Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 14 Jul 2016 08:18:12 +0000 (UTC)
From:      "Andrey A. Chernov" <ache@FreeBSD.org>
To:        src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org
Subject:   svn commit: r302820 - in head/lib/libc: gen locale regex stdio
Message-ID:  <201607140818.u6E8ICRo001964@repo.freebsd.org>

next in thread | raw e-mail | index | archive | help
Author: ache
Date: Thu Jul 14 08:18:12 2016
New Revision: 302820
URL: https://svnweb.freebsd.org/changeset/base/302820

Log:
  Back out non-collating [a-z] ranges.
  Instead of changing whole course to another POSIX-permitted way
  for consistency and uniformity I decide to completely ignore missing
  regex fucntionality and concentrace on fixing bugs in what we have now,
  too many small obstacles instead, counting ports.

Modified:
  head/lib/libc/gen/fnmatch.c
  head/lib/libc/gen/glob.c
  head/lib/libc/locale/collate.h
  head/lib/libc/locale/collcmp.c
  head/lib/libc/regex/regcomp.c
  head/lib/libc/stdio/vfscanf.c

Modified: head/lib/libc/gen/fnmatch.c
==============================================================================
--- head/lib/libc/gen/fnmatch.c	Thu Jul 14 08:15:13 2016	(r302819)
+++ head/lib/libc/gen/fnmatch.c	Thu Jul 14 08:18:12 2016	(r302820)
@@ -63,6 +63,8 @@ __FBSDID("$FreeBSD$");
 #include <wchar.h>
 #include <wctype.h>
 
+#include "collate.h"
+
 #define	EOS	'\0'
 
 #define RANGE_MATCH     1
@@ -236,6 +238,8 @@ rangematch(const char *pattern, wchar_t 
 	wchar_t c, c2;
 	size_t pclen;
 	const char *origpat;
+	struct xlocale_collate *table =
+		(struct xlocale_collate*)__get_locale()->components[XLC_COLLATE];
 
 	/*
 	 * A bracket expression starting with an unquoted circumflex
@@ -290,7 +294,11 @@ rangematch(const char *pattern, wchar_t 
 			if (flags & FNM_CASEFOLD)
 				c2 = towlower(c2);
 
-			if (c <= test && test <= c2)
+			if (table->__collate_load_error ?
+			    c <= test && test <= c2 :
+			       __wcollate_range_cmp(table, c, test) <= 0
+			    && __wcollate_range_cmp(table, test, c2) <= 0
+			   )
 				ok = 1;
 		} else if (c == test)
 			ok = 1;

Modified: head/lib/libc/gen/glob.c
==============================================================================
--- head/lib/libc/gen/glob.c	Thu Jul 14 08:15:13 2016	(r302819)
+++ head/lib/libc/gen/glob.c	Thu Jul 14 08:18:12 2016	(r302820)
@@ -92,6 +92,8 @@ __FBSDID("$FreeBSD$");
 #include <unistd.h>
 #include <wchar.h>
 
+#include "collate.h"
+
 /*
  * glob(3) expansion limits. Stop the expansion if any of these limits
  * is reached. This caps the runtime in the face of DoS attacks. See
@@ -802,6 +804,8 @@ match(Char *name, Char *pat, Char *paten
 {
 	int ok, negate_range;
 	Char c, k;
+	struct xlocale_collate *table =
+		(struct xlocale_collate*)__get_locale()->components[XLC_COLLATE];
 
 	while (pat < patend) {
 		c = *pat++;
@@ -826,7 +830,11 @@ match(Char *name, Char *pat, Char *paten
 				++pat;
 			while (((c = *pat++) & M_MASK) != M_END)
 				if ((*pat & M_MASK) == M_RNG) {
-					if (CHAR(c) <= CHAR(k) && CHAR(k) <= CHAR(pat[1]))
+					if (table->__collate_load_error ?
+					    CHAR(c) <= CHAR(k) && CHAR(k) <= CHAR(pat[1]) :
+					       __wcollate_range_cmp(table, CHAR(c), CHAR(k)) <= 0
+					    && __wcollate_range_cmp(table, CHAR(k), CHAR(pat[1])) <= 0
+					   )
 						ok = 1;
 					pat += 2;
 				} else if (c == k)

Modified: head/lib/libc/locale/collate.h
==============================================================================
--- head/lib/libc/locale/collate.h	Thu Jul 14 08:15:13 2016	(r302819)
+++ head/lib/libc/locale/collate.h	Thu Jul 14 08:18:12 2016	(r302820)
@@ -128,7 +128,8 @@ int	__collate_load_tables(const char *);
 int	__collate_equiv_value(locale_t, const wchar_t *, size_t);
 void	_collate_lookup(struct xlocale_collate *,const wchar_t *, int *, int *,
 	int, const int **);
-int	__collate_range_cmp(int, int);
+int	__collate_range_cmp(struct xlocale_collate *, char, char);
+int	__wcollate_range_cmp(struct xlocale_collate *, wchar_t, wchar_t);
 size_t	_collate_wxfrm(struct xlocale_collate *, const wchar_t *, wchar_t *,
 	size_t);
 size_t	_collate_sxfrm(struct xlocale_collate *, const wchar_t *, char *,

Modified: head/lib/libc/locale/collcmp.c
==============================================================================
--- head/lib/libc/locale/collcmp.c	Thu Jul 14 08:15:13 2016	(r302819)
+++ head/lib/libc/locale/collcmp.c	Thu Jul 14 08:18:12 2016	(r302820)
@@ -33,13 +33,15 @@
 __FBSDID("$FreeBSD$");
 
 #include <string.h>
+#include <wchar.h>
+#include <xlocale.h>
 #include "collate.h"
 
 /*
  * Compare two characters using collate
  */
 
-int __collate_range_cmp(int c1, int c2)
+int __collate_range_cmp(struct xlocale_collate *table, char c1, char c2)
 {
 	char s1[2], s2[2];
 
@@ -47,5 +49,20 @@ int __collate_range_cmp(int c1, int c2)
 	s1[1] = '\0';
 	s2[0] = c2;
 	s2[1] = '\0';
-	return (strcoll(s1, s2));
+	struct _xlocale l = {{0}};
+	l.components[XLC_COLLATE] = (struct xlocale_component *)table;
+	return (strcoll_l(s1, s2, &l));
+}
+
+int __wcollate_range_cmp(struct xlocale_collate *table, wchar_t c1, wchar_t c2)
+{
+	wchar_t s1[2], s2[2];
+
+	s1[0] = c1;
+	s1[1] = L'\0';
+	s2[0] = c2;
+	s2[1] = L'\0';
+	struct _xlocale l = {{0}};
+	l.components[XLC_COLLATE] = (struct xlocale_component *)table;
+	return (wcscoll_l(s1, s2, &l));
 }

Modified: head/lib/libc/regex/regcomp.c
==============================================================================
--- head/lib/libc/regex/regcomp.c	Thu Jul 14 08:15:13 2016	(r302819)
+++ head/lib/libc/regex/regcomp.c	Thu Jul 14 08:18:12 2016	(r302820)
@@ -51,9 +51,12 @@ __FBSDID("$FreeBSD$");
 #include <limits.h>
 #include <stdlib.h>
 #include <regex.h>
+#include <runetype.h>
 #include <wchar.h>
 #include <wctype.h>
 
+#include "collate.h"
+
 #include "utils.h"
 #include "regex2.h"
 
@@ -764,6 +767,9 @@ p_b_term(struct parse *p, cset *cs)
 {
 	char c;
 	wint_t start, finish;
+	wint_t i;
+	struct xlocale_collate *table =
+		(struct xlocale_collate*)__get_locale()->components[XLC_COLLATE];
 
 	/* classify what we've got */
 	switch ((MORE()) ? PEEK() : '\0') {
@@ -811,8 +817,18 @@ p_b_term(struct parse *p, cset *cs)
 		if (start == finish)
 			CHadd(p, cs, start);
 		else {
-			(void)REQUIRE(start <= finish, REG_ERANGE);
-			CHaddrange(p, cs, start, finish);
+			if (table->__collate_load_error) {
+				(void)REQUIRE((uch)start <= (uch)finish, REG_ERANGE);
+				CHaddrange(p, cs, start, finish);
+			} else {
+				(void)REQUIRE(__wcollate_range_cmp(table, start, finish) <= 0, REG_ERANGE);
+				for (i = 0; i <= UCHAR_MAX; i++) {
+					if (   __wcollate_range_cmp(table, start, i) <= 0
+					    && __wcollate_range_cmp(table, i, finish) <= 0
+					   )
+						CHadd(p, cs, i);
+				}
+			}
 		}
 		break;
 	}

Modified: head/lib/libc/stdio/vfscanf.c
==============================================================================
--- head/lib/libc/stdio/vfscanf.c	Thu Jul 14 08:15:13 2016	(r302819)
+++ head/lib/libc/stdio/vfscanf.c	Thu Jul 14 08:18:12 2016	(r302820)
@@ -53,6 +53,7 @@ __FBSDID("$FreeBSD$");
 #include <wctype.h>
 #include "un-namespace.h"
 
+#include "collate.h"
 #include "libc_private.h"
 #include "local.h"
 #include "xlocale_private.h"
@@ -815,7 +816,9 @@ match_failure:
 static const u_char *
 __sccl(char *tab, const u_char *fmt)
 {
-	int c, n, v;
+	int c, n, v, i;
+	struct xlocale_collate *table =
+		(struct xlocale_collate*)__get_locale()->components[XLC_COLLATE];
 
 	/* first `clear' the whole table */
 	c = *fmt++;		/* first char hat => negated scanset */
@@ -868,15 +871,29 @@ doswitch:
 			 * we just stored in the table (c).
 			 */
 			n = *fmt;
-			if (n == ']' || n < c) {
+			if (n == ']'
+			    || (table->__collate_load_error ? n < c :
+				__wcollate_range_cmp(table, n, c) < 0
+			       )
+			   ) {
 				c = '-';
 				break;	/* resume the for(;;) */
 			}
 			fmt++;
-			do {		/* fill in the range */
-				tab[++c] = v;
-			} while (c < n);
+			/* fill in the range */
+			if (table->__collate_load_error) {
+				do {
+					tab[++c] = v;
+				} while (c < n);
+			} else {
+				for (i = 0; i < 256; i ++)
+					if (__wcollate_range_cmp(table, c, i) < 0 &&
+					    __wcollate_range_cmp(table, i, n) <= 0
+					   )
+						tab[i] = v;
+			}
 #if 1	/* XXX another disgusting compatibility hack */
+			c = n;
 			/*
 			 * Alas, the V7 Unix scanf also treats formats
 			 * such as [a-c-e] as `the letters a through e'.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201607140818.u6E8ICRo001964>