fix: override ICD10CM prefix expansion to use icd10data.com

[kevinschaper]

Summary

• The upstream prefixmaps library maps ICD10CM CURIEs to https://icd.codes/icd10cm/, but that domain is dead (returning 403s)
• Override the canonical URI prefix to use http://purl.bioontology.org/ontology/ICD10CM/ instead, which redirects to BioPortal's ICD10CM class pages
• Uses direct prefix_map mutation because add_prefix(merge=True) only adds the new URL as a synonym rather than replacing the canonical prefix

Upstream fix should go into linkml/prefixmaps and/or biopragmatics/bioregistry.

[netlify]

✅ Deploy Preview for monarch-app canceled.

Name	Link
🔨 Latest commit	e0651f7
🔍 Latest deploy log	https://app.netlify.com/projects/monarch-app/deploys/69973bbb18b10a0007141801

[claude]

PR Review: fix: override ICD10CM prefix expansion to use icd10data.com

Overall: Looks good — clear fix for a real breakage, well-commented. One suggestion below.

---

What the change does

Directly patches converter.prefix_map["ICD10CM"] to replace the dead icd.codes URL with https://www.icd10data.com/search?s=, matching what bioregistry.io resolves to.

---

Positive observations

• The comment clearly explains why add_prefix(merge=True) isn't sufficient (it only adds a synonym rather than replacing the canonical prefix). That context is valuable for future maintainers.
• The fix is minimal and targeted — no unnecessary changes.
• The PR description calls out that the proper fix belongs upstream in linkml/prefixmaps / biopragmatics/bioregistry, which is the right framing for a workaround.

---

One concern: no test case added

test_curie_service.py has parametrized tests for many prefixes but none for ICD10CM. This means:

1. There's no automated verification that the override actually works today.
2. If the curies/prefixmaps library changes behavior (e.g., fixes ICD10CM upstream to a different URL, or changes how prefix_map is exposed), the regression won't be caught.

Consider adding a test case:

("ICD10CM:J45.20", "www.icd10data.com/search?s=J45.20"),

(following the existing pattern of checking expected in expanded_curie)

---

Minor note: search-URL semantics

The target URL (https://www.icd10data.com/search?s=) is a search endpoint rather than a canonical entity resolver, so ICD10CM:J45.20 expands to a search result page rather than a stable, bookmarkable entity page. This matches bioregistry.io's behavior and is a practical improvement over a 403, but worth noting in case a more stable direct-link resolver becomes available in the future.

---

No blockers — the fix addresses a real user-facing issue. Adding the test case would strengthen confidence in the workaround.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 73.52%. Comparing base (e32ef46) to head (e0651f7).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1260   +/-   ##
=======================================
  Coverage   73.51%   73.52%           
=======================================
  Files          95       95           
  Lines        3417     3418    +1     
=======================================
+ Hits         2512     2513    +1     
  Misses        905      905           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[claude]

PR Review: fix: override ICD10CM prefix expansion

Overall: Straightforward, well-motivated fix. One suggestion below.

Note: A previous review comment from this bot had an incorrect URL for the override target — the actual URL used is purl.bioontology.org, not icd10data.com. This comment supersedes it.

---

What the change does

Directly patches converter.prefix_map["ICD10CM"] to replace the dead icd.codes canonical URL with http://purl.bioontology.org/ontology/ICD10CM/, which redirects to BioPortal's ICD10CM class pages.

---

Positive observations

• The inline comment clearly explains why add_prefix(merge=True) isn't sufficient (it only adds as a synonym rather than replacing the canonical prefix). That context is valuable for future maintainers.
• The fix is minimal and targeted — 3 lines, no unnecessary changes.
• The PR description correctly frames this as an upstream workaround and identifies linkml/prefixmaps and biopragmatics/bioregistry as the right places for a permanent fix.

---

Minor: PR title vs. actual target URL

The PR title says "use icd10data.com" but the code and commit message use purl.bioontology.org. The PR description is accurate — just a title inconsistency worth cleaning up if the PR is updated.

---

Missing test case

backend/tests/unit/test_curie_service.py has parametrized expansion tests for many prefixes but none for ICD10CM. Adding one would:

1. Confirm the override works today.
2. Catch regressions if the upstream library changes prefix_map semantics or fixes ICD10CM to a different URL.

Suggested addition to the parametrize list (following the existing pattern):

("ICD10CM:J45.20", "purl.bioontology.org/ontology/ICD10CM/J45.20"),

---

No blockers — the fix addresses a real user-facing issue with a clear workaround and good documentation.