andreamust / harte-library Goto Github PK

N and X symbols are not recognized by the parser

A cleanup of the harte parser utility is needed in order to get rid of the unused code.

See for instance F##:dim7, Abb:maj7

similar to Issue #8
related to Issue #6

As for now, the Harte constructor takes a harte chord as input. This hinders the possibility of using the library in a different way which is not for parsing harte chord.
For example, this limits the possibility of instantiating a chord from music21.chord and port it to harte.

should be treated the same as power.

related to Issue #6

Shorthand 6 has not been implemented, but usage has been found in some annotations.
Further details on which should be its intervals are required (is it the same as maj6? or a shorthand for 7(6)?)

Realted to Issue #6

Method __deepcopy__ hasn't been implemented and it throws an error when trying to initialize the Chord object (the chord is missing)

This is fixed by #17 which explains the details - I'm just creating an issue for visibility in case anyone else runs into this in the meantime

the convert_interval function yields A5 for a b5 degree and d5 for a #5 degree.

see PR #5 (related to Issue #3) for automated test on chord coverage.
Some chords are wrong annotations from ChoCo itself though.

Further issues will be opened and linked to this one for each fix type.

Should be probably be treated the same as hdim7 is.
Further details need to be checked to confirm.

related to Issue #6

Contains F- while it should contain F#. Also it would be nice if quality returned augmented (or suspended, but this appears never to be returned by music21.chord.Chord anyway).

c = Harte('Bb:(4,#5,b7)')
c.quality
Out[6]: 'other'
c.pitches
Out[7]: 
(<music21.pitch.Pitch B-4>,
 <music21.pitch.Pitch E-5>,
 <music21.pitch.Pitch F-5>,
 <music21.pitch.Pitch A-5>)

c2 = Harte('C#:min(b7)/b1')
c2.pitches
Out[6]: 
(<music21.pitch.Pitch C##3>,
 <music21.pitch.Pitch C#4>,
 <music21.pitch.Pitch E4>,
 <music21.pitch.Pitch G#4>,
 <music21.pitch.Pitch B4>)
c2.prettify()
Out[7]: 'C#:min7(b1)/b1'
c2.unwrap_shorthand()
Out[8]: ['b1', 'b3', '5', 'b7']

I might be wrong, and I don't know this chord, but I don't see where the C##3 comes from? Shouldn't the /b1 in the bass become C4 as first pitch? The chord occurs in biab-internet-corpus_1004.jams in the ChoCo dataset.

c = Harte('C:hdim7')

Gives

(<music21.pitch.Pitch C4>,
 <music21.pitch.Pitch E-4>,
 <music21.pitch.Pitch G#4>,
 <music21.pitch.Pitch B-4>)

Unless I'm mistaken, it should give

(<music21.pitch.Pitch C4>,
 <music21.pitch.Pitch E-4>,
 <music21.pitch.Pitch G-4>,
 <music21.pitch.Pitch B-4>)

Extend the shorthand mapping with the superset provided by the JAMS library.

If bass is not equal to the root note, the prettify method does not return the bass note.

should be treated the same as pedal probably, to check

related to Issue #6

Add test cases and setup related GitHub actions.

See for instance Fb:maj/##4 or C#:min7/bb1

related to Issue #6

Add encoder and decoder functionalities for DL applications.

sometimes notated also as "X"

from Issue #6