[codex] SQLite node store read scaling and UI materialized views - #24
Conversation
- add shared NodeStore query descriptors with plan metadata - add storage pushdown hooks for NodeStore-backed readers - add SQLite scalar property sidecar, rebuild support, and batch hydration - add conservative SQL candidate planning with JS verification - route data bridges and Electron IPC storage through safe paginated paths - update exploration checklist and focused coverage for indexes and IPC storage
- record stable NodeQuery descriptor stats for SQLite query planning - add opt-in bounded partial scalar indexes for hot property filters - expose adaptive index options through SQLite adapter factory exports - document Phase 3 status and migration checklist updates
- add bounded parity metadata for SQLite-backed NodeQuery plans - compare SQL results against shared JS descriptor semantics for small scopes - log high-severity diagnostics when SQL candidates miss expected nodes - mark the Phase 2 verification checklist item complete
- time SQLite candidate queries through shared diagnostics helpers - attach used indexes, plan details, and index inventory to query plans - run ANALYZE after adaptive index creation before PRAGMA optimize - update the exploration checklist for diagnostics integration
- track adaptive index estimated bytes and indexed rows in schema metadata - drop stale or least-recently-used adaptive indexes when budgets are exceeded - skip adaptive index creation when disk or write-row budgets cannot fit - add opt-in query and adaptive-index debug logging - update exploration checklists for Phase 3 budget enforcement
- filter NodeStore get, list, and query reads through read authorization - disable storage query pushdown when a read policy evaluator is active - apply pagination after read filtering so hidden rows cannot affect pages - redact query plan counts to visible nodes for auth-sensitive reads - add regression coverage and update the exploration checklist
- add cached FTS5 and R-Tree virtual-table probes to sqlite diagnostics - expose storage capability metadata on SQLite-backed NodeStore query plans - guard existing FTS helpers with runtime FTS5 detection - check off the capability-detection validation item in the exploration
- add synthetic SQLite NodeStore benchmark fixtures for 1k, 10k, 100k, and 1M node scales - add Vitest benchmark coverage for SQL pagination and scalar candidate reads - cover JS-verified property sorts in the benchmark harness - measure scalar sidecar writes, adaptive indexes, and write amplification - check off the performance and mutation benchmark validation items
- create SQLite R-Tree side tables lazily when runtime support is detected - compile spatial query descriptors into R-Tree candidate joins - keep exact spatial semantics in the shared JavaScript verifier - maintain spatial rows after registered mappings see later node writes - cover unsupported sql.js fallback and native R-Tree acceleration
- formalize token-prefix search descriptors for node queries - use SQLite FTS5 as a candidate accelerator when available - keep exact search filtering in the shared JS descriptor matcher - disable unsafe fallback pagination before search filtering - cover sql.js fallback and native FTS maintenance
- add stable materializedView query options for database views - store disposable node ID and ordinal result caches in SQLite - hydrate current NodeState pages from materialized IDs - invalidate cached view results after node writes and deletes - cover refresh, cache-hit pagination, and invalidation behavior
- Route canonical row pages through NodeStore.query descriptors. - Add offset and materializedView options to database row queries. - Opt Electron and web table/board views into stable cache keys. - Cover descriptor propagation and visible-view switching in tests. - Document the completed UI integration.
- Rename the SQLite read-scaling exploration from [_] to [x]. - Preserve the completed implementation notes and checklist state.
There was a problem hiding this comment.
Your free trial has ended. If you'd like to continue receiving code reviews, you can add a payment method here.
|
Caution Review failedPull request was closed or merged during review 📝 WalkthroughWalkthroughThis PR implements a comprehensive SQLite-backed query descriptor system for xNet's ChangesSQLite Read Scaling & Query System
🎯 4 (Complex) | ⏱️ ~60 minutes
✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
|
|
Preview: https://xnet.fyi/pr/24/app/ |
Summary
Validation
pnpm typecheckpnpm test(389 files, 5588 passed, 3 skipped)pnpm --filter xnet-desktop buildpnpm --filter xnet-web buildSummary by CodeRabbit
New Features
Documentation