refactor(multimorphic) adopt shared JsonLdProductParser

[jkeeley2073]

Summary

Strict-subset follow-up to PR #42. MultimorphicProductExtractor deletes its private 140-line JSON-LD walker and routes through the shared PinballWizard.Infrastructure.Scraping.JsonLd.JsonLdProductParser that JJP and BoF already consume.

Net: −173 lines / +16 lines in MultimorphicProductExtractor.cs. The class now mirrors the post-#42 BofProductExtractor template — same Extract skeleton, same fallback order (product?.Name ?? og:title ?? h1), same EditionInfo shape — modulo the per-storefront prefix / slug rule / DiscoveredOn token.

The class-level <remarks> block that named a "future PR" to extract the shared helper is replaced with a forward reference to JsonLdProductParser. BofProductExtractor's docstring also drops the parenthetical "(when PR #39 lands)" qualifier now that Multimorphic actually consumes the parser.

Why it matters: validates the shared parser against a third real consumer in production code. The test suite from PR #42 already pinned every shape (including the simultaneous-flat-and-nested case unique to Multimorphic), but a third consumer is the signal that the abstraction generalises cleanly.

Test Plan

• dotnet build → 0 warnings, 0 errors
• dotnet test → 430 / 430 passing
• All 27 existing Multimorphic tests pass without modification, including:
  • Extract_RealMultimorphicShape_BuildsRecord (end-to-end live-shape capture)
  • Extract_NestedPriceSpecOnly_StillReadsPrice
  • Extract_GraphWrappedJsonLd_AlsoWorks
  • Extract_MalformedJsonLd_FallsThroughCleanly
• Behavior equivalence verified by inspection: deleted private parser was a near-identical copy of the now-shared JsonLdProductParser; pre/post comparison against git show HEAD:... confirms GameRecord construction at lines 74-88 is byte-identical.

Out of Scope

• JJP cleanup follow-up. Local review surfaced two pre-existing JJP drift items (not introduced by this PR): JjpProductExtractor.ExtractSlug is missing the ArgumentNullException.ThrowIfNull(productUrl) guard that BoF and Multimorphic both have, and JjpProductExtractor.NormalizeAvailability is private while BoF / Multimorphic make it public (the latter is test-driven, the former is a real defect). Tracked as a separate ~3-line PR to keep this one strictly scoped.

Checklist

• CI is green (build + test + coverage + CodeQL + sanitization)
• PR title follows the Conventional Commits format above
• If this is a new architectural decision, an ADR has been added under docs/adr/ — N/A
• If user-visible behavior changes, README.md and/or docs/ are updated in the same PR — N/A (no behavior change)
• If a memory in ~/.claude/projects/c--projects-PinballWizard/memory/ is now stale, it has been updated or removed in the same PR
• No TODO / FIXME / commented-out code committed
• No new entries in <NoWarn> without a comment explaining why and the removal criterion

Pre-push self-audit

Step 0 — /local-review (qualitative)

• Ran /local-review and addressed every 🔴 finding before push
• Local review outcome: 0 🔴 / 1 ⚠️ / 9 categories ✅
  • ⚠️ deferred: pre-existing JJP drift (missing ExtractSlug null-guard; NormalizeAvailability visibility inconsistency). Justification: not introduced by this PR; addressing it would expand a strict-subset adoption PR into a sibling-cleanup PR. Tracked as a separate follow-up (see "Out of Scope").

Step 1 — Mechanical checklist

• Every new *Options property has at least one real getter call in src/ — N/A (no new options)
• Sibling-diffed against the closest existing implementation (BoF) — Multimorphic now mirrors the BoF template
• No bare catch { } — only scoped catch (JsonException) (preserved behavior)
• New ISourceScraper? — N/A (no new scraper); SourceAliasContractTests still passes
• Tests assert behavior, not just structure
• Build is zero-warning
• git log -1 --format='%an <%ae>' shows personal noreply, not work email