test+docs: harden and document the historical-import preserveAudit seam (#3493 follow-ups)
Follow-ups to #3493 / #3549 / #3556, no behavior change:
- Real-SQLite end-to-end test (
@objectstack/runtime): wires the real ObjectQL engine to the realSqlDriver(better-sqlite3) and proves thepreserveAuditseam the mock/in-memory drivers structurally cannot — thatcontext.preserveAuditthreads throughbuildDriverOptionsand defeats the SQL driver'supdated_atforce-stamp on both the forward historical write and the undo restore, and that a businessreadonlyfield (closed_at) survives the engine strip all the way to disk. Includes the #3549 undo capstone (restore-under-preserveAudit rolls the timeline back; a plain restore re-stamps now — the bug #3556 fixed). - Docs: the
treatAsHistoricalschema describe (@objectstack/spec→ regeneratedreferences/api/export.mdx) now documents all three effects — FSM skip (#3479), audit-timeline preservation (#3493), and undo mirroring (#3556) — instead of only the state-machine half; andprotocol/objectql/state-machine.mdxgains a bullet on the symmetric undo behavior.
Describe-string + regenerated docs + a new test only — releases nothing.