On Thu, 2010-06-17 at 08:27 -0700, JD wrote: > > On 06/17/2010 12:50 AM, Panu Matilainen wrote: > > On Wed, 16 Jun 2010, seth vidal wrote: > > > >> On Wed, 2010-06-16 at 09:44 -0400, Adam Jackson wrote: > >>> On Wed, 2010-06-16 at 08:53 -0400, seth vidal wrote: > >>>> On Wed, 2010-06-16 at 00:30 -0700, JD wrote: > >>>>> Greetings all, > >>>>> Is there a way to query the database to list the packages > >>>>> that depend on a given package? > >>>>> I checked the man page, and I do not see such an option. > >>>>> I see the option --requires, which is great - but would like the > >>>>> converse of --requires, such as: --required_by > >>>> > >>>> rpm -q --whatrequires pkgname will tell you what specifically requires > >>>> that pkg name - but not all the things that pkg provides. > >>> > >>> For that, you would do: > >>> > >>> % rpm --quiet -q --whatrequires $(rpm -q --provides pkgname) | sort -u > >>> > >>> The uniquifying step being because rpm will print the list of consumers > >>> for _each_ thing provided by pkgname. > >> > >> but you'll need to do every file in the pkg, too. > >> > >> b/c of file-requires. > > > > One possibility is "abusing" --test with erasure, eg: > > $ rpm -e --test <pkg> > > > > To get just the depending package names something like this works: > > $ rpm -e --test <pkg> 2>&1 | tail -n +2 |awk '{print $NF'} > > > > - Panu - > Yes - that works nicely > rpm -e --test libguestfs 2>&1 | tail -n +2 |awk '{print $NF'} > libguestfs-java-1:1.2.9-2.fc13.i686 > ocaml-libguestfs-1:1.2.9-2.fc13.i686 > perl-libguestfs-1:1.2.9-2.fc13.i686 > guestfish-1:1.2.9-2.fc13.i686 > python-libguestfs-1:1.2.9-2.fc13.i686 > ruby-libguestfs-1:1.2.9-2.fc13.i686 > libguestfs-java-1:1.2.9-2.fc13.i686 > ocaml-libguestfs-1:1.2.9-2.fc13.i686 > perl-libguestfs-1:1.2.9-2.fc13.i686 > guestfish-1:1.2.9-2.fc13.i686 > libguestfs-tools-1:1.2.9-2.fc13.i686 > libguestfs-javadoc-1:1.2.9-2.fc13.i686 > libguestfs-java-devel-1:1.2.9-2.fc13.i686 > python-libguestfs-1:1.2.9-2.fc13.i686 > ruby-libguestfs-1:1.2.9-2.fc13.i686 > > But why is the name slighly mangled with the insertion of 1: > into the name? What's the usefulness of that? > that's the package epoch. 3 values make up the package versioning: epoch version release that's the epoch. -sv _______________________________________________ Rpm-list mailing list Rpm-list@xxxxxxxxxxxxx http://lists.rpm.org/mailman/listinfo/rpm-list