Loading...
--- Libc/Libc-763.12/regex/FreeBSD/regcomp.c.patch
+++ /dev/null
@@ -1,514 +0,0 @@
---- regcomp.c.orig 2010-06-21 14:05:04.000000000 -0700
-+++ regcomp.c 2010-06-21 14:23:51.000000000 -0700
-@@ -39,6 +39,8 @@ static char sccsid[] = "@(#)regcomp.c 8.
- #include <sys/cdefs.h>
- __FBSDID("$FreeBSD: src/lib/libc/regex/regcomp.c,v 1.36 2007/06/11 03:05:54 delphij Exp $");
-
-+#include "xlocale_private.h"
-+
- #include <sys/types.h>
- #include <stdio.h>
- #include <string.h>
-@@ -69,6 +71,9 @@ struct parse {
- sopno ssize; /* malloced strip size (allocated) */
- sopno slen; /* malloced strip length (used) */
- int ncsalloc; /* number of csets allocated */
-+#if __DARWIN_UNIX03
-+ int zerorepeats;
-+#endif /* __DARWIN_UNIX03 */
- struct re_guts *g;
- # define NPAREN 10 /* we need to remember () 1-9 for back refs */
- sopno pbegin[NPAREN]; /* -> ( ([0] unused) */
-@@ -93,7 +98,7 @@ static void p_b_cclass(struct parse *p,
- static void p_b_eclass(struct parse *p, cset *cs);
- static wint_t p_b_symbol(struct parse *p);
- static wint_t p_b_coll_elem(struct parse *p, wint_t endc);
--static wint_t othercase(wint_t ch);
-+static wint_t othercase(wint_t ch, locale_t loc);
- static void bothcases(struct parse *p, wint_t ch);
- static void ordinary(struct parse *p, wint_t ch);
- static void nonnewline(struct parse *p);
-@@ -104,7 +109,7 @@ static void freeset(struct parse *p, cse
- static void CHadd(struct parse *p, cset *cs, wint_t ch);
- static void CHaddrange(struct parse *p, cset *cs, wint_t min, wint_t max);
- static void CHaddtype(struct parse *p, cset *cs, wctype_t wct);
--static wint_t singleton(cset *cs);
-+static wint_t singleton(cset *cs, locale_t loc);
- static sopno dupl(struct parse *p, sopno start, sopno finish);
- static void doemit(struct parse *p, sop op, size_t opnd);
- static void doinsert(struct parse *p, sop op, size_t opnd, sopno pos);
-@@ -222,10 +227,14 @@ regcomp(regex_t * __restrict preg,
- p->end = p->next + len;
- p->error = 0;
- p->ncsalloc = 0;
-+#if __DARWIN_UNIX03
-+ p->zerorepeats = 0;
-+#endif /* __DARWIN_UNIX03 */
- for (i = 0; i < NPAREN; i++) {
- p->pbegin[i] = 0;
- p->pend[i] = 0;
- }
-+ g->loc = __current_locale();
- g->sets = NULL;
- g->ncsets = 0;
- g->cflags = cflags;
-@@ -302,8 +311,12 @@ p_ere(struct parse *p,
- conc = HERE();
- while (MORE() && (c = PEEK()) != '|' && c != stop)
- p_ere_exp(p);
-+#if __DARWIN_UNIX03
-+ if (!p->zerorepeats) REQUIRE(HERE() != conc, REG_EMPTY); /* require nonempty */
-+ else p->zerorepeats--;
-+#else
- (void)REQUIRE(HERE() != conc, REG_EMPTY); /* require nonempty */
--
-+#endif
- if (!EAT('|'))
- break; /* NOTE BREAK OUT */
-
-@@ -410,7 +423,7 @@ p_ere_exp(struct parse *p)
- ordinary(p, wc);
- break;
- case '{': /* okay as ordinary except if digit follows */
-- (void)REQUIRE(!MORE() || !isdigit((uch)PEEK()), REG_BADRPT);
-+ (void)REQUIRE(!MORE() || !isdigit_l((uch)PEEK(), p->g->loc), REG_BADRPT);
- /* FALLTHROUGH */
- default:
- p->next--;
-@@ -424,7 +437,7 @@ p_ere_exp(struct parse *p)
- c = PEEK();
- /* we call { a repetition if followed by a digit */
- if (!( c == '*' || c == '+' || c == '?' ||
-- (c == '{' && MORE2() && isdigit((uch)PEEK2())) ))
-+ (c == '{' && MORE2() && isdigit_l((uch)PEEK2(), p->g->loc)) ))
- return; /* no repetition, we're done */
- NEXT();
-
-@@ -453,7 +466,7 @@ p_ere_exp(struct parse *p)
- case '{':
- count = p_count(p);
- if (EAT(',')) {
-- if (isdigit((uch)PEEK())) {
-+ if (isdigit_l((uch)PEEK(), p->g->loc)) {
- count2 = p_count(p);
- (void)REQUIRE(count <= count2, REG_BADBR);
- } else /* single number with comma */
-@@ -474,7 +487,7 @@ p_ere_exp(struct parse *p)
- return;
- c = PEEK();
- if (!( c == '*' || c == '+' || c == '?' ||
-- (c == '{' && MORE2() && isdigit((uch)PEEK2())) ) )
-+ (c == '{' && MORE2() && isdigit_l((uch)PEEK2(), p->g->loc)) ) )
- return;
- SETERROR(REG_BADRPT);
- }
-@@ -486,7 +499,12 @@ p_ere_exp(struct parse *p)
- static void
- p_str(struct parse *p)
- {
-+#if __DARWIN_UNIX03
-+ if (!p->zerorepeats) REQUIRE(MORE(), REG_EMPTY);
-+ else p->zerorepeats--;
-+#else /* !__DARWIN_UNIX03 */
- (void)REQUIRE(MORE(), REG_EMPTY);
-+#endif /* __DARWIN_UNIX03 */
- while (MORE())
- ordinary(p, WGETNEXT());
- }
-@@ -525,8 +543,12 @@ p_bre(struct parse *p,
- p->g->iflags |= USEEOL;
- p->g->neol++;
- }
--
-+#if __DARWIN_UNIX03
-+ if (!p->zerorepeats) REQUIRE(HERE() != start, REG_EMPTY); /* require nonempty */
-+ else p->zerorepeats--;
-+#else /* !__DARWIN_UNIX03 */
- (void)REQUIRE(HERE() != start, REG_EMPTY); /* require nonempty */
-+#endif /* __DARWIN_UNIX03 */
- }
-
- /*
-@@ -599,12 +621,22 @@ p_simp_re(struct parse *p,
- i = (c&~BACKSL) - '0';
- assert(i < NPAREN);
- if (p->pend[i] != 0) {
-+#if __DARWIN_UNIX03
-+ int skip = 1;
-+#endif /* __DARWIN_UNIX03 */
- assert(i <= p->g->nsub);
- EMIT(OBACK_, i);
- assert(p->pbegin[i] != 0);
- assert(OP(p->strip[p->pbegin[i]]) == OLPAREN);
- assert(OP(p->strip[p->pend[i]]) == ORPAREN);
-+#if __DARWIN_UNIX03
-+ if (OP(p->strip[p->pbegin[i]+skip]) == OBOL) {
-+ skip++; /* don't dup anchor in subexp */
-+ }
-+ (void) dupl(p, p->pbegin[i]+skip, p->pend[i]);
-+#else /* !__DARWIN_UNIX03 */
- (void) dupl(p, p->pbegin[i]+1, p->pend[i]);
-+#endif /* __DARWIN_UNIX03 */
- EMIT(O_BACK, i);
- } else
- SETERROR(REG_ESUBREG);
-@@ -627,9 +659,10 @@ p_simp_re(struct parse *p,
- INSERT(OQUEST_, pos);
- ASTERN(O_QUEST, pos);
- } else if (EATTWO('\\', '{')) {
-+ (void)REQUIRE(MORE(), REG_EBRACE);
- count = p_count(p);
- if (EAT(',')) {
-- if (MORE() && isdigit((uch)PEEK())) {
-+ if (MORE() && isdigit_l((uch)PEEK(), p->g->loc)) {
- count2 = p_count(p);
- (void)REQUIRE(count <= count2, REG_BADBR);
- } else /* single number with comma */
-@@ -659,7 +692,7 @@ p_count(struct parse *p)
- int count = 0;
- int ndigits = 0;
-
-- while (MORE() && isdigit((uch)PEEK()) && count <= DUPMAX) {
-+ while (MORE() && isdigit_l((uch)PEEK(), p->g->loc) && count <= DUPMAX) {
- count = count*10 + (GETNEXT() - '0');
- ndigits++;
- }
-@@ -697,10 +730,22 @@ p_bracket(struct parse *p)
- cs->icase = 1;
- if (EAT('^'))
- cs->invert = 1;
-+#if __DARWIN_UNIX03
-+ if (PEEK2() != '-' && PEEK2() != ']') { /* Don't eat '-' or ']' if they're part of ranges
-+ * but do process [^-] */
- if (EAT(']'))
- CHadd(p, cs, ']');
- else if (EAT('-'))
- CHadd(p, cs, '-');
-+ }
-+ if (MORE() && !SEETWO('-',']')) /* Parse RE []-'] */
-+ p_b_term(p, cs);
-+#else /* !__DARWIN_UNIX03 */
-+ if (EAT(']'))
-+ CHadd(p, cs, ']');
-+ else if (EAT('-'))
-+ CHadd(p, cs, '-');
-+#endif /* __DARWIN_UNIX03 */
- while (MORE() && PEEK() != ']' && !SEETWO('-', ']'))
- p_b_term(p, cs);
- if (EAT('-'))
-@@ -713,7 +758,7 @@ p_bracket(struct parse *p)
- if (cs->invert && p->g->cflags®_NEWLINE)
- cs->bmp['\n' >> 3] |= 1 << ('\n' & 7);
-
-- if ((ch = singleton(cs)) != OUT) { /* optimize singleton sets */
-+ if ((ch = singleton(cs, p->g->loc)) != OUT) { /* optimize singleton sets */
- ordinary(p, ch);
- freeset(p, cs);
- } else
-@@ -737,8 +782,16 @@ p_b_term(struct parse *p, cset *cs)
- c = (MORE2()) ? PEEK2() : '\0';
- break;
- case '-':
-+#if __DARWIN_UNIX03
-+ if (PEEK2() != '-') { /* Allow [---] */
-+ SETERROR(REG_ERANGE);
-+ return; /* NOTE RETURN */
-+ } else
-+ c = '-';
-+#else /* !__DARWIN_UNIX03 */
- SETERROR(REG_ERANGE);
- return; /* NOTE RETURN */
-+#endif /* __DARWIN_UNIX03 */
- break;
- default:
- c = '\0';
-@@ -759,7 +812,11 @@ p_b_term(struct parse *p, cset *cs)
- NEXT2();
- (void)REQUIRE(MORE(), REG_EBRACK);
- c = PEEK();
-+#if __DARWIN_UNIX03
-+ REQUIRE(c != '-', REG_ECOLLATE); /* allow [=]=] */
-+#else /* !__DARWIN_UNIX03 */
- (void)REQUIRE(c != '-' && c != ']', REG_ECOLLATE);
-+#endif /* __DARWIN_UNIX03 */
- p_b_eclass(p, cs);
- (void)REQUIRE(MORE(), REG_EBRACK);
- (void)REQUIRE(EATTWO('=', ']'), REG_ECOLLATE);
-@@ -778,14 +835,14 @@ p_b_term(struct parse *p, cset *cs)
- if (start == finish)
- CHadd(p, cs, start);
- else {
-- if (__collate_load_error) {
-+ if (p->g->loc->__collate_load_error) {
- (void)REQUIRE((uch)start <= (uch)finish, REG_ERANGE);
- CHaddrange(p, cs, start, finish);
- } else {
-- (void)REQUIRE(__collate_range_cmp(start, finish) <= 0, REG_ERANGE);
-+ (void)REQUIRE(__collate_range_cmp(start, finish, p->g->loc) <= 0, REG_ERANGE);
- for (i = 0; i <= UCHAR_MAX; i++) {
-- if ( __collate_range_cmp(start, i) <= 0
-- && __collate_range_cmp(i, finish) <= 0
-+ if ( __collate_range_cmp(start, i, p->g->loc) <= 0
-+ && __collate_range_cmp(i, finish, p->g->loc) <= 0
- )
- CHadd(p, cs, i);
- }
-@@ -807,7 +864,7 @@ p_b_cclass(struct parse *p, cset *cs)
- wctype_t wct;
- char clname[16];
-
-- while (MORE() && isalpha((uch)PEEK()))
-+ while (MORE() && isalpha_l((uch)PEEK(), p->g->loc))
- NEXT();
- len = p->next - sp;
- if (len >= sizeof(clname) - 1) {
-@@ -816,7 +873,7 @@ p_b_cclass(struct parse *p, cset *cs)
- }
- memcpy(clname, sp, len);
- clname[len] = '\0';
-- if ((wct = wctype(clname)) == 0) {
-+ if ((wct = wctype_l(clname, p->g->loc)) == 0) {
- SETERROR(REG_ECTYPE);
- return;
- }
-@@ -826,14 +883,38 @@ p_b_cclass(struct parse *p, cset *cs)
- /*
- - p_b_eclass - parse an equivalence-class name and deal with it
- == static void p_b_eclass(struct parse *p, cset *cs);
-- *
-- * This implementation is incomplete. xxx
- */
- static void
- p_b_eclass(struct parse *p, cset *cs)
- {
-- wint_t c;
--
-+ char *sp = p->next;
-+ int len, ec;
-+ mbstate_t mbs;
-+ int *newequiv_classes;
-+ wint_t c;
-+
-+ while (MORE() && !SEETWO('=', ']'))
-+ NEXT();
-+ if (!MORE()) {
-+ SETERROR(REG_EBRACK);
-+ return;
-+ }
-+ len = p->next - sp;
-+ memset(&mbs, 0, sizeof(mbs));
-+ ec = __collate_equiv_class(sp, len, &mbs, p->g->loc);
-+ if (ec > 0) {
-+ newequiv_classes = realloc(cs->equiv_classes,
-+ (cs->nequiv_classes + 1) * sizeof(*cs->equiv_classes));
-+ if (newequiv_classes == NULL) {
-+ SETERROR(REG_ESPACE);
-+ return;
-+ }
-+ cs->equiv_classes = newequiv_classes;
-+ cs->equiv_classes[cs->nequiv_classes++] = ec;
-+ return;
-+ }
-+ /* not an equivalence class, so fallback to a collating element */
-+ p->next = sp;
- c = p_b_coll_elem(p, '=');
- CHadd(p, cs, c);
- }
-@@ -866,10 +947,10 @@ p_b_coll_elem(struct parse *p,
- wint_t endc) /* name ended by endc,']' */
- {
- char *sp = p->next;
-- struct cname *cp;
-+ const struct cname *cp;
- int len;
- mbstate_t mbs;
-- wchar_t wc;
-+ wchar_t wbuf[16];
- size_t clen;
-
- while (MORE() && !SEETWO(endc, ']'))
-@@ -883,9 +964,10 @@ p_b_coll_elem(struct parse *p,
- if (strncmp(cp->name, sp, len) == 0 && cp->name[len] == '\0')
- return(cp->code); /* known name */
- memset(&mbs, 0, sizeof(mbs));
-- if ((clen = mbrtowc(&wc, sp, len, &mbs)) == len)
-- return (wc); /* single character */
-- else if (clen == (size_t)-1 || clen == (size_t)-2)
-+ clen = __collate_collating_symbol(wbuf, 16, sp, len, &mbs, p->g->loc);
-+ if (clen == 1)
-+ return (*wbuf); /* single character */
-+ else if (clen == (size_t)-1)
- SETERROR(REG_ILLSEQ);
- else
- SETERROR(REG_ECOLLATE); /* neither */
-@@ -894,16 +976,16 @@ p_b_coll_elem(struct parse *p,
-
- /*
- - othercase - return the case counterpart of an alphabetic
-- == static char othercase(int ch);
-+ == static char othercase(wint_t ch, locale_t loc);
- */
- static wint_t /* if no counterpart, return ch */
--othercase(wint_t ch)
-+othercase(wint_t ch, locale_t loc)
- {
-- assert(iswalpha(ch));
-- if (iswupper(ch))
-- return(towlower(ch));
-- else if (iswlower(ch))
-- return(towupper(ch));
-+ assert(iswalpha_l(ch, loc));
-+ if (iswupper_l(ch, loc))
-+ return(towlower_l(ch, loc));
-+ else if (iswlower_l(ch, loc))
-+ return(towupper_l(ch, loc));
- else /* peculiar, but could happen */
- return(ch);
- }
-@@ -923,10 +1005,10 @@ bothcases(struct parse *p, wint_t ch)
- size_t n;
- mbstate_t mbs;
-
-- assert(othercase(ch) != ch); /* p_bracket() would recurse */
-+ assert(othercase(ch, p->g->loc) != ch); /* p_bracket() would recurse */
- p->next = bracket;
- memset(&mbs, 0, sizeof(mbs));
-- n = wcrtomb(bracket, ch, &mbs);
-+ n = wcrtomb_l(bracket, ch, &mbs, p->g->loc);
- assert(n != (size_t)-1);
- bracket[n] = ']';
- bracket[n + 1] = '\0';
-@@ -946,7 +1028,7 @@ ordinary(struct parse *p, wint_t ch)
- {
- cset *cs;
-
-- if ((p->g->cflags®_ICASE) && iswalpha(ch) && othercase(ch) != ch)
-+ if ((p->g->cflags®_ICASE) && iswalpha_l(ch, p->g->loc) && othercase(ch, p->g->loc) != ch)
- bothcases(p, ch);
- else if ((ch & OPDMASK) == ch)
- EMIT(OCHAR, ch);
-@@ -1012,10 +1094,22 @@ repeat(struct parse *p,
- switch (REP(MAP(from), MAP(to))) {
- case REP(0, 0): /* must be user doing this */
- DROP(finish-start); /* drop the operand */
-+#if __DARWIN_UNIX03
-+ p->zerorepeats++;
-+#endif /* __DARWIN_UNIX03 */
- break;
-+ case REP(0, INF): /* as x{1,}? */
-+#if __DARWIN_UNIX03
-+ /* this case does not require the (y|) trick, noKLUDGE */
-+ /* Just like * =+? */
-+ INSERT(OPLUS_, start);
-+ ASTERN(O_PLUS, start);
-+ INSERT(OQUEST_, start);
-+ ASTERN(O_QUEST, start);
-+ break;
-+#endif /* __DARWIN_UNIX03 */
- 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);
-@@ -1029,6 +1123,10 @@ repeat(struct parse *p,
- /* done */
- break;
- case REP(1, N): /* as x?x{1,n-1} */
-+#if __DARWIN_UNIX03
-+ INSERT(OQUEST_, start);
-+ ASTERN(O_QUEST, start);
-+#else /* !__DARWIN_UNIX03 */
- /* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
- INSERT(OCH_, start);
- ASTERN(OOR1, start);
-@@ -1036,6 +1134,7 @@ repeat(struct parse *p,
- EMIT(OOR2, 0); /* offset very wrong... */
- AHEAD(THERE()); /* ...so fix it */
- ASTERN(O_CH, THERETHERE());
-+#endif /* __DARWIN_UNIX03 */
- copy = dupl(p, start+1, finish+1);
- assert(copy == finish+4);
- repeat(p, copy, 1, to-1);
-@@ -1071,7 +1170,7 @@ wgetnext(struct parse *p)
- size_t n;
-
- memset(&mbs, 0, sizeof(mbs));
-- n = mbrtowc(&wc, p->next, p->end - p->next, &mbs);
-+ n = mbrtowc_l(&wc, p->next, p->end - p->next, &mbs, p->g->loc);
- if (n == (size_t)-1 || n == (size_t)-2) {
- SETERROR(REG_ILLSEQ);
- return (0);
-@@ -1139,12 +1238,12 @@ freeset(struct parse *p, cset *cs)
- - returning it if so, otherwise returning OUT.
- */
- static wint_t
--singleton(cset *cs)
-+singleton(cset *cs, locale_t loc)
- {
- wint_t i, s, n;
-
- for (i = n = 0; i < NC; i++)
-- if (CHIN(cs, i)) {
-+ if (CHIN(cs, i, loc)) {
- n++;
- s = i;
- }
-@@ -1178,9 +1277,9 @@ CHadd(struct parse *p, cset *cs, wint_t
- cs->wides[cs->nwides++] = ch;
- }
- if (cs->icase) {
-- if ((nch = towlower(ch)) < NC)
-+ if ((nch = towlower_l(ch, p->g->loc)) < NC)
- cs->bmp[nch >> 3] |= 1 << (nch & 7);
-- if ((nch = towupper(ch)) < NC)
-+ if ((nch = towupper_l(ch, p->g->loc)) < NC)
- cs->bmp[nch >> 3] |= 1 << (nch & 7);
- }
- }
-@@ -1219,7 +1318,7 @@ CHaddtype(struct parse *p, cset *cs, wct
- wctype_t *newtypes;
-
- for (i = 0; i < NC; i++)
-- if (iswctype(i, wct))
-+ if (iswctype_l(i, wct, p->g->loc))
- CHadd(p, cs, i);
- newtypes = realloc(cs->types, (cs->ntypes + 1) *
- sizeof(*cs->types));
-@@ -1391,6 +1490,7 @@ findmust(struct parse *p, struct re_guts
- char buf[MB_LEN_MAX];
- size_t clen;
- mbstate_t mbs;
-+ struct __xlocale_st_runelocale *rl = p->g->loc->__lc_ctype;
-
- /* avoid making error situations worse */
- if (p->error != 0)
-@@ -1401,8 +1501,8 @@ findmust(struct parse *p, struct re_guts
- * multibyte character strings, but it's safe for at least
- * UTF-8 (see RFC 3629).
- */
-- if (MB_CUR_MAX > 1 &&
-- strcmp(_CurrentRuneLocale->__encoding, "UTF-8") != 0)
-+ if (rl->__mb_cur_max > 1 &&
-+ strcmp(rl->_CurrentRuneLocale.__encoding, "UTF-8") != 0)
- return;
-
- /* find the longest OCHAR sequence in strip */
-@@ -1418,7 +1518,7 @@ findmust(struct parse *p, struct re_guts
- memset(&mbs, 0, sizeof(mbs));
- newstart = scan - 1;
- }
-- clen = wcrtomb(buf, OPND(s), &mbs);
-+ clen = wcrtomb_l(buf, OPND(s), &mbs, p->g->loc);
- if (clen == (size_t)-1)
- goto toohard;
- newlen += clen;
-@@ -1537,7 +1637,7 @@ findmust(struct parse *p, struct re_guts
- while (cp < g->must + g->mlen) {
- while (OP(s = *scan++) != OCHAR)
- continue;
-- clen = wcrtomb(cp, OPND(s), &mbs);
-+ clen = wcrtomb_l(cp, OPND(s), &mbs, p->g->loc);
- assert(clen != (size_t)-1);
- cp += clen;
- }