DataFusion weekly project plan (Andrew Lamb) - July 1, 2024

[alamb]

Follow on to #11106

My (personal) North ⭐ : 1000 projects are built using DataFusion 📈

It would be great for other contributors to DataFusion to to make them visible somehow as well 🙏 (feel free to copy / modify the format of this ticket)

Externally Facing News

• Meetup in SF went great thanks to @cpard and @emgeee Invitation: Monday June 24, 2024 5:00PM PDT: DataFusion Meetup in San Francisco #10800
• Talk at MSR went well: Presentation about DataFusion at Microsoft GSL (Gray Systems Lab) #10969

Feature Highlights

• feat: Conditionally allow to keep partition_by columns when using PARTITIONED BY enhancement #11107 with @hveiga
• Support COPY TO Externally Defined File Formats, add FileType trait #11060 from @devinjdangelo (and fast follow on docs: add example for custom file format with COPY TO #11174 docs from @tshauck)
• @Blizzara continues to churn through making substrait better: feat: Support Map type in Substrait conversions #11129
• @2010YOUY01 found a bunch of corner cases in SQLancer Implement SQLancer (a end-to-end SQL fuzz testing library) #11030 and then @LorrensP-2158466, @jonahgao and others fixed corner cases like Overflow bug in POW scalar function (found by SQLancer)  #11075

My (personal) plans for this week

• July 10, 2024 ASF Board Report #10282
• Blog post with DataFusion Jan - June 2024 #9602
• SQLParser release: Release sqlparser-rs version 0.48.0 sqlparser-rs/sqlparser-rs#1296
• Release arrow-rs / parquet minor version 52.1.0 arrow-rs#5905
• Faster multi-column aggregation with @jayzhan211 Improve performance for grouping by variable length columns (strings) #9403 (Optionally display schema in explain plan #11177 was a step towards this)

Project Queue (list of future projects)

• Return TableProviderFilterPushDown::Exact when Parquet Pushdown Enabled #4028

Projects I plan to help actively help review / plan

• Help user defined SQL planner API Discussion: make it easier for specify SQL --> function translation #10534 with @jayzhan211 and @samuelcolvin, e.g. Introduce user defined SQL planner API #11180)
• Help push USAD over the line ([Epic] Unify AggregateFunction Interface (remove built in list of AggregateFunction s), improve the system #8708 -- e.g. Convert Min and Max to UDAF  #10943 . Thanks to @edmondop)
• [EPIC] Continued correct and improved extracting Parquet statistics into ArrayRefs  #10922 (I hope to finish the last few items, with @efredine, @dharanad and others)
• [EPIC] Implement StringViewArray and BinaryViewArray arrow-rs#5374 and [Epic] Implement support for StringView in DataFusion #10918 with @XiangpengHao and others

Algorithm for (my) prioritizing PR reviews

Note there are many committers who can and do review and merge PRs, so this is not the priorities of the project as a whole, just the approximate algorithm I use to prioritize my own time.

Priority:

1. Bug fixes (where something is just incorrect), especially regressions (where it used to work and now does not)
2. Improvements directly related to features needed for InfluxDB (my employer)
3. Documentation and test improvements (I view these as very strategically important)
4. PRs that I think are strategically important
5. Other new features / additions to functionality (note this is the lowest)

The top strategically important projects in my head are:

• Anything that makes it easier to use DataFusion as a user (docs, examples, better APIs, etc), including Getting started guide for new users (who want to use DataFusion in their project)
• Anything that improves DataFusion's quality such as bug fixes, and improved / expanded tests, etc
• Making function packages modular and easier to mix/match
• Improved performance (with benchmarks)

Thus, if you are interested in contributing to DataFusion and are interested in a fast turn around time I would recommend looking into bug fixes / test improvements / documentation or the projects named above.

If you propose adding new functionality, especially if the PR is large/complex and not connected to a wider need, the review cycle will likely be longer. You can make it a shorter cycle by looking at the comments on other recent PRs and following the same model (e.g. ensure there are tests in sqllogictest for example, the CI passes, includes documentation, etc)

Background

The idea of this ticket is make my plans for DataFusion visible, largely for my own personal organizational needs, but also to:

1. Communicate / coordinate in the community
2. Help provide an interesting summary of what is happening in DataFusion this week

[alamb]

Review Queue

DataFusion

• [draft] Add LogicalType, try to support user-defined types #11160
• Implement prettier SQL unparsing (more human readable) #11186
• Enhance short circuit handling in CommonSubexprEliminate #11197

Arrow

• feat(5851): ArrowWriter memory usage arrow-rs#5967
• Implement directly build byte view array on top of parquet buffer arrow-rs#5972
• Add size statistics to ParquetMetaData introduced in PARQUET-2261 arrow-rs#5486

[alamb]

Review Queue

DataFusion

• Implement prettier SQL unparsing (more human readable) #11186
• Enhance short circuit handling in CommonSubexprEliminate #11197
• Add Optimizer Sanity Checker, improve sortedness equivalence properties #11196

Arrow:

• Fix 5592: Colon (:) in in object_store::path::{Path} is not handled on Windows arrow-rs#5830

[alamb]

Review Queue

DataFusion

• Introduce the DynamicFileCatalog in datafusion-catalog #11035
• Feat: Implement hf:// / "hugging face" integration in datafusion-cli #10792

[samuelcolvin]

I'd love to get apache/arrow-rs#5994 merged and released fairly soon if at all possible 🤞 🙏 .

[alamb]

Review Queue

Arrow

• Add ParquetMetadataWriter allow ad-hoc encoding of ParquetMetadata arrow-rs#6000
• Do not write ColumnIndex for null columns when not writing page statistics arrow-rs#6011
• Add size statistics to ParquetMetaData introduced in PARQUET-2261 arrow-rs#5486
• Make display of interval types more pretty arrow-rs#6006

DataFusion

• Introduce the DynamicFileCatalog in datafusion-catalog #11035
• Extract parquet statistics for StructArray #11289
• Feat: Implement hf:// / "hugging face" integration in datafusion-cli #10792

[alamb]

Next week in #11334