Introducing new handlers to maintain an rpm dep chain. #15
Conversation
|
|
||
| # TODO - look for any further special psuedo-error handling cases here. | ||
| import pprint | ||
| pprint.pprint(body) |
There was a problem hiding this comment.
do we want to keep this in for now?
There was a problem hiding this comment.
I should probably change it to log.something(body).
I'd like to keep it because I'm not sure what kind of error PDC might throw back at me. If it is a new error I've never seen before, I'd like to know the details in the logs so I can make an appropriate patch to handle it in the future.
|
Ok, nothing widly standing out for me, but I must confess I'm not sure to follow everything and I might have been running low on coffee when I reviewed this one |
| return ['buildsys.tag'] | ||
|
|
||
| def can_handle(self, msg): | ||
| if not msg['topic'].endswith('buildsys.tag'): |
There was a problem hiding this comment.
I'm wondering if it'd be a good idea to use 'self.topic_suffixes' instead of hardcoding 'buildsys.tag'. Not a big deal for now, but as things grow, it might be nice to just update one location for what is supported in this handler.
| pdcupdater.utils.ensure_release_exists(pdc, release_id, release) | ||
|
|
||
| koji_relationships = self._yield_koji_relationships(pdc, tag) | ||
| for parent, type, child in koji_relationships: |
There was a problem hiding this comment.
Same as above, can we rename type to something like relationship_type to not override the builtin type function.
|
@ralphbean looks good. I didn't look too closely at the tests though. Just a few comments to address and then go ahead and merge it. |
|
OK, after lots of refactoring, this is much faster. It used to take ~1 hour to process a single build. Now it takes ~1 minute. :) |
ralphbean
force-pushed
the
feature/rpm-dep-chain
branch
from
fd8a913
to
a136836
Compare
This is the first part of the puzzle for https://fedoraproject.org/wiki/User:Ralph/Drafts/Infrastructure/Factory2/ModellingDeps