Re: StGIT and repo-config

From: Catalin Marinas <catalin.marinas@gmail.com>
Date: Jan 26, 2007, 22:51 UTC
Message-ID: <b0943d9e0701261451y58062564yddad0a7471e0eff6@mail.gmail.com>
In-Reply-To: <20070126175302.GG4083@nan92-1-81-57-214-146.fbx.proxad.net>

On 26/01/07, Yann Dirson <ydirson@altern.org> wrote:

Show 9 quoted lines

> On Fri, Jan 26, 2007 at 12:58:38AM +0100, Johannes Schindelin wrote:
> > On Thu, 25 Jan 2007, Catalin Marinas wrote:
> > > The dictionary should also be pre-populated with the default values
> > > (only in memory, not going to the config file on disk).
>
> Pre-filling with default value has a drawback: it will create lots of
> useless entries, especially for per-branch settings.  I'd rather let
> the accessors return the default value when needed.  We can still
> group all defaults in a single dictionnary.

What I meant is that it should pre-fill only with the default entries needed by StGIT (probably around 10, see stgit.config.config_setup()). The rest should be filled in when reading the config files.

Show 8 quoted lines

> > The most important point (to me) which came out of the discussion: It is
> > not at all easy, or straight-forward, to handle multi-vars, i.e. multiple
> > values for the same key.
>
> Right, at least for filling a dictionnary.  We would need to declare
> multi-valued parameters as such, which basically means we must only
> try to read those config items we know about, which has all sorts of
> consequences for config.py :)

But are multi-vars in the same file currently used? There is a clear situation when a key is written in both the global and the per-repository config files. From the StGIT point of view, it should only care about the most specific one, i.e. the one in the per-repo config and this would be the last one listed by git-repo-config (and therefore overriding any preceding dictionary entry).

> It would seem reasonable to start without a cache dictionnary, at
> least for now.  After all, there are not so many config items to know
> about in a single stgit run, so IIMHO we're only going to notice a
> difference for the time needed to run the testsuite.

I'd like to avoid calling git-repo-config for every config option, even if this wouldn't be noticeable (initially, but the tool would evolve). If we don't care about multi-vars, as I said above, a dictionary cache would make sense.

-- 
Catalin