Thread (1 message) 1 message, 1 author, 2016-06-15

Re: [PATCH/RFC] git-svn: don't create master if another head exists

From: Junio C Hamano <hidden>
Date: 2016-06-15 22:54:10

Eric Wong [off-list ref] writes:
Marcin Owsiany [off-list ref] wrote:
quoted
git-svn insists on creating the "master" head (unless it exists) on every
"fetch". While it is useful that it gets created initially (users expect this
git convention), some users find it annoying that it gets recreated, especially
when they would like the git branch names to follow SVN repository branch
names. More background in
http://thread.gmane.org/gmane.comp.version-control.git/115030

Make git-svn skip the "master" creation if there is another head ref pointing
to the same place. This means "master" does get created on initial "clone" but
does not get recreated once a user deletes it.
Sounds reasonable to me.  Thanks for following up on this after all this
time :)
quoted
Signed-off-by: Marcin Owsiany <redacted>
---
 git-svn.perl |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/git-svn.perl b/git-svn.perl
index 0b074c4..90f3d06 100755
--- a/git-svn.perl
+++ b/git-svn.perl
@@ -1613,6 +1613,8 @@ sub post_fetch_checkout {
 	}
 
 	my $valid_head = verify_ref('HEAD^0');
+	my @heads_commits = eval { command(qw(show-ref --heads --hash)) };
+	return if $valid_head and grep { $_ eq $valid_head } @heads_commits;
I (and I believe most git hackers) prefer C-style "&&" for boolean
expressions:

	return if $valid_head && grep { $_ eq $valid_head } @heads_commits;

"and" is lower precedence and best reserved for control flow:

	-f $file and print "File: $file exists!\n";

There's no logical difference in this case, but "&&" is probably easier for
C programmers to read.

I'll just swap "and" for "&&" and push unless there's any objection from
you.
I personally do not like either (I would generally avoid the
statement modifiers unless the condition is really trivial and it is
about an exceptional case) from the style point of view, but more
importantly, I do not understand what the new check is trying to do,
and I am not convinced if it is the whole solution to "I do not want
'master', as I am using something else" problem (please see my two
review messages).
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help