Tracking Issue: Extension Types

[connortsui20]

This is a further continuation of #6618

Extension Types

Now that things have more or less been figured out, we can actually start implementing extension arrays. This tracking issue will simply link together other tracking issues and PRs related to extension types.

See the Extension Types RFC for more details.

(Potential) Future Extension Types

• DateTimeParts (Primitive)
• Vector (FixedSizeList)
• Matrix (FixedSizeList)
• Tensor (FixedSizeList)
• UUID (Do we need to add FixedSizeBinary as a canonical type?)
• JSON (UTF8 and also maybe Variant?)
• PDX: https://arxiv.org/pdf/2503.04422v1 (FixedSizeList)
• Tags: Efficient encoding for List<Categorical> #5772 (comment) (ListView<Utf8>)
• Struct but with protobuf-style field numbers (Struct)
• Probably lots more!

Steps / History

• Tracking Issue: Tensor Types #6865
• Tracking Issue: UUID Extension Type #6854
• Pluggable Compressor #7018

Unresolved Questions

TODO