lance-format / lance-format/lance
[MemTable] Make unenforced primary key columns ordered
Nobody has claimed this yet.
- Dominant language
- Rust
- Stars
- 7.1k
- Forks
- 852
- Avg merge
- 3d 18h
- Merged PRs (30d)
- 272
Description
#4002 introduced the unenforced primary key concept, but the key is unordered because we just record a primary key boolean at the field level. @majin1102 brought up this in https://github.com/lancedb/lance/pull/4207#discussion_r2208168187.
I thought about this a bit, it's probably better to make it ordered for future potential use cases. For example, we could build a multi-column b-tree index for primary key columns instead one one for each column, and in that case the order of the column would affect the effectiveness of the b-tree.
Contributor guide
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 at the MemTable implementation and the unenforced primary-key metadata introduced by #4002; review the ordering concern in PR #4207. Done means primary-key columns retain a defined order rather than only a field-level boolean, with behavior suitable for future multi-column index use.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- rust
- Domain
- databases
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100