Rob Crittenden <rcrit...@redhat.com> wrote: > Rob Crittenden wrote: > > Jan Zelený wrote: > >> Jan Zelený<jzel...@redhat.com> wrote: > >>> Rob Crittenden<rcrit...@redhat.com> wrote: > >>>> Yi found a tricky way to remove required attributes that aren't > >>>> required > >>>> in the schema. The problem was we weren't enforcing parameter.required > >>>> in mods (because it was enforcing that every variable with required be > >>>> provided). > >>>> > >>>> I added a new check routine that is executed after setattr/addattr > >>>> does its work and verifies that no required parameters get skipped. > >>>> > >>>> ticket 852 > >>>> > >>>> rob > >>> > >>> Looks fine, works as expected. ACK > >>> > >>> I'm just not sure whether is is necessary to call the function twice > >>> - once > >>> on self.params and once on self.obj.params (I get the latter one, but > >>> I'm > >>> not sure whether the former one is necessary). > > > > Hmm, you may be right. I did it in case any of self.params had a > > requires on it, but since this is a mod operation then I think by > > definition it can't. > > > >>> Jan > >> > >> One more thing - I'm not sure whether it is necessary to add the check > >> to LDAPCreate - I tried to create role with empty description and it > >> failed as > >> expected. > > > > I think you're. I did it to prevent something like this: > > > > # ipa group-add --desc='foo' --setattr description='' foo > > > > but it is already handled. > > > > I'll work up a new patch. > > > > rob > > Updated patch attached. > > rob
ack Jan _______________________________________________ Freeipa-devel mailing list Freeipa-devel@redhat.com https://www.redhat.com/mailman/listinfo/freeipa-devel