Move runfolder analysis classes out of scripts/runfolder.py into seperate files

Move runfolder analysis classes out of scripts/runfolder.py into seperate files
Also rename runfolder.py to runfolder

This was a really annoying patch to make, I wanted to do two major things,
be able to construct the runfolder configuration extracting classes
from the xml file I was creating, and to make unit tests to make sure
all the code was at least somewhat correct.

Writing all of the xml serialization code was really annoying and dull,
there was probably some nifty metaprogramming way of solving it, but
I didn't feel like figuring it out, as I really need to move on to
more important parts of the project.

I wanted to rename runfolder.py to runfolder as the solexa pipeline code
has a runfolder.py (and if anyone has a better name for the script that's
supposed to dump the runfolder xml file, let me know).

Also in working on the xml serialization code, I extended the serialization
for the eland files, this version now dumps the genome_map and the
eland statistics, like reads, match counts and the like. It does
mean that the --archive mode will take longer, but it also means
I'll have enough information to generate the run statistics later.

Now I might have to redo this if we figure out if we should be handling
the realign files instead.