Corelli : krn versus DCML

[MarkGotham]

As noted with @malcolmsailor there's a discrepancy between how krn and DCML count Corelli movements.

These don't work (aren't present on kern.humdrum) but would if the movements matched up: https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n2-02.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n2-04.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n4-04.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n10-05.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n10-04.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n10-01.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n1-04.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n1-02.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n6-06.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n6-07.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n6-05.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n6-01.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n6-04.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n7-05.krn&f=kern https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op4&file=op4n7-02.krn&f=kern

Summary:

• the URL format is now correct
• these files are indeed not present on kern.humdrum
• they may result from movements combinations (e.g., DCML have 4; Craig has 3 with two of them joined). This is typical for short Baroque movements.

Proposal:

• only solution I see would be:
  • download a local (not doing for licence reasons) and
  • split movements (doable, but a bit complex)
• do not implement that "solution" do that (do nothing) for now because ...
  • we also have working DCML remote score options in these case
• simply comment out krn sources throughout for now (remove but easy to reinstate)
• Revisit if someone desperately wants this functionality or if the kern hosting pattern changes (both unlikely) .

[MarkGotham]

To clarify

• keep metadata for op3 (all work)
• remove for:
  • op1 (one fail: https://kern.humdrum.org/cgi-bin/ksdata?l=musedata/corelli/op1&file=op01n03d.krn&f=kern)
  • and op4 (as listed above)

[malcolmsailor]

Since DCML has scores too, I think removing the humdrum scores makes the most sense. Also the DCML analyses should align with the DCML scores whereas we don't know if that's true for the humdrum scores.