Patchwork [4,of,6,V3] merge: consider successor changesets for a bare update

login
register
mail settings
Submitter Sean Farley
Date Jan. 16, 2014, 12:34 a.m.
Message ID <97551cab431afbaa04f8.1389832459@laptop.local>
Download mbox | patch
Permalink /patch/3335/
State Superseded
Commit 95b9c6149e1711d4871772dc819d63252164da25
Headers show

Comments

Sean Farley - Jan. 16, 2014, 12:34 a.m.
# HG changeset patch
# User Sean Farley <sean.michael.farley@gmail.com>
# Date 1389825678 21600
#      Wed Jan 15 16:41:18 2014 -0600
# Node ID 97551cab431afbaa04f85ba0278e800b026dc74d
# Parent  c545df40daa10ff9200c81ba566db6329175280b
merge: consider successor changesets for a bare update

Previously, a bare update would ignore any successor changesets thus
potentially leaving you on an obsolete head. This happens commonly when there
is an old bookmark that hasn't been moved forward which is the motivating
reason for this patch series.

Now, we will check for successor changesets if two conditions hold: 1) we are
doing a bare update 2) *and* we are currently on an obsolete head.

If we are in this situation, then we calculate the branchtip of the successor
set and update to that changeset.

Tests coverage has been added.
Pierre-Yves David - Jan. 16, 2014, 1:11 a.m.
On 01/15/2014 04:34 PM, Sean Farley wrote:
> # HG changeset patch
> # User Sean Farley <sean.michael.farley@gmail.com>
> # Date 1389825678 21600
> #      Wed Jan 15 16:41:18 2014 -0600
> # Node ID 97551cab431afbaa04f85ba0278e800b026dc74d
> # Parent  c545df40daa10ff9200c81ba566db6329175280b
> merge: consider successor changesets for a bare update
>
> Previously, a bare update would ignore any successor changesets thus
> potentially leaving you on an obsolete head. This happens commonly when there
> is an old bookmark that hasn't been moved forward which is the motivating
> reason for this patch series.
>
> Now, we will check for successor changesets if two conditions hold: 1) we are
> doing a bare update 2) *and* we are currently on an obsolete head.
>
> If we are in this situation, then we calculate the branchtip of the successor
> set and update to that changeset.
>
> Tests coverage has been added.
>
> diff --git a/mercurial/merge.py b/mercurial/merge.py
> --- a/mercurial/merge.py
> +++ b/mercurial/merge.py
> @@ -694,10 +694,32 @@ def update(repo, node, branchmerge, forc
>               except error.RepoLookupError:
>                   if wc.branch() == "default": # no default branch!
>                       node = repo.lookup("tip") # update to tip
>                   else:
>                       raise util.Abort(_("branch %s not found") % wc.branch())
> +
> +            if p1.obsolete() and not p1.children():
> +                # allow updating to successors
> +                successors = obsolete.successorssets(repo, p1.node())
> +
> +                # check for divergence
> +                if len(successors) > 1:
> +                    raise util.Abort(_("candidate changeset is divergent"),
> +                                     hint=_("use 'hg evolve' to resolve"
> +                                            "divergence"))

(1) evolve is not part of core mercurial, do not talk about it in error 
message yet

(2) you are one an old obsolete precursors of the actual divergent 
changesets, hg evolve will not solve the divergence from you anyway

(3) we could pick one of the two and go there (update then warning you 
are one a divergent changeset) the same we do when descendant have two 
heads.

> +
> +                # we're not divergent, so we only have one element
> +                successors = successors[0]

Remember that earlier part in the movie when the old sage fooled you 
with false information about `[()]`?

You probably want a unittest for that case too. That would have help you 
to spot that betrayal.

> +
> +                # if successors is empty here, then it has been pruned and
> +                # there is nothing to consider updating to
> +                if successors:
> +                    # get the max revision for the given successors set
> +                    unfil = [repo.unfiltered()[n].rev() for n in successors]

Why are you using unfiltered repo here ?

> +                    node = repo[max(unfil)].node()
> +                    pa = p1

What about repo.revs("max(%ln)", successors)[0] ?
Pierre-Yves David - Jan. 16, 2014, 1:14 a.m.
On 01/15/2014 04:34 PM, Sean Farley wrote:
> +                # if successors is empty here, then it has been pruned and
> +                # there is nothing to consider updating to
Well, we could consider considering the first non obsolete parent. of 
the successor of the first parent with one.
This could(must!) be done in another patches.

Patch

diff --git a/mercurial/merge.py b/mercurial/merge.py
--- a/mercurial/merge.py
+++ b/mercurial/merge.py
@@ -694,10 +694,32 @@  def update(repo, node, branchmerge, forc
             except error.RepoLookupError:
                 if wc.branch() == "default": # no default branch!
                     node = repo.lookup("tip") # update to tip
                 else:
                     raise util.Abort(_("branch %s not found") % wc.branch())
+
+            if p1.obsolete() and not p1.children():
+                # allow updating to successors
+                successors = obsolete.successorssets(repo, p1.node())
+
+                # check for divergence
+                if len(successors) > 1:
+                    raise util.Abort(_("candidate changeset is divergent"),
+                                     hint=_("use 'hg evolve' to resolve"
+                                            "divergence"))
+
+                # we're not divergent, so we only have one element
+                successors = successors[0]
+
+                # if successors is empty here, then it has been pruned and
+                # there is nothing to consider updating to
+                if successors:
+                    # get the max revision for the given successors set
+                    unfil = [repo.unfiltered()[n].rev() for n in successors]
+                    node = repo[max(unfil)].node()
+                    pa = p1
+
         overwrite = force and not branchmerge
 
         p2 = repo[node]
         if pa is None:
             pa = p1.ancestor(p2)
diff --git a/tests/test-update-branches.t b/tests/test-update-branches.t
--- a/tests/test-update-branches.t
+++ b/tests/test-update-branches.t
@@ -227,10 +227,18 @@  the bookmark (issue4015)
   $ hg book bm
   moving bookmark 'bm' forward from 6efa171f091b
   $ hg bookmarks
    * bm                        5:ff252e8273df
 
+Test that 4 is detected as the no-argument destination from 3
+  $ hg up --quiet 0          # we should be able to update to 3 directly
+  $ hg up --quiet --hidden 3 # but not implemented yet.
+  $ hg up
+  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
+  $ hg id
+  d047485b3896+ (b1)
+
 Test that 5 is detected as a valid destination from 1
   $ hg up --quiet 0          # we should be able to update to 3 directly
   $ hg up --quiet --hidden 3 # but not implemented yet.
   $ hg up 5
   1 files updated, 0 files merged, 0 files removed, 0 files unresolved