[PATCH 1 of 3] rebase: add a 'revtodo' constant to

[Mads Kiilerich]

On 12/02/2014 09:32 PM, Pierre-Yves David wrote:
> # HG changeset patch
> # User Pierre-Yves David <pierre-yves.david at fb.com>
> # Date 1417542380 28800
> #      Tue Dec 02 09:46:20 2014 -0800
> # Node ID 0e50f872225ac6082796871f236c05c2e51a4d25
> # Parent  6b2953028561f9b49fb79e975d68e85e8666f13a
> rebase: add a 'revtodo' constant to

to what?

> The state mapping is using '-1' to mark revision that have not been rebased yet.
> We introduce and use a constant for that purpose. This will help emphasising the
> fact the value mean something else than nullrev.
>
> diff --git a/hgext/rebase.py b/hgext/rebase.py
> --- a/hgext/rebase.py
> +++ b/hgext/rebase.py


> @@ -808,11 +809,11 @@ def inrebase(repo, originalwd, state):
>   
>       return False
>   
>   def abort(repo, originalwd, target, state):
>       'Restore the repository to its original state'
> -    dstates = [s for s in state.values() if s > nullrev]
> +    dstates = [s for s in state.values() if s > revtodo]

I am not sure revtodo is more correct here than nullrev. If you don't 
want to rely on nullrev=0-1 then I think it would be more correct to use 
 >= 0 to filter out all normal revision numbers. That is also what is 
used in clearrebased.

> @@ -831,11 +832,11 @@ def abort(repo, originalwd, target, stat
>           # Update away from the rebase if necessary
>           if inrebase(repo, originalwd, state):
>               merge.update(repo, repo[originalwd].rev(), False, True, False)
>   
>           # Strip from the first rebased revision
> -        rebased = filter(lambda x: x > -1 and x != target, state.values())
> +        rebased = filter(lambda x: x > revtodo and x != target, state.values())

Again, when we don't rely on revtodo==nullrev=-1, wouldn't it be more 
correct to use >= 0?

/Mads