Re: Description-less packages file

[Andreas Tille <andreas@an3as.eu>]

[Ansgar as OP in d-d-a in CC - please tell me where to direct this kind
 of questions to]

Hi,

when trying to import the new Translation files into UDD I realised that
the files are containing useless duplicates with no reason.  It might
possibly not harm current applications but it might perfectly be a
result of an unwanted effect and thus we should care.  I realised this
problem when working on a new importer which has a constraint like this:

   PRIMARY KEY (package, release, language, description, description_md5)

I get unimportable data sets for instance these:

$ wget -q http://ftp.debian.org/debian/dists/sid/main/i18n/Translation-en.bz2

$ bzgrep -B1 "ea55aa04a24ea847c61e4865c135347b" Translation-en.bz2
Package: gnome-control-center-dev
Description-md5: ea55aa04a24ea847c61e4865c135347b
--
Package: gnome-control-center-dev
Description-md5: ea55aa04a24ea847c61e4865c135347b

$ bzgrep -B1 "0cf5a8771cafcd619336f37af75b9a7c" Translation-en.bz2
Package: gnutls-doc
Description-md5: 0cf5a8771cafcd619336f37af75b9a7c
--
Package: gnutls-doc
Description-md5: 0cf5a8771cafcd619336f37af75b9a7c

$ bzgrep -B1 "2150078ff0e4ffed278ad048818feef8" Translation-en.bz2
Package: calendar-google-provider
Description-md5: 2150078ff0e4ffed278ad048818feef8
--
Package: calendar-google-provider
Description-md5: 2150078ff0e4ffed278ad048818feef8


These are three examples out of 8 I found when importing the current data set.

Any chance to prevent these duplicates?

Kind regards

       Andreas.

-- 
http://fam-tille.de