Please note that the type identifiers in Taskotron are all lowercase (it's not obvious from the documentation, because it's an enum class documentation, but the values are lowercase strings), maybe this could be the same? So koji_build, bodhi_update and pagure_pr.
type
koji_build
bodhi_update
pagure_pr
1 new commit added
Adjusted :)
This happens a lot, in reality, though https://taskotron.fedoraproject.org/resultsdb/results?testcases:like=org.centos.prod.ci.pipeline.*
Might be worth caching the data, as this gets called for the same result_id many times in sequence.
These aren't taskotron results, so you're not interested in these :)
Good idea, we can add dogpile in there
This method is called for any/all result_id though, from rats/lib/tasks.py lines 20/21, IIUIC. Not that this is any problem for how this works (it is IMO right), I just wanted to make sure the comment is not misleading for anybody that goes to optimize/restructure the code in the future :)
Would a rephrase to This should never happen in theory for taskotron's messages work or should I make it longer?
This should never happen in theory for taskotron's messages
2 new commits added
Comment adjusted and caching added.
Do we want anything else?
3 new commits added
5 new commits added
Ok, this has been open for review for a while now but it seems it won't be reviewed :(
Since this application hasn't been released yet, I'm inclined to just go ahead and merge it, so we can move further ahead with it and test it in stg :)
Pull-Request has been merged by pingou