RemoteDBConnection should support passing exist_ok to create_table, just
like LanceDBConnection (the non-remote form) does. It can support this
by passing 'exist_ok' as the mode parameter.
Implement parallel execution of multiple embedding functions using
std:🧵:scope to improve performance when a table has multiple
embedding columns.
Key changes:
- Add compute_embeddings_parallel() helper method to WithEmbeddings
- Use fast path for single embeddings (no threading overhead)
- Use scoped threads for parallel execution of multiple embeddings
- Add comprehensive tests including parallelization timing verification
- Update WithEmbeddings documentation
Performance improvements:
- I/O-bound embeddings (OpenAI, Bedrock): High benefit from concurrent
API calls
- CPU-bound embeddings (sentence-transformers): Medium benefit from core
utilization
- Single embedding: No overhead (fast path)
Closes TODO on line 266 in rust/lancedb/src/embeddings.rs
Aesthetic and styling fixes to the SDK reference docs:
- [x] Improve readability of LanceDB in the header
- [x] Make header more compact, and consistent in gradient color with
the main website/docs
- [x] Updated favicon to match with the docs page
- [x] Enable permalink display to allow users to get anchor links to
each function/method
- [x] Point readers to the main docs at
[docs.lancedb.com](https://docs.lancedb.com)
Convert test_table_names to test both remote and local connections.
This PR also includes some miscellaneous improvements in
src/test_utils/connection.rs. It starts a thread to drain stdout from
the server process. It adds the
PRINT_LANCEDB_TEST_CONNECTION_SCRIPT_OUTPUT environment variable, which
optionally displays server stdout.
Fix a bash conditional in run_with_test_connection.sh.
The page_token and limit parameters for table_names() are supported by
both local storage and LanceDB Cloud, not just Cloud as the docstring
incorrectly stated.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
Adds `Table.to_lance()` and `Table.to_polars()` methods (non-abstract
methods, defaulting to `NotImplementedError`) so type checkers like
mypy, pyright and ty don’t flag them as unknown attributes on `Table`.
Not making these abstract methods should keep existing remote/other
`Table` implementations instantiable.
This is non-breaking change to existing functionality and is purely for
the purpose of pleasing static type-checkers like mypy, ty and pyright.
<img width="626" height="134" alt="image"
src="https://github.com/user-attachments/assets/f4619bca-a882-432b-bd23-ae8f189ff9e3"
/>
Issues found during integration tests:
1. describe_namespace should use POST
2. service needs to access the underlying namespace to be able to do
operations like create_empty_table directly, or get credentials in
isolated paths like a remote take
## Summary
- update all lance crates to v1.0.0 using the helper script (fallbacks
to the v1.0.0 tag)
- refresh Cargo.lock to pull the new release
- add script fallback to retry with the git tag when a crates.io release
is unavailable
## Testing
- cargo clippy --workspace --tests --all-features -- -D warnings
- cargo fmt --all
Tag: https://github.com/lance-format/lance/releases/tag/v1.0.0
---------
Co-authored-by: Jack Ye <yezhaoqin@gmail.com>
## Summary
- infer integer vector columns as float32 when any value exceeds uint8
range or is negative
- keep uint8 for integer vectors within range and nulls only
- add sync/async tests covering large integer vector inference
## Testing
- ./.venv/bin/pytest python/python/tests/test_table.py -k
"large_int_vectors"
In #2845 we ported the lancedb integration in lance-namespace to
lancedb. But that is too specific to RestNamespace. We can improve the
user entry point so that we can put local mode and future version of the
Flight SQL-based LanceDB server all behind this single
`LanceDbNamespaceClientBuilder` API.
Also I renamed `namespace` to `namesapceClient` to avoid confusion with
the namespace path.
After the refactoring on both client and server side, we should have the
ability to fully use lance REST namespace to call into LanceDB cloud and
enterprise. We can avoid having a JNI implementation (which today does
not really do anything except for vending a connection object), and just
use lance-core's RestNamespace.
We will at this moment have a LanceDbRestNamespaceBuilder to allow users
to more easily build the RestNamespace to talk to LanceDB Cloud or
Enterprise endpoint.
In the future, we could extend this further to also support the local
mode through DirectoryNamespace. That will be a separated PR.
Adds IVF_SQ index config through Rust core and Python bindings, plus
alias names IvfHnswSq/Pq for backward compatibility. Updates
remote/table helpers and types to accept the new index type. Includes
tests covering IVF SQ creation and alias usage.
1. Use generated models in lance-namespace for request response models
to avoid multiple layers of conversions
2. Make sure the API is consistent with the namespace spec
3. Deprecate the table_names API in favor of the list_tables API in
namespace that allows full pagination support without the need to have
sorted table names
4. Add describe_namespace API which was a miss in the original
implementation
Currently a table in a namespace is still backed with a `NativeTable`,
which means after getting the location of the table and optional storage
options override from `namespace.describe_table`, all things work like a
normal local table. However, namespace also supports `query_table`,
which is exactly the same API as remote table. This PR adds a
`server_side_query` capability, when enabled, it runs the query by
calling `namespace.query_table`. For namespace that implements the
operation (e.g. REST namespace), this could hit a backend server that
could execute the query faster (e.g. using a distributed engine).
We have very low download stats for mac x86, and also latest github
runners for mac are all arm, so it makes sense at this point to
deprecate x86 support in general.
Add support for enabling stable row IDs when creating tables via the
`new_table_enable_stable_row_ids` storage option.
Stable row IDs ensure that row identifiers remain constant after
compaction, update, delete, and merge operations. This is useful for
materialized views and other use cases that need to track source rows
across these operations.
The option can be set at two levels:
- Connection level: applies to all tables created with that connection
- Table level: per-table override via create_table storage_options
🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: Claude <noreply@anthropic.com>
## Summary
- bump all Lance crates to v1.0.0-beta.16 via ci/set_lance_version.py
- refresh Cargo.lock (reqwest/opendal/etc.) to satisfy the new release
## Verification
- cargo clippy --workspace --tests --all-features -- -D warnings
- cargo fmt --all
Triggered by
[refs/tags/v1.0.0-beta.16](https://github.com/lance-format/lance/releases/tag/v1.0.0-beta.16)
---------
Co-authored-by: Jack Ye <yezhaoqin@gmail.com>
## Summary
- bump all Lance crates to 1.0.0-beta.14 via ci/set_lance_version.py
- refresh Cargo.lock to capture new transitive requirements
- verified `cargo clippy --workspace --tests --all-features -- -D
warnings` and `cargo fmt --all`
Triggered by refs/tags/v1.0.0-beta.14
---------
Co-authored-by: Jack Ye <yezhaoqin@gmail.com>
This tries to fix#2771. It is not a complete fix because
`lance-namespace-impls` uses `lance` which has its default features
enabled. Thus, to close#2771, the lance repo also needs an update.
The `dir-*` features are enabled by the respective remote feature
(`aws`, `gcp`, `azure`, `oss`).
The `rest` feature is enabled via `remote`.
clippy::string_to_string has been replaced by implicit_clone, so lancedb
will raise a build error in Rust 1.91. This PR suppresses it.
---
**This PR was primarily authored with Codex using GPT-5-Codex and then
hand-reviewed by me. I AM responsible for every change made in this PR.
I aimed to keep it aligned with our goals, though I may have missed
minor issues. Please flag anything that feels off, I'll fix it
quickly.**
Signed-off-by: Xuanwo <github@xuanwo.io>
