[Chicago-talk] What module will we attack?
frag at ripco.com
Wed Aug 11 12:38:24 CDT 2004
On Wed, 11 Aug 2004, Andy Lester wrote:
> On Wed, Aug 11, 2004 at 11:14:48AM -0500, Shawn Carroll (shawn.c.carroll at gmail.com) wrote:
> > http://pjcj.sytes.net/cpancover/Date-Calc-5.3/coverage.html
> > Not great, but not bad. Nearly no pod, which I find odd....
There's almost no inline POD because the module is almost entirely written
in C. The pod is distributed as a separate file, Date/Calc.pod. I'm
surprised Devel::Cover doesn't identify these conditions.
> Do you understand the module? On how the API should work? I'm no whiz on
> date-related stuff, and some of it looks pretty hairy. Also, do we have
> buy-in from the author?
I use Date::Calc all the time, so I'm comfortable with it. But if we find
errors due to testing, proposing patches will be a bit of a hassle.
Also, if part of the idea is getting under the hood of a major Perl module
to see some Perl wizardary, then that isn't really going to happen unless
we want to focus on DynaLoader, XS and all that.
How about Date::Manip? It's Perl-only, and while it's got about 71%
coverage and a number of test scripts already, it's not complete yet.
Another one that would be feasible is HTML::Tree. (Author: SBURKE, %42
coverage). Or Scalar::List::Utils. (GBARR, %36 coverage.) Or we can go
for broke: CGI.pm! (70% coverage.)
-- Mike F.
More information about the Chicago-talk