email interface into a module?

[Craig]

Max Kanat-Alexander wrote:
> On Wed, 28 Nov 2007 16:39:07 -0800 Craig <craig5 at pobox.com> wrote:
>> So, what if the bulk of the "email_in.pl" code was moved into a module
>> (Bugzilla::EmailIn???)?
> 
> 	It might work. Actually, a lot of the code needs to be
> re-written to use Bugzilla::Bug anyhow. It's possible that a lot of the
> code could go into Bugzilla::Bug and be generally useful.

As a "first pass", what if all of the subroutines went into a module
(Bugzilla::Email.pm) and everything below "Main Script" stayed in
email_in.pl?

At least then it is broken up and email_in.pl has to instantiate a
package and make method calls. And this minimizes the amount of changes.

Then we can worry about moving code between modules after that.

I have just started reading the developer doc... why doesn't email_in.pl
use taint mode (-T)?

email_in.pl has "-w" and "use warnings". Isn't that redundant? If the
version of perl being used doesn't support "use warnings", then it's
going to die anyway, right?


> 	Something like Bugzilla::Bug::update_from_params might be good
> to have. We have to finish bringing process_bug.cgi into Bugzilla::Bug
> first, though (which we're almost done with, in HEAD).

Are you refering to the sections of Bug.pm that are preceded by:
	"Temporary hack until all of process_bug uses update()."