Humdrum: section-structured kern file parses to 8 of 69 measures (silent truncation) #1987

Environment: music21 10.5.0, Python 3.14, macOS.

Summary: the Humdrum importer parses a section-structured kern file to a small fraction of its measures. The kernscores transcription of Chopin's Mazurka Op. 41 No. 2 (mazurka41-2.krn: 69 numbered measures, section labels *>A, *>B, ... with an expansion list) parses to a score whose parts contain 8 measures, numbered 1–8 — the remaining ~61 source measures are silently absent. Content addressed by source measure number (e.g. a text annotation at source measure 64) has no corresponding location in the parse at all.

Expected: either the full expanded sequence of sections, or (if expansion is unsupported) the sections in source order — not a silent 8-measure fragment.

Reproduction:

from music21 import converter, stream
sc = converter.parse("mazurka41-2.krn")   # kernscores chopin-mazurkas
part = sc.parts[0]
print(len(part.getElementsByClass(stream.Measure)))   # 8; the source has 69 "=N" barlines

Source file: the kernscores chopin-mazurkas set (craigsapp/chopin-mazurkas on GitHub), kern/mazurka41-2.krn.