feat(od-ontology): pull alignment symbol-table from lance-graph-callcenter

[AdaWorldAPI]

What

Phase-2 of the repatriation (per REPATRIATION-FRAME.md shipped in #12,
following the schema pull in #13). Pulled the OWL pivot alignment from
lance_graph_callcenter::odoo_alignment — ~350 LOC of muscle memory + classifier

• lookup, leaving the runtime-hydration types (OgitFamilyTable, FamilyEntry)
  in lance-graph as generic infrastructure.

Under the compiler-AST frame (#13): this module IS the symbol table /
type-resolver for the Odoo language frontend — Odoo class name → OGAR
(family, slot) identity + owl:equivalentClass pivot URI + DOLCE marker.

What lands

Surface	What	Empowerment vs muscle memory
OwlPivot struct	(pivot_uri, family, slot, dolce)	Empowerment shape — universal across source languages
dolce_odoo	suffix-pattern → DOLCE marker classifier	Empowerment mechanism + Odoo muscle memory (specific suffix patterns)
ODOO_SEED	15 rows: res.partner → fibo:LegalEntity, account.move → fibo:Transaction, …	Muscle memory — pure Odoo specifics
6 family constants	0x61 BillingCore, 0x62 SMBAccounting, 0x64 ProductCatalog, 0x80 SmbFoundryCustomer, 0x81 SmbFoundryInvoice, 0x90 HRFoundation	Muscle memory — restated bytes from lance-graph's family registry
resolve_odoo	symbol-table lookup with product.* prefix fallback	Empowerment — universal pattern

What stays in lance-graph (explicitly NOT pulled)

• OgitFamilyTable runtime hydration — generic graph spine infrastructure
• FamilyEntry / OwlCharacteristics / SchemaKind — lance-graph's table types
• family_default_style returning StyleCluster — Phase 3 push to OGAR as a Class capability

What the invariant pins surfaced (real findings)

14 invariant pins in tests/alignment_pin.rs, all green — and writing the
cross-validation surfaced two real findings the lance-graph original didn't
have tests to catch:

Finding 1 — the lance-graph dolce_odoo classifier was incomplete

The seed row declares product.pricelist is Abstract (a pricelist IS a
pricing rule, not a persistent product). The lance-graph classifier swept it
into Endurant because its product.* rule matched first. Added the
pricelist exception; the classifier now agrees with the seed. lance-graph's
own tests didn't catch this because they only spot-tested product.template
not product.pricelist.

Finding 2 — OGAR's OdooPort covers the commerce arm only

The cross-axis identity check (seeded_classes_have_compatible_ogar_identity)
proves the alignment table and OGAR's identity codebook agree about which Odoo
classes are first-class citizens. But OdooPort has 9 aliases (the
commerce arm: account.move, sale.order, account.move.line,
sale.order.line, account.tax, res.partner, account.payment,
res.currency, account.analytic.line) — the seed table has 15 rows
covering product + accounting + HR basins too.

Surfaced as informational eprintln (= candidate list for the next OdooPort
PR); the hard pin asserts only the commerce-arm intersection. The "gap" IS
the surface for a future PR, not an error today.

informational: 11 of 15 seeded classes have no canonical OGAR classid yet
(= candidates for the next OdooPort PR — product / accounting / HR basins):
["account.account", "account.account.template", "hr.contract", "hr.department",
 "hr.employee", "hr.job", "product.pricelist", "product.pricelist.item",
 "product.product", "product.template", "uom.uom"]

Tests

130+ test results green across the suite: 39 lib + 14 alignment-pin + 10
corpus-pin + 13 typed-lift + 10 ogar-parallel + 7 ogar-convergence + 7 bin +
others. clippy on my new code clean (pre-existing doc_markdown warnings on
older files untouched). lib.rs / alignment.rs / alignment_pin.rs fmt-clean.

Builds on:

• docs(od-ontology): repatriation frame — lance-graph → (OGAR + odoo-rs + ruff) #12 (the repatriation frame — north star)
• docs(od-ontology): pull SPO schema from lance-graph + name the compiler-AST frame #13 (Phase 1: SPO predicate-schema pull + compiler-AST frame)

Next phases per the frame:

• Phase 2 continued: pull odoo_blueprint/{l1..l15}.rs (14K LOC) when ready
• Phase 3: push empowerment shapes (OwlPivot, suffix-classifier pattern,
  family_default_style → StyleCluster inheritance) to OGAR as universal
  Class capabilities so other source languages reuse them
• Phase 4: producer-side OGAR-class detection in ruff
• Phase 5: lance-graph becomes generic by subtraction

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, you can upgrade your account or add credits to your account and enable them for code reviews in your settings.