The currently available parsers, the default Haskell-based parser by W. Bas de Haas and John Ashley Burgoyne and MARL's JAMS parser have a focus on audio content retrieval, as is evidenced by the fact that they focus on timestamps and give results that do not support queries about musical structure.

To remedy this, I have decided to adopt a symbolic-music-based approach rather than a signal-based approach.

This is a parser designed for corpus-based musicology and music theory, and as such, is designed for those who have goals that involve:

• Chord rhythm given by position in meter, hypermeter, and phrase instead of just timestamps
• Chord reduction and/or transcription
• Pitch class vectors instead of just chord labels
• Aggregation by section, beat class, as well as other musically salient features

Only spits out chords + beat positions until I actually need more features implemented at the parsing rather than the data processing level. > sbt run