git-svn revision support #169
Conversation
|
I haven't reviewed this yet; in the way of that is that I don't have a good git-svn repo to test with (at hand, I do have archives laying about). Don't worry about the XIBs. They don't scare me. Ideally, I think the column should be hidden (unavailable) unless the data is actually present; and probably not shown by default unless there is actually a git-svn remote configured; as in my experience, the commits survive indefinitely after a project has migrated entirely to git. |
This patch adds parsing of the output of git log to extract the SVN revision number tied to each commit. The revision number is part of the git-svn-id string.
This patch adds the ability to query the repo for the presence of any SVN remotes. This is determined by whether the string "svn-remote" is present in the repository's .git/config file.
This patch causes the "Git SVN Revision" table column to be removed when a repository has no SVN remote.
|
Okay, so this should bring the pull request up to date with your comments. I've rebased against the current master and added the auto-hiding of the column for repositories with no SVN remote configured. I think there's probably a better way to go about this, but I'm not familiar enough with GitX's source to know if that's true or not. |
|
Commits look good; I'll just run it over my test repos and this should be good to make it into the next build. |
|
|
@kfb FYI, Objective-Git does provide a facility for reading all remote properties; it's just through the GTConfiguration, not GTRemote; see my changes in PBGitRepository.m if you're curious :-) |
Add a "raw" mode to the search filters
I do a lot of reviewing of code hosted in an SVN repo; developers will often reference the SVN revision number of changes. This patch adds another column to the history view that shows the SVN revision number, extracted from the
git-svn-idline. This allows for quickly mapping between git and SVN commits.In quick testing, performance doesn't seem to be heavily impacted, but if others find differently it could be an off-by-default option.
(Apologies also for the XIB diff spam, I'm never sure which of the bits IB changes are actually required, so I played it safe and committed the whole thing)