"David Glasser" <glasser_at_davidglasser.net> writes:
> *well*. I would be tempted to write the whole function without the
> whole "optimize things if everything is readable by checking
> readable_paths != paths" thing, since I think it would be a whole lot
> simpler to always make a new readable_paths array (in the part above
> your hash), and to cache all the return values from the authz function
> in the first half in some other hash. Seems simpler than the twisty
> conditional logic you have here, and the cost of copying one array of
> strings shouldn't be too high. ie:
I know what you mean. I took that approach from the existing code in
that function, which already worked in the complex way (hey, always
blame the other one! He did it first!). Testing the change is my
priority now; I'd rather not rewrite the logic until there's a test.
Thanks for looking at this so quickly,
-K
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe_at_subversion.tigris.org
For additional commands, e-mail: dev-help_at_subversion.tigris.org
Received on 2008-01-10 06:12:17 CET