pirl-unc / tsarina / 27096776595

Build # 27096776595

Build Type

push

github

Committed by web-flow

Commit Message

Add versioned reference-data CLI: tsarina data sources (#100) (#100)

The HPA/NCBI curation source files (RNA consensus, IHC normal_tissue,
NCBI gene_info) were fetched ad-hoc by scripts/cta_sources.py with no
version tracking and no CLI surface -- and the upstream mirrors are
release-inconsistent (the 'latest' www mirror serves the RNA consensus
but not normal_tissue; only the pinned v23 archive carries both as a
matched pair). Regenerating the bundled table from mismatched releases
is exactly how the table's mixed-provenance drift crept in.

Add tsarina/reference_data.py: a versioned download/cache layer with a
JSON manifest (version, url, sha256, bytes, timestamp), a registry of the
three reference datasets, and a pinned DEFAULT_HPA_VERSION=v23. Surface
it as 'tsarina data sources {list,download,path}' -- the HPA/NCBI
counterpart of the hitlist-backed 'tsarina data' (IEDB/CEDAR/viral),
mirroring pyensembl's install/list ergonomics.

scripts/cta_sources.py now delegates to it (one shared, version-stamped
cache; adds normal_tissue_path). Network-free tests cover download/zip
extraction, manifest recording, cache reuse vs --force, version
resolution/defaults, and status reporting. 407 tests pass.

Minor bump: new public module + CLI subcommand.

Coverage Stats

2575 of 3598 relevant lines covered (71.57%)

0.72 hits per line

Source File

Press 'n' to go to next uncovered line, 'b' for previous

Source Not Available