Message-ID: <d5081a7f4f6d425c28dd69d1e09b2e78f149e726.camel@j-davis.com>
Subject: 17: lc_collate_is_c() when LC_CTYPE != LC_COLLATE
From: Jeff Davis <pgsql@j-davis.com>
To: pgsql-bugs@postgresql.org
Date: Wed, 06 Nov 2024 09:35:27 -0800
Content-Type: multipart/mixed; boundary="=-HSSOAUrGFybixha/2fXb"
User-Agent: Evolution 3.44.4-0ubuntu2 
MIME-Version: 1.0
Archived-At: 
 <https://www.postgresql.org/message-id/d5081a7f4f6d425c28dd69d1e09b2e78f149e726.camel%40j-davis.com>
Precedence: bulk

--=-HSSOAUrGFybixha/2fXb
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Due to a very unfortunate copy-and-paste error in 2d819a08a1, for the
libc provider, lc_collate_is_c() is checking the environment's LC_CTYPE
rather than LC_COLLATE :(

This sort of works if LC_COLLATE=3DC and LC_CTYPE=3Den_US.utf8 (or another
real locale), because lc_collate_is_c() returns false and it falls
through to strcoll(), which will observe LC_COLLATE.

But if LC_COLLATE=3Den_US.utf8 and LC_CTYPE=3DC, then it's wrong results.=
=C2=A0

By good fortune, I believe that the latter combination is fairly rare,
which may explain why it evaded manual testing until now. But it's
still pretty bad.

Trivial patch attached, which I intend to commit shortly to 17 only.
The bug does not exist in master due to recent refactoring.

Regards,
	Jeff Davis


--=-HSSOAUrGFybixha/2fXb
Content-Disposition: attachment;
	filename="v1-0001-Fix-lc_collate_is_c-when-LC_COLLATE-LC_CTYPE.patch"
Content-Type: text/x-patch;
	name="v1-0001-Fix-lc_collate_is_c-when-LC_COLLATE-LC_CTYPE.patch";
	charset="UTF-8"
Content-Transfer-Encoding: base64

RnJvbSAxNmY0NmFkODVhMzdkYTU5MGM4YmZiNDA4MWRlZTM4MDU3OGY2ODk2IE1vbiBTZXAgMTcg
MDA6MDA6MDAgMjAwMQpGcm9tOiBKZWZmIERhdmlzIDxqZWZmQGotZGF2aXMuY29tPgpEYXRlOiBU
dWUsIDUgTm92IDIwMjQgMTc6Mzg6MjIgLTA4MDAKU3ViamVjdDogW1BBVENIIHYxXSBGaXggbGNf
Y29sbGF0ZV9pc19jKCkgd2hlbiBMQ19DT0xMQVRFICE9IExDX0NUWVBFLgoKLS0tCiBzcmMvYmFj
a2VuZC91dGlscy9hZHQvcGdfbG9jYWxlLmMgfCA0ICsrLS0KIDEgZmlsZSBjaGFuZ2VkLCAyIGlu
c2VydGlvbnMoKyksIDIgZGVsZXRpb25zKC0pCgpkaWZmIC0tZ2l0IGEvc3JjL2JhY2tlbmQvdXRp
bHMvYWR0L3BnX2xvY2FsZS5jIGIvc3JjL2JhY2tlbmQvdXRpbHMvYWR0L3BnX2xvY2FsZS5jCmlu
ZGV4IGNkNWFkMWU3MDUuLjRjODVhMDFiMjggMTAwNjQ0Ci0tLSBhL3NyYy9iYWNrZW5kL3V0aWxz
L2FkdC9wZ19sb2NhbGUuYworKysgYi9zcmMvYmFja2VuZC91dGlscy9hZHQvcGdfbG9jYWxlLmMK
QEAgLTEzNzAsOSArMTM3MCw5IEBAIGxjX2NvbGxhdGVfaXNfYyhPaWQgY29sbGF0aW9uKQogCQl9
CiAJCWVsc2UgaWYgKGRlZmF1bHRfbG9jYWxlLnByb3ZpZGVyID09IENPTExQUk9WSURFUl9MSUJD
KQogCQl7Ci0JCQlsb2NhbGVwdHIgPSBzZXRsb2NhbGUoTENfQ1RZUEUsIE5VTEwpOworCQkJbG9j
YWxlcHRyID0gc2V0bG9jYWxlKExDX0NPTExBVEUsIE5VTEwpOwogCQkJaWYgKCFsb2NhbGVwdHIp
Ci0JCQkJZWxvZyhFUlJPUiwgImludmFsaWQgTENfQ1RZUEUgc2V0dGluZyIpOworCQkJCWVsb2co
RVJST1IsICJpbnZhbGlkIExDX0NPTExBVEUgc2V0dGluZyIpOwogCQl9CiAJCWVsc2UKIAkJCWVs
b2coRVJST1IsICJ1bmV4cGVjdGVkIGNvbGxhdGlvbiBwcm92aWRlciAnJWMnIiwKLS0gCjIuMzQu
MQoK


--=-HSSOAUrGFybixha/2fXb--