[austin-barrington]

Stacked on #133. Implements docs/superpowers/plans/2026-11-03-query-analysis-module.md.

timeseriesql::analysis now holds the pure InfluxQL AST facts the domain needs:

• output column names;
• aggregate detection;
• the TOP/BOTTOM/SAMPLE selector descriptor;
• the merge column names.

domain::rollup, domain::sharding::query_merge and domain::sharding::multi_selector now import from it instead of to_clickhouse. The translator re-exports the moved items, so the generated SQL is unchanged.

The move is pure: a reviewer compared every moved body line by line, and only visibility changed. New tests pin output names, that window transforms are not aggregates, and the merge column literals.

Gate: fmt, clippy -D warnings, 883 lib tests, 373 integration tests and 54 CLI tests pass.

[cursor]

PR Summary

Low Risk
Pure refactor with line-for-line moves and translator re-exports; no intended behavior change to query translation or merge semantics.

Overview
Introduces timeseriesql::analysis as the home for backend-agnostic InfluxQL AST facts: output column naming (select_output_field_name, Influx-style aliases), true aggregate detection (excluding bare window transforms), TOP/BOTTOM/SAMPLE MultiSelector metadata, and shared merge column constants __bucket / __tns.

Domain code in rollup, sharding::query_merge, and sharding::multi_selector now imports these from analysis instead of to_clickhouse, so rollup and shard merge logic no longer depend on the ClickHouse translator.

The moved logic is lifted out of to_clickhouse/select, aggregates, and selectors; the translator re-exports the same public items and pulls in shared helpers (e.g. expr_contains_raw_transform) so generated SQL behavior should be unchanged. analysis adds unit tests for output names, window transforms vs aggregates, and merge column literals.

^{Reviewed by Cursor Bugbot for commit 88a2840. Configure here.}

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: bb52c930-66ad-405e-94ef-5139fa612dcb)