Cleaned up code as recommended by pylint and yapf, without major overhauls. I am still overriding some of the 'too-many-X' messages in pylint, but I think that rewriting the code to eliminate those warnings will make the parsers more difficult to understand. Parsing big text files is nasty, and pylint doesn't like it.
Merged the BEL csv parsers into a single file that uses an internal dictionary to convert between JIS encoding and western text encoding. This eliminates all duplicate code in the BEL csv parsers and simplifies maintenance
Two major objectives in this PR: