«

I am working on a module that when given a CPAN distribution, will return
which modules the distribtion requires (by parsing the Makefile.PL using
Module::MakefilePL:arse if the META.yml file is not available).

A companion module will actually try to fetch the distribution information
from CPAN (since it can query a .meta file if it exists before downloading an
entire distribution).

Because it is more closely tied to CPAN and works with distributions rather
than modules, I am thinking of calling it CPAN:istribution:epends.

Comments or suggestions about the namespace would be appreciated.

FYI--I am writing yet another module to do this because existing modules
either do not work on my platform or they require preqrequisite modules to be
installed. Basically they either compile the module or run
Makefile.PL/Build.PL and see what modules are asked for. I'd rather use a
safer (and I think faster) method to partially-parse the Makefile.PL if no
META.yml is available.

Some tests using all the latest distribution on CPAN show that it works for
all but a handful of distributions.

This is actually part of a larger project to show test results (organized by
platform) of module dependencies. If there is a failure in the dependency
tree, authors can check for this.

I've also posted this question on Perl Monks at
<http://perlmonks.org/index.pl?node_id=374747>

Hi,

CPAN:istribution:epends is good, but how about
CPAN:istribution::Resolv ?

Robert Rothenberg wrote:

> Because it is more closely tied to CPAN and works with distributions
> rather than modules, I am thinking of calling it
> CPAN:istribution:epends.
>
> Comments or suggestions about the namespace would be appreciated.

The one recommendation I remember from many, many years ago was that in
the vein of the data structure being the real core of an object system,
that you name the module after a noun, rather than a verb. So 'Depends'
would be good (as short for 'dependencies'), and 'Distribution' is
better than 'Distribute'.

Krishna Sethuraman