Page MenuHomeFreeBSD
Files F115976567

D43149.id131715.diff
No OneTemporary
Actions

D43149.id131715.diff
View Options

diff --git a/lib/libc/regex/regcomp.c b/lib/libc/regex/regcomp.c
--- a/lib/libc/regex/regcomp.c
+++ b/lib/libc/regex/regcomp.c
@@ -1586,17 +1586,32 @@
{
wint_t i, s, n;
+ /* Exclude the complicated cases we don't want to deal with */
+ if (cs->nranges != 0 || cs->ntypes != 0 || cs->icase != 0)
+ return (OUT);
+
+ if (cs->nwides > 1)
+ return (OUT);
+
+ /* Count the number of characters present in the bitmap */
for (i = n = 0; i < NC; i++)
if (CHIN(cs, i)) {
n++;
s = i;
}
- if (n == 1)
- return (s);
- if (cs->nwides == 1 && cs->nranges == 0 && cs->ntypes == 0 &&
- cs->icase == 0)
+
+ if (n > 1)
+ return (OUT);
+
+ if (n == 1) {
+ if (cs->nwides == 0)
+ return (s);
+ else
+ return (OUT);
+ }
+ if (cs->nwides == 1)
return (cs->wides[0]);
- /* Don't bother handling the other cases. */
+
return (OUT);
}
diff --git a/lib/libc/tests/regex/multibyte.sh b/lib/libc/tests/regex/multibyte.sh
--- a/lib/libc/tests/regex/multibyte.sh
+++ b/lib/libc/tests/regex/multibyte.sh
@@ -1,4 +1,3 @@
-
atf_test_case bmpat
bmpat_head()
{
@@ -45,8 +44,50 @@
echo $c | atf_check -o "inline:$c\n" sed -ne "/$a/Ip"
}
+atf_test_case mbset cleanup
+mbset_head()
+{
+ atf_set "descr" "Check multibyte sets matching"
+}
+mbset_body()
+{
+ export LC_CTYPE="C.UTF-8"
+
+ # This involved an erroneously implemented optimization which reduces
+ # single-element sets to an exact match with a single codepoint.
+ # Match sets record small-codepoint characters in a bitmap and
+ # large-codepoint characters in an array; the optimization would falsely
+ # trigger if either the bitmap or the array was a singleton, ignoring
+ # the members of the other side of the set.
+ #
+ # To exercise this, we construct sets which have one member of one side
+ # and one or more of the other, and verify that all members can be
+ # found.
+ printf "a" > mbset; atf_check -o not-empty sed -ne '/[aà]/p' mbset
+ printf "à" > mbset; atf_check -o not-empty sed -ne '/[aà]/p' mbset
+ printf "a" > mbset; atf_check -o not-empty sed -ne '/[aàá]/p' mbset
+ printf "à" > mbset; atf_check -o not-empty sed -ne '/[aàá]/p' mbset
+ printf "á" > mbset; atf_check -o not-empty sed -ne '/[aàá]/p' mbset
+ printf "à" > mbset; atf_check -o not-empty sed -ne '/[abà]/p' mbset
+ printf "a" > mbset; atf_check -o not-empty sed -ne '/[abà]/p' mbset
+ printf "b" > mbset; atf_check -o not-empty sed -ne '/[abà]/p' mbset
+ printf "a" > mbset; atf_check -o not-empty sed -Ene '/[aà]/p' mbset
+ printf "à" > mbset; atf_check -o not-empty sed -Ene '/[aà]/p' mbset
+ printf "a" > mbset; atf_check -o not-empty sed -Ene '/[aàá]/p' mbset
+ printf "à" > mbset; atf_check -o not-empty sed -Ene '/[aàá]/p' mbset
+ printf "á" > mbset; atf_check -o not-empty sed -Ene '/[aàá]/p' mbset
+ printf "à" > mbset; atf_check -o not-empty sed -Ene '/[abà]/p' mbset
+ printf "a" > mbset; atf_check -o not-empty sed -Ene '/[abà]/p' mbset
+ printf "b" > mbset; atf_check -o not-empty sed -Ene '/[abà]/p' mbset
+}
+mbset_cleanup()
+{
+ rm -f mbset
+}
+
atf_init_test_cases()
{
atf_add_test_case bmpat
atf_add_test_case icase
+ atf_add_test_case mbset
}

File Metadata

Mime Type
text/plain
Expires
Fri, May 2, 3:07 AM (16 h, 59 m)
Storage Engine
blob
Storage Format
Raw Data
Storage Handle
17896954
Default Alt Text
D43149.id131715.diff (3 KB)

Event Timeline