added a function that can be uncommented to obtain a lot of useful debug info about every query executed

Added functions to allow immediate registration of all mapping pairs in the database.  Makes mapping more memory intensive but much faster for some applications.

Parser for RefSeq .gpff files. Reads accession, description, sequence, parses dependent xrefs for mRNA (will be found in .fna files) , pubmed, medline.

fixed elusive memory leak: key space of parallel hash grew indefinately (one key per query performed).  Should have used delete function rather than setting to undef.

Parser for each file is now stored directly in the source_url table rather than hardcoded in BaseParser.

Check if an xref exists when inserting, if not use the new ID, if it does exist, use the existing xref_id.

avoid a perl warning in three/five_prime_utr when dealing with pseudogene, and improved out of date documentation

Split source table into 2; one for "high level" source name and the other (source_url) to contain the list of actual files we download from.

Added UNIQUE constraint for accession & source to xref table.

Beginning of a parser for RefSeq xrefs. Just creates xrefs and primary xref for protein RefSeq files - still need to parse and somehow incorporate mRNA files

Allowed for file extensions when guessing source ID from file name.