search functionality in translate.fpo requires translate-toolkit, having an unpackaged python-enchant dependency.
skvidal okay - well we need python-enchant in epel for translate-toolkit
skvidal okay, not in cvs
skvidal looking in bugzilla
skvidal okay - so - it's roozbeh's package
skvidal nothing in bugzilla about a branch request
Re-implemented search using msggrep instead of the translate toolkit, so the dependency is no longer there.
This (of course) brought up new bugs. Most obvious is that we get [http://translate.fedoraproject.org/search/?s=foo&language=POT&msgid=True&msgstr=True&ignore-case=true&highlight=true a bunch of warnings about the locale charset evnironment] during the execution of msgmerge:
msggrep: warning: Locale charset "ANSI_X3.4-1968" is different from
input file charset "UTF-8".
Output of 'msggrep' might be incorrect.
Possible workarounds are:
- Set LC_ALL to a locale with encoding UTF-8.
- Convert the translation catalog to ASCII using 'msgconv',
then apply 'msggrep',
then convert back to UTF-8 using 'msgconv'.
Any pointers for help or patches greatly appreciated.
roozbeh, would you mind branching python-enchant for EPEL[4-5]? We need it for http://translate.fedoraproject.org/
Doah, I guess we don't need it. (Sorry Roozbeh)
Glezos, are those warnings safe to ignore?
Any idea where the local charset warning is coming from? Its not the local charset we're using on the servers :)
Closing this bug as python-enchant isn't needed any more.
I've removed search completely until I find the time to implement it properly.
to comment on this ticket.