Repository navigation
Expand SQL plan coverage - #226
Merged
Merged
Conversation
vogelsgesang
force-pushed
the
avogelsgesang-update-plan-dumps-duckdb
branch
7 times, most recently
from
October 24, 2026 01:29
3c525ef to
d5bdeb6
Compare
Broaden the repository SQL fixtures and checked-in plans so the example gallery exercises more relational operators, optimizer shapes, and PGQ constructs across the current database versions available in the plan-dumper environment. The SQL coverage includes all set-operation variants with nested and n-ary patterns, layered and multiply referenced CTEs, scalar and correlated subqueries, TABLESAMPLE, and DML for INSERT, UPDATE, and DELETE including RETURNING, FROM, and USING forms. DML fixtures are organized in their own directory, and the redundant UNION ALL fixture is removed. Hyper-only PGQ fixtures use meaningful TPC-H part, supplier, and partsupp relationships and cover labels, projected properties, vertex-local and edge-local WHERE predicates, graph-level filtering, optional edges, ONE ROW PER STEP, and optimizer-step plans with SQL-position annotations. Refresh the available plan dumps and index metadata for PostgreSQL, DuckDB, Hyper, Umbra, and CedarDB. CedarDB now uses v2026-10-22, and Umbra was refreshed from the umbradb/umbra:26.10 image. Retain structural changes while removing measurement-only timing, memory, scheduler, and pipeline-metric churn with filter-timing-diff.py. CedarDB pipeline-profile comparisons also ignore nondeterministic operator-list and parallelism ordering. Record TABLESAMPLE as unsupported for Umbra, CedarDB, and MariaDB. MariaDB now authenticates through the local default Unix socket, but its ANALYZE FORMAT=JSON dump consistently stalls on TPC-H Q9 before completing, so its existing plans remain unchanged pending an engine-side or runner-side timeout fix. Update loader expectations for the refreshed plan formats and keep regression coverage for DuckDB, Hyper, and the timing-diff filter.
vogelsgesang
force-pushed
the
avogelsgesang-update-plan-dumps-duckdb
branch
from
October 24, 2026 01:55
d5bdeb6 to
7baf60d
Compare
vogelsgesang
commented
Oct 24, 2026
vogelsgesang
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add dumped plans so we exercise more relational operators, optimizer shapes, and PGQ constructs.
The SQL coverage now includes:
The Hyper-only PGQ fixtures use meaningful TPC-H part, supplier, and partsupp relationships. They cover labels, projected properties, vertex-local and edge-local WHERE predicates, graph-level filtering, optional edges, ONE ROW PER STEP, and optimizer-step plans with SQL-position annotations.