pkglistgen: Support working on git branches #3089
Conversation
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #3089 +/- ##
==========================================
- Coverage 28.02% 28.01% -0.02%
==========================================
Files 86 86
Lines 14981 14987 +6
==========================================
Hits 4199 4199
- Misses 10782 10788 +6 ☔ View full report in Codecov by Sentry. |
pkglistgen/tool.py
Outdated
| git_url = split[0] | ||
| args = ['git', 'clone', '--recurse-submodules', git_url, cache_dir] | ||
| if len(split) == 2: | ||
| args.append('--branch') |
There was a problem hiding this comment.
--single-branch as well (unconditionally)?
There was a problem hiding this comment.
we can do so ... added another commit for that as it is independend of the branch selection.
|
Should it also handle if the URL (any part) changes while the cache dir already exists? |
|
yes, but that can get complex as also the default branch can change server side... maybe we should do that with a second pull request, to avoid blocking Marco's request. |
That only impacts the initial clone though, so I don't see how that would be complex? If it's ok to discard the cache dir on mismatches, then |
|
Without support for cache invalidation it would currently keep pulling and pushing the wrong URL |
adrianschroeter
force-pushed
the
support_branches
branch
2 times, most recently
from
954969b
to
e378573
Compare
|
I added a 90% switch solution here, but it does not cover the situation when no branch is specified, but default branch switches server side. I reverted the --single-branch clone option therefore to be able to switch later on. |
pkglistgen/tool.py
Outdated
| if fragment: | ||
| # FIXME: this is not takeing account default branches, which may change server side | ||
| logging.debug(subprocess.check_output( | ||
| ['git', 'switch', fragment], cwd=cache_dir, encoding='utf-8')) |
There was a problem hiding this comment.
git switch is experimental and might not be available on ibs-slsa-botmaster
There was a problem hiding this comment.
Needs to be fragment[0] anyway
There was a problem hiding this comment.
the git on botmaster knows git switch
There was a problem hiding this comment.
also git on ibs-slsa-botmaster knows it
adrianschroeter
force-pushed
the
support_branches
branch
from
e378573
to
2e79964
Compare
|
@adrianschroeter @Vogtinator we need this for SLFO stagings. Is there any action to be taken or can this be merged? |
|
it is fine from my side. Please speak up if something needs to be changed. |
Following scmsync syntax where a branch can be specified behind the URL fragment hash Co-authored-by: Fabian Vogt <fabian@ritter-vogt.de>
Vogtinator
force-pushed
the
support_branches
branch
from
2e79964
to
feb1e3f
Compare
Following scmsync syntax where a branch can be specified behind the URL fragment hash