Strip aborts: no node!

David Soria Parra sn_ at gmx.net
Mon Nov 3 12:21:18 CST 2008


Sorry for that late answer.

So first, thank you for the patch. Actually I guess ignoring the 
LookupError is not quite what you want as this would not just ignore the 
error during strip, but actually every LookupError during pulls, etc. So 
frankly, I think there are better ways to fix that issue. I'm currently 
investigating the issue and I'll come up with a patch for that in the 
next days. But thank you very much for your observations. It's quite a 
good point to start from.

David

Andrei Vermel wrote:
> Following patch fixes the problem for me.
> 
> # HG changeset patch
> # User Andrei Vermel <avermel at mail.ru>
> # Date 1225557354 -10800
> # Node ID 112b0f35370495de4d7889c2c2f1ef5cf126b898
> # Parent  30300b14ecd6499467866204ca6ddd664b75f4d6
> Fix strip to abort with 'no node' when stripping a sibling of a current
> changeset
> 
> diff --git a/hgext/bookmarks.py b/hgext/bookmarks.py
> --- a/hgext/bookmarks.py
> +++ b/hgext/bookmarks.py
> @@ -202,7 +202,8 @@
>          def addchangegroup(self, source, srctype, url, emptyok=False):
>              try:
>                  onode = repo.changectx('.').node()
> -            except RepoError, inst:
> +            except (RepoError, LookupError), inst:
> +                onode = None
>                  pass
>  
>              result = super(bookmark_repo, self).addchangegroup(
> 



More information about the Mercurial-devel mailing list