#1 Downloads same file many times

[Nick Urbanik]

I found that perlpodder was downloading the same file many times. I
had a look at the code, and wondered how it ever worked at all.

I edited it, and it now works for me. I also removed many cases where
the code was incorrect, and figured that I should submit this back so
at least the poor people who might try to use this won't be so
disappointed.

Bugs fixed:

1. Fixed the download many times bug
2. Made it work with use strict, uncovering typos.
3. We check to see the URL begins with http:// instead of checking to
see if it contains one of the letters h, t, t, p (Oh, deary me :-)
4. Remove many temporary variables introduced for no discernable
reason
5. Avoid confusing bandwidth with size :-)
6. Regularise random indentation somewhat
7. Remove a few gratuitous chomps where of no use.
8. Check that open, close work.
9. Use grep properly, instead of as a regular expression test against
one assembled string.
10. Made it a little less sad to look at.

There are a few other minor enhancements, such as making successful
execution silent, adding a --debug option, make constants upper case,
make code more Perl-like.

[Nick Urbanik]

A patch that contains a rewrite to fix the various bugs.