Re: svn commit: r26006 - in trunk/subversion/bindings/swig: include ruby/libsvn_swig_ruby ruby/svn

Hi,

> changed the apis. I just did a diff between the 1.4.x branch and
> trunk and I think update_editor has a new required argument (perhaps
> just a rename, though?).

Ah, it's not good... Sorry.
We should define update_editor2 for Wc.get_update_editor3
and keep update_editor API. Could you fix my faults?

> > Of cause we should validate arguments of hash argument
> > version methods (xxx_yyy2). We'll need to use the following
> > method:
> >
> > class Hash
> > def assert_valid_keys(*valid_keys)
> > unknown_keys = keys - [valid_keys].flatten
> > unless unknown_keys.empty?
> > raise(ArgumentError, "Unknown key(s): #{unknown_keys.join(", ")}")
> > end
> > end
> > end
> >
> > # The above method is quoted from ActiveSupport.
> >
>
> I'm not familiar with this. Is it just to ensure that there were no
> typos in the passed in options?

Yes.

> I was thinking it might be nice just
> to ignore unrecognized options. I haven't given it a lot of thought,
> but it seemed like clients might be able to create their own options
> hashes, and their might be room for reuse if we ignored unrecognized
> keys.

In my experience, there aren't many cases to reuse option
hash but there are many typos in scripts. :<
I think a case that we trust our script's logic doesn't have
any problems but the script doesn't work is too costly (we
will use much time and be damaged our mind). And most of
causes of the case are caused by a typo!

So I want to validate hash options. We can remove validation
after many people request.

Thanks,

--
kou
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org