feat(OGC-1054): add safe Analyzer Type mapping

[pmanko]

Summary

Deliver OE-M2, the OpenELIS half of the OGC-1054 safe mapping checkpoint. Add
one reusable Analyzer Type mapping editor, catalog-safe test/result bindings,
auditable confirmation, and complete removal of superseded per-analyzer mapping
and classifier paths.

Scope

• one linkable mapping editor under Analyzer Types; Verify returns to it and no
  per-analyzer editor remains
• every profile source row remains visible with raw code, normalized identity,
  match state, catalog selection, and explicit exclusion where allowed
• complete active Test search by name, code, or LOINC with unique suggestions
• qualitative mappings accept only active Result Options owned by the selected
  Test; value and label are derived server-side and survive reload
• separate confirmation of bindings and human-readable Bridge profile
  recognition with actor, time, row IDs, revision, and fingerprints
• stale confirmation and affected-analyzer/update-available behavior after a
  shared revision or binding change
• explicit Duplicate Profile and Update shared scope with no implicit analyzer
  movement
• remove the duplicate field-mapping/custom-field admin UI and the complete
  AnalyzerQcRule runtime, schema, payload, readiness, and UI path
• amend the Westgard specification so it retains operational QC without
  restoring an OpenELIS analyzer classifier or a narrow competing “MVP” label

Stack

• Base: OE-M1 feat(OGC-1054): build Bridge-backed Analyzer Types #4056
• Required producer: Bridge #47,
  pinned by the ordinary Bridge submodule at
  d826d6ac3eb55c63c3b74649facfdf3b2c6e87b1
• Analyzer-mock: java error fix to begin beta testing #40; no M2 companion PR is needed
• Authoritative roadmap: docs(OGC-1054): establish authoritative analyzer roadmap #4049

Strict order is BR-M2 -> OE-M2. On publication, the roadmap moves M2 to [x]
and M3 to the sole [*] checkpoint.

TDD evidence

Service, persistence, migration, consumer-contract, removal-guard, RTL/router,
and UI-story behavior was introduced in red-green commits. The final
specification guard first failed on the obsolete per-analyzer QC story and then
passed after the source specification, plan, and tasks were amended.

Validation

• OpenELIS analyzer/analyzer-import/QC suite: 898 passed, 0 failures/errors
• focused M2 RTL: 8 files, 60 passed
• full frontend suite: 1,210 passed, 8 skipped
• UI-only isolated Playwright story: 4 passed; desktop/mobile captures and
  confirmation timestamp inspected; no application console errors
• mvn spotless:check, full Prettier check, lint, production build, and both
  Playwright guards: passed
• Bridge mvn verify: 632 passed, 3 skipped
• analyzer-mock M2 profile-contract selection: 144 passed
• openelis-work@main visual baseline refreshed at 5fc8963b; implementation
  preserves the mapping hierarchy and result-choice intent and remains usable on
  mobile where the prototype itself overflows beside its fixed sidebar

Known inherited gates

• repository-wide TypeScript checking remains red: 1,754 errors versus 1,791 on
  OE-M1. M2 introduces no errors in its new files and removes the obsolete Copy
  Mappings declaration error; this PR does not claim that global gate is green.
• retained QC listener tests include an old fixture that omits a DAO mock and
  logs caught null errors while passing. M3 linked operational-QC acceptance
  must correct and strengthen that fixture before claiming clean QC evidence.

Architecture guard

Bridge remains the profile and analyzer-runtime owner. OpenELIS owns local
clinical bindings, verification/audit, activation, and separate operational QC.
There is no copied profile authority, per-analyzer mapping override,
AnalyzerQcRule, QcRun, hidden classifier fallback, dual writer, duplicate
queue, or OpenELIS FILE poller.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: f9bb30a7e0

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".