Which lesser-known CPAN modules have you found to be real gems?
Please post links to the modules like this
[`Example::Module`][Example::Module] [Example::Module]: http://search.cpan.org/perldoc/Example::Module
Or post a link to the version control repository.
A short description would be good to include, too.
How about "the builtin function set"?
I can't count the number of times I've struggled reading code littered with unfamiliar modules and APIs just to discover it was doing the equivalent of glob() or grep() or unpack(). There's lots of value in CPAN, but familiarity with the basic language is more important than extensive library knowledge IMHO.
It may be very well known to some, but I have come across many, many Perl developers and companies who have never heard of it yet are constantly dealing with their own in-house solutions or outdated technology.
It's worth spreading around whenever possible, and if one new person stumbles upon this link, my job here is done.
As a bonus, the dependency list will undoubtedly turn you on to a slew of underrated modules that turn out to be real gems!
Check out chromatic's blog post, Aspects of a Novice-Friendly Distribution. He lists a few really great modules that he suggestions should be provided to Perl novices:
If you're not using Moose by now, you're missing out.
I don't know how 'obscure' it really is, but I find it to be a fantastic debugging tool. It lets you expose almost everything there is in a complex data structure, and formats it all nicely. As an added bonus, its output has the handy property of being the Perl code required to rebuild said data structure with an eval.
My vote is for Convert::Binary::C.
Simply point it at some C header files, and you can then easily pack and unpack structures with amazing ease. This module allows you to configure settings so that it can accommodate big-endian 64 bit systems and little-endian 8 bitters, and everything in between.
You can also use hooks to handle serializing and deserializing data into perl types. For example, if you have fixed point numbers that you want to treat as floats in your perl code, you can set a hook on the typedef, so that translation is automatic.
This module is brilliant. I do embedded development in C for ARM and 8051 systems and this module makes my life much easier. But anyone who needs to work with big ugly C structures should use this module. It's so much more powerful and easier to use than pack/unpack templates.
I'm a fan of the Inline modules for embedding foreign (i.e., C++, Java, etc.) code in your Perl scripts. It is so much easier to build a hybrid application than it used to be.
Here is a list of modules I use "regularly" which maybe classed as underrated:
Continuity offers a different paradigm to doing web applications in Perl and definitely deserves more attention (methinks!).
/I3az/
File::ShareDir is something which would be useful to a lot of people if they knew about it.
Git::CPAN::Patch It will allow you to easily (1) download a CPAN distro, and all of the earlier versions, (2) convert it into a git repository, and (3) patch it and (4) send the patch as an RT bug report via rt-email.
You can even send multiple patches, this is great for something like doc-bugs because it makes the processes massively simpler.
XXX, This module uses a very simply api to dump a data structure, and the output is YAML. Which IMHO is many times easier to read than the Perl-syntax Data::Dumper gives. You can also make a quick serialized-file db with the dump which can then be brought into other languages.
See both XXX::XXX() and XXX::YYY() gotta have them both. Great tools!
HTML::Seamstress
This module for HTML "templating" provides the cleanest and strongest separation Index between Perl code and HTML. It is the Perl analogue of XMLC for Java.
Unlike all other solutions which resort to mini-languages, Seamstress weaves HTML and object-oriented Perl together using no additional technologies.
Nice question, thanks.
Found myself using a lot: