Matching between metadata from the Million Song Dataset and metadata from large online audio collection.

Run "python mk_flaclist_index.py" to parse a compressed list flaclist.txt.gz, with each line describing a known track with tab-separated fields for artist, album, title, duration (sec), archivename, filepath for each track. The parsed result is written to a "whoosh" index in ./WCDindexdir .

Run "python qry_flaclist_index.py" to query the index with information from each line of a list MSD-all-artist-release-title.txt, with tab-separated fields for artist, album, title, duration (sec), and MSD_ID. Of the items in the index that (fuzzily) match the artist and title (and album if there are any), choose the one whose duration is closest to the one in the index (if any), then write out a tab-separated description to MSD-to-WCD.txt with MSDid MSDduration WCDduration MSDartist MSDalbum MSDname WCDartist WCDalbum WCDtitle WCDarch WCDfile for each MSD item. Where no match is found, WCDduration = 0, WCDartst = "NO_MATCH", and the other WCD fields are empty.