[PATCH 1 of 4] histedit: add a method to cleanup nodes safely

[Jun Wu]

Excerpts from Yuya Nishihara's message of 2017-03-17 22:32:45 +0900:
> On Mon, 13 Mar 2017 22:36:16 -0700, Jun Wu wrote:
> > # HG changeset patch
> > # User Jun Wu <quark at fb.com>
> > # Date 1489464645 25200
> > #      Mon Mar 13 21:10:45 2017 -0700
> > # Node ID b4cf155f7a41ebf314407000f6948716ae0a64e2
> > # Parent  3d3109339b57341b333c1112beb41dd281fa944a
> > # Available At https://bitbucket.org/quark-zju/hg-draft 
> > #              hg pull https://bitbucket.org/quark-zju/hg-draft  -r b4cf155f7a41
> > histedit: add a method to cleanup nodes safely
> 
> The series looks good to me, but one nit.
> 
> > diff --git a/hgext/histedit.py b/hgext/histedit.py
> > --- a/hgext/histedit.py
> > +++ b/hgext/histedit.py
> > @@ -1582,4 +1582,29 @@ def cleanupnode(ui, repo, name, nodes):
> >              repair.strip(ui, repo, c)
> >  
> > +def safecleanupnode(ui, repo, name, nodes):
> > +    """strip or obsolete nodes
> > +
> > +    nodes could be either a set or dict which maps to replacements.
> > +    nodes could be unknown (outside the repo).
> > +    """
> > +    supportsmarkers = obsolete.isenabled(repo, obsolete.createmarkersopt)
> > +    if supportsmarkers:
> > +        if util.safehasattr(nodes, 'get'):
> > +            # nodes is a dict-like mapping
> > +            # use unfiltered repo for successors in case they are hidden
> > +            urepo = repo.unfiltered()
> > +            marker = lambda x: (repo[x], (urepo[n] for n in nodes.get(x, ())))
>                                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 
> This isn't a tuple but a generator. Is that by intention?

Yes. Since obsolete.createmarkers only requires an "iterable", I think a
generator may make the space complexity from O(N) to O(1). Another reason is
the code fits in one line.