pirl-unc / mhcgnomes / 33460742237
90%

DEFAULT BRANCH: main
Ran
Jobs 1
Files 41
Run time 1min
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
coverage: 89.198% (+0.02%) from 89.182%
33460742237

push

github

web-flow 
Support the hyphenated class shorthand (#150)

* Curate prefix provenance from the IPD-MHC species tables

Marks 50 entries `designated`, each citing the IPD-MHC group species
table it came from. The designated count goes 11 -> 61 and the
unestablished tail 163 -> 113.

Eight fetches rather than 163 lookups: the group tables list every
species and its designation on one page each. DLA (10 rows), CeLA (33),
SLA (3), BoLA (4), OLA (2), RT1 (2), CLA (1), ELA (1). Each read
verbatim rather than from a summary, and every row checked against our
prefix before marking.

Zero of the 56 species IPD lists in these groups is missing from our
ontology, and the prefixes agree except in five places. Two are #112's
documented resolution -- Caau belongs to the goldfish in published use
and Hyam to a minnow, so the canid and the bottlenose whale keep their
generated prefixes and hold the IPD code as context-only.

The other three are new: Delphinapterus leucas, Neophocaena
asiaeorientalis and Orcinus orca carry generated 4+4 prefixes while IPD
designates Dele, Neas and Oror -- which already resolve to the right
species, with nothing else claiming them. So we emit an invented name in
preference to the published one. Filed as #146, since swapping them
changes normalized output.

The remaining 113 are the harder half. The eight pinned in the canary
test are group-level _LA codes, and IPD publishes no group page for any
of them -- /group/FLA/ is a 404 -- so establishing those means the
nomenclature reports rather than a species table.

A new test pins the citations: every IPD group named in a species.yaml
URL must be one of the eleven IPD publishes. Its first draft failed on
FISH and CHICKEN, real groups I had left out, so it caught my error
rather than a data error.

0 of 11,558 corpus names change.

Claude-Session: https://claude.ai/code/session_012s4siLj2Vm33eNMGjNSazH

* Prefer IPD's designation over a prefix we generated

Delphinapterus leucas, Neophocaena asiaeorien... (continued)

4137 of 4638 relevant lines covered (89.2%)

0.89 hits per line

Coverage Regressions

Lines Coverage ∆ File
42
93.94
 0.04% parser.py
Jobs
ID Job ID Ran Files Coverage
1 33460742237.1 41
89.2
 GitHub Action Run
Source Files on build 33460742237