Revision 0d182c079b200633b234b9a8216aba155384f21b

[Ka-Hing Cheung]

On Sat, 2007-09-08 at 22:19 -0500, Richard Laager wrote:
> Is there some reason that purple_str_to_time() couldn't be extended to
> accept that format? If not, then is your timezone code a direct
> copy-and-paste from that function? If so, could we abstract that portion
> out into an internal function? If nothing else, you can at least compare
> your code to that function if you haven't already (and I don't mean to
> suggest that you didn't--I didn't look at the code closely).

I didn't extend purple_str_to_time because I don't want to make its
logic more complicated than it already is. The way it's written also
makes it very forgiving, which I think is bad because it can give a
seemingly valid time that's actually way off (for example, if the year
is missing, it will happily return the rest with the current year tagged
on). I don't know if there are things depending on this forgiving
behavior.

We can certainly abstract the timezone part out, I just didn't want to
because I don't fully understand it.

-khc