New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Repomanage tests, add srpm to modules, remove redundant packages (RhBug:1804720) #825
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can you please also add a scenario for case when there are both modular and non-modular packages of the same name in the repo?
Also, this is not related to the bug, but since you are also adding the basic tests for repomanage, could you also add a scenarios for the case when there are multiple files for the newest nevra and test that repomanage --new
will output both files and --repomanage new -k 2
will output both files and the next one? Because this makes repomanage different from just returning the latest packages.
|
||
Scenario: basic functionality of repomanage --new works without repodata | ||
Given I copy repository "dnf-ci-thirdparty-updates" for modification | ||
And I delete directory "/{context.dnf.repos[dnf-ci-thirdparty-updates].path}/repodata" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Although it works now, the first '/' shouldn't be there.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It actually has to be there, this way the path start with //
which means it doesn't get prepended with installroot, that is what we need since the {context.dnf.repos[dnf-ci-thirdparty-updates].path}
already contains the installroot path.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I really dislike this btw. Thinking about it maybe we could prepend installroot only if the path doesn't start with a format placeholder. Because I think it almost always should if it's not an installroot path.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I agree, that could work, paths not starting with format placeholder really shouldn't be pointing to the host and thus can be prepended.
But this pr is complex enough, so I don't want to do this kind of change here.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
You are right, my bad.
...have-tests/fixtures/specs/dnf-ci-fedora-modular/postgresql-9.6.8-1.module_1710+b535a823.spec
Outdated
Show resolved
Hide resolved
Rpm nodejs-1:8.14.0-1.module_2030+42747d41.x86_64 was not part of any module despite its name.
This cleans up repomanage output
I have added two new tests with duplicate packages, it turned out there was actually a bug in there, good catch. 👍
Changed repo to |
Well, now I just hope this indeed was the intended behavior. Thanks for adding the tests, looks good to me. |
While trying to have saner repomanage output I have noticed there are several unused packages which only make things more complicated.
There are also various fixes for modular metadata and modular packages.
https://bugzilla.redhat.com/show_bug.cgi?id=1804720