Skip to content

chore(deps): bump datafusion 40.1.0 -> 52.3.0#253

Closed
bvolpato wants to merge 1 commit intosubstrait-io:mainfrom
bvolpato:bvolpato/bump-datafusion-52.3.0
Closed

chore(deps): bump datafusion 40.1.0 -> 52.3.0#253
bvolpato wants to merge 1 commit intosubstrait-io:mainfrom
bvolpato:bvolpato/bump-datafusion-52.3.0

Conversation

@bvolpato
Copy link
Copy Markdown
Member

Summary

Bumps datafusion (PyDataFusion) from 40.1.0 to 52.3.0 — a 12-version jump.

Changes

  • requirements.txt: datafusion 40.1.052.3.0
  • requirements-unlocked.txt: datafusion 40.1.052.3.0

No code changes needed — the substrait producer/consumer API (ds.serde.serialize_to_plan, ds.serde.deserialize_bytes, ds.consumer.from_substrait_plan) remains compatible.

Testing

Verified import and round-trip (serialize → deserialize → execute) works with DataFusion 52.3.0.

Bumps the datafusion Python package from 40.1.0 to 52.3.0.
The substrait producer/consumer API remains compatible.
@bvolpato
Copy link
Copy Markdown
Member Author

Closing: this bump requires regenerating all DataFusion test snapshots (producer plans + consumer results) and support matrix CSVs. The snapshot changes are massive (~8.6M lines across 287 files) because DataFusion's query results changed across 12 major versions (40 → 52). This needs a dedicated effort to:

  1. Regenerate producer snapshots (DataFusion plans)
  2. Regenerate consumer result snapshots across all producer×consumer combinations
  3. Update the support matrix CSVs

Will revisit with a proper approach — likely running the full regeneration in CI.

@bvolpato bvolpato closed this Mar 17, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Development

Successfully merging this pull request may close these issues.

1 participant