   1.214 +		match = 1;
   1.215 +		break;
   1.216 +	    }
   1.217 +	}
   1.218 +    }
   1.219 +    return (cset->exclude ? !match : match);    
   1.220 +}
   1.221 +
   1.222 +/*
   1.223 + *----------------------------------------------------------------------
   1.224 + *
   1.225 + * ReleaseCharSet --
   1.226 + *
   1.227 + *	Free the storage associated with a character set.
   1.228 + *
   1.229 + * Results:
   1.230 + *	None.
   1.231 + *
   1.232 + * Side effects:
   1.233 + *	None.
   1.234 + *
   1.235 + *----------------------------------------------------------------------
   1.236 + */
   1.237 +
   1.238 +static void
   1.239 +ReleaseCharSet(cset)
   1.240 +    CharSet *cset;
   1.241 +{
   1.242 +    ckfree((char *)cset->chars);
   1.243 +    if (cset->ranges) {
   1.244 +	ckfree((char *)cset->ranges);
   1.245 +    }
   1.246 +}
   1.247 +
   1.248 +/*
   1.249 + *----------------------------------------------------------------------
   1.250 + *
   1.251 + * ValidateFormat --
   1.252 + *
   1.253 + *	Parse the format string and verify that it is properly formed
   1.254 + *	and that there are exactly enough variables on the command line.
   1.255 + *
   1.256 + * Results:
   1.257 + *	A standard Tcl result.
   1.258 + *
   1.259 + * Side effects:
   1.260 + *	May place an error in the interpreter result.
   1.261 + *
   1.262 + *----------------------------------------------------------------------
   1.263 + */
   1.264 +
   1.265 +static int
   1.266 +ValidateFormat(interp, format, numVars, totalSubs)
   1.267 +    Tcl_Interp *interp;		/* Current interpreter. */
   1.268 +    char *format;		/* The format string. */
   1.269 +    int numVars;		/* The number of variables passed to the
   1.270 +				 * scan command. */
   1.271 +    int *totalSubs;		/* The number of variables that will be
   1.272 +				 * required. */
   1.273 +{
   1.274 +#define STATIC_LIST_SIZE 16
   1.275 +    int gotXpg, gotSequential, value, i, flags;
   1.276 +    char *end;
   1.277 +    Tcl_UniChar ch;
   1.278 +    int staticAssign[STATIC_LIST_SIZE];
   1.279 +    int *nassign = staticAssign;
   1.280 +    int objIndex, xpgSize, nspace = STATIC_LIST_SIZE;
   1.281 +    char buf[TCL_UTF_MAX+1];
   1.282 +
   1.283 +    /*
   1.284 +     * Initialize an array that records the number of times a variable
   1.285 +     * is assigned to by the format string.  We use this to detect if
   1.286 +     * a variable is multiply assigned or left unassigned.
   1.287 +     */
   1.288 +
   1.289 +    if (numVars > nspace) {
   1.290 +	nassign = (int*)ckalloc(sizeof(int) * numVars);
   1.291 +	nspace = numVars;
   1.292 +    }
   1.293 +    for (i = 0; i < nspace; i++) {
   1.294 +	nassign[i] = 0;
   1.295 +    }
   1.296 +
   1.297 +    xpgSize = objIndex = gotXpg = gotSequential = 0;
   1.298 +
   1.299 +    while (*format != '\0') {
   1.300 +	format += Tcl_UtfToUniChar(format, &ch);
   1.301 +
   1.302 +	flags = 0;
   1.303 +
   1.304 +	if (ch != '%') {
   1.305 +	    continue;
   1.306 +	}
   1.307 +	format += Tcl_UtfToUniChar(format, &ch);
   1.308 +	if (ch == '%') {
   1.309 +	    continue;
   1.310 +	}
   1.311 +	if (ch == '*') {
   1.312 +	    flags |= SCAN_SUPPRESS;
   1.313 +	    format += Tcl_UtfToUniChar(format, &ch);
   1.314 +	    goto xpgCheckDone;
   1.315 +	}
   1.316 +
   1.317 +	if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */
   1.318 +	    /*
   1.319 +	     * Check for an XPG3-style %n$ specification.  Note: there
   1.320 +	     * must not be a mixture of XPG3 specs and non-XPG3 specs
   1.321 +	     * in the same format string.
   1.322 +	     */
   1.323 +
   1.324 +	    value = strtoul(format-1, &end, 10); /* INTL: "C" locale. */
   1.325 +	    if (*end != '$') {
   1.326 +		goto notXpg;
   1.327 +	    }
   1.328 +	    format = end+1;
   1.329 +	    format += Tcl_UtfToUniChar(format, &ch);
   1.330 +	    gotXpg = 1;
   1.331 +	    if (gotSequential) {
   1.332 +		goto mixedXPG;
   1.333 +	    }
   1.334 +	    objIndex = value - 1;
   1.335 +	    if ((objIndex < 0) || (numVars && (objIndex >= numVars))) {
   1.336 +		goto badIndex;
   1.337 +	    } else if (numVars == 0) {
   1.338 +		/*
   1.339 +		 * In the case where no vars are specified, the user can
   1.340 +		 * specify %9999$ legally, so we have to consider special
   1.341 +		 * rules for growing the assign array.  'value' is
   1.342 +		 * guaranteed to be > 0.
   1.343 +		 */
   1.344 +		xpgSize = (xpgSize > value) ? xpgSize : value;
   1.345 +	    }
   1.346 +	    goto xpgCheckDone;
   1.347 +	}
   1.348 +
   1.349 +	notXpg:
   1.350 +	gotSequential = 1;
   1.351 +	if (gotXpg) {
   1.352 +	    mixedXPG:
   1.353 +	    Tcl_SetResult(interp,
   1.354 +		    "cannot mix \"%\" and \"%n$\" conversion specifiers",
   1.355 +		    TCL_STATIC);
   1.356 +	    goto error;
   1.357 +	}
   1.358 +
   1.359 +	xpgCheckDone:
   1.360 +	/*
   1.361 +	 * Parse any width specifier.
   1.362 +	 */
   1.363 +
   1.364 +	if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */
   1.365 +	    value = strtoul(format-1, &format, 10); /* INTL: "C" locale. */
   1.366 +	    flags |= SCAN_WIDTH;
   1.367 +	    format += Tcl_UtfToUniChar(format, &ch);
   1.368 +	}
   1.369 +
   1.370 +	/*
   1.371 +	 * Handle any size specifier.
   1.372 +	 */
   1.373 +
   1.374 +	switch (ch) {
   1.375 +	case 'l':
   1.376 +	case 'L':
   1.377 +	    flags |= SCAN_LONGER;
   1.378 +	case 'h':
   1.379 +	    format += Tcl_UtfToUniChar(format, &ch);
   1.380 +	}
   1.381 +
   1.382 +	if (!(flags & SCAN_SUPPRESS) && numVars && (objIndex >= numVars)) {
   1.383 +	    goto badIndex;
   1.384 +	}
   1.385 +
   1.386 +	/*
   1.387 +	 * Handle the various field types.
   1.388 +	 */
   1.389 +
   1.390 +	switch (ch) {
   1.391 +	    case 'c':
   1.392 +                if (flags & SCAN_WIDTH) {
   1.393 +		    Tcl_SetResult(interp,
   1.394 +			    "field width may not be specified in %c conversion",
   1.395 +			    TCL_STATIC);
   1.396 +		    goto error;
   1.397 +                }
   1.398 +		/*
   1.399 +		 * Fall through!
   1.400 +		 */
   1.401 +	    case 'n':
   1.402 +	    case 's':
   1.403 +		if (flags & SCAN_LONGER) {
   1.404 +		invalidLonger:
   1.405 +		    buf[Tcl_UniCharToUtf(ch, buf)] = '\0';
   1.406 +		    Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
   1.407 +			   "'l' modifier may not be specified in %", buf,
   1.408 +			   " conversion", NULL);
   1.409 +		    goto error;
   1.410 +		}
   1.411 +		/*
   1.412 +		 * Fall through!
   1.413 +		 */
   1.414 +	    case 'd':
   1.415 +	    case 'e':
   1.416 +	    case 'f':
   1.417 +	    case 'g':
   1.418 +	    case 'i':
   1.419 +	    case 'o':
   1.420 +	    case 'u':
   1.421 +	    case 'x':
   1.422 + 		break;
   1.423 +		/*
   1.424 +		 * Bracket terms need special checking
   1.425 +		 */
   1.426 +	    case '[':
   1.427 +		if (flags & SCAN_LONGER) {
   1.428 +		    goto invalidLonger;
   1.429 +		}
   1.430 +		if (*format == '\0') {
   1.431 +		    goto badSet;
   1.432 +		}
   1.433 +		format += Tcl_UtfToUniChar(format, &ch);
   1.434 +		if (ch == '^') {
   1.435 +		    if (*format == '\0') {
   1.436 +			goto badSet;
   1.437 +		    }
   1.438 +		    format += Tcl_UtfToUniChar(format, &ch);
   1.439 +		}
   1.440 +		if (ch == ']') {
   1.441 +		    if (*format == '\0') {
   1.442 +			goto badSet;
   1.443 +		    }
   1.444 +		    format += Tcl_UtfToUniChar(format, &ch);
   1.445 +		}
   1.446 +		while (ch != ']') {
   1.447 +		    if (*format == '\0') {
   1.448 +			goto badSet;
   1.449 +		    }
   1.450 +		    format += Tcl_UtfToUniChar(format, &ch);
   1.451 +		}
   1.452 +		break;
   1.453 +	    badSet:
   1.454 +		Tcl_SetResult(interp, "unmatched [ in format string",
   1.455 +			TCL_STATIC);
   1.456 +		goto error;
   1.457 +	    default:
   1.458 +	    {
   1.459 +		char buf[TCL_UTF_MAX+1];
   1.460 +
   1.461 +		buf[Tcl_UniCharToUtf(ch, buf)] = '\0';
   1.462 +		Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
   1.463 +			"bad scan conversion character \"", buf, "\"", NULL);
   1.464 +		goto error;
   1.465 +	    }
   1.466 +	}
   1.467 +	if (!(flags & SCAN_SUPPRESS)) {
   1.468 +	    if (objIndex >= nspace) {
   1.469 +		/*
   1.470 +		 * Expand the nassign buffer.  If we are using XPG specifiers,
   1.471 +		 * make sure that we grow to a large enough size.  xpgSize is
   1.472 +		 * guaranteed to be at least one larger than objIndex.
   1.473 +		 */
   1.474 +		value = nspace;
   1.475 +		if (xpgSize) {
   1.476 +		    nspace = xpgSize;
   1.477 +		} else {
   1.478 +		    nspace += STATIC_LIST_SIZE;
   1.479 +		}
   1.480 +		if (nassign == staticAssign) {
   1.481 +		    nassign = (void *)ckalloc(nspace * sizeof(int));
   1.482 +		    for (i = 0; i < STATIC_LIST_SIZE; ++i) {
   1.483 +			nassign[i] = staticAssign[i];
   1.484 +		    }
   1.485 +		} else {
   1.486 +		    nassign = (void *)ckrealloc((void *)nassign,
   1.487 +			    nspace * sizeof(int));
   1.488 +		}
   1.489 +		for (i = value; i < nspace; i++) {
   1.490 +		    nassign[i] = 0;
   1.491 +		}
   1.492 +	    }
   1.493 +	    nassign[objIndex]++;
   1.494 +	    objIndex++;
   1.495 +	}
   1.496 +    }
   1.497 +
   1.498 +    /*
   1.499 +     * Verify that all of the variable were assigned exactly once.
   1.500 +     */
   1.501 +
   1.502 +    if (numVars == 0) {
   1.503 +	if (xpgSize) {
   1.504 +	    numVars = xpgSize;
   1.505 +	} else {
   1.506 +	    numVars = objIndex;
   1.507 +	}
   1.508 +    }
   1.509 +    if (totalSubs) {
   1.510 +	*totalSubs = numVars;
   1.511 +    }
   1.512 +    for (i = 0; i < numVars; i++) {
   1.513 +	if (nassign[i] > 1) {
   1.514 +	    Tcl_SetResult(interp, "variable is assigned by multiple \"%n$\" conversion specifiers", TCL_STATIC);
   1.515 +	    goto error;
   1.516 +	} else if (!xpgSize && (nassign[i] == 0)) {
   1.517 +	    /*
   1.518 +	     * If the space is empty, and xpgSize is 0 (means XPG wasn't
   1.519 +	     * used, and/or numVars != 0), then too many vars were given
   1.520 +	     */
   1.521 +	    Tcl_SetResult(interp, "variable is not assigned by any conversion specifiers", TCL_STATIC);
   1.522 +	    goto error;
   1.523 +	}
   1.524 +    }
   1.525 +
   1.526 +    if (nassign != staticAssign) {
   1.527 +	ckfree((char *)nassign);
   1.528 +    }
   1.529 +    return TCL_OK;
   1.530 +
   1.531 +    badIndex:
   1.532 +    if (gotXpg) {
   1.533 +	Tcl_SetResult(interp, "\"%n$\" argument index out of range",
   1.534 +		TCL_STATIC);
   1.535 +    } else {
   1.536 +	Tcl_SetResult(interp, 
   1.537 +		"different numbers of variable names and field specifiers",
   1.538 +		TCL_STATIC);
   1.539 +    }
   1.540 +
   1.541 +    error:
   1.542 +    if (nassign != staticAssign) {
   1.543 +	ckfree((char *)nassign);
   1.544 +    }
   1.545 +    return TCL_ERROR;
   1.546 +#undef STATIC_LIST_SIZE
   1.547 +}
   1.548 +
   1.549 +/*
   1.550 + *----------------------------------------------------------------------
   1.551 + *
   1.552 + * Tcl_ScanObjCmd --
   1.553 + *
   1.554 + *	This procedure is invoked to process the "scan" Tcl command.
   1.555 + *	See the user documentation for details on what it does.
   1.556 + *
   1.557 + * Results:
   1.558 + *	A standard Tcl result.
   1.559 + *
   1.560 + * Side effects:
   1.561 + *	See the user documentation.
   1.562 + *
   1.563 + *----------------------------------------------------------------------
   1.564 + */
   1.565 +
   1.566 +	/* ARGSUSED */
   1.567 +int
   1.568 +Tcl_ScanObjCmd(dummy, interp, objc, objv)
   1.569 +    ClientData dummy;    	/* Not used. */
   1.570 +    Tcl_Interp *interp;		/* Current interpreter. */
   1.571 +    int objc;			/* Number of arguments. */
   1.572 +    Tcl_Obj *CONST objv[];	/* Argument objects. */
   1.573 +{
   1.574 +    char *format;
   1.575 +    int numVars, nconversions, totalVars = -1;
   1.576 +    int objIndex, offset, i, result, code;
   1.577 +    long value;
   1.578 +    char *string, *end, *baseString;
   1.579 +    char op = 0;
   1.580 +    int base = 0;
   1.581 +    int underflow = 0;
   1.582 +    size_t width;
   1.583 +    long (*fn)() = NULL;
   1.584 +#ifndef TCL_WIDE_INT_IS_LONG
   1.585 +    Tcl_WideInt (*lfn)() = NULL;
   1.586 +    Tcl_WideInt wideValue;
   1.587 +#endif
   1.588 +    Tcl_UniChar ch, sch;
   1.589 +    Tcl_Obj **objs = NULL, *objPtr = NULL;
   1.590 +    int flags;
   1.591 +    char buf[513];			  /* Temporary buffer to hold scanned
   1.592 +					   * number strings before they are
   1.593 +					   * passed to strtoul. */
   1.594 +
   1.595 +    if (objc < 3) {
   1.596 +        Tcl_WrongNumArgs(interp, 1, objv,
   1.597 +		"string format ?varName varName ...?");
   1.598 +	return TCL_ERROR;
   1.599 +    }
   1.600 +
   1.601 +    format = Tcl_GetStringFromObj(objv[2], NULL);
   1.602 +    numVars = objc-3;
   1.603 +
   1.604 +    /*
   1.605 +     * Check for errors in the format string.
   1.606 +     */
   1.607 +    
   1.608 +    if (ValidateFormat(interp, format, numVars, &totalVars) == TCL_ERROR) {
   1.609 +	return TCL_ERROR;
   1.610 +    }
   1.611 +
   1.612 +    /*
   1.613 +     * Allocate space for the result objects.
   1.614 +     */
   1.615 +
   1.616 +    if (totalVars > 0) {
   1.617 +	objs = (Tcl_Obj **) ckalloc(sizeof(Tcl_Obj*) * totalVars);
   1.618 +	for (i = 0; i < totalVars; i++) {
   1.619 +	    objs[i] = NULL;
   1.620 +	}
   1.621 +    }
   1.622 +
   1.623 +    string = Tcl_GetStringFromObj(objv[1], NULL);
   1.624 +    baseString = string;
   1.625 +
   1.626 +    /*
   1.627 +     * Iterate over the format string filling in the result objects until
   1.628 +     * we reach the end of input, the end of the format string, or there
   1.629 +     * is a mismatch.
   1.630 +     */
   1.631 +
   1.632 +    objIndex = 0;
   1.633 +    nconversions = 0;
   1.634 +    while (*format != '\0') {
   1.635 +	format += Tcl_UtfToUniChar(format, &ch);
   1.636 +
   1.637 +	flags = 0;
   1.638 +
   1.639 +	/*
   1.640 +	 * If we see whitespace in the format, skip whitespace in the string.
   1.641 +	 */
   1.642 +
   1.643 +	if (Tcl_UniCharIsSpace(ch)) {
   1.644 +	    offset = Tcl_UtfToUniChar(string, &sch);
   1.645 +	    while (Tcl_UniCharIsSpace(sch)) {
   1.646 +		if (*string == '\0') {
   1.647 +		    goto done;
   1.648 +		}
   1.649 +		string += offset;
   1.650 +		offset = Tcl_UtfToUniChar(string, &sch);
   1.651 +	    }
   1.652 +	    continue;
   1.653 +	}
   1.654 +	    
   1.655 +	if (ch != '%') {
   1.656 +	    literal:
   1.657 +	    if (*string == '\0') {
   1.658 +		underflow = 1;
   1.659 +		goto done;
   1.660 +	    }
   1.661 +	    string += Tcl_UtfToUniChar(string, &sch);
   1.662 +	    if (ch != sch) {
   1.663 +		goto done;
   1.664 +	    }
   1.665 +	    continue;
   1.666 +	}
   1.667 +
   1.668 +	format += Tcl_UtfToUniChar(format, &ch);
   1.669 +	if (ch == '%') {
   1.670 +	    goto literal;
   1.671 +	}
   1.672 +
   1.673 +	/*
   1.674 +	 * Check for assignment suppression ('*') or an XPG3-style
   1.675 +	 * assignment ('%n$').
   1.676 +	 */
   1.677 +
   1.678 +	if (ch == '*') {
   1.679 +	    flags |= SCAN_SUPPRESS;
   1.680 +	    format += Tcl_UtfToUniChar(format, &ch);
   1.681 +	} else if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */
   1.682 +	    value = strtoul(format-1, &end, 10); /* INTL: "C" locale. */
   1.683 +	    if (*end == '$') {
   1.684 +		format = end+1;
   1.685 +		format += Tcl_UtfToUniChar(format, &ch);
   1.686 +		objIndex = (int) value - 1;
   1.687 +	    }
   1.688 +	}
   1.689 +
   1.690 +	/*
   1.691 +	 * Parse any width specifier.
   1.692 +	 */
   1.693 +
   1.694 +	if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */
   1.695 +	    width = strtoul(format-1, &format, 10); /* INTL: "C" locale. */
   1.696 +	    format += Tcl_UtfToUniChar(format, &ch);
   1.697 +	} else {
   1.698 +	    width = 0;
   1.699 +	}
   1.700 +
   1.701 +	/*
   1.702 +	 * Handle any size specifier.
   1.703 +	 */
   1.704 +
   1.705 +	switch (ch) {
   1.706 +	case 'l':
   1.707 +	case 'L':
   1.708 +	    flags |= SCAN_LONGER;
   1.709 +	    /*
   1.710 +	     * Fall through so we skip to the next character.
   1.711 +	     */
   1.712 +	case 'h':
   1.713 +	    format += Tcl_UtfToUniChar(format, &ch);
   1.714 +	}
   1.715 +
   1.716 +	/*
   1.717 +	 * Handle the various field types.
   1.718 +	 */
   1.719 +
   1.720 +	switch (ch) {
   1.721 +	    case 'n':
   1.722 +		if (!(flags & SCAN_SUPPRESS)) {
   1.723 +		    objPtr = Tcl_NewIntObj(string - baseString);
   1.724 +		    Tcl_IncrRefCount(objPtr);
   1.725 +		    objs[objIndex++] = objPtr;
   1.726 +		}
   1.727 +		nconversions++;
   1.728 +		continue;
   1.729 +
   1.730 +	    case 'd':
   1.731 +		op = 'i';
   1.732 +		base = 10;
   1.733 +		fn = (long (*)())strtol;
   1.734 +#ifndef TCL_WIDE_INT_IS_LONG
   1.735 +		lfn = (Tcl_WideInt (*)())strtoll;
   1.736 +#endif
   1.737 +		break;
   1.738 +	    case 'i':
   1.739 +		op = 'i';
   1.740 +		base = 0;
   1.741 +		fn = (long (*)())strtol;
   1.742 +#ifndef TCL_WIDE_INT_IS_LONG
   1.743 +		lfn = (Tcl_WideInt (*)())strtoll;
   1.744 +#endif
   1.745 +		break;
   1.746 +	    case 'o':
   1.747 +		op = 'i';
   1.748 +		base = 8;
   1.749 +		fn = (long (*)())strtoul;
   1.750 +#ifndef TCL_WIDE_INT_IS_LONG
   1.751 +		lfn = (Tcl_WideInt (*)())strtoull;
   1.752 +#endif
   1.753 +		break;
   1.754 +	    case 'x':
   1.755 +		op = 'i';
   1.756 +		base = 16;
   1.757 +		fn = (long (*)())strtoul;
   1.758 +#ifndef TCL_WIDE_INT_IS_LONG
   1.759 +		lfn = (Tcl_WideInt (*)())strtoull;
   1.760 +#endif
   1.761 +		break;
   1.762 +	    case 'u':
   1.763 +		op = 'i';
   1.764 +		base = 10;
   1.765 +		flags |= SCAN_UNSIGNED;
   1.766 +		fn = (long (*)())strtoul;
   1.767 +#ifndef TCL_WIDE_INT_IS_LONG
   1.768 +		lfn = (Tcl_WideInt (*)())strtoull;
   1.769 +#endif
   1.770 +		break;
   1.771 +
   1.772 +	    case 'f':
   1.773 +	    case 'e':
   1.774 +	    case 'g':
   1.775 +		op = 'f';
   1.776 +		break;
   1.777 +
   1.778 +	    case 's':
   1.779 +		op = 's';
   1.780 +		break;
   1.781 +
   1.782 +	    case 'c':
   1.783 +		op = 'c';
   1.784 +		flags |= SCAN_NOSKIP;
   1.785 +		break;
   1.786 +	    case '[':
   1.787 +		op = '[';
   1.788 +		flags |= SCAN_NOSKIP;
   1.789 +		break;
   1.790 +	}
   1.791 +
   1.792 +	/*
   1.793 +	 * At this point, we will need additional characters from the
   1.794 +	 * string to proceed.
   1.795 +	 */
   1.796 +
   1.797 +	if (*string == '\0') {
   1.798 +	    underflow = 1;
   1.799 +	    goto done;
   1.800 +	}
   1.801 +	
   1.802 +	/*
   1.803 +	 * Skip any leading whitespace at the beginning of a field unless
   1.804 +	 * the format suppresses this behavior.
   1.805 +	 */
   1.806 +
   1.807 +	if (!(flags & SCAN_NOSKIP)) {
   1.808 +	    while (*string != '\0') {
   1.809 +		offset = Tcl_UtfToUniChar(string, &sch);
   1.810 +		if (!Tcl_UniCharIsSpace(sch)) {
   1.811 +		    break;
   1.812 +		}
   1.813 +		string += offset;
   1.814 +	    }
   1.815 +	    if (*string == '\0') {
   1.816 +		underflow = 1;
   1.817 +		goto done;
   1.818 +	    }
   1.819 +	}
   1.820 +
   1.821 +	/*
   1.822 +	 * Perform the requested scanning operation.
   1.823 +	 */
   1.824 +	
   1.825 +	switch (op) {
   1.826 +	    case 's':
   1.827 +		/*
   1.828 +		 * Scan a string up to width characters or whitespace.
   1.829 +		 */
   1.830 +
   1.831 +		if (width == 0) {
   1.832 +		    width = (size_t) ~0;
   1.833 +		}
   1.834 +		end = string;
   1.835 +		while (*end != '\0') {
   1.836 +		    offset = Tcl_UtfToUniChar(end, &sch);
   1.837 +		    if (Tcl_UniCharIsSpace(sch)) {
   1.838 +			break;
   1.839 +		    }
   1.840 +		    end += offset;
   1.841 +		    if (--width == 0) {
   1.842 +			break;
   1.843 +		    }
   1.844 +		}
   1.845 +		if (!(flags & SCAN_SUPPRESS)) {
   1.846 +		    objPtr = Tcl_NewStringObj(string, end-string);
   1.847 +		    Tcl_IncrRefCount(objPtr);
   1.848 +		    objs[objIndex++] = objPtr;
   1.849 +		}
   1.850 +		string = end;
   1.851 +		break;
   1.852 +
   1.853 +	    case '[': {
   1.854 +		CharSet cset;
   1.855 +
   1.856 +		if (width == 0) {
   1.857 +		    width = (size_t) ~0;
   1.858 +		}
   1.859 +		end = string;
   1.860 +
   1.861 +		format = BuildCharSet(&cset, format);
   1.862 +		while (*end != '\0') {
   1.863 +		    offset = Tcl_UtfToUniChar(end, &sch);
   1.864 +		    if (!CharInSet(&cset, (int)sch)) {
   1.865 +			break;
   1.866 +		    }
   1.867 +		    end += offset;
   1.868 +		    if (--width == 0) {
   1.869 +			break;
   1.870 +		    }
   1.871 +		}
   1.872 +		ReleaseCharSet(&cset);
   1.873 +
   1.874 +		if (string == end) {
   1.875 +		    /*
   1.876 +		     * Nothing matched the range, stop processing
   1.877 +		     */
   1.878 +		    goto done;
   1.879 +		}
   1.880 +		if (!(flags & SCAN_SUPPRESS)) {
   1.881 +		    objPtr = Tcl_NewStringObj(string, end-string);
   1.882 +		    Tcl_IncrRefCount(objPtr);
   1.883 +		    objs[objIndex++] = objPtr;
   1.884 +		}
   1.885 +		string = end;
   1.886 +		
   1.887 +		break;
   1.888 +	    }
   1.889 +	    case 'c':
   1.890 +		/*
   1.891 +		 * Scan a single Unicode character.
   1.892 +		 */
   1.893 +
   1.894 +		string += Tcl_UtfToUniChar(string, &sch);
   1.895 +		if (!(flags & SCAN_SUPPRESS)) {
   1.896 +		    objPtr = Tcl_NewIntObj((int)sch);
   1.897 +		    Tcl_IncrRefCount(objPtr);
   1.898 +		    objs[objIndex++] = objPtr;
   1.899 +		}
   1.900 +		break;
   1.901 +
   1.902 +	    case 'i':
   1.903 +		/*
   1.904 +		 * Scan an unsigned or signed integer.
   1.905 +		 */
   1.906 +
   1.907 +		if ((width == 0) || (width > sizeof(buf) - 1)) {
   1.908 +		    width = sizeof(buf) - 1;
   1.909 +		}
   1.910 +		flags |= SCAN_SIGNOK | SCAN_NODIGITS | SCAN_NOZERO;
   1.911 +		for (end = buf; width > 0; width--) {
   1.912 +		    switch (*string) {
   1.913 +			/*
   1.914 +			 * The 0 digit has special meaning at the beginning of
   1.915 +			 * a number.  If we are unsure of the base, it
   1.916 +			 * indicates that we are in base 8 or base 16 (if it is
   1.917 +			 * followed by an 'x').
   1.918 +			 *
   1.919 +			 * 8.1 - 8.3.4 incorrectly handled 0x... base-16
   1.920 +			 * cases for %x by not reading the 0x as the
   1.921 +			 * auto-prelude for base-16. [Bug #495213]
   1.922 +			 */
   1.923 +			case '0':
   1.924 +			    if (base == 0) {
   1.925 +				base = 8;
   1.926 +				flags |= SCAN_XOK;
   1.927 +			    }
   1.928 +			    if (base == 16) {
   1.929 +				flags |= SCAN_XOK;
   1.930 +			    }
   1.931 +			    if (flags & SCAN_NOZERO) {
   1.932 +				flags &= ~(SCAN_SIGNOK | SCAN_NODIGITS
   1.933 +					| SCAN_NOZERO);
   1.934 +			    } else {
   1.935 +				flags &= ~(SCAN_SIGNOK | SCAN_XOK
   1.936 +					| SCAN_NODIGITS);
   1.937 +			    }
   1.938 +			    goto addToInt;
   1.939 +
   1.940 +			case '1': case '2': case '3': case '4':
   1.941 +			case '5': case '6': case '7':
   1.942 +			    if (base == 0) {
   1.943 +				base = 10;
   1.944 +			    }
   1.945 +			    flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS);
   1.946 +			    goto addToInt;
   1.947 +
   1.948 +			case '8': case '9':
   1.949 +			    if (base == 0) {
   1.950 +				base = 10;
   1.951 +			    }
   1.952 +			    if (base <= 8) {
   1.953 +				break;
   1.954 +			    }
   1.955 +			    flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS);
   1.956 +			    goto addToInt;
   1.957 +
   1.958 +			case 'A': case 'B': case 'C':
   1.959 +			case 'D': case 'E': case 'F': 
   1.960 +			case 'a': case 'b': case 'c':
   1.961 +			case 'd': case 'e': case 'f':
   1.962 +			    if (base <= 10) {
   1.963 +				break;
   1.964 +			    }
   1.965 +			    flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS);
   1.966 +			    goto addToInt;
   1.967 +
   1.968 +			case '+': case '-':
   1.969 +			    if (flags & SCAN_SIGNOK) {
   1.970 +				flags &= ~SCAN_SIGNOK;
   1.971 +				goto addToInt;
   1.972 +			    }
   1.973 +			    break;
   1.974 +
   1.975 +			case 'x': case 'X':
   1.976 +			    if ((flags & SCAN_XOK) && (end == buf+1)) {
   1.977 +				base = 16;
   1.978 +				flags &= ~SCAN_XOK;
   1.979 +				goto addToInt;
   1.980 +			    }
   1.981 +			    break;
   1.982 +		    }
   1.983 +
   1.984 +		    /*
   1.985 +		     * We got an illegal character so we are done accumulating.
   1.986 +		     */
   1.987 +
   1.988 +		    break;
   1.989 +
   1.990 +		    addToInt:
   1.991 +		    /*
   1.992 +		     * Add the character to the temporary buffer.
   1.993 +		     */
   1.994 +
   1.995 +		    *end++ = *string++;
   1.996 +		    if (*string == '\0') {
   1.997 +			break;
   1.998 +		    }
   1.999 +		}
  1.1000 +
  1.1001 +		/*
  1.1002 +		 * Check to see if we need to back up because we only got a
  1.1003 +		 * sign or a trailing x after a 0.
  1.1004 +		 */
  1.1005 +
  1.1006 +		if (flags & SCAN_NODIGITS) {
  1.1007 +		    if (*string == '\0') {
  1.1008 +			underflow = 1;
  1.1009 +		    }
  1.1010 +		    goto done;
  1.1011 +		} else if (end[-1] == 'x' || end[-1] == 'X') {
  1.1012 +		    end--;
  1.1013 +		    string--;
  1.1014 +		}
  1.1015 +
  1.1016 +
  1.1017 +		/*
  1.1018 +		 * Scan the value from the temporary buffer.  If we are
  1.1019 +		 * returning a large unsigned value, we have to convert it back
  1.1020 +		 * to a string since Tcl only supports signed values.
  1.1021 +		 */
  1.1022 +
  1.1023 +		if (!(flags & SCAN_SUPPRESS)) {
  1.1024 +		    *end = '\0';
  1.1025 +#ifndef TCL_WIDE_INT_IS_LONG
  1.1026 +		    if (flags & SCAN_LONGER) {
  1.1027 +			wideValue = (Tcl_WideInt) (*lfn)(buf, NULL, base);
  1.1028 +			if ((flags & SCAN_UNSIGNED) && (wideValue < 0)) {
  1.1029 +			    /* INTL: ISO digit */
  1.1030 +			    sprintf(buf, "%" TCL_LL_MODIFIER "u",
  1.1031 +				    (Tcl_WideUInt)wideValue);
  1.1032 +			    objPtr = Tcl_NewStringObj(buf, -1);
  1.1033 +			} else {
  1.1034 +			    objPtr = Tcl_NewWideIntObj(wideValue);
  1.1035 +			}
  1.1036 +		    } else {
  1.1037 +#endif /* !TCL_WIDE_INT_IS_LONG */
  1.1038 +			value = (long) (*fn)(buf, NULL, base);
  1.1039 +			if ((flags & SCAN_UNSIGNED) && (value < 0)) {
  1.1040 +			    sprintf(buf, "%lu", value); /* INTL: ISO digit */
  1.1041 +			    objPtr = Tcl_NewStringObj(buf, -1);
  1.1042 +			} else if ((flags & SCAN_LONGER)
  1.1043 +				|| (unsigned long) value > UINT_MAX) {
  1.1044 +			    objPtr = Tcl_NewLongObj(value);
  1.1045 +			} else {
  1.1046 +			    objPtr = Tcl_NewIntObj(value);
  1.1047 +			}
  1.1048 +#ifndef TCL_WIDE_INT_IS_LONG
  1.1049 +		    }
  1.1050 +#endif
  1.1051 +		    Tcl_IncrRefCount(objPtr);
  1.1052 +		    objs[objIndex++] = objPtr;
  1.1053 +		}
  1.1054 +
  1.1055 +		break;
  1.1056 +
  1.1057 +	    case 'f':
  1.1058 +		/*
  1.1059 +		 * Scan a floating point number
  1.1060 +		 */
  1.1061 +
  1.1062 +		if ((width == 0) || (width > sizeof(buf) - 1)) {
  1.1063 +		    width = sizeof(buf) - 1;
  1.1064 +		}
  1.1065 +		flags &= ~SCAN_LONGER;
  1.1066 +		flags |= SCAN_SIGNOK | SCAN_NODIGITS | SCAN_PTOK | SCAN_EXPOK;
  1.1067 +		for (end = buf; width > 0; width--) {
  1.1068 +		    switch (*string) {
  1.1069 +			case '0': case '1': case '2': case '3':
  1.1070 +			case '4': case '5': case '6': case '7':
  1.1071 +			case '8': case '9':
  1.1072 +			    flags &= ~(SCAN_SIGNOK | SCAN_NODIGITS);
  1.1073 +			    goto addToFloat;
  1.1074 +			case '+': case '-':
  1.1075 +			    if (flags & SCAN_SIGNOK) {
  1.1076 +				flags &= ~SCAN_SIGNOK;
  1.1077 +				goto addToFloat;
  1.1078 +			    }
  1.1079 +			    break;
  1.1080 +			case '.':
  1.1081 +			    if (flags & SCAN_PTOK) {
  1.1082 +				flags &= ~(SCAN_SIGNOK | SCAN_PTOK);
  1.1083 +				goto addToFloat;
  1.1084 +			    }
  1.1085 +			    break;
  1.1086 +			case 'e': case 'E':
  1.1087 +			    /*
  1.1088 +			     * An exponent is not allowed until there has
  1.1089 +			     * been at least one digit.
  1.1090 +			     */
  1.1091 +
  1.1092 +			    if ((flags & (SCAN_NODIGITS | SCAN_EXPOK))
  1.1093 +				    == SCAN_EXPOK) {
  1.1094 +				flags = (flags & ~(SCAN_EXPOK|SCAN_PTOK))
  1.1095 +				    | SCAN_SIGNOK | SCAN_NODIGITS;
  1.1096 +				goto addToFloat;
  1.1097 +			    }
  1.1098 +			    break;
  1.1099 +		    }
  1.1100 +
  1.1101 +		    /*
  1.1102 +		     * We got an illegal character so we are done accumulating.
  1.1103 +		     */
  1.1104 +
  1.1105 +		    break;
  1.1106 +
  1.1107 +		    addToFloat:
  1.1108 +		    /*
  1.1109 +		     * Add the character to the temporary buffer.
  1.1110 +		     */
  1.1111 +
  1.1112 +		    *end++ = *string++;
  1.1113 +		    if (*string == '\0') {
  1.1114 +			break;
  1.1115 +		    }
  1.1116 +		}
  1.1117 +
  1.1118 +		/*
  1.1119 +		 * Check to see if we need to back up because we saw a
  1.1120 +		 * trailing 'e' or sign.
  1.1121 +		 */
  1.1122 +
  1.1123 +		if (flags & SCAN_NODIGITS) {
  1.1124 +		    if (flags & SCAN_EXPOK) {
  1.1125 +			/*
  1.1126 +			 * There were no digits at all so scanning has
  1.1127 +			 * failed and we are done.
  1.1128 +			 */
  1.1129 +			if (*string == '\0') {
  1.1130 +			    underflow = 1;
  1.1131 +			}
  1.1132 +			goto done;
  1.1133 +		    }
  1.1134 +
  1.1135 +		    /*
  1.1136 +		     * We got a bad exponent ('e' and maybe a sign).
  1.1137 +		     */
  1.1138 +
  1.1139 +		    end--;
  1.1140 +		    string--;
  1.1141 +		    if (*end != 'e' && *end != 'E') {
  1.1142 +			end--;
  1.1143 +			string--;
  1.1144 +		    }
  1.1145 +		}
  1.1146 +
  1.1147 +		/*
  1.1148 +		 * Scan the value from the temporary buffer.
  1.1149 +		 */
  1.1150 +
  1.1151 +		if (!(flags & SCAN_SUPPRESS)) {
  1.1152 +		    double dvalue;
  1.1153 +		    *end = '\0';
  1.1154 +		    dvalue = strtod(buf, NULL);
  1.1155 +		    objPtr = Tcl_NewDoubleObj(dvalue);
  1.1156 +		    Tcl_IncrRefCount(objPtr);
  1.1157 +		    objs[objIndex++] = objPtr;
  1.1158 +		}
  1.1159 +		break;
  1.1160 +	}
  1.1161 +	nconversions++;
  1.1162 +    }
  1.1163 +
  1.1164 +    done:
  1.1165 +    result = 0;
  1.1166 +    code = TCL_OK;
  1.1167 +
  1.1168 +    if (numVars) {
  1.1169 +	/*
  1.1170 +	 * In this case, variables were specified (classic scan)
  1.1171 +	 */
  1.1172 +	for (i = 0; i < totalVars; i++) {
  1.1173 +	    if (objs[i] != NULL) {
  1.1174 +		Tcl_Obj *tmpPtr;
  1.1175 +		
  1.1176 +		result++;
  1.1177 +		tmpPtr = Tcl_ObjSetVar2(interp, objv[i+3], NULL, objs[i], 0);
  1.1178 +		Tcl_DecrRefCount(objs[i]);
  1.1179 +		if (tmpPtr == NULL) {
  1.1180 +		    Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
  1.1181 +			    "couldn't set variable \"",
  1.1182 +			    Tcl_GetString(objv[i+3]), "\"", (char *) NULL);
  1.1183 +		    code = TCL_ERROR;
  1.1184 +		}
  1.1185 +	    }
  1.1186 +	}
  1.1187 +    } else {
  1.1188 +	/*
  1.1189 +	 * Here no vars were specified, we want a list returned (inline scan)
  1.1190 +	 */
  1.1191 +	objPtr = Tcl_NewObj();
  1.1192 +	for (i = 0; i < totalVars; i++) {
  1.1193 +	    if (objs[i] != NULL) {
  1.1194 +		Tcl_ListObjAppendElement(NULL, objPtr, objs[i]);
  1.1195 +		Tcl_DecrRefCount(objs[i]);
  1.1196 +	    } else {
  1.1197 +		/*
  1.1198 +		 * More %-specifiers than matching chars, so we
  1.1199 +		 * just spit out empty strings for these
  1.1200 +		 */
  1.1201 +		Tcl_ListObjAppendElement(NULL, objPtr, Tcl_NewObj());
  1.1202 +	    }
  1.1203 +	}
  1.1204 +    }
  1.1205 +    if (objs != NULL) {
  1.1206 +	ckfree((char*) objs);
  1.1207 +    }
  1.1208 +    if (code == TCL_OK) {
  1.1209 +	if (underflow && (nconversions == 0)) {
  1.1210 +	    if (numVars) {
  1.1211 +		objPtr = Tcl_NewIntObj(-1);
  1.1212 +	    } else {
  1.1213 +		if (objPtr) {
  1.1214 +		    Tcl_SetListObj(objPtr, 0, NULL);
  1.1215 +		} else {
  1.1216 +		    objPtr = Tcl_NewObj();
  1.1217 +		}
  1.1218 +	    }
  1.1219 +	} else if (numVars) {
  1.1220 +	    objPtr = Tcl_NewIntObj(result);
  1.1221 +	}
  1.1222 +	Tcl_SetObjResult(interp, objPtr);
  1.1223 +    }
  1.1224 +    return code;
  1.1225 +}