When using the 'report' subcommand of fedora-openqa-schedule,
currently you have to pass job IDs. If something went wrong
with submitting the results for a build at the time the tests
ran and you want to do it manually, or you need to test this
for some reason, it'd obviously be nice to be able to report
by build ID as well. This allows for that, in a pretty light
weight way: you can simply pass build identifiers as well as
job IDs in the list of things to be reported. We figure any
value which is just digits is a job ID, and anything else is
a build ID. There isn't any error checking here, so if you
pass an invalid build ID that doesn't exist it's going to sit
there pointlessly and wait for it to exist for 8 hours, but
since this is something that's really only used by devs/admins
I figure it's not really necessary to over-complicate it.
Details
Details
Try reporting (well, use -ns so there's no actual
wiki submission) single job IDs, single builds, multiples of
each, and mixed lists, make sure it all works as expected.
Diff Detail
Diff Detail
- Repository
- rOPENQA fedora_openqa
- Lint
Automatic diff as part of commit; lint not applicable. - Unit
Automatic diff as part of commit; unit tests not applicable.