Summary
Finish the last PostgreSQL conversion gaps: temporal_ordered via LAG/LEAD, prepared output format, and value-modifier transforms.
Motivation
The Postgres backend is broadly shipped; a few constructs still degrade or error. Closing them improves hunt and convert fidelity for Timescale/Postgres archives.
Proposed approach
- Implement temporal_ordered windowing with LAG/LEAD.
- Add prepared output format.
- Cover remaining value-modifier transforms with goldens.
Out of scope
- New non-Postgres backends.
- Daemon-side SQL execution (hunt run is separate).
Tasks
References
- Postgres backend module in
rsigma-convert
- Existing backend golden suite
Summary
Finish the last PostgreSQL conversion gaps: temporal_ordered via LAG/LEAD, prepared output format, and value-modifier transforms.
Motivation
The Postgres backend is broadly shipped; a few constructs still degrade or error. Closing them improves hunt and convert fidelity for Timescale/Postgres archives.
Proposed approach
Out of scope
Tasks
References
rsigma-convert