datafusion-contrib / datafusion-contrib/datafusion-distributed
[dynamic filtering] display dynamic filters for producers
Nobody has claimed this yet.
- Dominant language
- Rust
- Stars
- 139
- Forks
- 67
- Avg merge
- 3d 1h
- Merged PRs (30d)
- 35
Description
In #623, we added a way to display dynamic filters for all consumers. Some producers, ex. SortExec display dynamic filters. We should support this.
Unlike consumers, this one is hard to implement because we only show one SortExec for the stage rather than 1 per task. Consumers were easier to do because we already display a different variant per task.
┌───── DistributedExec
│ SortPreservingMergeExec: [MinTemp@0 DESC], fetch=10
│ [Stage 1] => NetworkCoalesceExec: output_partitions=6, input_tasks=2
└──────────────────────────────────────────────────
┌───── Stage 1 ── tasks=2, partitions=6
│ SortExec: TopK(fetch=10), expr=[MinTemp@0 DESC], preserve_partitioning=[true], filter=[MinTemp@0 IS NULL OR MinTemp@0 > 17.2]
│ DistributedLeafExec:
│ t0: DataSourceExec: file_groups={3 groups: [[/testdata/weather/result-000000.parquet:<int>..<int>], [/testdata/weather/result-000000.parquet:<int>..<int>, /testdata/weather/result-000001.parquet:<int>..<int>], [/testdata/weather/result-000002.parquet:<int>..<int>]]}, projection=[MinTemp], file_type=parquet, predicate=DynamicFilter [ MinTemp@0 IS NULL OR MinTemp@0 > 17.2 ], dynamic_rg_pruning=eligible, pruning_predicate=MinTemp_null_count@0 > 0 OR MinTemp_null_count@0 != row_count@2 AND MinTemp_max@1 > 17.2, required_guarantees=[]
│ t1: DataSourceExec: file_groups={3 groups: [[/testdata/weather/result-000000.parquet:<int>..<int>], [/testdata/weather/result-000001.parquet:<int>..<int>, /testdata/weather/result-000002.parquet:<int>..<int>], [/testdata/weather/result-000002.parquet:<int>..<int>]]}, projection=[MinTemp], file_type=parquet, predicate=DynamicFilter [ MinTemp@0 IS NULL OR MinTemp@0 > 17.2 ], dynamic_rg_pruning=eligible, pruning_predicate=MinTemp_null_count@0 > 0 OR MinTemp_null_count@0 != row_count@2 AND MinTemp_max@1 > 17.2, required_guarantees=[]
└──────────────────────────────────────────────────
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reading the dynamic-filter display work from #623 and trace how consumer variants are rendered per task. Focus on the producer-side SortExec and the stage-level representation, including SortPreservingMergeExec and NetworkCoalesceExec. Done means producer dynamic filters appear correctly even when only one SortExec is shown for the stage.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- rust
- Domain
- distributed-systems
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 48/100