[PATCH] Move alias into core

Brendan Cully brendan at kublai.com
Wed May 20 11:15:05 CDT 2009


On Wednesday, 20 May 2009 at 09:15, Greg Ward wrote:
> On Wed, May 20, 2009 at 3:24 AM, Brendan Cully <brendan at kublai.com> wrote:
> > On Monday, 18 May 2009 at 19:49, Brendan Cully wrote:
> >> # HG changeset patch
> >> # User Brendan Cully <brendan at kublai.com>
> >> # Date 1242701320 25200
> >> # Node ID ff2c73f240604842f214e44e595726357c22f2ee
> >> # Parent  252232621165917755727729c7f0b9a1f1263668
> >> Move alias into core.
> >
> > Here's an improved version with better error reporting.
> [...]
> > +echo '% bad aliases'
> > +hg heads
> 
> IIUC, this call is just to provoke the "bad alias" warnings, right?
> And those warnings will happen on any hg invocation?  Then how about
> just "hg > /dev/null"?  That'll test that the warnings are printed to
> stderr, that they are correct, and won't clutter up the test output
> with irrelevant stuff.

Not quite. The heads call is doing double duty to check how aliases
that shadow existing commands behave, and here the "heads = status"
definition should mean that hg heads actually produces status output.

> As for the format of the warnings: how about "warning: ..." rather
> than "***: ...".  That's pretty standard.  (Hmmm: why does ui.warn()
> do that for us?)

I'll remove the ***. I'd like this to be automated too, but I don't
think ui.warn cleanups belong in the alias patch.


More information about the Mercurial-devel mailing list