brukeropus
SpectraSherpa's bounded native Bruker OPUS reader adapts the read-only binary
file layout and data/status pairing from brukeropus 1.4.3, tag v1.4.3,
commit af5a508cef7de8089acd27a215d644ab451257dd.
- Upstream: https://github.com/joshduran/brukeropus
- Copyright: © 2024 Josh Duran
- License: MIT
- Source distribution SHA-256:
0e67c27d6dcc8fbe06e8c56eec616aec9d55457ae06666520e7bcdc5c91b59cb
Only the read-only file-parser concepts needed for one-dimensional OPUS data are adapted. SpectraSherpa does not include the upstream DDE control surface. Its parser independently adds bounded reads, checked offsets and allocation limits, typed multi-asset results, exact source digests, ambiguity rejection, and explicit refusal of series/3-D blocks.
Conformance-corpus boundary
Nine redistribution-qualified OPUS acquisitions are bundled and therefore run
in every checkout: one from Open Specy (open-specy.md) and eight from
opusreader2 (opusreader2.md), covering singular, duplicate-pairing, and
several out-of-scope block-type layouts. Three additional exact-hash OPUS
acquisitions cover the two-asset reference and six-asset sample layouts, but
their upstream corpus does not state redistribution terms; SpectraSherpa
records them as external qualification inputs and does not ship their bytes.
Tests that need those three files may skip in ordinary CI, but the files
remain mandatory under MONO-2026-08-13-051 for C8/M4.20 release
qualification. This is not evidence for unlisted OPUS layouts.