Tcl Source Code

2011-12-23  Jan Nijtmans  <[email protected]>

	* generic/tclUtf.c:     [Bug 3464428] string is graph \u0120 is wrong
	* generic/tclUniData.c:
	* generic/regc_locale.c:
	* tests/utf.test:
	* tools/uniParse.tcl:   clean up some unused stuff, and be more robust

.nf
\fBalpha\fR	A letter. 
\fBupper\fR	An upper-case letter. 
\fBlower\fR	A lower-case letter. 
\fBdigit\fR	A decimal digit. 
\fBxdigit\fR	A hexadecimal digit. 
\fBalnum\fR	An alphanumeric (letter or digit). 
\fBprint\fR	An alphanumeric (same as alnum).
\fBblank\fR	A space or tab character.
\fBspace\fR	A character producing white space in displayed text. 
\fBpunct\fR	A punctuation character. 
\fBgraph\fR	A character with a visible representation. 
\fBcntrl\fR	A control character. 
.fi
.RE

/*
 * Utility functions for handling cvecs
 * This file is #included by regcomp.c.
 *
 * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
 * 
 * Development of this software was funded, in part, by Cray Research Inc.,
 * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
 * Corporation, none of whom are responsible for the results.  The author
 * thanks all of them. 
 * 
 * Redistribution and use in source and binary forms -- with or without
 * modification -- are permitted for any purpose, provided that
 * redistributions in source form retain this entire copyright notice and
 * indicate the origin and nature of any modifications.
 * 
 * I'd appreciate being given credit for this package in the documentation
 * of software which uses it, but that is not a requirement.
 * 
 * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
 * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
 * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL
 * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
 * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
 * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,

{
    assert(cv->nranges < cv->rangespace);
    cv->ranges[cv->nranges*2] = (chr)from;
    cv->ranges[cv->nranges*2 + 1] = (chr)to;
    cv->nranges++;
}

/*
 - addmcce - add an MCCE to a cvec
 ^ static VOID addmcce(struct cvec *, chr *, chr *);
 */
static VOID
addmcce(cv, startp, endp)
    struct cvec *cv;			/* character vector */
    chr *startp;			/* beginning of text */
    chr *endp;				/* just past end of text */
{
    int len;
    int i;
    chr *s;
    chr *d;

    if (startp == NULL && endp == NULL) {
	return;
    }
    len = endp - startp;
    assert(len > 0);
    assert(cv->nchrs + len < cv->chrspace - cv->nmccechrs);
    assert(cv->nmcces < cv->mccespace);
    d = &cv->chrs[cv->chrspace - cv->nmccechrs - len - 1];
    cv->mcces[cv->nmcces++] = d;
    for (s = startp, i = len; i > 0; s++, i--) {
	*d++ = *s;
    }
    *d++ = 0;				/* endmarker */
    assert(d == &cv->chrs[cv->chrspace - cv->nmccechrs]);
    cv->nmccechrs += len + 1;
}

/*
 - haschr - does a cvec contain this chr?
 ^ static int haschr(struct cvec *, pchr);
 */
static int				/* predicate */
haschr(cv, c)
    struct cvec *cv;			/* character vector */

	}
    }
    return 0;
}

/*
 - getcvec - get a cvec, remembering it as v->cv
 ^ static struct cvec *getcvec(struct vars *, int, int, int);
 */
static struct cvec *
getcvec(v, nchrs, nranges, nmcces)
    struct vars *v;			/* context */
    int nchrs;				/* to hold this many chrs... */
    int nranges;			/* ... and this many ranges... */
    int nmcces;				/* ... and this many MCCEs */
{
    if (v->cv != NULL && nchrs <= v->cv->chrspace &&
	    nranges <= v->cv->rangespace && nmcces <= v->cv->mccespace) {
	return clearcvec(v->cv);
    }

    if (v->cv != NULL) {
	freecvec(v->cv);
    }
    v->cv = newcvec(nchrs, nranges, nmcces);
    if (v->cv == NULL) {
	ERR(REG_ESPACE);
    }

    return v->cv;
}

 *
 * See the file "license.terms" for information on usage and redistribution of
 * this file, and for a DISCLAIMER OF ALL WARRANTIES.
 */

/* ASCII character-name table */

static struct cname {
    CONST char *name;
    CONST char code;
} cnames[] = {
    {"NUL",		'\0'},
    {"SOH",		'\001'},
    {"STX",		'\002'},
    {"ETX",		'\003'},

    0x2e2f, 0x3005, 0x3006, 0x303b, 0x303c, 0xa62a, 0xa62b, 0xa790, 0xa791,
    0xa8fb, 0xa9cf, 0xaa7a, 0xaab1, 0xaab5, 0xaab6, 0xaac0, 0xaac2, 0xfb1d,
    0xfb3e, 0xfb40, 0xfb41, 0xfb43, 0xfb44
};

#define NUM_ALPHA_CHAR (sizeof(alphaCharTable)/sizeof(chr))


















/*
 * Unicode: decimal digit characters.
 */

static CONST crange digitRangeTable[] = {
    {0x0030, 0x0039}, {0x0660, 0x0669}, {0x06f0, 0x06f9}, {0x07c0, 0x07c9},
    {0x0966, 0x096f}, {0x09e6, 0x09ef}, {0x0a66, 0x0a6f}, {0x0ae6, 0x0aef},

#define NUM_UPPER_CHAR (sizeof(upperCharTable)/sizeof(chr))

/*
 * Unicode: unicode print characters excluding space.
 */

static CONST crange graphRangeTable[] = {
    {0x0021, 0x007e}, {0x00a0, 0x00ac}, {0x00ae, 0x0377}, {0x037a, 0x037e},
    {0x0384, 0x038a}, {0x038e, 0x03a1}, {0x03a3, 0x0527}, {0x0531, 0x0556},
    {0x0559, 0x055f}, {0x0561, 0x0587}, {0x0591, 0x05c7}, {0x05d0, 0x05ea},
    {0x05f0, 0x05f4}, {0x0606, 0x061b}, {0x061e, 0x06dc}, {0x06de, 0x070d},
    {0x0710, 0x074a}, {0x074d, 0x07b1}, {0x07c0, 0x07fa}, {0x0800, 0x082d},
    {0x0830, 0x083e}, {0x0840, 0x085b}, {0x0900, 0x0977}, {0x0979, 0x097f},
    {0x0981, 0x0983}, {0x0985, 0x098c}, {0x0993, 0x09a8}, {0x09aa, 0x09b0},
    {0x09b6, 0x09b9}, {0x09bc, 0x09c4}, {0x09cb, 0x09ce}, {0x09df, 0x09e3},

    {0x0ea1, 0x0ea3}, {0x0ead, 0x0eb9}, {0x0ebb, 0x0ebd}, {0x0ec0, 0x0ec4},
    {0x0ec8, 0x0ecd}, {0x0ed0, 0x0ed9}, {0x0f00, 0x0f47}, {0x0f49, 0x0f6c},
    {0x0f71, 0x0f97}, {0x0f99, 0x0fbc}, {0x0fbe, 0x0fcc}, {0x0fce, 0x0fda},
    {0x1000, 0x10c5}, {0x10d0, 0x10fc}, {0x1100, 0x1248}, {0x124a, 0x124d},
    {0x1250, 0x1256}, {0x125a, 0x125d}, {0x1260, 0x1288}, {0x128a, 0x128d},
    {0x1290, 0x12b0}, {0x12b2, 0x12b5}, {0x12b8, 0x12be}, {0x12c2, 0x12c5},
    {0x12c8, 0x12d6}, {0x12d8, 0x1310}, {0x1312, 0x1315}, {0x1318, 0x135a},
    {0x135d, 0x137c}, {0x1380, 0x1399}, {0x13a0, 0x13f4}, {0x1400, 0x169c},
    {0x16a0, 0x16f0}, {0x1700, 0x170c}, {0x170e, 0x1714}, {0x1720, 0x1736},
    {0x1740, 0x1753}, {0x1760, 0x176c}, {0x176e, 0x1770}, {0x1780, 0x17b3},
    {0x17b6, 0x17dd}, {0x17e0, 0x17e9}, {0x17f0, 0x17f9}, {0x1800, 0x180e},
    {0x1810, 0x1819}, {0x1820, 0x1877}, {0x1880, 0x18aa}, {0x18b0, 0x18f5},
    {0x1900, 0x191c}, {0x1920, 0x192b}, {0x1930, 0x193b}, {0x1944, 0x196d},
    {0x1970, 0x1974}, {0x1980, 0x19ab}, {0x19b0, 0x19c9}, {0x19d0, 0x19da},
    {0x19de, 0x1a1b}, {0x1a1e, 0x1a5e}, {0x1a60, 0x1a7c}, {0x1a7f, 0x1a89},
    {0x1a90, 0x1a99}, {0x1aa0, 0x1aad}, {0x1b00, 0x1b4b}, {0x1b50, 0x1b7c},
    {0x1b80, 0x1baa}, {0x1bae, 0x1bb9}, {0x1bc0, 0x1bf3}, {0x1bfc, 0x1c37},
    {0x1c3b, 0x1c49}, {0x1c4d, 0x1c7f}, {0x1cd0, 0x1cf2}, {0x1d00, 0x1de6},
    {0x1dfc, 0x1f15}, {0x1f18, 0x1f1d}, {0x1f20, 0x1f45}, {0x1f48, 0x1f4d},
    {0x1f50, 0x1f57}, {0x1f5f, 0x1f7d}, {0x1f80, 0x1fb4}, {0x1fb6, 0x1fc4},
    {0x1fc6, 0x1fd3}, {0x1fd6, 0x1fdb}, {0x1fdd, 0x1fef}, {0x1ff2, 0x1ff4},
    {0x1ff6, 0x1ffe}, {0x2000, 0x200a}, {0x2010, 0x2029}, {0x202f, 0x205f},
    {0x2074, 0x208e}, {0x2090, 0x209c}, {0x20a0, 0x20b9}, {0x20d0, 0x20f0},
    {0x2100, 0x2189}, {0x2190, 0x23f3}, {0x2400, 0x2426}, {0x2440, 0x244a},
    {0x2460, 0x26ff}, {0x2701, 0x27ca}, {0x27ce, 0x2b4c}, {0x2b50, 0x2b59},
    {0x2c00, 0x2c2e}, {0x2c30, 0x2c5e}, {0x2c60, 0x2cf1}, {0x2cf9, 0x2d25},
    {0x2d30, 0x2d65}, {0x2d7f, 0x2d96}, {0x2da0, 0x2da6}, {0x2da8, 0x2dae},
    {0x2db0, 0x2db6}, {0x2db8, 0x2dbe}, {0x2dc0, 0x2dc6}, {0x2dc8, 0x2dce},
    {0x2dd0, 0x2dd6}, {0x2dd8, 0x2dde}, {0x2de0, 0x2e31}, {0x2e80, 0x2e99},
    {0x2e9b, 0x2ef3}, {0x2f00, 0x2fd5}, {0x2ff0, 0x2ffb}, {0x3000, 0x303f},
    {0x3041, 0x3096}, {0x3099, 0x30ff}, {0x3105, 0x312d}, {0x3131, 0x318e},
    {0x3190, 0x31ba}, {0x31c0, 0x31e3}, {0x31f0, 0x321e}, {0x3220, 0x32fe},
    {0x3300, 0x4db5}, {0x4dc0, 0x9fcb}, {0xa000, 0xa48c}, {0xa490, 0xa4c6},
    {0xa4d0, 0xa62b}, {0xa640, 0xa673}, {0xa67c, 0xa697}, {0xa6a0, 0xa6f7},
    {0xa700, 0xa78e}, {0xa7a0, 0xa7a9}, {0xa7fa, 0xa82b}, {0xa830, 0xa839},
    {0xa840, 0xa877}, {0xa880, 0xa8c4}, {0xa8ce, 0xa8d9}, {0xa8e0, 0xa8fb},
    {0xa900, 0xa953}, {0xa95f, 0xa97c}, {0xa980, 0xa9cd}, {0xa9cf, 0xa9d9},

/*
 *	End of auto-generated Unicode character ranges declarations.
 */

#define	CH	NOCELT

/*
 - nmcces - how many distinct MCCEs are there?
 ^ static int nmcces(struct vars *);
 */
static int
nmcces(v)
    struct vars *v;			/* context */
{
    /*
     * No multi-character collating elements defined at the moment.
     */
    return 0;
}

/*
 - nleaders - how many chrs can be first chrs of MCCEs?
 ^ static int nleaders(struct vars *);
 */
static int
nleaders(v)
    struct vars *v;			/* context */
{
    return 0;
}

/*
 - allmcces - return a cvec with all the MCCEs of the locale
 ^ static struct cvec *allmcces(struct vars *, struct cvec *);
 */
static struct cvec *
allmcces(v, cv)
    struct vars *v;			/* context */
    struct cvec *cv;			/* this is supposed to have enough room */
{
    return clearcvec(cv);
}

/*
 - element - map collating-element name to celt
 ^ static celt element(struct vars *, CONST chr *, CONST chr *);
 */
static celt
element(v, startp, endp)
    struct vars *v;			/* context */

    if (a != b && !before(a, b)) {
	ERR(REG_ERANGE);
	return NULL;
    }

    if (!cases) {		/* easy version */
	cv = getcvec(v, 0, 1, 0);
	NOERRN();
	addrange(cv, a, b);
	return cv;
    }

    /*
     * When case-independent, it's hard to decide when cvec ranges are usable,
     * so for now at least, we won't try. We allocate enough space for two
     * case variants plus a little extra for the two title case variants.
     */

    nchrs = (b - a + 1)*2 + 4;

    cv = getcvec(v, nchrs, 0, 0);
    NOERRN();

    for (c=a; c<=b; c++) {
	addchr(cv, c);
	lc = Tcl_UniCharToLower((chr)c);
	uc = Tcl_UniCharToUpper((chr)c);
	tc = Tcl_UniCharToTitle((chr)c);

    struct cvec *cv;

    /*
     * Crude fake equivalence class for testing.
     */

    if ((v->cflags&REG_FAKE) && c == 'x') {
	cv = getcvec(v, 4, 0, 0);
	addchr(cv, (chr)'x');
	addchr(cv, (chr)'y');
	if (cases) {
	    addchr(cv, (chr)'X');
	    addchr(cv, (chr)'Y');
	}
	return cv;
    }

    /*
     * Otherwise, none.
     */

    if (cases) {
	return allcases(v, c);
    }
    cv = getcvec(v, 1, 0, 0);
    assert(cv != NULL);
    addchr(cv, (chr)c);
    return cv;
}

/*
 - cclass - supply cvec for a character class

     * Extract the class name
     */

    len = endp - startp;
    Tcl_DStringInit(&ds);
    np = Tcl_UniCharToUtfDString(startp, (int)len, &ds);

    /*
     * Remap lower and upper to alpha if the match is case insensitive.
     */

    if (cases && len == 5 && (strncmp("lower", np, 5) == 0
	    || strncmp("upper", np, 5) == 0)) {
	np = "alpha";
    }

    /*
     * Map the name to the corresponding enumerated value.
     */

    index = -1;
    for (namePtr=classNames,i=0 ; *namePtr!=NULL ; namePtr++,i++) {
	if ((strlen(*namePtr) == len) && (strncmp(*namePtr, np, len) == 0)) {
	    index = i;
	    break;
	}
    }
    Tcl_DStringFree(&ds);
    if (index == -1) {
	ERR(REG_ECTYPE);
	return NULL;
    }









    /*
     * Now compute the character class contents.
     */

    switch((enum classes) index) {
    case CC_PRINT:
    case CC_ALNUM:
	cv = getcvec(v, NUM_ALPHA_CHAR, NUM_DIGIT_RANGE + NUM_ALPHA_RANGE, 0);
	if (cv) {
	    for (i=0 ; (size_t)i<NUM_ALPHA_CHAR ; i++) {
		addchr(cv, alphaCharTable[i]);
	    }
	    for (i=0 ; (size_t)i<NUM_ALPHA_RANGE ; i++) {
		addrange(cv, alphaRangeTable[i].start,
			alphaRangeTable[i].end);
	    }
	    for (i=0 ; (size_t)i<NUM_DIGIT_RANGE ; i++) {
		addrange(cv, digitRangeTable[i].start,
			digitRangeTable[i].end);
	    }
	}
	break;
    case CC_ALPHA:
	cv = getcvec(v, NUM_ALPHA_CHAR, NUM_ALPHA_RANGE, 0);
	if (cv) {
	    for (i=0 ; (size_t)i<NUM_ALPHA_RANGE ; i++) {
		addrange(cv, alphaRangeTable[i].start,
			alphaRangeTable[i].end);
	    }
	    for (i=0 ; (size_t)i<NUM_ALPHA_CHAR ; i++) {
		addchr(cv, alphaCharTable[i]);
	    }
	}
	break;
    case CC_ASCII:
	cv = getcvec(v, 0, 1, 0);
	if (cv) {
	    addrange(cv, 0, 0x7f);
	}
	break;
    case CC_BLANK:
	cv = getcvec(v, 2, 0, 0);
	addchr(cv, '\t');
	addchr(cv, ' ');
	break;
    case CC_CNTRL:
	cv = getcvec(v, 0, 2, 0);
	addrange(cv, 0x0, 0x1f);

	addrange(cv, 0x7f, 0x9f);






	break;
    case CC_DIGIT:
	cv = getcvec(v, 0, NUM_DIGIT_RANGE, 0);
	if (cv) {
	    for (i=0 ; (size_t)i<NUM_DIGIT_RANGE ; i++) {
		addrange(cv, digitRangeTable[i].start,
			digitRangeTable[i].end);
	    }
	}
	break;
    case CC_PUNCT:
	cv = getcvec(v, NUM_PUNCT_CHAR, NUM_PUNCT_RANGE, 0);
	if (cv) {
	    for (i=0 ; (size_t)i<NUM_PUNCT_RANGE ; i++) {
		addrange(cv, punctRangeTable[i].start,
			punctRangeTable[i].end);
	    }
	    for (i=0 ; (size_t)i<NUM_PUNCT_CHAR ; i++) {
		addchr(cv, punctCharTable[i]);
	    }
	}
	break;
    case CC_XDIGIT:
	/*
	 * This is a 3 instead of (NUM_DIGIT_RANGE+2) because I've no idea how
	 * to define the digits 'a' through 'f' in non-western locales. The
	 * concept is quite possibly non portable, or only used in contextx
	 * where the characters used would be the western ones anyway!
	 * Whatever is actually the case, the number of ranges is fixed (until
	 * someone comes up with a better arrangement!)
	 */

	cv = getcvec(v, 0, 3, 0);
	if (cv) {
	    addrange(cv, '0', '9');
	    addrange(cv, 'a', 'f');
	    addrange(cv, 'A', 'F');
	}
	break;
    case CC_SPACE:
	cv = getcvec(v, NUM_SPACE_CHAR, NUM_SPACE_RANGE, 0);
	if (cv) {
	    for (i=0 ; (size_t)i<NUM_SPACE_RANGE ; i++) {
		addrange(cv, spaceRangeTable[i].start,
			spaceRangeTable[i].end);
	    }
	    for (i=0 ; (size_t)i<NUM_SPACE_CHAR ; i++) {
		addchr(cv, spaceCharTable[i]);
	    }
	}
	break;
    case CC_LOWER:
	cv  = getcvec(v, NUM_LOWER_CHAR, NUM_LOWER_RANGE, 0);
	if (cv) {
	    for (i=0 ; (size_t)i<NUM_LOWER_RANGE ; i++) {
		addrange(cv, lowerRangeTable[i].start,
			lowerRangeTable[i].end);
	    }
	    for (i=0 ; (size_t)i<NUM_LOWER_CHAR ; i++) {
		addchr(cv, lowerCharTable[i]);
	    }
	}
	break;
    case CC_UPPER:
	cv  = getcvec(v, NUM_UPPER_CHAR, NUM_UPPER_RANGE, 0);
	if (cv) {
	    for (i=0 ; (size_t)i<NUM_UPPER_RANGE ; i++) {
		addrange(cv, upperRangeTable[i].start,
			upperRangeTable[i].end);
	    }
	    for (i=0 ; (size_t)i<NUM_UPPER_CHAR ; i++) {
		addchr(cv, upperCharTable[i]);
	    }
	}
	break;



















    case CC_GRAPH:
	cv  = getcvec(v, NUM_GRAPH_CHAR, NUM_GRAPH_RANGE, 0);
	if (cv) {
	    for (i=0 ; (size_t)i<NUM_GRAPH_RANGE ; i++) {
		addrange(cv, graphRangeTable[i].start,
			graphRangeTable[i].end);
	    }
	    for (i=0 ; (size_t)i<NUM_GRAPH_CHAR ; i++) {
		addchr(cv, graphCharTable[i]);

    chr lc, uc, tc;

    lc = Tcl_UniCharToLower((chr)c);
    uc = Tcl_UniCharToUpper((chr)c);
    tc = Tcl_UniCharToTitle((chr)c);

    if (tc != uc) {
	cv = getcvec(v, 3, 0, 0);
	addchr(cv, tc);
    } else {
	cv = getcvec(v, 2, 0, 0);
    }
    addchr(cv, lc);
    if (lc != uc) {
	addchr(cv, uc);
    }
    return cv;
}

/*
 * re_*comp and friends - compile REs
 * This file #includes several others (see the bottom).
 *
 * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
 * 
 * Development of this software was funded, in part, by Cray Research Inc.,
 * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
 * Corporation, none of whom are responsible for the results.  The author
 * thanks all of them. 
 * 
 * Redistribution and use in source and binary forms -- with or without
 * modification -- are permitted for any purpose, provided that
 * redistributions in source form retain this entire copyright notice and
 * indicate the origin and nature of any modifications.
 * 
 * I'd appreciate being given credit for this package in the documentation
 * of software which uses it, but that is not a requirement.
 * 
 * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
 * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
 * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL
 * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
 * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
 * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,

static VOID word _ANSI_ARGS_((struct vars *, int, struct state *, struct state *));
static int scannum _ANSI_ARGS_((struct vars *));
static VOID repeat _ANSI_ARGS_((struct vars *, struct state *, struct state *, int, int));
static VOID bracket _ANSI_ARGS_((struct vars *, struct state *, struct state *));
static VOID cbracket _ANSI_ARGS_((struct vars *, struct state *, struct state *));
static VOID brackpart _ANSI_ARGS_((struct vars *, struct state *, struct state *));
static chr *scanplain _ANSI_ARGS_((struct vars *));
static VOID leaders _ANSI_ARGS_((struct vars *, struct cvec *));
static VOID onechr _ANSI_ARGS_((struct vars *, pchr, struct state *, struct state *));
static VOID dovec _ANSI_ARGS_((struct vars *, struct cvec *, struct state *, struct state *));
static celt nextleader _ANSI_ARGS_((struct vars *, pchr, pchr));
static VOID wordchrs _ANSI_ARGS_((struct vars *));
static struct subre *subre _ANSI_ARGS_((struct vars *, int, int, struct state *, struct state *));
static VOID freesubre _ANSI_ARGS_((struct vars *, struct subre *));
static VOID freesrnode _ANSI_ARGS_((struct vars *, struct subre *));

static VOID dumpcstate _ANSI_ARGS_((int, struct carc *, struct cnfa *, FILE *));
#endif
/* === regc_cvec.c === */
static struct cvec *newcvec _ANSI_ARGS_((int, int, int));
static struct cvec *clearcvec _ANSI_ARGS_((struct cvec *));
static VOID addchr _ANSI_ARGS_((struct cvec *, pchr));
static VOID addrange _ANSI_ARGS_((struct cvec *, pchr, pchr));
static VOID addmcce _ANSI_ARGS_((struct cvec *, chr *, chr *));
static int haschr _ANSI_ARGS_((struct cvec *, pchr));
static struct cvec *getcvec _ANSI_ARGS_((struct vars *, int, int, int));
static VOID freecvec _ANSI_ARGS_((struct cvec *));
/* === regc_locale.c === */
static int nmcces _ANSI_ARGS_((struct vars *));
static int nleaders _ANSI_ARGS_((struct vars *));
static struct cvec *allmcces _ANSI_ARGS_((struct vars *, struct cvec *));
static celt element _ANSI_ARGS_((struct vars *, CONST chr *, CONST chr *));
static struct cvec *range _ANSI_ARGS_((struct vars *, celt, celt, int));
static int before _ANSI_ARGS_((celt, celt));
static struct cvec *eclass _ANSI_ARGS_((struct vars *, celt, int));
static struct cvec *cclass _ANSI_ARGS_((struct vars *, CONST chr *, CONST chr *, int));
static struct cvec *allcases _ANSI_ARGS_((struct vars *, pchr));
static int cmp _ANSI_ARGS_((CONST chr *, CONST chr *, size_t));

	g->nlacons = 0;
	ZAPCNFA(g->search);
	v->nfa = newnfa(v, v->cm, (struct nfa *)NULL);
	CNOERR();
	v->cv = newcvec(100, 20, 10);
	if (v->cv == NULL)
		return freev(v, REG_ESPACE);
	i = nmcces(v);
	if (i > 0) {
		v->mcces = newcvec(nleaders(v), 0, i);
		CNOERR();
		v->mcces = allmcces(v, v->mcces);
		leaders(v, v->mcces);
		addmcce(v->mcces, (chr *)NULL, (chr *)NULL);	/* dummy */
	}
	CNOERR();

	/* parsing */
	lexstart(v);			/* also handles prefixes */
	if ((v->cflags&REG_NLSTOP) || (v->cflags&REG_NLANCH)) {
		/* assign newline a unique color */
		v->nlcolor = subcolor(v->cm, newline());

	delsub(v->nfa, left, right);
	assert(left->nouts == 0);
	freestate(v->nfa, left);
	assert(right->nins == 0);
	freestate(v->nfa, right);
}
			
/*
 - brackpart - handle one item (or range) within a bracket expression
 ^ static VOID brackpart(struct vars *, struct state *, struct state *);
 */
static VOID
brackpart(v, lp, rp)
struct vars *v;

	assert(SEE(END) || ISERR());
	NEXT();

	return endp;
}

/*
 - leaders - process a cvec of collating elements to also include leaders
 * Also gives all characters involved their own colors, which is almost
 * certainly necessary, and sets up little disconnected subNFA.
 ^ static VOID leaders(struct vars *, struct cvec *);
 */
static VOID
leaders(v, cv)
struct vars *v;
struct cvec *cv;
{
	int mcce;
	chr *p;
	chr leader;
	struct state *s;
	struct arc *a;

	v->mccepbegin = newstate(v->nfa);
	v->mccepend = newstate(v->nfa);
	NOERR();

	for (mcce = 0; mcce < cv->nmcces; mcce++) {
		p = cv->mcces[mcce];
		leader = *p;
		if (!haschr(cv, leader)) {
			addchr(cv, leader);
			s = newstate(v->nfa);
			newarc(v->nfa, PLAIN, subcolor(v->cm, leader),
							v->mccepbegin, s);
			okcolors(v->nfa, v->cm);
		} else {
			a = findarc(v->mccepbegin, PLAIN,
						GETCOLOR(v->cm, leader));
			assert(a != NULL);
			s = a->to;
			assert(s != v->mccepend);
		}
		p++;
		assert(*p != 0 && *(p+1) == 0);	/* only 2-char MCCEs for now */
		newarc(v->nfa, PLAIN, subcolor(v->cm, *p), s, v->mccepend);
		okcolors(v->nfa, v->cm);
	}
}

/*
 - onechr - fill in arcs for a plain character, and possible case complements
 * This is mostly a shortcut for efficient handling of the common case.
 ^ static VOID onechr(struct vars *, pchr, struct state *, struct state *);
 */
static VOID
onechr(v, c, lp, rp)

	color co;
	struct cvec *leads;
	struct arc *a;
	struct arc *pa;		/* arc in prototype */
	struct state *s;
	struct state *ps;	/* state in prototype */

	/* need a place to store leaders, if any */
	if (nmcces(v) > 0) {
		assert(v->mcces != NULL);
		if (v->cv2 == NULL || v->cv2->nchrs < v->mcces->nchrs) {
			if (v->cv2 != NULL)
				free(v->cv2);
			v->cv2 = newcvec(v->mcces->nchrs, 0, v->mcces->nmcces);
			NOERR();
			leads = v->cv2;
		} else
			leads = clearcvec(v->cv2);
	} else
		leads = NULL;

	/* first, get the ordinary characters out of the way */
	for (p = cv->chrs, i = cv->nchrs; i > 0; p++, i--) {
		ch = *p;
		if (!ISCELEADER(v, ch))
			newarc(v->nfa, PLAIN, subcolor(v->cm, ch), lp, rp);
		else {

	}
	g = (struct guts *)re->re_guts;
	if (g->magic != GUTSMAGIC)
		fprintf(f, "bad guts magic number (0x%x not 0x%x)\n", g->magic,
								GUTSMAGIC);

	fprintf(f, "\n\n\n========= DUMP ==========\n");
	fprintf(f, "nsub %d, info 0%lo, csize %d, ntree %d\n", 
		re->re_nsub, re->re_info, re->re_csize, g->ntree);

	dumpcolors(&g->cmap, f);
	if (!NULLCNFA(g->search)) {
		printf("\nsearch:\n");
		dumpcnfa(&g->search, f);
	}

#include "tclUniData.c"

/*
 * The following macros are used for fast character category tests.  The
 * x_BITS values are shifted right by the category value to determine whether
 * the given category is included in the set.
 */ 

#define ALPHA_BITS ((1 << UPPERCASE_LETTER) | (1 << LOWERCASE_LETTER) \
    | (1 << TITLECASE_LETTER) | (1 << MODIFIER_LETTER) | (1 << OTHER_LETTER))



#define DIGIT_BITS (1 << DECIMAL_DIGIT_NUMBER)

#define SPACE_BITS ((1 << SPACE_SEPARATOR) | (1 << LINE_SEPARATOR) \
    | (1 << PARAGRAPH_SEPARATOR))

#define CONNECTOR_BITS (1 << CONNECTOR_PUNCTUATION)

#define PRINT_BITS (ALPHA_BITS | DIGIT_BITS | SPACE_BITS | \
	    (1 << NON_SPACING_MARK) | (1 << ENCLOSING_MARK) | \
	    (1 << COMBINING_SPACING_MARK) | (1 << LETTER_NUMBER) | \
	    (1 << OTHER_NUMBER) | (1 << CONNECTOR_PUNCTUATION) | \
	    (1 << DASH_PUNCTUATION) | (1 << OPEN_PUNCTUATION) | \
	    (1 << CLOSE_PUNCTUATION) | (1 << INITIAL_QUOTE_PUNCTUATION) | \
	    (1 << FINAL_QUOTE_PUNCTUATION) | (1 << OTHER_PUNCTUATION) | \
	    (1 << MATH_SYMBOL) | (1 << CURRENCY_SYMBOL) | \
	    (1 << MODIFIER_SYMBOL) | (1 << OTHER_SYMBOL))

#define PUNCT_BITS ((1 << CONNECTOR_PUNCTUATION) | \
	    (1 << DASH_PUNCTUATION) | (1 << OPEN_PUNCTUATION) | \
	    (1 << CLOSE_PUNCTUATION) | (1 << INITIAL_QUOTE_PUNCTUATION) | \
	    (1 << FINAL_QUOTE_PUNCTUATION) | (1 << OTHER_PUNCTUATION))








/*
 * Unicode characters less than this value are represented by themselves 
 * in UTF-8 strings. 
 */

#define UNICODE_SELF	0x80

/*
 * The following structures are used when mapping between Unicode (UCS-2)
 * and UTF-8.

 *	The return values is the number of bytes in the Utf character "ch".
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */
 
INLINE static int
UtfCount(ch)
    int ch;			/* The Tcl_UniChar whose size is returned. */
{
    if ((ch > 0) && (ch < UNICODE_SELF)) {
	return 1;
    }

 * Tcl_UniCharToUtf --
 *
 *	Store the given Tcl_UniChar as a sequence of UTF-8 bytes in the
 *	provided buffer.  Equivalent to Plan 9 runetochar().
 *
 * Results:
 *	The return values is the number of bytes in the buffer that
 *	were consumed.  
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */
 
INLINE int
Tcl_UniCharToUtf(ch, str)
    int ch;			/* The Tcl_UniChar to be stored in the
				 * buffer. */
    char *str;			/* Buffer in which the UTF-8 representation
				 * of the Tcl_UniChar is stored.  Buffer must
				 * be large enough to hold the UTF-8 character

 *	end of dsPtr.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */
 
char *
Tcl_UniCharToUtfDString(wString, numChars, dsPtr)
    CONST Tcl_UniChar *wString;	/* Unicode string to convert to UTF-8. */
    int numChars;		/* Length of Unicode string in Tcl_UniChars
				 * (must be >= 0). */
    Tcl_DString *dsPtr;		/* UTF-8 representation of string is
				 * appended to this previously initialized

 *	number of bytes from the UTF-8 string that were consumed.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */
 
int
Tcl_UtfToUniChar(str, chPtr)
    register CONST char *str;	 /* The UTF-8 string. */
    register Tcl_UniChar *chPtr; /* Filled with the Tcl_UniChar represented
				  * by the UTF-8 string. */
{
    register int byte;
    
    /*
     * Unroll 1 to 3 byte UTF-8 sequences, use loop to handle longer ones.
     */

    byte = *((unsigned char *) str);
    if (byte < 0xC0) {
	/*

	return 1;
    } else if (byte < 0xF0) {
	if (((str[1] & 0xC0) == 0x80) && ((str[2] & 0xC0) == 0x80)) {
	    /*
	     * Three-byte-character lead byte followed by two trail bytes.
	     */

	    *chPtr = (Tcl_UniChar) (((byte & 0x0F) << 12) 
		    | ((str[1] & 0x3F) << 6) | (str[2] & 0x3F));
	    return 3;
	}
	/*
	 * A three-byte-character lead-byte not followed by two trail-bytes
	 * represents itself.
	 */

 * Tcl_NumUtfChars --
 *
 *	Returns the number of characters (not bytes) in the UTF-8 string,
 *	not including the terminating NULL byte.  This is equivalent to
 *	Plan 9 utflen() and utfnlen().
 *
 * Results:
 *	As above.  
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */
 
int 
Tcl_NumUtfChars(str, len)
    register CONST char *str;	/* The UTF-8 string to measure. */
    int len;			/* The length of the string in bytes, or -1
				 * for strlen(string). */
{
    Tcl_UniChar ch;
    register Tcl_UniChar *chPtr = &ch;

CONST char *
Tcl_UtfFindFirst(string, ch)
    CONST char *string;		/* The UTF-8 string to be searched. */
    int ch;			/* The Tcl_UniChar to search for. */
{
    int len;
    Tcl_UniChar find;
    
    while (1) {
	len = TclUtfToUniChar(string, &find);
	if (find == ch) {
	    return string;
	}
	if (*string == '\0') {
	    return NULL;

Tcl_UtfFindLast(string, ch)
    CONST char *string;		/* The UTF-8 string to be searched. */
    int ch;			/* The Tcl_UniChar to search for. */
{
    int len;
    Tcl_UniChar find;
    CONST char *last;
	
    last = NULL;
    while (1) {
	len = TclUtfToUniChar(string, &find);
	if (find == ch) {
	    last = string;
	}
	if (*string == '\0') {

 *	the UTF-8 string.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */
 
CONST char *
Tcl_UtfNext(str) 
    CONST char *str;		    /* The current location in the string. */
{
    Tcl_UniChar ch;

    return str + TclUtfToUniChar(str, &ch);
}


    CONST char *str;		    /* The current location in the string. */
    CONST char *start;		    /* Pointer to the beginning of the
				     * string, to avoid going backwards too
				     * far. */
{
    CONST char *look;
    int i, byte;
    
    str--;
    look = str;
    for (i = 0; i < TCL_UTF_MAX; i++) {
	if (look < start) {
	    if (str < start) {
		str = start;
	    }
	    break;
	}
	byte = *((unsigned char *) look);
	if (byte < 0x80) {
	    break;
	}
	if (byte >= 0xC0) {
	    return look;
	}
	look--;
    }
    return str;
}
	
/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UniCharAtIndex --
 *
 *	Returns the Unicode character represented at the specified
 *	character (not byte) position in the UTF-8 string.
 *
 * Results:
 *	As above.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */
 
Tcl_UniChar
Tcl_UniCharAtIndex(src, index)
    register CONST char *src;	/* The UTF-8 string to dereference. */
    register int index;		/* The position of the desired character. */
{
    Tcl_UniChar ch;

CONST char *
Tcl_UtfAtIndex(src, index)
    register CONST char *src;	/* The UTF-8 string. */
    register int index;		/* The position of the desired character. */
{
    Tcl_UniChar ch;
    
    while (index > 0) {
	index--;
	src += TclUtfToUniChar(src, &ch);
    }
    return src;
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UtfBackslash --
 *
 *	Figure out how to handle a backslash sequence.
 *
 * Results:
 *	Stores the bytes represented by the backslash sequence in dst and
 *	returns the number of bytes written to dst.  At most TCL_UTF_MAX
 *	bytes are written to dst; dst must have been large enough to accept
 *	those bytes.  If readPtr isn't NULL then it is filled in with a
 *	count of the number of bytes in the backslash sequence.  
 *
 * Side effects:
 *	The maximum number of bytes it takes to represent a Unicode
 *	character in UTF-8 is guaranteed to be less than the number of
 *	bytes used to express the backslash sequence that represents
 *	that Unicode character.  If the target buffer into which the
 *	caller is going to store the bytes that represent the Unicode

	upChar = Tcl_UniCharToUpper(ch);

	/*
	 * To keep badly formed Utf strings from getting inflated by
	 * the conversion (thereby causing a segfault), only copy the
	 * upper case char to dst if its size is <= the original char.
	 */
	
	if (bytes < UtfCount(upChar)) {
	    memcpy(dst, src, (size_t) bytes);
	    dst += bytes;
	} else {
	    dst += Tcl_UniCharToUtf(upChar, dst);
	}
	src += bytes;

int
Tcl_UtfToLower(str)
    char *str;			/* String to convert in place. */
{
    Tcl_UniChar ch, lowChar;
    char *src, *dst;
    int bytes;
    
    /*
     * Iterate over the string until we hit the terminating null.
     */

    src = dst = str;
    while (*src) {
	bytes = TclUtfToUniChar(src, &ch);
	lowChar = Tcl_UniCharToLower(ch);

	/*
	 * To keep badly formed Utf strings from getting inflated by
	 * the conversion (thereby causing a segfault), only copy the
	 * lower case char to dst if its size is <= the original char.
	 */
	
	if (bytes < UtfCount(lowChar)) {
	    memcpy(dst, src, (size_t) bytes);
	    dst += bytes;
	} else {
	    dst += Tcl_UniCharToUtf(lowChar, dst);
	}
	src += bytes;

int
Tcl_UtfToTitle(str)
    char *str;			/* String to convert in place. */
{
    Tcl_UniChar ch, titleChar, lowChar;
    char *src, *dst;
    int bytes;
    
    /*
     * Capitalize the first character and then lowercase the rest of the
     * characters until we get to a null.
     */

    src = dst = str;

 */

int
Tcl_UniCharLen(str)
    CONST Tcl_UniChar *str;	/* Unicode string to find length of. */
{
    int len = 0;
    
    while (*str != '\0') {
	len++;
	str++;
    }
    return len;
}


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

int
Tcl_UniCharIsAlnum(ch)
    int ch;			/* Unicode character to test. */
{
    register int category = (GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK);

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

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

int
Tcl_UniCharIsAlpha(ch)
    int ch;			/* Unicode character to test. */
{
    register int category = (GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK);
    return ((ALPHA_BITS >> category) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsControl --
 *

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

int
Tcl_UniCharIsControl(ch)
    int ch;			/* Unicode character to test. */
{
    return ((GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK) == CONTROL);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsDigit --
 *

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

int
Tcl_UniCharIsDigit(ch)
    int ch;			/* Unicode character to test. */
{
    return ((GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK)
	    == DECIMAL_DIGIT_NUMBER);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsGraph --
 *

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

int
Tcl_UniCharIsGraph(ch)
    int ch;			/* Unicode character to test. */
{
    register int category = (GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK);
    return (((PRINT_BITS >> category) & 1) && (ch != ' '));
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsLower --
 *

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

int
Tcl_UniCharIsLower(ch)
    int ch;			/* Unicode character to test. */
{
    return ((GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK) == LOWERCASE_LETTER);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsPrint --
 *

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

int
Tcl_UniCharIsPrint(ch)
    int ch;			/* Unicode character to test. */
{
    register int category = (GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK);
    return ((PRINT_BITS >> category) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsPunct --
 *

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

int
Tcl_UniCharIsPunct(ch)
    int ch;			/* Unicode character to test. */
{
    register int category = (GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK);
    return ((PUNCT_BITS >> category) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsSpace --
 *

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

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

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

    if (ch < 0x80) {
	return isspace(UCHAR(ch)); /* INTL: ISO space */
    } else {
	category = (GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK);
	return ((SPACE_BITS >> category) & 1);
    }
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsUpper --

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

int
Tcl_UniCharIsUpper(ch)
    int ch;			/* Unicode character to test. */
{
    return ((GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK) == UPPERCASE_LETTER);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsWordChar --
 *

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

int
Tcl_UniCharIsWordChar(ch)
    int ch;			/* Unicode character to test. */
{
    register int category = (GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK);

    return (((ALPHA_BITS | DIGIT_BITS | CONNECTOR_BITS) >> category) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharCaseMatch --
 *

Tcl_UniCharCaseMatch(string, pattern, nocase)
    CONST Tcl_UniChar *string;	/* Unicode String. */
    CONST Tcl_UniChar *pattern;	/* Pattern, which may contain special
				 * characters. */
    int nocase;			/* 0 for case sensitive, 1 for insensitive */
{
    Tcl_UniChar ch1, p;
    
    while (1) {
	p = *pattern;
	
	/*
	 * See if we're at the end of both the pattern and the string.  If
	 * so, we succeeded.  If we're at the end of the pattern but not at
	 * the end of the string, we failed.
	 */
	
	if (p == 0) {
	    return (*string == 0);
	}
	if ((*string == 0) && (p != '*')) {
	    return 0;
	}

	/*
	 * Check for a "*" as the next pattern character.  It matches any
	 * substring.  We handle this by skipping all the characters up to the
	 * next matching one in the pattern, and then calling ourselves
	 * recursively for each postfix of string, until either we match or we
	 * reach the end of the string.
	 */
	
	if (p == '*') {
	    /*
	     * Skip all successive *'s in the pattern
	     */
	    while (*(++pattern) == '*') {}
	    p = *pattern;
	    if (p == 0) {

	}

	/*
	 * Check for a "[" as the next pattern character.  It is followed
	 * by a list of characters that are acceptable, or by a range
	 * (two characters separated by "-").
	 */
	
	if (p == '[') {
	    Tcl_UniChar startChar, endChar;

	    pattern++;
	    ch1 = (nocase ? Tcl_UniCharToLower(*string) : *string);
	    string++;
	    while (1) {

	/*
	 * Check for a "*" as the next pattern character.  It matches any
	 * substring.  We handle this by skipping all the characters up to the
	 * next matching one in the pattern, and then calling ourselves
	 * recursively for each postfix of string, until either we match or we
	 * reach the end of the string.
	 */
	
	if (p == '*') {
	    /*
	     * Skip all successive *'s in the pattern
	     */
	    while (*(++pattern) == '*') {}
	    if (pattern == patternEnd) {
		return 1;

	}

	/*
	 * Check for a "[" as the next pattern character.  It is followed
	 * by a list of characters that are acceptable, or by a range
	 * (two characters separated by "-").
	 */
	
	if (p == '[') {
	    Tcl_UniChar ch1, startChar, endChar;

	    pattern++;
	    ch1 = (nocase ? Tcl_UniCharToLower(*string) : *string);
	    string++;
	    while (1) {

test utf-21.4 {TclUniCharIsGraph} {
    # [Bug 3464428]
    string is graph \u0120
} {1}
test utf-21.5 {unicode graph char in regc_locale.c} {
    # [Bug 3464428]
    regexp {^[[:graph:]]+$} \u0120




























} {1}

test utf-22.1 {TclUniCharIsWordChar} {
    string wordend "xyz123_bar fg" 0
} 10
test utf-22.2 {TclUniCharIsWordChar} {
    string wordend "x\u5080z123_bar\u203c fg" 0

 *	is automatically generated by the tools/uniClass.tcl script
 *	and used in generic/regc_locale.c.  Do not modify by hand.
 */
"

foreach {type desc} {
    alpha "alphabetic characters"

    digit "decimal digit characters"
    punct "punctuation characters"
    space "white space characters"
    lower "lowercase characters"
    upper "uppercase characters"
    graph "unicode print characters excluding space"
} {