Remove dead PMC full-text code from sources/pmid.py (superseded by PMCFullTextProvider)

[cmungall]

Follow-up from the review of #48.

PMIDSource.fetch no longer fetches PMC full text — that logic now lives in PMCFullTextProvider (etl/fulltext/pmc.py). The following methods in src/linkml_reference_validator/etl/sources/pmid.py are now dead (no live caller; _fetch_pmc_fulltext is only defined, never invoked):

• _fetch_pmc_fulltext (~L257)
• _get_pmcid (~L283)
• _fetch_pmc_xml (~L323)
• _fetch_pmc_html (~L359)

Scope

• Delete the four methods (lines ~257–390).
• Remove tests/test_sources.py::test_get_pmcid_handles_entrez_error (it covers the dead _get_pmcid).
• Add an equivalent error-path test for PMCFullTextProvider._resolve_pmcid (Entrez elink failure → None) so PMC-ID resolution error handling stays covered.
• Drop the stale references in src/linkml_reference_validator/plugins/REVIEW.md.

Deferred out of #48 to keep that PR focused (it was already flagged as a follow-up in the PR description, and noted again in the #48 review).

[github-actions]

Summary

Type: Follow-up cleanup / dead-code removal
Blocked by: PR #48 (must be merged before this work begins)
Labels: enhancement, good first issue

---

Context

This issue is a deferred cleanup task spun out of the review of PR #48. PR #48 refactors PMC full-text retrieval by introducing a dedicated PMCFullTextProvider class in etl/fulltext/pmc.py. After that refactor, PMIDSource.fetch (in sources/pmid.py) will no longer call _fetch_pmc_fulltext directly — making four private methods in that class dead code.

Current codebase state (pre-#48 merge)

As of now (main, commit fcb9efa), these four methods still exist and _fetch_pmc_fulltext is still called live at line 141 of pmid.py. This issue should only be worked after PR #48 lands.

Once #48 is merged, the following will be dead:

Method	~Lines	What it does
_fetch_pmc_fulltext	~261–286	Orchestrates PMCID lookup → XML → HTML fallback
_get_pmcid	~287–325	Calls Entrez elink to resolve PMID → PMCID
_fetch_pmc_xml	~327–361	Fetches and parses PMC XML via Entrez efetch
_fetch_pmc_html	~363–394	Scrapes PMC HTML as a fallback

Together these span roughly lines 261–394, the last ~130 lines of pmid.py (which is currently 394 lines total).

Scope of work

1. Delete the four dead methods from sources/pmid.py
2. Remove tests/test_sources.py::test_get_pmcid_handles_entrez_error — it covers the now-deleted _get_pmcid (currently at line ~288)
3. Add an equivalent error-path test for PMCFullTextProvider._resolve_pmcid covering the Entrez elink failure → None path, so that PMC-ID resolution error handling remains tested
4. Drop stale PMC full-text references in src/linkml_reference_validator/plugins/REVIEW.md (lines ~34, 96, 232, 250–255)

Notes

• The REVIEW.md currently documents PMC full-text fetching as untested, which will become moot once the methods are removed and the new PMCFullTextProvider has its own tests.
• Good first issue: the scope is well-contained, the dead methods are clearly identified, and the test replacement is a straightforward pattern match of the existing error-path test.
• Per project conventions, the new test for PMCFullTextProvider._resolve_pmcid should use a doctest-friendly form where possible (see CLAUDE.md).