feat(meta): vendor OO-LD 1.0.0-rc.4 and rc.5

[simontaurus]

The newest tracked meta-schema was 1.0.0-rc.3. meta_store resolves the default to the highest tracked version, so pip install oold[validation] && oold validate ran against rc.3's 73 rules and could not see the 19 added since, including OOLD-SCH-96a3, the root @context requirement that rc.4 led with.

After: the default resolves to 1.0.0-rc.5, 92 rules, 47 machine-checkable.

Three consequences, each handled

A withdrawn keyword breaks coverage for older versions. The fixture corpus is refreshed from the newest tag, so it cannot exercise x-oold-reverse-default-properties, which the vocabulary removed, while 1.0.0-rc.1 still declares it. VOCAB_EXEMPT does not apply: that covers keywords outside the x-oold- prefix, and this one has it. Added VOCAB_WITHDRAWN as the separate concept and excluded it from vocabulary_coverage.

Probe schemas needed @context. Three tests build a minimal document to check the dialect resolves; from rc.4 a document requires @context as well as $id. The comment already noted the $id case for the same reason.

no_context.schema.json is now refused outright. It is a deliberately broken fixture and the meta-schema rejects it, which is correct. The test's subject is the round trip, so it still asserts roundtrip.generated is OK and coverage WARNs, and now asserts schema.meta is the only failure rather than that there are none.

Fixture refresh also picked up the http://schema.org correction in tests/data/oold/UiAnnotations.schema.json (OO-LD/oold-schema#179).

Verified: 504 passed, 7 skipped; oold validate over the specification's examples/ reports 380 passing checks against 1.0.0-rc.5.

[github-actions]

Release preview

Merging this PR would release v1.1.0 (current: v1.0.3).

Changelog preview (truncated)

## v1.1.0 (2026-09-27)

### Features

- **meta**: Vendor OO-LD 1.0.0-rc.4 and rc.5
  ([`fae8ed0`](https://github.com/OO-LD/oold-python/commit/fae8ed07142e4af41215a77792ee40aa10e34661))

_{Preview via python-semantic-release and conventional commits.}

[github-actions]

📊 Benchmark Results

Click to see benchmark comparison

📊 Benchmark Comparison (threshold: 1.3x)
============================================================

➖ Unchanged (within threshold):
  ➖ test_simple_dict_document_store: 0.0020s → 0.0020s (+0.9%)
  ➖ test_sqlite_document_store: 0.0021s → 0.0021s (-0.1%)
  ➖ test_local_sparql_store: 0.0387s → 0.0391s (+0.9%)
  ➖ test_oneof_subschema: 0.0616s → 0.0616s (-0.0%)
  ➖ test_enum_docstrings: 0.0521s → 0.0524s (+0.7%)
  ➖ test_subclass_inheritance: 0.0553s → 0.0579s (+4.6%)
  ➖ test_class_hierarchy: 0.0523s → 0.0531s (+1.6%)
  ➖ test_core[v1]: 0.0369s → 0.0366s (-0.6%)
  ➖ test_core[v2]: 0.0446s → 0.0455s (+2.0%)
  ➖ test_schema_generation[v1]: 0.0017s → 0.0018s (+5.3%)
  ➖ test_schema_generation[v2]: 0.0038s → 0.0038s (-0.1%)
  ➖ test_simple_json: 0.0007s → 0.0007s (+0.5%)
  ➖ test_complex_graph: 0.0016s → 0.0017s (+0.9%)

============================================================
Summary: 0 regressions, 0 improvements, 13 unchanged
============================================================

✅ No significant performance regressions

Threshold: 1.3x (30% slower triggers a regression warning)

Note: Benchmarks are informational only and won't fail the build.

_{💡 Tip: Download the benchmark-results artifact for detailed JSON data}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!