On Sep 14, 2012, at 8:31 AM, Elan Ruusamäe wrote:
> On 14.09.2012 15:18, baggins wrote:
>> @@ -0,0 +1,11 @@
>> +--- rpm-5.4.10/macros/macros.in~ 2012-09-12 13:52:19.305575500 +0200
>> ++++ rpm-5.4.10/macros/macros.in 2012-09-14 14:15:34.115044785 +0200
>> +@@ -778,7 +778,7 @@
>> + %pattern_Name ^[A-Za-z0-9][A-Za-z0-9+._-]*$
>> + %pattern_Epoch ^[0-9]+$
>> + %pattern_Version ^[A-Za-z0-9+._~]+$
>> +-%pattern_Release ^[A-Za-z0-9+._]+$
>> ++%pattern_Release ^[A-Za-z0-9+._@{}%]+$
>> + %pattern_Disttag ^[A-Za-z0-9]+$
>> + %pattern_Repotag ^[A-Za-z0-9]+$
>> + %pattern_Distepoch ^[A-Za-z0-9.]+$
>
> i'm wondering, are these patterns locale insensitive?
>
The specific answer is likely not locale insensitive, because PCRE is locale
sensitive.
The patterns are to catch malicious odd characters in patterns that script
kiddies blindly
extract from packages directly into scripts.
Meanwhile the *.spec is in the C locale (and all the usual localizations --
while permitted --
are ignored).
Of course there is noting stopping anyone from littering *.spec files with any
octets they wish, and the usual
Sh*t happens.
will ensue,
> i hope they are, otherwise under LC_ALL=et_EE letters 'tuv' won't be matched
>
*shrug*
Running rpmbuild in anything other than the C locale has no known benefit or
usage case.
73 de Jeff
_______________________________________________
pld-devel-en mailing list
[email protected]
http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
