regcomp.c

regex-spencer-3.8-src.zip

				count2 = INFINITY;
		} else		/* just a single number */
			count2 = count;
		repeat(p, pos, count, count2);
		if (!EATTWO('\\', '}')) {	/* error heuristics */
			while (MORE() && !SEETWO('\\', '}'))
				NEXT();
			REQUIRE(MORE(), REG_EBRACE);
			SETERROR(REG_BADBR);
		}
	} else if (c == (unsigned char)'$')	/* $ (but not \$) ends it */
		return(1);

	return(0);
}

/*
 - p_count - parse a repetition count
 == static int p_count(register struct parse *p);
 */
static int			/* the value */
p_count(p)
register struct parse *p;
{
	register int count = 0;
	register int ndigits = 0;

	while (MORE() && isdigit(PEEK()) && count <= DUPMAX) {
		count = count*10 + (GETNEXT() - '0');
		ndigits++;
	}

	REQUIRE(ndigits > 0 && count <= DUPMAX, REG_BADBR);
	return(count);
}

/*
 - p_bracket - parse a bracketed character list
 == static void p_bracket(register struct parse *p);
 *
 * Note a significant property of this code:  if the allocset() did SETERROR,
 * no set operations are done.
 */
static void
p_bracket(p)
register struct parse *p;
{
	register cset *cs = allocset(p);
	register int invert = 0;

	/* Dept of Truly Sickening Special-Case Kludges */
	if (p->next + 5 < p->end && strncmp(p->next, "[:<:]]", 6) == 0) {
		EMIT(OBOW, 0);
		NEXTn(6);
		return;
	}
	if (p->next + 5 < p->end && strncmp(p->next, "[:>:]]", 6) == 0) {
		EMIT(OEOW, 0);
		NEXTn(6);
		return;
	}

	if (EAT('^'))
		invert++;	/* make note to invert set at end */
	if (EAT(']'))
		CHadd(cs, ']');
	else if (EAT('-'))
		CHadd(cs, '-');
	while (MORE() && PEEK() != ']' && !SEETWO('-', ']'))
		p_b_term(p, cs);
	if (EAT('-'))
		CHadd(cs, '-');
	MUSTEAT(']', REG_EBRACK);

	if (p->error != 0)	/* don't mess things up further */
		return;

	if (p->g->cflags&REG_ICASE) {
		register int i;
		register int ci;

		for (i = p->g->csetsize - 1; i >= 0; i--)
			if (CHIN(cs, i) && isalpha(i)) {
				ci = othercase(i);
				if (ci != i)
					CHadd(cs, ci);
			}
		if (cs->multis != NULL)
			mccase(p, cs);
	}
	if (invert) {
		register int i;

		for (i = p->g->csetsize - 1; i >= 0; i--)
			if (CHIN(cs, i))
				CHsub(cs, i);
			else
				CHadd(cs, i);
		if (p->g->cflags&REG_NEWLINE)
			CHsub(cs, '\n');
		if (cs->multis != NULL)
			mcinvert(p, cs);
	}

	assert(cs->multis == NULL);		/* xxx */

	if (nch(p, cs) == 1) {		/* optimize singleton sets */
		ordinary(p, firstch(p, cs));
		freeset(p, cs);
	} else
		EMIT(OANYOF, freezeset(p, cs));
}

/*
 - p_b_term - parse one term of a bracketed character list
 == static void p_b_term(register struct parse *p, register cset *cs);
 */
static void
p_b_term(p, cs)
register struct parse *p;
register cset *cs;
{
	register char c;
	register char start, finish;
	register int i;

	/* classify what we've got */
	switch ((MORE()) ? PEEK() : '\0') {
	case '[':
		c = (MORE2()) ? PEEK2() : '\0';
		break;
	case '-':
		SETERROR(REG_ERANGE);
		return;			/* NOTE RETURN */
		break;
	default:
		c = '\0';
		break;
	}

	switch (c) {
	case ':':		/* character class */
		NEXT2();
		REQUIRE(MORE(), REG_EBRACK);
		c = PEEK();
		REQUIRE(c != '-' && c != ']', REG_ECTYPE);
		p_b_cclass(p, cs);
		REQUIRE(MORE(), REG_EBRACK);
		REQUIRE(EATTWO(':', ']'), REG_ECTYPE);
		break;
	case '=':		/* equivalence class */
		NEXT2();
		REQUIRE(MORE(), REG_EBRACK);
		c = PEEK();
		REQUIRE(c != '-' && c != ']', REG_ECOLLATE);
		p_b_eclass(p, cs);
		REQUIRE(MORE(), REG_EBRACK);
		REQUIRE(EATTWO('=', ']'), REG_ECOLLATE);
		break;
	default:		/* symbol, ordinary character, or range */
/* xxx revision needed for multichar stuff */
		start = p_b_symbol(p);
		if (SEE('-') && MORE2() && PEEK2() != ']') {
			/* range */
			NEXT();
			if (EAT('-'))
				finish = '-';
			else
				finish = p_b_symbol(p);
		} else
			finish = start;
/* xxx what about signed chars here... */
		REQUIRE(start <= finish, REG_ERANGE);
		for (i = start; i <= finish; i++)
			CHadd(cs, i);
		break;
	}
}

/*
 - p_b_cclass - parse a character-class name and deal with it
 == static void p_b_cclass(register struct parse *p, register cset *cs);
 */
static void
p_b_cclass(p, cs)
register struct parse *p;
register cset *cs;
{
	register char *sp = p->next;
	register struct cclass *cp;
	register size_t len;
	register char *u;
	register char c;

	while (MORE() && isalpha(PEEK()))
		NEXT();
	len = p->next - sp;
	for (cp = cclasses; cp->name != NULL; cp++)
		if (strncmp(cp->name, sp, len) == 0 && cp->name[len] == '\0')
			break;
	if (cp->name == NULL) {
		/* oops, didn't find it */
		SETERROR(REG_ECTYPE);
		return;
	}

	u = cp->chars;
	while ((c = *u++) != '\0')
		CHadd(cs, c);
	for (u = cp->multis; *u != '\0'; u += strlen(u) + 1)
		MCadd(p, cs, u);
}

/*
 - p_b_eclass - parse an equivalence-class name and deal with it
 == static void p_b_eclass(register struct parse *p, register cset *cs);
 *
 * This implementation is incomplete. xxx
 */
static void
p_b_eclass(p, cs)
register struct parse *p;
register cset *cs;
{
	register char c;

	c = p_b_coll_elem(p, '=');
	CHadd(cs, c);
}

/*
 - p_b_symbol - parse a character or [..]ed multicharacter collating symbol
 == static char p_b_symbol(register struct parse *p);
 */
static char			/* value of symbol */
p_b_symbol(p)
register struct parse *p;
{
	register char value;

	REQUIRE(MORE(), REG_EBRACK);
	if (!EATTWO('[', '.'))
		return(GETNEXT());

	/* collating symbol */
	value = p_b_coll_elem(p, '.');
	REQUIRE(EATTWO('.', ']'), REG_ECOLLATE);
	return(value);
}

/*
 - p_b_coll_elem - parse a collating-element name and look it up
 == static char p_b_coll_elem(register struct parse *p, int endc);
 */
static char			/* value of collating element */
p_b_coll_elem(p, endc)
register struct parse *p;
int endc;			/* name ended by endc,']' */
{
	register char *sp = p->next;
	register struct cname *cp;
	register int len;

	while (MORE() && !SEETWO(endc, ']'))
		NEXT();
	if (!MORE()) {
		SETERROR(REG_EBRACK);
		return(0);
	}
	len = p->next - sp;
	for (cp = cnames; cp->name != NULL; cp++)
		if (strncmp(cp->name, sp, len) == 0 && cp->name[len] == '\0')
			return(cp->code);	/* known name */
	if (len == 1)
		return(*sp);	/* single character */
	SETERROR(REG_ECOLLATE);			/* neither */
	return(0);
}

/*
 - othercase - return the case counterpart of an alphabetic
 == static char othercase(int ch);
 */
static char			/* if no counterpart, return ch */
othercase(ch)
int ch;
{
	assert(isalpha(ch));
	if (isupper(ch))
		return(tolower(ch));
	else if (islower(ch))
		return(toupper(ch));
	else			/* peculiar, but could happen */
		return(ch);
}

/*
 - bothcases - emit a dualcase version of a two-case character
 == static void bothcases(register struct parse *p, int ch);
 *
 * Boy, is this implementation ever a kludge...
 */
static void
bothcases(p, ch)
register struct parse *p;
int ch;
{
	register char *oldnext = p->next;
	register char *oldend = p->end;
	char bracket[3];

	assert(othercase(ch) != ch);	/* p_bracket() would recurse */
	p->next = bracket;
	p->end = bracket+2;
	bracket[0] = ch;
	bracket[1] = ']';
	bracket[2] = '\0';
	p_bracket(p);
	assert(p->next == bracket+2);
	p->next = oldnext;
	p->end = oldend;
}

/*
 - ordinary - emit an ordinary character
 == static void ordinary(register struct parse *p, register int ch);
 */
static void
ordinary(p, ch)
register struct parse *p;
register int ch;
{
	register cat_t *cap = p->g->categories;

	if ((p->g->cflags&REG_ICASE) && isalpha(ch) && othercase(ch) != ch)
		bothcases(p, ch);
	else {
		EMIT(OCHAR, (unsigned char)ch);
		if (cap[ch] == 0)
			cap[ch] = p->g->ncategories++;
	}
}

/*
 - nonnewline - emit REG_NEWLINE version of OANY
 == static void nonnewline(register struct parse *p);
 *
 * Boy, is this implementation ever a kludge...
 */
static void
nonnewline(p)
register struct parse *p;
{
	register char *oldnext = p->next;
	register char *oldend = p->end;
	char bracket[4];

	p->next = bracket;
	p->end = bracket+3;
	bracket[0] = '^';
	bracket[1] = '\n';
	bracket[2] = ']';
	bracket[3] = '\0';
	p_bracket(p);
	assert(p->next == bracket+3);
	p->next = oldnext;
	p->end = oldend;
}

/*
 - repeat - generate code for a bounded repetition, recursively if needed
 == static void repeat(register struct parse *p, sopno start, int from, int to);
 */
static void
repeat(p, start, from, to)
register struct parse *p;
sopno start;			/* operand from here to end of strip */
int from;			/* repeated from this number */
int to;				/* to this number of times (maybe INFINITY) */
{
	register sopno finish = HERE();
#	define	N	2
#	define	INF	3
#	define	REP(f, t)	((f)*8 + (t))
#	define	MAP(n)	(((n) <= 1) ? (n) : ((n) == INFINITY) ? INF : N)
	register sopno copy;

	if (p->error != 0)	/* head off possible runaway recursion */
		return;

	assert(from <= to);

	switch (REP(MAP(from), MAP(to))) {
	case REP(0, 0):			/* must be user doing this */
		DROP(finish-start);	/* drop the operand */
		break;
	case REP(0, 1):			/* as x{1,1}? */
	case REP(0, N):			/* as x{1,n}? */
	case REP(0, INF):		/* as x{1,}? */
		/* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
		INSERT(OCH_, start);		/* offset is wrong... */
		repeat(p, start+1, 1, to);
		ASTERN(OOR1, start);
		AHEAD(start);			/* ... fix it */
		EMIT(OOR2, 0);
		AHEAD(THERE());
		ASTERN(O_CH, THERETHERE());
		break;
	case REP(1, 1):			/* trivial case */
		/* done */
		break;
	case REP(1, N):			/* as x?x{1,n-1} */
		/* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
		INSERT(OCH_, start);
		ASTERN(OOR1, start);
		AHEAD(start);
		EMIT(OOR2, 0);			/* offset very wrong... */
		AHEAD(THERE());			/* ...so fix it */
		ASTERN(O_CH, THERETHERE());
		copy = dupl(p, start+1, finish+1);
		assert(copy == finish+4);
		repeat(p, copy, 1, to-1);
		break;
	case REP(1, INF):		/* as x+ */
		INSERT(OPLUS_, start);
		ASTERN(O_PLUS, start);
		break;
	case REP(N, N):			/* as xx{m-1,n-1} */
		copy = dupl(p, start, finish);
		repeat(p, copy, from-1, to-1);
		break;
	case REP(N, INF):		/* as xx{n-1,INF} */
		copy = dupl(p, start, finish);
		repeat(p, copy, from-1, to);
		break;
	default:			/* "can't happen" */
		SETERROR(REG_ASSERT);	/* just in case */
		break;
	}
}

/*
 - seterr - set an error condition
 == static int seterr(register struct parse *p, int e);
 */
static int			/* useless but makes type checking happy */
seterr(p, e)
register struct parse *p;
int e;
{
	if (p->error == 0)	/* keep earliest error condition */
		p->error = e;
	p->next = nuls;		/* try to bring things to a halt */
	p->end = nuls;
	return(0);		/* make the return value well-defined */
}

/*
 - allocset - allocate a set of characters for []
 == static cset *allocset(register struct parse *p);
 */
static cset *
allocset(p)
register struct parse *p;
{
	register int no = p->g->ncsets++;
	register size_t nc;
	register size_t nbytes;
	register cset *cs;
	register size_t css = (size_t)p->g->csetsize;
	register int i;

	if (no >= p->ncsalloc) {	/* need another column of space */
		p->ncsalloc += CHAR_BIT;
		nc = p->ncsalloc;
		assert(nc % CHAR_BIT == 0);
		nbytes = nc / CHAR_BIT * css;
		if (p->g->sets == NULL)
			p->g->sets = (cset *)malloc(nc * sizeof(cset));
		else
			p->g->sets = (cset *)realloc((char *)p->g->sets,
							nc * sizeof(cset));
		if (p->g->setbits == NULL)
			p->g->setbits = (uch *)malloc(nbytes);
		else {
			p->g->setbits = (uch *)realloc((char *)p->g->setbits,
								nbytes);
			/* xxx this isn't right if setbits is now NULL */
			for (i = 0; i < no; i++)
				p->g->sets[i].ptr = p->g->setbits + css*(i/CHAR_BIT);
		}
		if (p->g->sets != NULL && p->g->setbits != NULL)
			(void) memset((char *)p->g->setbits + (nbytes - css),
								0, css);
		else {
			no = 0;
			SETERROR(REG_ESPACE);
			/* caller's responsibility not to do set ops */
		}
	}

	assert(p->g->sets != NULL);	/* xxx */
	cs = &p->g->sets[no];
	cs->ptr = p->g->setbits + css*((no)/CHAR_BIT);
	cs->mask = 1 << ((no) % CHAR_BIT);
	cs->hash = 0;
	cs->smultis = 0;
	cs->multis = NULL;

	return(cs);
}

/*
 - freeset - free a now-unused set
 == static void freeset(register struct parse *p, register cset *cs);
 */
static void
freeset(p, cs)
register struct parse *p;
register cset *cs;
{
	register int i;
	register cset *top = &p->g->sets[p->g->ncsets];
	register size_t css = (size_t)p->g->csetsize;

	for (i = 0; i < css; i++)
		CHsub(cs, i);
	if (cs == top-1)	/* recover only the easy case */
		p->g->ncsets--;
}

/*
 - freezeset - final processing on a set of characters
 == static int freezeset(register struct parse *p, register cset *cs);
 *