Re: reviving a git-svn clone
- From
- martin f krafft <madduck@madduck.net>
- Date
- Apr 21, 2009, 08:14 UTC
- Message-ID
- <20090421081457.GC28007@lapse.rw.madduck.net>
- In-Reply-To
- <1240296712.3424.8.camel@maia.lan>
also sprach Sam Vilain <sam@vilain.net> [2009.04.21.0851 +0200]:
Show 6 quoted lines
> eg for this use case something like this to copy the
> refs/remotes/origin/* refs to refs/remotes/upstream/*:
>
> git for-each-ref --format="%(refname)" refs/remotes/origin/* |
> perl -le '$x = $_; $x =~ s{origin}{upstream};
> system "git update-ref $x $_"'Let me see if I parse this right: you want me to put the same SHA into refs/remotes/upstream/trunk as there is in refs/remotes/origin/upstream (since the upstream branch tracks the trunk)?
> It's a bit hard to test without access to the upstream svn!
Sorry, I should have been more careful. You can access the upstream SVN at http://svn.tuxonice.net/svn/hibernate-script . I am on a train, so I cannot verify that, but the exact URL is listed on the http://svn.tuxonice.net/ website.
One reason why I was a bit hesitant to change the URL from svn+ssh to http was because my study of the code base suggested, that git-svn actually matches the git-svn-id in the last commit message with the URL in the svn-remote config stanza, and recurses the commits parents until it finds a match. The error I am seeing actually seems to occur when git-svn recursed all the way to the first commit without a match.
Then again, I may well have misparsed the Perl code, so feel free to ignore me.
Many thanks,
-- martin | http://madduck.net/ | http://two.sentenc.es/ all of you that believe in telekinetics, raise my hand! spamtraps: madduck.bogus@madduck.net