mirror of
https://github.com/lancedb/lancedb.git
synced 2026-10-28 22:46:14 +00:00
lancedb-gatefixer[bot]
dcbf5ce66a
SQL descriptor access returns NULL for every field of a NULL Blob v2 value, so `payload.size = 0` matches only empty blobs and `payload.size IS NULL` matches NULL blobs. DataFusion 54.1.0 extracts struct children without combining their validity with the parent struct. Lance descriptors keep nullness in the parent and can store zero-filled children beneath NULL blobs. The adapter derives the descriptor schema from Lance's default projection, marks descriptor children nullable, and masks them before SQL evaluation. Schema derivation requires no query request, and the mask handles struct and list nesting while preserving existing child nulls and value buffers without fetching blob bytes. Filter classification follows the referenced struct field path. Blob predicates stay above the mask; scalar siblings such as `extra.note` retain exact filter pushdown even when `extra.clip` is a blob. **Merge hold:** as requested in the [review discussion](https://github.com/lancedb/lancedb/pull/4571#issuecomment-6253443077), blob SQL INSERT compatibility remains pending a separate write-mode design. DataFusion uses the provider's single schema for reads and INSERT coercion, so the descriptor schema still rejects blob write structs. This follow-up does not restore those INSERTs; the merge should wait until the write contract is settled. Validation: - Reproduced the original NULL blob returning `kind = 0` before enabling the mask. - Reproduced both the extra remote query and lost nested sibling pushdown before the follow-up fixes. - All 441 DataFusion adapter tests passed after merging main at `e6789a41`, including SQL projections, filters, mixed predicates, LIMIT, COUNT, schema agreement, sliced arrays, nested structs, list variants, and EXPLAIN assertions for scalar sibling pushdown. The blob fixtures use the current four-field write format, and NULL-blob filtering is also covered with row IDs and row addresses enabled. - Remote constructor regression passed: one describe request and no query request, with nullable descriptor fields preserved. - `cargo fmt --all` - `cargo clippy --locked --profile ci -p lancedb --features remote --tests --examples -- -D warnings` Fixes #4569 <!-- lance-gatekeeper-fix:v1 agent=2ca7389efd48d11a8debe9111abeb00e generation=1 --> --------- Co-authored-by: Gatefixer <313497061+lancedb-gatefixer[bot]@users.noreply.github.com>
The Multimodal Lakehouse for AI
How to Install ✦ Detailed Documentation ✦ Tutorials and Recipes ✦ Contributors
The ultimate multimodal data platform for AI/ML applications.
LanceDB is designed for fast, scalable, and production-ready vector search. It is built on top of the Lance columnar format. You can store, index, and search over petabytes of multimodal data and vectors with ease. LanceDB is a central location where developers can build, train and analyze their AI workloads.
Demo: Multimodal Search by Keyword, Vector or with SQL
Star LanceDB to get updates!
⭐ Click here ⭐ to see how fast we're growing!
Key Features:
- Fast Vector Search: Search billions of vectors in milliseconds with state-of-the-art indexing.
- Comprehensive Search: Support for vector similarity search, full-text search and SQL.
- Multimodal Support: Store, query and filter vectors, metadata and multimodal data (text, images, videos, point clouds, and more).
- Advanced Features: Zero-copy, automatic versioning, manage versions of your data without needing extra infrastructure. GPU support in building vector index.
Products:
- Open Source & Local: 100% open source, runs locally or in your cloud. No vendor lock-in.
- Cloud and Enterprise: Production-scale vector search with no servers to manage. Complete data sovereignty and security.
Ecosystem:
- Columnar Storage: Built on the Lance columnar format for efficient storage and analytics.
- Seamless Integration: Python, Node.js, Rust, and REST APIs for easy integration. Native Python and Javascript/Typescript support.
- Rich Ecosystem: Integrations with LangChain 🦜️🔗, LlamaIndex 🦙, Apache-Arrow, Pandas, Polars, DuckDB and more on the way.
How to Install:
Follow the Quickstart doc to set up LanceDB locally.
API & SDK: We also support Python, Typescript and Rust SDKs
| Interface | Documentation |
|---|---|
| Python SDK | https://lancedb.github.io/lancedb/python/python/ |
| Typescript SDK | https://lancedb.github.io/lancedb/js/globals/ |
| Rust SDK | https://docs.rs/lancedb/latest/lancedb/index.html |
| REST API | https://docs.lancedb.com/api-reference/rest |
Join Us and Contribute
We welcome contributions from everyone! Whether you're a developer, researcher, or just someone who wants to help out.
If you have any suggestions or feature requests, please feel free to open an issue on GitHub or discuss it on our Discord server.
Check out the GitHub Issues if you would like to work on the features that are planned for the future. If you have any suggestions or feature requests, please feel free to open an issue on GitHub.
Contributors
Stay in Touch With Us