[nathan-stender]

Summary

Two small cleanups from a TODO sweep of the repo. No behavior change — comments, one prose typo, and the generated table that mirrors it.

1. Record the ASM molecular weight gap with the tracked convention

The Agilent OpenLab CDS GPC/SEC support added in #1269 reports the molecular weight averages (Mn, Mw, Mz, Mz+1, Mv, Mp, PD) as peak custom information, because ASM has no molecular weight concept at any level — not on the peak, not on the processed data document, nowhere. That was the right call, but the reason was only written as prose, so it didn't turn up in a git grep "TODO(ASM gaps)" alongside the other nine tracked gaps in the repo.

This adds the marker where the affected fields are defined. Note that every existing TODO(ASM gaps) lives in a schema mapper; this one can't, because the liquid chromatography mapper passes Peak.custom_info straight through and has no molecular-weight-specific line to annotate. The parser's constants file is where the substitution actually happens.

2. Fix "signifcant" -> "significant"

In the parser maturation level descriptions in get_table_contents(), plus the generated SUPPORTED_INSTRUMENT_SOFTWARE.adoc (regenerated with hatch run scripts:update-instrument-table, which changed only that line — no other drift).

Testing

• hatch run lint — clean (ruff, black, mypy: no issues in 795 source files)
• hatch run test_all.py3.10:pytest tests/ -q — 1520 passed

tests/parser_factory_test.py::test_table_contents asserts the checked-in .adoc matches the generated table, so the two typo fixes have to land together — and that test passing confirms they did.

Deliberately not included

The same sweep turned up items that need a decision rather than a mechanical edit, so they're left alone:

• jsonschema < 4.18.0 pin (pyproject.toml) — the TODO(nstender) there wants refs stripped from schemas before use to work around a performance regression. Confirmed as still needed, not touching it.
• 4x TODO(switch-statement) — all say "use switch statement once Benchling can use 3.10 syntax". requires-python = ">=3.10" here suggests this is unblocked, but the wording refers to a Benchling-internal constraint, so it needs one confirmation from someone who'd know before the conversions are worth making.
• 3x duplicated "should we move this inside of measurements?" in the tapestation and nanodrop parsers — a structural question about where the data belongs, not a cleanup.
• 4x NaN-handling TODOs across agilent_gen5, cfxmaestro, softmax_pro, and appbio_quantstudio — these look like they want a shared helper, which overlaps with open PR fix: Luminex xPONENT - report NaN assay bead counts as NaN via error document #1268 (it defines get_bead_count_error locally in the Luminex parser). Worth settling as one design decision while that PR is open.

🤖 Generated with Claude Code