sydraQL readiness checklist
This page tracks the current v0 sydraQL surface as it exists today. It is a readiness document, not a greenfield wishlist.
See also:
Spec readiness
- A basic statement surface exists (
SELECT,INSERT,DELETE,EXPLAIN) - Function metadata exists for a non-trivial builtin set
- Grammar and precedence are documented as a stable user-facing contract for the current alpha subset
- Error codes/messages are documented as a stable public API
- Practical examples cover the main compiled read path, selector tag reads, and explicit fallback behavior
Implementation readiness
- Lexer, parser, validator, planner, optimizer, and operator pipeline are present
- HTTP execution path exists at
POST /api/v1/sydraql - End-to-end scan/filter/project/aggregate/sort/limit flow exists for the implemented subset
- Selector
tag_filtersyntax is explicitly de-scoped from thev0.4.0public subset - Function execution coverage matches the current registry breadth
- Rollup/downsampling planning is implemented rather than stubbed
- Quotas and limits are enforced as a clear runtime contract
API readiness
- The HTTP endpoint exists and returns columns, rows, and stats
- Query stats contain enough detail for basic debugging
- Error responses are documented and normalized as a stable contract
- Practical client examples cover both success and failure cases
Testing & tooling readiness
- Unit coverage exists in the query pipeline modules
- Translator fixtures exist for the SQL bridge
- Golden suites for parser/plan/exec are in place
- Fuzz smoke tests run in CI via the main
zig build testlane - Query benchmarks are published for the hot paths we care about
Docs readiness
- The docs site explains the role of sydraQL in the system
- Source reference pages cover the major query modules
- The high-level docs present the supported subset cleanly enough for external users
- Readiness, backlog, and roadmap pages are reviewed regularly enough to stay truthful