Parsing License Headers

[robertburrelldonkin]

ATM RAT uses a very basic header parsing system based. this reads one
line at a time and then feeds this line to all the license matchers to
see if they match.

going forward, i'm thinking about switching to using nio. so, rather
than passing in a line at a time this means passing in a charsequence
for an initial region (not sure how big as yet) of the document. for
most resources this sequence will be backed by a charbuffer but
strings will be ok too.

opinions?

- robert