Format python

Signed-off-by: Xuanwo <github@xuanwo.io>

.bumpversion.toml

@@ -1,5 +1,5 @@
 [tool.bumpversion]
-current_version = "0.19.1-beta.1"
+current_version = "0.21.1-beta.1"
 parse = """(?x)
     (?P<major>0|[1-9]\\d*)\\.
     (?P<minor>0|[1-9]\\d*)\\.

.github/workflows/java.yml

@@ -35,6 +35,9 @@ jobs:
       - uses: Swatinem/rust-cache@v2
         with:
           workspaces: java/core/lancedb-jni
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
+        with:
+          components: rustfmt
       - name: Run cargo fmt
         run: cargo fmt --check
         working-directory: ./java/core/lancedb-jni
@@ -68,6 +71,9 @@ jobs:
       - uses: Swatinem/rust-cache@v2
         with:
           workspaces: java/core/lancedb-jni
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
+        with:
+          components: rustfmt
       - name: Run cargo fmt
         run: cargo fmt --check
         working-directory: ./java/core/lancedb-jni
@@ -110,4 +116,3 @@ jobs:
           -Djdk.reflect.useDirectMethodHandle=false \
           -Dio.netty.tryReflectionSetAccessible=true"
           JAVA_HOME=$JAVA_17 mvn clean test
-  

.github/workflows/make-release-commit.yml

@@ -84,6 +84,7 @@ jobs:
         run: |
           pip install bump-my-version PyGithub packaging
           bash ci/bump_version.sh ${{ inputs.type }} ${{ inputs.bump-minor }} v $COMMIT_BEFORE_BUMP
+          bash ci/update_lockfiles.sh --amend
       - name: Push new version tag
         if: ${{ !inputs.dry_run }}
         uses: ad-m/github-push-action@master
@@ -92,11 +93,3 @@ jobs:
           github_token: ${{ secrets.LANCEDB_RELEASE_TOKEN }}
           branch: ${{ github.ref }}
           tags: true
-      - uses: ./.github/workflows/update_package_lock
-        if: ${{ !inputs.dry_run && inputs.other }}
-        with:
-          github_token: ${{ secrets.GITHUB_TOKEN }}
-      - uses: ./.github/workflows/update_package_lock_nodejs
-        if: ${{ !inputs.dry_run && inputs.other }}
-        with:
-          github_token: ${{ secrets.GITHUB_TOKEN }}

.github/workflows/nodejs.yml

@@ -47,6 +47,9 @@ jobs:
       run: |
         sudo apt update
         sudo apt install -y protobuf-compiler libssl-dev
+    - uses: actions-rust-lang/setup-rust-toolchain@v1
+      with:
+        components: rustfmt, clippy
     - name: Lint
       run: |
         cargo fmt --all -- --check
@@ -113,7 +116,7 @@ jobs:
         set -e
         npm ci
         npm run docs
-        if ! git diff --exit-code; then
+        if ! git diff --exit-code -- . ':(exclude)Cargo.lock'; then
           echo "Docs need to be updated"
           echo "Run 'npm run docs', fix any warnings, and commit the changes."
           exit 1

.github/workflows/npm-publish.yml

@@ -505,6 +505,8 @@ jobs:
     name: vectordb NPM Publish
     needs: [node, node-macos, node-linux-gnu, node-windows]
     runs-on: ubuntu-latest
+    permissions:
+      contents: write
     # Only runs on tags that matches the make-release action
     if: startsWith(github.ref, 'refs/tags/v')
     steps:
@@ -537,6 +539,20 @@ jobs:
         # We need to deprecate the old package to avoid confusion.
         # Each time we publish a new version, it gets undeprecated.
         run: npm deprecate vectordb "Use @lancedb/lancedb instead."
+      - name: Checkout
+        uses: actions/checkout@v4
+        with:
+          ref: main
+      - name: Update package-lock.json
+        run: |
+          git config user.name 'Lance Release'
+          git config user.email 'lance-dev@lancedb.com'
+          bash ci/update_lockfiles.sh
+      - name: Push new commit
+        uses: ad-m/github-push-action@master
+        with:
+          github_token: ${{ secrets.LANCEDB_RELEASE_TOKEN }}
+          branch: main
       - name: Notify Slack Action
         uses: ravsamhq/notify-slack-action@2.3.0
         if: ${{ always() }}
@@ -546,21 +562,3 @@ jobs:
           notification_title: "{workflow} is failing"
         env:
           SLACK_WEBHOOK_URL: ${{ secrets.ACTION_MONITORING_SLACK }}
-
-  update-package-lock:
-    if: startsWith(github.ref, 'refs/tags/v')
-    needs: [release]
-    runs-on: ubuntu-latest
-    permissions:
-      contents: write
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-        with:
-          ref: main
-          token: ${{ secrets.LANCEDB_RELEASE_TOKEN }}
-          fetch-depth: 0
-          lfs: true
-      - uses: ./.github/workflows/update_package_lock
-        with:
-          github_token: ${{ secrets.GITHUB_TOKEN }}

.github/workflows/run_tests/action.yml

@@ -24,8 +24,8 @@ runs:
     - name: pytest (with integration)
       shell: bash
       if: ${{ inputs.integration == 'true' }}
-      run: pytest -m "not slow" -x -v --durations=30 python/python/tests
+      run: pytest -m "not slow" -vv --durations=30 python/python/tests
     - name: pytest (no integration tests)
       shell: bash
       if: ${{ inputs.integration != 'true' }}
-      run: pytest -m "not slow and not s3_test" -x -v --durations=30 python/python/tests
+      run: pytest -m "not slow and not s3_test" -vv --durations=30 python/python/tests

.github/workflows/rust.yml

@@ -40,6 +40,9 @@ jobs:
         with:
           fetch-depth: 0
           lfs: true
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
+        with:
+          components: rustfmt, clippy
       - uses: Swatinem/rust-cache@v2
         with:
           workspaces: rust
@@ -160,8 +163,8 @@ jobs:
     strategy:
       matrix:
         target:
-        - x86_64-pc-windows-msvc
-        - aarch64-pc-windows-msvc
+          - x86_64-pc-windows-msvc
+          - aarch64-pc-windows-msvc
     defaults:
       run:
         working-directory: rust/lancedb

.github/workflows/update_package_lock/action.yml

@@ -1,33 +0,0 @@
-name: update_package_lock
-description: "Update node's package.lock"
-
-inputs:
-  github_token:
-    required: true
-    description: "github token for the repo"
-
-runs:
-  using: "composite"
-  steps:
-    - uses: actions/setup-node@v3
-      with:
-        node-version: 20
-    - name: Set git configs
-      shell: bash
-      run: |
-        git config user.name 'Lance Release'
-        git config user.email 'lance-dev@lancedb.com'
-    - name: Update package-lock.json file
-      working-directory: ./node
-      run: |
-        npm install
-        git add package-lock.json
-        git commit -m "Updating package-lock.json"
-      shell: bash
-    - name: Push changes
-      if: ${{ inputs.dry_run }} == "false"
-      uses: ad-m/github-push-action@master
-      with:
-        github_token: ${{ inputs.github_token }}
-        branch: main
-        tags: true

.github/workflows/update_package_lock_nodejs/action.yml

@@ -1,33 +0,0 @@
-name: update_package_lock_nodejs
-description: "Update nodejs's package.lock"
-
-inputs:
-  github_token:
-    required: true
-    description: "github token for the repo"
-
-runs:
-  using: "composite"
-  steps:
-    - uses: actions/setup-node@v3
-      with:
-        node-version: 20
-    - name: Set git configs
-      shell: bash
-      run: |
-        git config user.name 'Lance Release'
-        git config user.email 'lance-dev@lancedb.com'
-    - name: Update package-lock.json file
-      working-directory: ./nodejs
-      run: |
-        npm install
-        git add package-lock.json
-        git commit -m "Updating package-lock.json"
-      shell: bash
-    - name: Push changes
-      if: ${{ inputs.dry_run }} == "false"
-      uses: ad-m/github-push-action@master
-      with:
-        github_token: ${{ inputs.github_token }}
-        branch: main
-        tags: true

Cargo.toml

@@ -21,55 +21,52 @@ categories = ["database-implementations"]
 rust-version = "1.78.0"
 
 [workspace.dependencies]
-lance = { "version" = "=0.27.0", "features" = ["dynamodb"], tag = "v0.27.0-beta.2", git="https://github.com/lancedb/lance.git" }
-lance-io = { version = "=0.27.0", tag = "v0.27.0-beta.2", git="https://github.com/lancedb/lance.git" }
-lance-index = { version = "=0.27.0", tag = "v0.27.0-beta.2", git="https://github.com/lancedb/lance.git" }
-lance-linalg = { version = "=0.27.0", tag = "v0.27.0-beta.2", git="https://github.com/lancedb/lance.git" }
-lance-table = { version = "=0.27.0", tag = "v0.27.0-beta.2", git="https://github.com/lancedb/lance.git" }
-lance-testing = { version = "=0.27.0", tag = "v0.27.0-beta.2", git="https://github.com/lancedb/lance.git" }
-lance-datafusion = { version = "=0.27.0", tag = "v0.27.0-beta.2", git="https://github.com/lancedb/lance.git" }
-lance-encoding = { version = "=0.27.0", tag = "v0.27.0-beta.2", git="https://github.com/lancedb/lance.git" }
+lance = { "version" = "=0.31.1", tag="v0.31.1-beta.2", git="https://github.com/lancedb/lance.git", features = ["dynamodb"] }
+lance-io = { "version" = "=0.31.1", tag="v0.31.1-beta.2", git="https://github.com/lancedb/lance.git" }
+lance-index = { "version" = "=0.31.1", tag="v0.31.1-beta.2", git="https://github.com/lancedb/lance.git" }
+lance-linalg = { "version" = "=0.31.1", tag="v0.31.1-beta.2", git="https://github.com/lancedb/lance.git" }
+lance-table = { "version" = "=0.31.1", tag="v0.31.1-beta.2", git="https://github.com/lancedb/lance.git" }
+lance-testing = { "version" = "=0.31.1", tag="v0.31.1-beta.2", git="https://github.com/lancedb/lance.git" }
+lance-datafusion = { "version" = "=0.31.1", tag="v0.31.1-beta.2", git="https://github.com/lancedb/lance.git" }
+lance-encoding = { "version" = "=0.31.1", tag="v0.31.1-beta.2", git="https://github.com/lancedb/lance.git" }
 # Note that this one does not include pyarrow
-arrow = { version = "54.1", optional = false }
-arrow-array = "54.1"
-arrow-data = "54.1"
-arrow-ipc = "54.1"
-arrow-ord = "54.1"
-arrow-schema = "54.1"
-arrow-arith = "54.1"
-arrow-cast = "54.1"
+arrow = { version = "55.1", optional = false }
+arrow-array = "55.1"
+arrow-data = "55.1"
+arrow-ipc = "55.1"
+arrow-ord = "55.1"
+arrow-schema = "55.1"
+arrow-arith = "55.1"
+arrow-cast = "55.1"
 async-trait = "0"
-datafusion = { version = "46.0", default-features = false }
-datafusion-catalog = "46.0"
-datafusion-common = { version = "46.0", default-features = false }
-datafusion-execution = "46.0"
-datafusion-expr = "46.0"
-datafusion-physical-plan = "46.0"
+datafusion = { version = "48.0", default-features = false }
+datafusion-catalog = "48.0"
+datafusion-common = { version = "48.0", default-features = false }
+datafusion-execution = "48.0"
+datafusion-expr = "48.0"
+datafusion-physical-plan = "48.0"
 env_logger = "0.11"
-half = { "version" = "=2.4.1", default-features = false, features = [
+half = { "version" = "2.6.0", default-features = false, features = [
     "num-traits",
 ] }
 futures = "0"
 log = "0.4"
 moka = { version = "0.12", features = ["future"] }
-object_store = "0.11.0"
+object_store = "0.12.0"
 pin-project = "1.0.7"
 snafu = "0.8"
 url = "2"
 num-traits = "0.2"
-rand = "0.8"
+rand = "0.9"
 regex = "1.10"
 lazy_static = "1"
 semver = "1.0.25"
-
 # Temporary pins to work around downstream issues
 # https://github.com/apache/arrow-rs/commit/2fddf85afcd20110ce783ed5b4cdeb82293da30b
-chrono = "=0.4.39"
+chrono = "=0.4.41"
 # https://github.com/RustCrypto/formats/issues/1684
 base64ct = "=1.6.0"
-
 # Workaround for: https://github.com/eira-fransham/crunchy/issues/13
 crunchy = "=0.2.2"
-
 # Workaround for: https://github.com/Lokathor/bytemuck/issues/306
 bytemuck_derive = ">=1.8.1, <1.9.0"

README.md

@@ -1,94 +1,97 @@
 <a href="https://cloud.lancedb.com" target="_blank">
   <img src="https://github.com/user-attachments/assets/92dad0a2-2a37-4ce1-b783-0d1b4f30a00c" alt="LanceDB Cloud Public Beta" width="100%" style="max-width: 100%;">
 </a>
-
 <div align="center">
-<p align="center">
 
-<picture>
-  <source media="(prefers-color-scheme: dark)" srcset="https://github.com/user-attachments/assets/ac270358-333e-4bea-a132-acefaa94040e">
-  <source media="(prefers-color-scheme: light)" srcset="https://github.com/user-attachments/assets/b864d814-0d29-4784-8fd9-807297c758c0">
-  <img alt="LanceDB Logo" src="https://github.com/user-attachments/assets/b864d814-0d29-4784-8fd9-807297c758c0" width=300>
-</picture>
+[![LanceDB](docs/src/assets/hero-header.png)](https://lancedb.com)
+[![Website](https://img.shields.io/badge/-Website-100000?style=for-the-badge&labelColor=645cfb&color=645cfb)](https://lancedb.com/)
+[![Blog](https://img.shields.io/badge/Blog-100000?style=for-the-badge&labelColor=645cfb&color=645cfb)](https://blog.lancedb.com/)
+[![Discord](https://img.shields.io/badge/-Discord-100000?style=for-the-badge&logo=discord&logoColor=white&labelColor=645cfb&color=645cfb)](https://discord.gg/zMM32dvNtd)
+[![Twitter](https://img.shields.io/badge/-Twitter-100000?style=for-the-badge&logo=x&logoColor=white&labelColor=645cfb&color=645cfb)](https://twitter.com/lancedb)
+[![LinkedIn](https://img.shields.io/badge/-LinkedIn-100000?style=for-the-badge&logo=linkedin&logoColor=white&labelColor=645cfb&color=645cfb)](https://www.linkedin.com/company/lancedb/)
 
-**Search More, Manage Less**
 
-<a href='https://github.com/lancedb/vectordb-recipes/tree/main' target="_blank"><img alt='LanceDB' src='https://img.shields.io/badge/VectorDB_Recipes-100000?style=for-the-badge&logo=LanceDB&logoColor=white&labelColor=645cfb&color=645cfb'/></a>
-<a href='https://lancedb.github.io/lancedb/' target="_blank"><img alt='lancdb' src='https://img.shields.io/badge/DOCS-100000?style=for-the-badge&logo=lancdb&logoColor=white&labelColor=645cfb&color=645cfb'/></a>
-[![Blog](https://img.shields.io/badge/Blog-12100E?style=for-the-badge&logoColor=white)](https://blog.lancedb.com/)
-[![Discord](https://img.shields.io/badge/Discord-%235865F2.svg?style=for-the-badge&logo=discord&logoColor=white)](https://discord.gg/zMM32dvNtd)
-[![Twitter](https://img.shields.io/badge/Twitter-%231DA1F2.svg?style=for-the-badge&logo=Twitter&logoColor=white)](https://twitter.com/lancedb)
-[![Gurubase](https://img.shields.io/badge/Gurubase-Ask%20LanceDB%20Guru-006BFF?style=for-the-badge)](https://gurubase.io/g/lancedb)
+<img src="docs/src/assets/lancedb.png" alt="LanceDB" width="50%">
 
-</p>
+# **The Multimodal AI Lakehouse**
 
-<img max-width="750px" alt="LanceDB Multimodal Search" src="https://github.com/lancedb/lancedb/assets/917119/09c5afc5-7816-4687-bae4-f2ca194426ec">
+[**How to Install** ](#how-to-install) ✦ [**Detailed Documentation**](https://lancedb.github.io/lancedb/) ✦ [**Tutorials and Recipes**](https://github.com/lancedb/vectordb-recipes/tree/main) ✦  [**Contributors**](#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.
 
-</p>
 </div>
 
-<hr />
+<br>
 
-LanceDB is an open-source database for vector-search built with persistent storage, which greatly simplifies retrieval, filtering and management of embeddings.
+## **Demo: Multimodal Search by Keyword, Vector or with SQL**
+<img max-width="750px" alt="LanceDB Multimodal Search" src="https://github.com/lancedb/lancedb/assets/917119/09c5afc5-7816-4687-bae4-f2ca194426ec">
 
-The key features of LanceDB include:
+## **Star LanceDB to get updates!**
 
-* Production-scale vector search with no servers to manage.
+<details>
+<summary>⭐ Click here ⭐  to see how fast we're growing!</summary>
+<picture>
+  <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=lancedb/lancedb&theme=dark&type=Date">
+  <img width="100%" src="https://api.star-history.com/svg?repos=lancedb/lancedb&theme=dark&type=Date">
+</picture>
+</details>
 
-* Store, query and filter vectors, metadata and multi-modal data (text, images, videos, point clouds, and more).
+## **Key Features**:
 
-* Support for vector similarity search, full-text search and SQL.
+- **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.
 
-* Native Python and Javascript/Typescript support.
+### **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.
 
-* Zero-copy, automatic versioning, manage versions of your data without needing extra infrastructure.
+### **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** 🦜️🔗](https://python.langchain.com/docs/integrations/vectorstores/lancedb/), [**LlamaIndex** 🦙](https://gpt-index.readthedocs.io/en/latest/examples/vector_stores/LanceDBIndexDemo.html), Apache-Arrow, Pandas, Polars, DuckDB and more on the way.
 
-* GPU support in building vector index(*).
+## **How to Install**:
 
-* Ecosystem integrations with [LangChain 🦜️🔗](https://python.langchain.com/docs/integrations/vectorstores/lancedb/), [LlamaIndex 🦙](https://gpt-index.readthedocs.io/en/latest/examples/vector_stores/LanceDBIndexDemo.html), Apache-Arrow, Pandas, Polars, DuckDB and more on the way.
+Follow the [Quickstart](https://lancedb.github.io/lancedb/basic/) doc to set up LanceDB locally. 
 
-LanceDB's core is written in Rust 🦀 and is built using <a href="https://github.com/lancedb/lance">Lance</a>, an open-source columnar format designed for performant ML workloads.
+**API & SDK:** We also support Python, Typescript and Rust SDKs
 
-## Quick Start
+| 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/introduction |
 
-**Javascript**
-```shell
-npm install @lancedb/lancedb
-```
+## **Join Us and Contribute**
 
-```javascript
-import * as lancedb from "@lancedb/lancedb";
+We welcome contributions from everyone! Whether you're a developer, researcher, or just someone who wants to help out. 
 
-const db = await lancedb.connect("data/sample-lancedb");
-const table = await db.createTable("vectors", [
-	{ id: 1, vector: [0.1, 0.2], item: "foo", price: 10 },
-	{ id: 2, vector: [1.1, 1.2], item: "bar", price: 50 },
-], {mode: 'overwrite'});
+If you have any suggestions or feature requests, please feel free to open an issue on GitHub or discuss it on our [**Discord**](https://discord.gg/G5DcmnZWKB) server.
+
+[**Check out the GitHub Issues**](https://github.com/lancedb/lancedb/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**
+
+<a href="https://github.com/lancedb/lancedb/graphs/contributors">
+  <img src="https://contrib.rocks/image?repo=lancedb/lancedb" />
+</a>
 
 
-const query = table.vectorSearch([0.1, 0.3]).limit(2);
-const results = await query.toArray();
+## **Stay in Touch With Us**
+<div align="center">
 
-// You can also search for rows by specific criteria without involving a vector search.
-const rowsByCriteria = await table.query().where("price >= 10").toArray();
-```
+</br>
 
-**Python**
-```shell
-pip install lancedb
-```
+[![Website](https://img.shields.io/badge/-Website-100000?style=for-the-badge&labelColor=645cfb&color=645cfb)](https://lancedb.com/)
+[![Blog](https://img.shields.io/badge/Blog-100000?style=for-the-badge&labelColor=645cfb&color=645cfb)](https://blog.lancedb.com/)
+[![Discord](https://img.shields.io/badge/-Discord-100000?style=for-the-badge&logo=discord&logoColor=white&labelColor=645cfb&color=645cfb)](https://discord.gg/zMM32dvNtd)
+[![Twitter](https://img.shields.io/badge/-Twitter-100000?style=for-the-badge&logo=x&logoColor=white&labelColor=645cfb&color=645cfb)](https://twitter.com/lancedb)
+[![LinkedIn](https://img.shields.io/badge/-LinkedIn-100000?style=for-the-badge&logo=linkedin&logoColor=white&labelColor=645cfb&color=645cfb)](https://www.linkedin.com/company/lancedb/)
 
-```python
-import lancedb
-
-uri = "data/sample-lancedb"
-db = lancedb.connect(uri)
-table = db.create_table("my_table",
-                         data=[{"vector": [3.1, 4.1], "item": "foo", "price": 10.0},
-                               {"vector": [5.9, 26.5], "item": "bar", "price": 20.0}])
-result = table.search([100, 100]).limit(2).to_pandas()
-```
-
-## Blogs, Tutorials & Videos
-* 📈 <a href="https://blog.lancedb.com/benchmarking-random-access-in-lance/">2000x better performance with Lance over Parquet</a>
-* 🤖 <a href="https://github.com/lancedb/vectordb-recipes/tree/main/examples/Youtube-Search-QA-Bot">Build a question and answer bot with LanceDB</a>
+</div>

ci/set_lance_version.py

@@ -0,0 +1,174 @@
+import argparse
+import sys
+import json
+
+
+def run_command(command: str) -> str:
+    """
+    Run a shell command and return stdout as a string.
+    If exit code is not 0, raise an exception with the stderr output.
+    """
+    import subprocess
+
+    result = subprocess.run(command, shell=True, capture_output=True, text=True)
+    if result.returncode != 0:
+        raise Exception(f"Command failed with error: {result.stderr.strip()}")
+    return result.stdout.strip()
+
+
+def get_latest_stable_version() -> str:
+    version_line = run_command("cargo info lance | grep '^version:'")
+    version = version_line.split(" ")[1].strip()
+    return version
+
+
+def get_latest_preview_version() -> str:
+    lance_tags = run_command(
+        "git ls-remote --tags https://github.com/lancedb/lance.git | grep 'refs/tags/v[0-9beta.-]\\+$'"
+    ).splitlines()
+    lance_tags = (
+        tag.split("refs/tags/")[1]
+        for tag in lance_tags
+        if "refs/tags/" in tag and "beta" in tag
+    )
+    from packaging.version import Version
+
+    latest = max(
+        (tag[1:] for tag in lance_tags if tag.startswith("v")), key=lambda t: Version(t)
+    )
+    return str(latest)
+
+
+def extract_features(line: str) -> list:
+    """
+    Extracts the features from a line in Cargo.toml.
+    Example: 'lance = { "version" = "=0.29.0", "features" = ["dynamodb"] }'
+    Returns: ['dynamodb']
+    """
+    import re
+
+    match = re.search(r'"features"\s*=\s*\[(.*?)\]', line)
+    if match:
+        features_str = match.group(1)
+        return [f.strip('"') for f in features_str.split(",")]
+    return []
+
+
+def update_cargo_toml(line_updater):
+    """
+    Updates the Cargo.toml file by applying the line_updater function to each line.
+    The line_updater function should take a line as input and return the updated line.
+    """
+    with open("Cargo.toml", "r") as f:
+        lines = f.readlines()
+
+    new_lines = []
+    for line in lines:
+        if line.startswith("lance"):
+            # Update the line using the provided function
+            new_lines.append(line_updater(line))
+        else:
+            # Keep the line unchanged
+            new_lines.append(line)
+
+    with open("Cargo.toml", "w") as f:
+        f.writelines(new_lines)
+
+
+def set_stable_version(version: str):
+    """
+    Sets lines to
+    lance = { "version" = "=0.29.0", "features" = ["dynamodb"] }
+    lance-io = "=0.29.0"
+    ...
+    """
+
+    def line_updater(line: str) -> str:
+        package_name = line.split("=", maxsplit=1)[0].strip()
+        features = extract_features(line)
+        if features:
+            return f'{package_name} = {{ "version" = "={version}", "features" = {json.dumps(features)} }}\n'
+        else:
+            return f'{package_name} = "={version}"\n'
+
+    update_cargo_toml(line_updater)
+
+
+def set_preview_version(version: str):
+    """
+    Sets lines to
+    lance = { "version" = "=0.29.0", "features" = ["dynamodb"], tag = "v0.29.0-beta.2", git="https://github.com/lancedb/lance.git" }
+    lance-io = { version = "=0.29.0", tag = "v0.29.0-beta.2", git="https://github.com/lancedb/lance.git" }
+    ...
+    """
+
+    def line_updater(line: str) -> str:
+        package_name = line.split("=", maxsplit=1)[0].strip()
+        features = extract_features(line)
+        base_version = version.split("-")[0]  # Get the base version without beta suffix
+        if features:
+            return f'{package_name} = {{ "version" = "={base_version}", "features" = {json.dumps(features)}, "tag" = "v{version}", "git" = "https://github.com/lancedb/lance.git" }}\n'
+        else:
+            return f'{package_name} = {{ "version" = "={base_version}", "tag" = "v{version}", "git" = "https://github.com/lancedb/lance.git" }}\n'
+
+    update_cargo_toml(line_updater)
+
+
+def set_local_version():
+    """
+    Sets lines to
+    lance = { path = "../lance/rust/lance", features = ["dynamodb"] }
+    lance-io = { path = "../lance/rust/lance-io" }
+    ...
+    """
+
+    def line_updater(line: str) -> str:
+        package_name = line.split("=", maxsplit=1)[0].strip()
+        features = extract_features(line)
+        if features:
+            return f'{package_name} = {{ "path" = "../lance/rust/{package_name}", "features" = {json.dumps(features)} }}\n'
+        else:
+            return f'{package_name} = {{ "path" = "../lance/rust/{package_name}" }}\n'
+
+    update_cargo_toml(line_updater)
+
+
+parser = argparse.ArgumentParser(description="Set the version of the Lance package.")
+parser.add_argument(
+    "version",
+    type=str,
+    help="The version to set for the Lance package. Use 'stable' for the latest stable version, 'preview' for latest preview version, or a specific version number (e.g., '0.1.0'). You can also specify 'local' to use a local path.",
+)
+args = parser.parse_args()
+
+if args.version == "stable":
+    latest_stable_version = get_latest_stable_version()
+    print(
+        f"Found latest stable version: \033[1mv{latest_stable_version}\033[0m",
+        file=sys.stderr,
+    )
+    set_stable_version(latest_stable_version)
+elif args.version == "preview":
+    latest_preview_version = get_latest_preview_version()
+    print(
+        f"Found latest preview version: \033[1mv{latest_preview_version}\033[0m",
+        file=sys.stderr,
+    )
+    set_preview_version(latest_preview_version)
+elif args.version == "local":
+    set_local_version()
+else:
+    # Parse the version number.
+    version = args.version
+    # Ignore initial v if present.
+    if version.startswith("v"):
+        version = version[1:]
+
+    if "beta" in version:
+        set_preview_version(version)
+    else:
+        set_stable_version(version)
+
+print("Updating lockfiles...", file=sys.stderr, end="")
+run_command("cargo metadata > /dev/null")
+print(" done.", file=sys.stderr)

ci/update_lockfiles.sh

@@ -0,0 +1,30 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+AMEND=false
+
+for arg in "$@"; do
+  if [[ "$arg" == "--amend" ]]; then
+    AMEND=true
+  fi
+done
+
+# This updates the lockfile without building
+cargo metadata --quiet > /dev/null
+
+pushd nodejs || exit 1
+npm install --package-lock-only --silent
+popd
+pushd node || exit 1
+npm install --package-lock-only --silent
+popd
+
+if git diff --quiet --exit-code; then
+  echo "No lockfile changes to commit; skipping amend."
+elif $AMEND; then
+  git add Cargo.lock nodejs/package-lock.json node/package-lock.json
+  git commit --amend --no-edit
+else
+  git add Cargo.lock nodejs/package-lock.json node/package-lock.json
+  git commit -m "Update lockfiles"
+fi

docs/mkdocs.yml

@@ -105,8 +105,7 @@ markdown_extensions:
 nav:
   - Home:
       - LanceDB: index.md
-      - 👉 Quickstart: quickstart.md
-      - 🏃🏼‍♂️ Basic Usage: basic.md
+      - 🏃🏼‍♂️ Quick start: basic.md
       - 📚 Concepts:
           - Vector search: concepts/vector_search.md
           - Indexing:
@@ -194,6 +193,7 @@ nav:
           - Pandas and PyArrow: python/pandas_and_pyarrow.md
           - Polars: python/polars_arrow.md
           - DuckDB: python/duckdb.md
+          - Datafusion: python/datafusion.md
           - LangChain:
               - LangChain 🔗: integrations/langchain.md
               - LangChain demo: notebooks/langchain_demo.ipynb
@@ -206,6 +206,7 @@ nav:
           - PromptTools: integrations/prompttools.md
           - dlt: integrations/dlt.md
           - phidata: integrations/phidata.md
+          - Genkit: integrations/genkit.md
       - 🎯 Examples:
           - Overview: examples/index.md
           - 🐍 Python:
@@ -238,9 +239,7 @@ nav:
           - 👾 JavaScript (lancedb): js/globals.md
           - 🦀 Rust: https://docs.rs/lancedb/latest/lancedb/
 
-  - Getting Started:
-      - Quickstart: quickstart.md
-      - Basic Usage: basic.md
+  - Quick start: basic.md
   - Concepts:
       - Vector search: concepts/vector_search.md
       - Indexing:
@@ -250,6 +249,7 @@ nav:
       - Data management: concepts/data_management.md
   - Guides:
       - Working with tables: guides/tables.md
+      - Working with SQL: guides/sql_querying.md
       - Building an ANN index: ann_indexes.md
       - Vector Search: search.md
       - Full-text search (native): fts.md
@@ -326,6 +326,7 @@ nav:
       - Pandas and PyArrow: python/pandas_and_pyarrow.md
       - Polars: python/polars_arrow.md
       - DuckDB: python/duckdb.md
+      - Datafusion: python/datafusion.md
       - LangChain 🦜️🔗↗: integrations/langchain.md
       - LangChain.js 🦜️🔗↗: https://js.langchain.com/docs/integrations/vectorstores/lancedb
       - LlamaIndex 🦙↗: integrations/llamaIndex.md
@@ -334,6 +335,7 @@ nav:
       - PromptTools: integrations/prompttools.md
       - dlt: integrations/dlt.md
       - phidata: integrations/phidata.md
+      - Genkit: integrations/genkit.md
   - Examples:
       - examples/index.md
       - 🐍 Python:

docs/overrides/partials/main.html

@@ -0,0 +1,5 @@
+{% extends "base.html" %}
+
+{% block announce %}
+  📚 Starting June 1st, 2025, please use <a href="https://lancedb.github.io/documentation" target="_blank" rel="noopener noreferrer">lancedb.github.io/documentation</a> for the latest docs.
+{% endblock %}

docs/src/ann_indexes.md

@@ -291,7 +291,7 @@ Product quantization can lead to approximately `16 * sizeof(float32) / 1 = 64` t
 
 `num_partitions` is used to decide how many partitions the first level `IVF` index uses.
 Higher number of partitions could lead to more efficient I/O during queries and better accuracy, but it takes much more time to train.
-On `SIFT-1M` dataset, our benchmark shows that keeping each partition 1K-4K rows lead to a good latency / recall.
+On `SIFT-1M` dataset, our benchmark shows that keeping each partition 4K-8K rows lead to a good latency / recall.
 
 `num_sub_vectors` specifies how many Product Quantization (PQ) short codes to generate on each vector. The number should be a factor of the vector dimension. Because
 PQ is a lossy compression of the original vector, a higher `num_sub_vectors` usually results in

docs/src/basic.md

@@ -1,4 +1,4 @@
-# Basic Usage
+# Quick start
 
 !!! info "LanceDB can be run in a number of ways:"
 

docs/src/guides/sql_querying.md

@@ -0,0 +1,68 @@
+You can use DuckDB and Apache Datafusion to query your LanceDB tables using SQL.
+This guide will show how to query Lance tables them using both.
+
+We will re-use the dataset [created previously](./pandas_and_pyarrow.md):
+
+```python
+import lancedb
+
+db = lancedb.connect("data/sample-lancedb")
+data = [
+    {"vector": [3.1, 4.1], "item": "foo", "price": 10.0},
+    {"vector": [5.9, 26.5], "item": "bar", "price": 20.0}
+]
+table = db.create_table("pd_table", data=data)
+```
+
+## Querying a LanceDB Table with DuckDb
+
+The `to_lance` method converts the LanceDB table to a `LanceDataset`, which is accessible to DuckDB through the Arrow compatibility layer.
+To query the resulting Lance dataset in DuckDB, all you need to do is reference the dataset by the same name in your SQL query.
+
+```python
+import duckdb
+
+arrow_table = table.to_lance()
+
+duckdb.query("SELECT * FROM arrow_table")
+```
+
+```
+┌─────────────┬─────────┬────────┐
+│   vector    │  item   │ price  │
+│   float[]   │ varchar │ double │
+├─────────────┼─────────┼────────┤
+│ [3.1, 4.1]  │ foo     │   10.0 │
+│ [5.9, 26.5] │ bar     │   20.0 │
+└─────────────┴─────────┴────────┘
+```
+
+## Querying a LanceDB Table with Apache Datafusion
+
+Have the required imports before doing any querying.
+
+=== "Python"
+
+    ```python
+    --8<-- "python/python/tests/docs/test_guide_tables.py:import-lancedb"
+    --8<-- "python/python/tests/docs/test_guide_tables.py:import-session-context"
+    --8<-- "python/python/tests/docs/test_guide_tables.py:import-ffi-dataset"
+    ```
+
+Register the table created with the Datafusion session context.
+
+=== "Python"
+
+    ```python
+    --8<-- "python/python/tests/docs/test_guide_tables.py:lance_sql_basic"
+    ```
+
+```
+┌─────────────┬─────────┬────────┐
+│   vector    │  item   │ price  │
+│   float[]   │ varchar │ double │
+├─────────────┼─────────┼────────┤
+│ [3.1, 4.1]  │ foo     │   10.0 │
+│ [5.9, 26.5] │ bar     │   20.0 │
+└─────────────┴─────────┴────────┘
+```

docs/src/guides/tables.md

@@ -765,7 +765,7 @@ This can be used to update zero to all rows depending on how many rows match the
         ];
         const tbl = await db.createTable("my_table", data)
 
-        await tbl.update({ 
+        await tbl.update({
             values: { vector: [10, 10] },
             where: "x = 2"
         });
@@ -787,9 +787,9 @@ This can be used to update zero to all rows depending on how many rows match the
         ];
         const tbl = await db.createTable("my_table", data)
 
-        await tbl.update({ 
-            where: "x = 2", 
-            values: { vector: [10, 10] } 
+        await tbl.update({
+            where: "x = 2",
+            values: { vector: [10, 10] }
         });
         ```
 

docs/src/guides/tables/merge_insert.md

@@ -71,6 +71,45 @@ with merge insert, enable both `when_matched_update_all()` and
     If a column is nullable, it can be omitted from input data and it will be
     considered `null`. Columns can also be provided in any order.
 
+### Conditional Updates
+
+You can add a `where` clause to `when_matched_update_all()` to only update rows 
+that meet certain conditions. When using the `where` parameter, you must prefix 
+column names with either `source.` (for the new data) or `target.` (for the 
+existing data) to specify which table you're referencing.
+
+=== "Python"
+
+    ```python
+    # Only update rows where the target's status is 'active'
+    table.merge_insert("id")
+        .when_matched_update_all(where="target.status = 'active'")
+        .when_not_matched_insert_all()
+        .execute(new_data)
+
+    # Only update if the new price is higher than the existing price
+    table.merge_insert("product_id")
+        .when_matched_update_all(where="source.price > target.price")
+        .when_not_matched_insert_all()
+        .execute(new_data)
+    ```
+
+=== "Typescript"
+
+    ```typescript
+    // Only update rows where the target's status is 'active'
+    await table.mergeInsert("id")
+        .whenMatchedUpdateAll({ where: "target.status = 'active'" })
+        .whenNotMatchedInsertAll()
+        .execute(newData);
+
+    // Only update if the new price is higher than the existing price
+    await table.mergeInsert("product_id")
+        .whenMatchedUpdateAll({ where: "source.price > target.price" })
+        .whenNotMatchedInsertAll()
+        .execute(newData);
+    ```
+
 ## Insert-if-not-exists
 
 To avoid inserting duplicate rows, you can use the insert-if-not-exists command.

docs/src/integrations/genkit.md

@@ -0,0 +1,183 @@
+### genkitx-lancedb
+This is a lancedb plugin for genkit framework. It allows you to use LanceDB for ingesting and rereiving data using genkit framework.
+
+![integration-banner-genkit](https://github.com/user-attachments/assets/a6cc28af-98e9-4425-b87c-7ab139bd7893)
+
+### Installation
+```bash
+pnpm install genkitx-lancedb
+```
+
+### Usage
+
+Adding LanceDB plugin to your genkit instance.
+
+```ts
+import { lancedbIndexerRef, lancedb, lancedbRetrieverRef, WriteMode } from 'genkitx-lancedb';
+import { textEmbedding004, vertexAI } from '@genkit-ai/vertexai';
+import { gemini } from '@genkit-ai/vertexai';
+import { z, genkit } from 'genkit';
+import { Document } from 'genkit/retriever';
+import { chunk } from 'llm-chunk';
+import { readFile } from 'fs/promises';
+import path from 'path';
+import pdf from 'pdf-parse/lib/pdf-parse';
+
+const ai = genkit({
+  plugins: [
+    // vertexAI provides the textEmbedding004 embedder
+    vertexAI(),
+
+    // the local vector store requires an embedder to translate from text to vector
+    lancedb([
+      {
+        dbUri: '.db', // optional lancedb uri, default to .db
+        tableName: 'table', // optional table name, default to table
+        embedder: textEmbedding004,
+      },
+    ]),
+  ],
+});
+```
+
+You can run this app with the following command:
+```bash
+genkit start -- tsx --watch src/index.ts
+```
+
+This'll add LanceDB as a retriever and indexer to the genkit instance. You can see it in the GUI view
+<img width="1710" alt="Screenshot 2025-05-11 at 7 21 05 PM" src="https://github.com/user-attachments/assets/e752f7f4-785b-4797-a11e-72ab06a531b7" />
+
+**Testing retrieval on a sample table**
+Let's see the raw retrieval results
+
+<img width="1710" alt="Screenshot 2025-05-11 at 7 21 05 PM" src="https://github.com/user-attachments/assets/b8d356ed-8421-4790-8fc0-d6af563b9657" />
+On running this query, you'll 5 results fetched from the lancedb table, where each result looks something like this:
+<img width="1417" alt="Screenshot 2025-05-11 at 7 21 18 PM" src="https://github.com/user-attachments/assets/77429525-36e2-4da6-a694-e58c1cf9eb83" />
+
+
+
+## Creating a custom RAG flow
+
+Now that we've seen how you can use LanceDB for in a genkit pipeline, let's refine the flow and create a RAG. A RAG flow will consist of an index and a retreiver with its outputs postprocessed an fed into an LLM for final response
+
+### Creating custom indexer flows
+You can also create custom indexer flows, utilizing more options and features provided by LanceDB.
+
+```ts
+export const menuPdfIndexer = lancedbIndexerRef({
+   // Using all defaults, for dbUri, tableName, and embedder, etc
+});
+
+const chunkingConfig = {
+  minLength: 1000,
+  maxLength: 2000,
+  splitter: 'sentence',
+  overlap: 100,
+  delimiters: '',
+} as any;
+
+
+async function extractTextFromPdf(filePath: string) {
+  const pdfFile = path.resolve(filePath);
+  const dataBuffer = await readFile(pdfFile);
+  const data = await pdf(dataBuffer);
+  return data.text;
+}
+
+export const indexMenu = ai.defineFlow(
+  {
+    name: 'indexMenu',
+    inputSchema: z.string().describe('PDF file path'),
+    outputSchema: z.void(),
+  },
+  async (filePath: string) => {
+    filePath = path.resolve(filePath);
+
+    // Read the pdf.
+    const pdfTxt = await ai.run('extract-text', () =>
+      extractTextFromPdf(filePath)
+    );
+
+    // Divide the pdf text into segments.
+    const chunks = await ai.run('chunk-it', async () =>
+      chunk(pdfTxt, chunkingConfig)
+    );
+
+    // Convert chunks of text into documents to store in the index.
+    const documents = chunks.map((text) => {
+      return Document.fromText(text, { filePath });
+    });
+
+    // Add documents to the index.
+    await ai.index({
+      indexer: menuPdfIndexer,
+      documents,
+      options: {
+        writeMode: WriteMode.Overwrite,
+      } as any
+    });
+  }
+);
+```
+
+<img width="1316" alt="Screenshot 2025-05-11 at 8 35 56 PM" src="https://github.com/user-attachments/assets/e2a20ce4-d1d0-4fa2-9a84-f2cc26e3a29f" />
+
+In your console, you can see the logs
+
+<img width="511" alt="Screenshot 2025-05-11 at 7 19 14 PM" src="https://github.com/user-attachments/assets/243f26c5-ed38-40b6-b661-002f40f0423a" />
+
+### Creating custom retriever flows
+You can also create custom retriever flows, utilizing more options and features provided by LanceDB.
+```ts
+export const menuRetriever = lancedbRetrieverRef({
+  tableName: "table", // Use the same table name as the indexer.
+  displayName: "Menu", // Use a custom display name.
+
+export const menuQAFlow = ai.defineFlow(
+  { name: "Menu", inputSchema: z.string(), outputSchema: z.string() },
+  async (input: string) => {
+    // retrieve relevant documents
+    const docs = await ai.retrieve({
+      retriever: menuRetriever,
+      query: input,
+      options: { 
+        k: 3,
+      },
+    });
+
+    const extractedContent = docs.map(doc => {
+      if (doc.content && Array.isArray(doc.content) && doc.content.length > 0) {
+        if (doc.content[0].media && doc.content[0].media.url) {
+          return doc.content[0].media.url;
+        }
+      }
+      return "No content found";
+    });
+
+    console.log("Extracted content:", extractedContent);
+
+    const { text } = await ai.generate({
+      model: gemini('gemini-2.0-flash'),
+      prompt: `
+You are acting as a helpful AI assistant that can answer 
+questions about the food available on the menu at Genkit Grub Pub.
+
+Use only the context provided to answer the question.
+If you don't know, do not make up an answer.
+Do not add or change items on the menu.
+
+Context:
+${extractedContent.join('\n\n')}
+
+Question: ${input}`,
+      docs,
+    });
+    
+    return text;
+  }
+);
+```
+Now using our retrieval flow, we can ask question about the ingsted PDF
+<img width="1306" alt="Screenshot 2025-05-11 at 7 18 45 PM" src="https://github.com/user-attachments/assets/86c66b13-7c12-4d5f-9d81-ae36bfb1c346" />
+

docs/src/js/classes/BooleanQuery.md

@@ -0,0 +1,53 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / BooleanQuery
+
+# Class: BooleanQuery
+
+Represents a full-text query interface.
+This interface defines the structure and behavior for full-text queries,
+including methods to retrieve the query type and convert the query to a dictionary format.
+
+## Implements
+
+- [`FullTextQuery`](../interfaces/FullTextQuery.md)
+
+## Constructors
+
+### new BooleanQuery()
+
+```ts
+new BooleanQuery(queries): BooleanQuery
+```
+
+Creates an instance of BooleanQuery.
+
+#### Parameters
+
+* **queries**: [[`Occur`](../enumerations/Occur.md), [`FullTextQuery`](../interfaces/FullTextQuery.md)][]
+    An array of (Occur, FullTextQuery objects) to combine.
+    Occur specifies whether the query must match, or should match.
+
+#### Returns
+
+[`BooleanQuery`](BooleanQuery.md)
+
+## Methods
+
+### queryType()
+
+```ts
+queryType(): FullTextQueryType
+```
+
+The type of the full-text query.
+
+#### Returns
+
+[`FullTextQueryType`](../enumerations/FullTextQueryType.md)
+
+#### Implementation of
+
+[`FullTextQuery`](../interfaces/FullTextQuery.md).[`queryType`](../interfaces/FullTextQuery.md#querytype)

docs/src/js/classes/MatchQuery.md

@@ -40,6 +40,7 @@ Creates an instance of MatchQuery.
     - `boost`: The boost factor for the query (default is 1.0).
     - `fuzziness`: The fuzziness level for the query (default is 0).
     - `maxExpansions`: The maximum number of terms to consider for fuzzy matching (default is 50).
+    - `operator`: The logical operator to use for combining terms in the query (default is "OR").
 
 * **options.boost?**: `number`
 
@@ -47,6 +48,8 @@ Creates an instance of MatchQuery.
 
 * **options.maxExpansions?**: `number`
 
+* **options.operator?**: [`Operator`](../enumerations/Operator.md)
+
 #### Returns
 
 [`MatchQuery`](MatchQuery.md)

docs/src/js/classes/MergeInsertBuilder.md

@@ -33,7 +33,7 @@ Construct a MergeInsertBuilder. __Internal use only.__
 ### execute()
 
 ```ts
-execute(data): Promise<MergeStats>
+execute(data, execOptions?): Promise<MergeResult>
 ```
 
 Executes the merge insert operation
@@ -42,11 +42,13 @@ Executes the merge insert operation
 
 * **data**: [`Data`](../type-aliases/Data.md)
 
+* **execOptions?**: `Partial`&lt;[`WriteExecutionOptions`](../interfaces/WriteExecutionOptions.md)&gt;
+
 #### Returns
 
-`Promise`&lt;[`MergeStats`](../interfaces/MergeStats.md)&gt;
+`Promise`&lt;[`MergeResult`](../interfaces/MergeResult.md)&gt;
 
-Statistics about the merge operation: counts of inserted, updated, and deleted rows
+the merge result
 
 ***
 

docs/src/js/classes/MultiMatchQuery.md

@@ -38,9 +38,12 @@ Creates an instance of MultiMatchQuery.
 * **options?**
     Optional parameters for the multi-match query.
     - `boosts`: An array of boost factors for each column (default is 1.0 for all).
+    - `operator`: The logical operator to use for combining terms in the query (default is "OR").
 
 * **options.boosts?**: `number`[]
 
+* **options.operator?**: [`Operator`](../enumerations/Operator.md)
+
 #### Returns
 
 [`MultiMatchQuery`](MultiMatchQuery.md)

docs/src/js/classes/PhraseQuery.md

@@ -19,7 +19,10 @@ including methods to retrieve the query type and convert the query to a dictiona
 ### new PhraseQuery()
 
 ```ts
-new PhraseQuery(query, column): PhraseQuery
+new PhraseQuery(
+   query,
+   column,
+   options?): PhraseQuery
 ```
 
 Creates an instance of `PhraseQuery`.
@@ -32,6 +35,12 @@ Creates an instance of `PhraseQuery`.
 * **column**: `string`
     The name of the column to search within.
 
+* **options?**
+    Optional parameters for the phrase query.
+    - `slop`: The maximum number of intervening unmatched positions allowed between words in the phrase (default is 0).
+
+* **options.slop?**: `number`
+
 #### Returns
 
 [`PhraseQuery`](PhraseQuery.md)

docs/src/js/classes/Table.md

@@ -40,7 +40,7 @@ Returns the name of the table
 ### add()
 
 ```ts
-abstract add(data, options?): Promise<void>
+abstract add(data, options?): Promise<AddResult>
 ```
 
 Insert records into this Table.
@@ -54,14 +54,17 @@ Insert records into this Table.
 
 #### Returns
 
-`Promise`&lt;`void`&gt;
+`Promise`&lt;[`AddResult`](../interfaces/AddResult.md)&gt;
+
+A promise that resolves to an object
+containing the new version number of the table
 
 ***
 
 ### addColumns()
 
 ```ts
-abstract addColumns(newColumnTransforms): Promise<void>
+abstract addColumns(newColumnTransforms): Promise<AddColumnsResult>
 ```
 
 Add new columns with defined values.
@@ -76,14 +79,17 @@ Add new columns with defined values.
 
 #### Returns
 
-`Promise`&lt;`void`&gt;
+`Promise`&lt;[`AddColumnsResult`](../interfaces/AddColumnsResult.md)&gt;
+
+A promise that resolves to an object
+containing the new version number of the table after adding the columns.
 
 ***
 
 ### alterColumns()
 
 ```ts
-abstract alterColumns(columnAlterations): Promise<void>
+abstract alterColumns(columnAlterations): Promise<AlterColumnsResult>
 ```
 
 Alter the name or nullability of columns.
@@ -96,7 +102,10 @@ Alter the name or nullability of columns.
 
 #### Returns
 
-`Promise`&lt;`void`&gt;
+`Promise`&lt;[`AlterColumnsResult`](../interfaces/AlterColumnsResult.md)&gt;
+
+A promise that resolves to an object
+containing the new version number of the table after altering the columns.
 
 ***
 
@@ -252,7 +261,7 @@ await table.createIndex("my_float_col");
 ### delete()
 
 ```ts
-abstract delete(predicate): Promise<void>
+abstract delete(predicate): Promise<DeleteResult>
 ```
 
 Delete the rows that satisfy the predicate.
@@ -263,7 +272,10 @@ Delete the rows that satisfy the predicate.
 
 #### Returns
 
-`Promise`&lt;`void`&gt;
+`Promise`&lt;[`DeleteResult`](../interfaces/DeleteResult.md)&gt;
+
+A promise that resolves to an object
+containing the new version number of the table
 
 ***
 
@@ -284,7 +296,7 @@ Return a brief description of the table
 ### dropColumns()
 
 ```ts
-abstract dropColumns(columnNames): Promise<void>
+abstract dropColumns(columnNames): Promise<DropColumnsResult>
 ```
 
 Drop one or more columns from the dataset
@@ -303,7 +315,10 @@ then call ``cleanup_files`` to remove the old files.
 
 #### Returns
 
-`Promise`&lt;`void`&gt;
+`Promise`&lt;[`DropColumnsResult`](../interfaces/DropColumnsResult.md)&gt;
+
+A promise that resolves to an object
+containing the new version number of the table after dropping the columns.
 
 ***
 
@@ -678,7 +693,7 @@ Return the table as an arrow table
 #### update(opts)
 
 ```ts
-abstract update(opts): Promise<void>
+abstract update(opts): Promise<UpdateResult>
 ```
 
 Update existing records in the Table
@@ -689,7 +704,10 @@ Update existing records in the Table
 
 ##### Returns
 
-`Promise`&lt;`void`&gt;
+`Promise`&lt;[`UpdateResult`](../interfaces/UpdateResult.md)&gt;
+
+A promise that resolves to an object containing
+the number of rows updated and the new version number
 
 ##### Example
 
@@ -700,7 +718,7 @@ table.update({where:"x = 2", values:{"vector": [10, 10]}})
 #### update(opts)
 
 ```ts
-abstract update(opts): Promise<void>
+abstract update(opts): Promise<UpdateResult>
 ```
 
 Update existing records in the Table
@@ -711,7 +729,10 @@ Update existing records in the Table
 
 ##### Returns
 
-`Promise`&lt;`void`&gt;
+`Promise`&lt;[`UpdateResult`](../interfaces/UpdateResult.md)&gt;
+
+A promise that resolves to an object containing
+the number of rows updated and the new version number
 
 ##### Example
 
@@ -722,7 +743,7 @@ table.update({where:"x = 2", valuesSql:{"x": "x + 1"}})
 #### update(updates, options)
 
 ```ts
-abstract update(updates, options?): Promise<void>
+abstract update(updates, options?): Promise<UpdateResult>
 ```
 
 Update existing records in the Table
@@ -745,10 +766,6 @@ repeatedly calilng this method.
 * **updates**: `Record`&lt;`string`, `string`&gt; \| `Map`&lt;`string`, `string`&gt;
     the
     columns to update
-    Keys in the map should specify the name of the column to update.
-    Values in the map provide the new value of the column.  These can
-    be SQL literal strings (e.g. "7" or "'foo'") or they can be expressions
-    based on the row being updated (e.g. "my_col + 1")
 
 * **options?**: `Partial`&lt;[`UpdateOptions`](../interfaces/UpdateOptions.md)&gt;
     additional options to control
@@ -756,7 +773,15 @@ repeatedly calilng this method.
 
 ##### Returns
 
-`Promise`&lt;`void`&gt;
+`Promise`&lt;[`UpdateResult`](../interfaces/UpdateResult.md)&gt;
+
+A promise that resolves to an object
+containing the number of rows updated and the new version number
+
+Keys in the map should specify the name of the column to update.
+Values in the map provide the new value of the column.  These can
+be SQL literal strings (e.g. "7" or "'foo'") or they can be expressions
+based on the row being updated (e.g. "my_col + 1")
 
 ***
 

docs/src/js/enumerations/FullTextQueryType.md

@@ -15,6 +15,14 @@ Enum representing the types of full-text queries supported.
 
 ## Enumeration Members
 
+### Boolean
+
+```ts
+Boolean: "boolean";
+```
+
+***
+
 ### Boost
 
 ```ts

docs/src/js/enumerations/Occur.md

@@ -0,0 +1,28 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / Occur
+
+# Enumeration: Occur
+
+Enum representing the occurrence of terms in full-text queries.
+
+- `Must`: The term must be present in the document.
+- `Should`: The term should contribute to the document score, but is not required.
+
+## Enumeration Members
+
+### Must
+
+```ts
+Must: "MUST";
+```
+
+***
+
+### Should
+
+```ts
+Should: "SHOULD";
+```

docs/src/js/enumerations/Operator.md

@@ -0,0 +1,28 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / Operator
+
+# Enumeration: Operator
+
+Enum representing the logical operators used in full-text queries.
+
+- `And`: All terms must match.
+- `Or`: At least one term must match.
+
+## Enumeration Members
+
+### And
+
+```ts
+And: "AND";
+```
+
+***
+
+### Or
+
+```ts
+Or: "OR";
+```

docs/src/js/globals.md

@@ -12,9 +12,12 @@
 ## Enumerations
 
 - [FullTextQueryType](enumerations/FullTextQueryType.md)
+- [Occur](enumerations/Occur.md)
+- [Operator](enumerations/Operator.md)
 
 ## Classes
 
+- [BooleanQuery](classes/BooleanQuery.md)
 - [BoostQuery](classes/BoostQuery.md)
 - [Connection](classes/Connection.md)
 - [Index](classes/Index.md)
@@ -34,13 +37,18 @@
 
 ## Interfaces
 
+- [AddColumnsResult](interfaces/AddColumnsResult.md)
 - [AddColumnsSql](interfaces/AddColumnsSql.md)
 - [AddDataOptions](interfaces/AddDataOptions.md)
+- [AddResult](interfaces/AddResult.md)
+- [AlterColumnsResult](interfaces/AlterColumnsResult.md)
 - [ClientConfig](interfaces/ClientConfig.md)
 - [ColumnAlteration](interfaces/ColumnAlteration.md)
 - [CompactionStats](interfaces/CompactionStats.md)
 - [ConnectionOptions](interfaces/ConnectionOptions.md)
 - [CreateTableOptions](interfaces/CreateTableOptions.md)
+- [DeleteResult](interfaces/DeleteResult.md)
+- [DropColumnsResult](interfaces/DropColumnsResult.md)
 - [ExecutableQuery](interfaces/ExecutableQuery.md)
 - [FragmentStatistics](interfaces/FragmentStatistics.md)
 - [FragmentSummaryStats](interfaces/FragmentSummaryStats.md)
@@ -54,7 +62,7 @@
 - [IndexStatistics](interfaces/IndexStatistics.md)
 - [IvfFlatOptions](interfaces/IvfFlatOptions.md)
 - [IvfPqOptions](interfaces/IvfPqOptions.md)
-- [MergeStats](interfaces/MergeStats.md)
+- [MergeResult](interfaces/MergeResult.md)
 - [OpenTableOptions](interfaces/OpenTableOptions.md)
 - [OptimizeOptions](interfaces/OptimizeOptions.md)
 - [OptimizeStats](interfaces/OptimizeStats.md)
@@ -65,7 +73,9 @@
 - [TableStatistics](interfaces/TableStatistics.md)
 - [TimeoutConfig](interfaces/TimeoutConfig.md)
 - [UpdateOptions](interfaces/UpdateOptions.md)
+- [UpdateResult](interfaces/UpdateResult.md)
 - [Version](interfaces/Version.md)
+- [WriteExecutionOptions](interfaces/WriteExecutionOptions.md)
 
 ## Type Aliases
 

docs/src/js/interfaces/AddColumnsResult.md

@@ -0,0 +1,15 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / AddColumnsResult
+
+# Interface: AddColumnsResult
+
+## Properties
+
+### version
+
+```ts
+version: number;
+```

docs/src/js/interfaces/AddResult.md

@@ -0,0 +1,15 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / AddResult
+
+# Interface: AddResult
+
+## Properties
+
+### version
+
+```ts
+version: number;
+```

docs/src/js/interfaces/AlterColumnsResult.md

@@ -0,0 +1,15 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / AlterColumnsResult
+
+# Interface: AlterColumnsResult
+
+## Properties
+
+### version
+
+```ts
+version: number;
+```

docs/src/js/interfaces/DeleteResult.md

@@ -0,0 +1,15 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / DeleteResult
+
+# Interface: DeleteResult
+
+## Properties
+
+### version
+
+```ts
+version: number;
+```

docs/src/js/interfaces/DropColumnsResult.md

@@ -0,0 +1,15 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / DropColumnsResult
+
+# Interface: DropColumnsResult
+
+## Properties
+
+### version
+
+```ts
+version: number;
+```

docs/src/js/interfaces/MergeResult.md

@@ -0,0 +1,39 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / MergeResult
+
+# Interface: MergeResult
+
+## Properties
+
+### numDeletedRows
+
+```ts
+numDeletedRows: number;
+```
+
+***
+
+### numInsertedRows
+
+```ts
+numInsertedRows: number;
+```
+
+***
+
+### numUpdatedRows
+
+```ts
+numUpdatedRows: number;
+```
+
+***
+
+### version
+
+```ts
+version: number;
+```

docs/src/js/interfaces/MergeStats.md

@@ -1,31 +0,0 @@
-[**@lancedb/lancedb**](../README.md) • **Docs**
-
-***
-
-[@lancedb/lancedb](../globals.md) / MergeStats
-
-# Interface: MergeStats
-
-## Properties
-
-### numDeletedRows
-
-```ts
-numDeletedRows: bigint;
-```
-
-***
-
-### numInsertedRows
-
-```ts
-numInsertedRows: bigint;
-```
-
-***
-
-### numUpdatedRows
-
-```ts
-numUpdatedRows: bigint;
-```

docs/src/js/interfaces/UpdateResult.md

@@ -0,0 +1,23 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / UpdateResult
+
+# Interface: UpdateResult
+
+## Properties
+
+### rowsUpdated
+
+```ts
+rowsUpdated: number;
+```
+
+***
+
+### version
+
+```ts
+version: number;
+```

docs/src/js/interfaces/WriteExecutionOptions.md

@@ -0,0 +1,26 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / WriteExecutionOptions
+
+# Interface: WriteExecutionOptions
+
+## Properties
+
+### timeoutMs?
+
+```ts
+optional timeoutMs: number;
+```
+
+Maximum time to run the operation before cancelling it.
+
+By default, there is a 30-second timeout that is only enforced after the
+first attempt. This is to prevent spending too long retrying to resolve
+conflicts. For example, if a write attempt takes 20 seconds and fails,
+the second attempt will be cancelled after 10 seconds, hitting the
+30-second timeout. However, a write that takes one hour and succeeds on the
+first attempt will not be cancelled.
+
+When this is set, the timeout is enforced on all attempts, including the first.

docs/src/notebooks/Multivector_on_LanceDB.ipynb

@@ -428,7 +428,7 @@
         "\n",
         "**Why?**  \n",
         "Embedding the UFO dataset and ingesting it into LanceDB takes **~2 hours on a T4 GPU**. To save time:  \n",
-        "- **Use the pre-prepared table with index created ** (provided below) to proceed directly to step7: search.  \n",
+        "- **Use the pre-prepared table with index created** (provided below) to proceed directly to **Step 7**: search.  \n",
         "- **Step 5a** contains the full ingestion code for reference (run it only if necessary).  \n",
         "- **Step 6** contains the details on creating the index on the multivector column"
       ]

docs/src/python/datafusion.md

@@ -0,0 +1,53 @@
+# Apache Datafusion
+
+In Python, LanceDB tables can also be queried with [Apache Datafusion](https://datafusion.apache.org/), an extensible query engine written in Rust that uses Apache Arrow as its in-memory format. This means you can write complex SQL queries to analyze your data in LanceDB.
+
+This integration is done via [Datafusion FFI](https://docs.rs/datafusion-ffi/latest/datafusion_ffi/), which provides a native integration between LanceDB and Datafusion.
+The Datafusion FFI allows to pass down column selections and basic filters to LanceDB, reducing the amount of scanned data when executing your query. Additionally, the integration allows streaming data from LanceDB tables which allows to do aggregation larger-than-memory.
+
+We can demonstrate this by first installing `datafusion` and `lancedb`.
+
+```shell
+pip install datafusion lancedb
+```
+
+We will re-use the dataset [created previously](./pandas_and_pyarrow.md):
+
+```python
+import lancedb
+
+from datafusion import SessionContext
+from lance import FFILanceTableProvider
+
+db = lancedb.connect("data/sample-lancedb")
+data = [
+    {"vector": [3.1, 4.1], "item": "foo", "price": 10.0},
+    {"vector": [5.9, 26.5], "item": "bar", "price": 20.0}
+]
+lance_table = db.create_table("lance_table", data)
+
+ctx = SessionContext()
+
+ffi_lance_table = FFILanceTableProvider(
+    lance_table.to_lance(), with_row_id=True, with_row_addr=True
+)
+ctx.register_table_provider("ffi_lance_table", ffi_lance_table)
+```
+
+The `to_lance` method converts the LanceDB table to a `LanceDataset`, which is accessible to Datafusion through the Datafusion FFI integration layer.
+To query the resulting Lance dataset in Datafusion, you first need to register the dataset with Datafusion and then just reference it by the same name in your SQL query.
+
+```python
+ctx.table("ffi_lance_table")
+ctx.sql("SELECT * FROM ffi_lance_table")
+```
+
+```
+┌─────────────┬─────────┬────────┬─────────────────┬─────────────────┐
+│   vector    │  item   │ price  │ _rowid          │ _rowaddr        │
+│   float[]   │ varchar │ double │ bigint unsigned │ bigint unsigned │
+├─────────────┼─────────┼────────┼─────────────────┼─────────────────┤
+│ [3.1, 4.1]  │ foo     │   10.0 │               0 │               0 │
+│ [5.9, 26.5] │ bar     │   20.0 │               1 │               1 │
+└─────────────┴─────────┴────────┴─────────────────┴─────────────────┘
+```

docs/src/quickstart.md

@@ -1,101 +0,0 @@
-
-# Getting Started with LanceDB: A Minimal Vector Search Tutorial
-
-Let's set up a LanceDB database, insert vector data, and perform a simple vector search. We'll use simple character classes like "knight" and "rogue" to illustrate semantic relevance.
-
-## 1. Install Dependencies
-
-Before starting, make sure you have the necessary packages:
-
-```bash
-pip install lancedb pandas numpy
-```
-
-## 2. Import Required Libraries
-
-```python
-import lancedb
-import pandas as pd
-import numpy as np
-```
-
-## 3. Connect to LanceDB
-
-You can use a local directory to store your database:
-
-```python
-db = lancedb.connect("./lancedb")
-```
-
-## 4. Create Sample Data
-
-Add sample text data and corresponding 4D vectors:
-
-```python
-data = pd.DataFrame([
-    {"id": "1", "vector": [1.0, 0.0, 0.0, 0.0], "text": "knight"},
-    {"id": "2", "vector": [0.9, 0.1, 0.0, 0.0], "text": "warrior"},
-    {"id": "3", "vector": [0.0, 1.0, 0.0, 0.0], "text": "rogue"},
-    {"id": "4", "vector": [0.0, 0.9, 0.1, 0.0], "text": "thief"},
-    {"id": "5", "vector": [0.5, 0.5, 0.0, 0.0], "text": "ranger"},
-])
-```
-
-## 5. Create a Table in LanceDB
-
-```python
-table = db.create_table("rpg_classes", data=data, mode="overwrite")
-```
-
-Let's see how the table looks:
-```python
-print(data)
-```
-
-| id | vector | text |
-|----|--------|------|
-| 1 | [1.0, 0.0, 0.0, 0.0] | knight |
-| 2 | [0.9, 0.1, 0.0, 0.0] | warrior |
-| 3 | [0.0, 1.0, 0.0, 0.0] | rogue |
-| 4 | [0.0, 0.9, 0.1, 0.0] | thief |
-| 5 | [0.5, 0.5, 0.0, 0.0] | ranger |
-
-
-
-## 6. Perform a Vector Search
-
-Search for the most similar character classes to our query vector:
-
-```python
-# Query as if we are searching for "rogue"
-results = table.search([0.95, 0.05, 0.0, 0.0]).limit(3).to_df()
-print(results)
-```
-
-This will return the top 3 closest classes to the vector, effectively showing how LanceDB can be used for semantic search.
-
-|  id  |          vector         |   text   | _distance |
-|------|------------------------|----------|-----------|
-|  3   | [0.0, 1.0, 0.0, 0.0]  |  rogue   |   0.00   |
-|  4   | [0.0, 0.9, 0.1, 0.0]  | thief |   0.02   |
-|  5   | [0.5, 0.5, 0.0, 0.0]  |  ranger  |   0.50   |
-
-Let's try searching for "knight"
-
-```python
-query_vector = [1.0, 0.0, 0.0, 0.0]
-results = table.search(query_vector).limit(3).to_pandas()
-print(results)
-```
-
-|  id  |          vector         |   text   | _distance |
-|------|------------------------|----------|-----------|
-|  1   | [1.0, 0.0, 0.0, 0.0]  |  knight  |   0.00   |
-|  2   | [0.9, 0.1, 0.0, 0.0]  | warrior  |   0.02   |
-|  5   | [0.5, 0.5, 0.0, 0.0]  |  ranger  |   0.50   |
-
-## Next Steps
-
-That's it - you just conducted vector search!
-
-For more beginner tips, check out the [Basic Usage](basic.md) guide.

docs/test/requirements.txt

@@ -7,3 +7,4 @@ tantivy==0.20.1
 --extra-index-url https://download.pytorch.org/whl/cpu
 torch
 polars>=0.19, <=1.3.0
+datafusion

java/core/pom.xml

@@ -8,7 +8,7 @@
     <parent>
         <groupId>com.lancedb</groupId>
         <artifactId>lancedb-parent</artifactId>
-        <version>0.19.1-beta.1</version>
+        <version>0.21.1-beta.1</version>
         <relativePath>../pom.xml</relativePath>
     </parent>
 

java/pom.xml

@@ -6,7 +6,7 @@
 
     <groupId>com.lancedb</groupId>
     <artifactId>lancedb-parent</artifactId>
-    <version>0.19.1-beta.1</version>
+    <version>0.21.1-beta.1</version>
     <packaging>pom</packaging>
 
     <name>LanceDB Parent</name>

node/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "vectordb",
-  "version": "0.19.1-beta.1",
+  "version": "0.21.1-beta.1",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "vectordb",
-      "version": "0.19.1-beta.1",
+      "version": "0.21.1-beta.1",
       "cpu": [
         "x64",
         "arm64"
@@ -52,11 +52,11 @@
         "uuid": "^9.0.0"
       },
       "optionalDependencies": {
-        "@lancedb/vectordb-darwin-arm64": "0.19.1-beta.1",
-        "@lancedb/vectordb-darwin-x64": "0.19.1-beta.1",
-        "@lancedb/vectordb-linux-arm64-gnu": "0.19.1-beta.1",
-        "@lancedb/vectordb-linux-x64-gnu": "0.19.1-beta.1",
-        "@lancedb/vectordb-win32-x64-msvc": "0.19.1-beta.1"
+        "@lancedb/vectordb-darwin-arm64": "0.21.1-beta.1",
+        "@lancedb/vectordb-darwin-x64": "0.21.1-beta.1",
+        "@lancedb/vectordb-linux-arm64-gnu": "0.21.1-beta.1",
+        "@lancedb/vectordb-linux-x64-gnu": "0.21.1-beta.1",
+        "@lancedb/vectordb-win32-x64-msvc": "0.21.1-beta.1"
       },
       "peerDependencies": {
         "@apache-arrow/ts": "^14.0.2",
@@ -326,71 +326,6 @@
         "@jridgewell/sourcemap-codec": "^1.4.10"
       }
     },
-    "node_modules/@lancedb/vectordb-darwin-arm64": {
-      "version": "0.19.1-beta.1",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-arm64/-/vectordb-darwin-arm64-0.19.1-beta.1.tgz",
-      "integrity": "sha512-Epvel0pF5TM6MtIWQ2KhqezqSSHTL3Wr7a2rGAwz6X/XY23i6DbMPpPs0HyeIDzDrhxNfE3cz3S+SiCA6xpR0g==",
-      "cpu": [
-        "arm64"
-      ],
-      "license": "Apache-2.0",
-      "optional": true,
-      "os": [
-        "darwin"
-      ]
-    },
-    "node_modules/@lancedb/vectordb-darwin-x64": {
-      "version": "0.19.1-beta.1",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-x64/-/vectordb-darwin-x64-0.19.1-beta.1.tgz",
-      "integrity": "sha512-hOiUSlIoISbiXytp46hToi/r6sF5pImAsfbzCsIq8ExDV4TPa8fjbhcIT80vxxOwc2mpSSK4HsVJYod95RSbEQ==",
-      "cpu": [
-        "x64"
-      ],
-      "license": "Apache-2.0",
-      "optional": true,
-      "os": [
-        "darwin"
-      ]
-    },
-    "node_modules/@lancedb/vectordb-linux-arm64-gnu": {
-      "version": "0.19.1-beta.1",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-arm64-gnu/-/vectordb-linux-arm64-gnu-0.19.1-beta.1.tgz",
-      "integrity": "sha512-/1JhGVDEngwrlM8o2TNW8G6nJ9U/VgHKAORmj/cTA7O30helJIoo9jfvUAUy+vZ4VoEwRXQbMI+gaYTg0l3MTg==",
-      "cpu": [
-        "arm64"
-      ],
-      "license": "Apache-2.0",
-      "optional": true,
-      "os": [
-        "linux"
-      ]
-    },
-    "node_modules/@lancedb/vectordb-linux-x64-gnu": {
-      "version": "0.19.1-beta.1",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-x64-gnu/-/vectordb-linux-x64-gnu-0.19.1-beta.1.tgz",
-      "integrity": "sha512-zNRGSSUt8nTJMmll4NdxhQjwxR8Rezq3T4dsRoiDts5ienMam5HFjYiZ3FkDZQo16rgq2BcbFuH1G8u1chywlg==",
-      "cpu": [
-        "x64"
-      ],
-      "license": "Apache-2.0",
-      "optional": true,
-      "os": [
-        "linux"
-      ]
-    },
-    "node_modules/@lancedb/vectordb-win32-x64-msvc": {
-      "version": "0.19.1-beta.1",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-win32-x64-msvc/-/vectordb-win32-x64-msvc-0.19.1-beta.1.tgz",
-      "integrity": "sha512-yV550AJGlsIFdm1KoHQPJ1TZx121ZXCIdebBtBZj3wOObIhyB/i0kZAtGvwjkmr7EYyfzt1EHZzbjSGVdehIAA==",
-      "cpu": [
-        "x64"
-      ],
-      "license": "Apache-2.0",
-      "optional": true,
-      "os": [
-        "win32"
-      ]
-    },
     "node_modules/@neon-rs/cli": {
       "version": "0.0.160",
       "resolved": "https://registry.npmjs.org/@neon-rs/cli/-/cli-0.0.160.tgz",

node/package.json

@@ -1,6 +1,6 @@
 {
   "name": "vectordb",
-  "version": "0.19.1-beta.1",
+  "version": "0.21.1-beta.1",
   "description": " Serverless, low-latency vector database for AI applications",
   "private": false,
   "main": "dist/index.js",
@@ -89,10 +89,10 @@
     }
   },
   "optionalDependencies": {
-    "@lancedb/vectordb-darwin-x64": "0.19.1-beta.1",
-    "@lancedb/vectordb-darwin-arm64": "0.19.1-beta.1",
-    "@lancedb/vectordb-linux-x64-gnu": "0.19.1-beta.1",
-    "@lancedb/vectordb-linux-arm64-gnu": "0.19.1-beta.1",
-    "@lancedb/vectordb-win32-x64-msvc": "0.19.1-beta.1"
+    "@lancedb/vectordb-darwin-x64": "0.21.1-beta.1",
+    "@lancedb/vectordb-darwin-arm64": "0.21.1-beta.1",
+    "@lancedb/vectordb-linux-x64-gnu": "0.21.1-beta.1",
+    "@lancedb/vectordb-linux-arm64-gnu": "0.21.1-beta.1",
+    "@lancedb/vectordb-win32-x64-msvc": "0.21.1-beta.1"
   }
 }

nodejs/Cargo.toml

@@ -1,7 +1,7 @@
 [package]
 name = "lancedb-nodejs"
 edition.workspace = true
-version = "0.19.1-beta.1"
+version = "0.21.1-beta.1"
 license.workspace = true
 description.workspace = true
 repository.workspace = true
@@ -30,6 +30,7 @@ log.workspace = true
 
 # Workaround for build failure until we can fix it.
 aws-lc-sys = "=0.28.0"
+aws-lc-rs = "=1.13.0"
 
 [build-dependencies]
 napi-build = "2.1"

nodejs/__test__/arrow.test.ts

@@ -592,14 +592,14 @@ describe.each([arrow15, arrow16, arrow17, arrow18])(
         ).rejects.toThrow("column vector was missing");
       });
 
-      it("will provide a nice error if run twice", async function () {
+      it("will skip embedding application if already applied", async function () {
         const records = sampleRecords();
         const table = await convertToTable(records, dummyEmbeddingConfig);
 
         // fromTableToBuffer will try and apply the embeddings again
-        await expect(
-          fromTableToBuffer(table, dummyEmbeddingConfig),
-        ).rejects.toThrow("already existed");
+        // but should skip since the column already has non-null values
+        const result = await fromTableToBuffer(table, dummyEmbeddingConfig);
+        expect(result.byteLength).toBeGreaterThan(0);
       });
     });
 

nodejs/__test__/table.test.ts

@@ -33,7 +33,13 @@ import {
   register,
 } from "../lancedb/embedding";
 import { Index } from "../lancedb/indices";
-import { instanceOfFullTextQuery } from "../lancedb/query";
+import {
+  BooleanQuery,
+  Occur,
+  Operator,
+  instanceOfFullTextQuery,
+} from "../lancedb/query";
+import exp = require("constants");
 
 describe.each([arrow15, arrow16, arrow17, arrow18])(
   "Given a table",
@@ -95,7 +101,9 @@ describe.each([arrow15, arrow16, arrow17, arrow18])(
     });
 
     it("should overwrite data if asked", async () => {
-      await table.add([{ id: 1 }, { id: 2 }]);
+      const addRes = await table.add([{ id: 1 }, { id: 2 }]);
+      expect(addRes).toHaveProperty("version");
+      expect(addRes.version).toBe(2);
       await table.add([{ id: 1 }], { mode: "overwrite" });
       await expect(table.countRows()).resolves.toBe(1);
     });
@@ -111,7 +119,11 @@ describe.each([arrow15, arrow16, arrow17, arrow18])(
       await table.add([{ id: 1 }]);
       expect(await table.countRows("id == 1")).toBe(1);
       expect(await table.countRows("id == 7")).toBe(0);
-      await table.update({ id: "7" });
+      const updateRes = await table.update({ id: "7" });
+      expect(updateRes).toHaveProperty("version");
+      expect(updateRes.version).toBe(3);
+      expect(updateRes).toHaveProperty("rowsUpdated");
+      expect(updateRes.rowsUpdated).toBe(1);
       expect(await table.countRows("id == 1")).toBe(0);
       expect(await table.countRows("id == 7")).toBe(1);
       await table.add([{ id: 2 }]);
@@ -338,15 +350,16 @@ describe("merge insert", () => {
       { a: 3, b: "y" },
       { a: 4, b: "z" },
     ];
-    const stats = await table
+    const mergeInsertRes = await table
       .mergeInsert("a")
       .whenMatchedUpdateAll()
       .whenNotMatchedInsertAll()
-      .execute(newData);
-
-    expect(stats.numInsertedRows).toBe(1n);
-    expect(stats.numUpdatedRows).toBe(2n);
-    expect(stats.numDeletedRows).toBe(0n);
+      .execute(newData, { timeoutMs: 10_000 });
+    expect(mergeInsertRes).toHaveProperty("version");
+    expect(mergeInsertRes.version).toBe(2);
+    expect(mergeInsertRes.numInsertedRows).toBe(1);
+    expect(mergeInsertRes.numUpdatedRows).toBe(2);
+    expect(mergeInsertRes.numDeletedRows).toBe(0);
 
     const expected = [
       { a: 1, b: "a" },
@@ -355,9 +368,9 @@ describe("merge insert", () => {
       { a: 4, b: "z" },
     ];
 
-    expect(
-      JSON.parse(JSON.stringify((await table.toArrow()).toArray())),
-    ).toEqual(expected);
+    const result = (await table.toArrow()).toArray().sort((a, b) => a.a - b.a);
+
+    expect(result.map((row) => ({ ...row }))).toEqual(expected);
   });
   test("conditional update", async () => {
     const newData = [
@@ -365,10 +378,12 @@ describe("merge insert", () => {
       { a: 3, b: "y" },
       { a: 4, b: "z" },
     ];
-    await table
+    const mergeInsertRes = await table
       .mergeInsert("a")
       .whenMatchedUpdateAll({ where: "target.b = 'b'" })
       .execute(newData);
+    expect(mergeInsertRes).toHaveProperty("version");
+    expect(mergeInsertRes.version).toBe(2);
 
     const expected = [
       { a: 1, b: "a" },
@@ -453,6 +468,20 @@ describe("merge insert", () => {
     res = res.sort((a, b) => a.a - b.a);
     expect(res).toEqual(expected);
   });
+
+  test("timeout", async () => {
+    const newData = [
+      { a: 2, b: "x" },
+      { a: 4, b: "z" },
+    ];
+    await expect(
+      table
+        .mergeInsert("a")
+        .whenMatchedUpdateAll()
+        .whenNotMatchedInsertAll()
+        .execute(newData, { timeoutMs: 0 }),
+    ).rejects.toThrow("merge insert timed out");
+  });
 });
 
 describe("When creating an index", () => {
@@ -530,6 +559,32 @@ describe("When creating an index", () => {
     rst = await tbl.query().limit(2).offset(1).nearestTo(queryVec).toArrow();
     expect(rst.numRows).toBe(1);
 
+    // test nprobes
+    rst = await tbl.query().nearestTo(queryVec).limit(2).nprobes(50).toArrow();
+    expect(rst.numRows).toBe(2);
+    rst = await tbl
+      .query()
+      .nearestTo(queryVec)
+      .limit(2)
+      .minimumNprobes(15)
+      .toArrow();
+    expect(rst.numRows).toBe(2);
+    rst = await tbl
+      .query()
+      .nearestTo(queryVec)
+      .limit(2)
+      .minimumNprobes(10)
+      .maximumNprobes(20)
+      .toArrow();
+    expect(rst.numRows).toBe(2);
+
+    expect(() => tbl.query().nearestTo(queryVec).minimumNprobes(0)).toThrow(
+      "Invalid input, minimum_nprobes must be greater than 0",
+    );
+    expect(() => tbl.query().nearestTo(queryVec).maximumNprobes(5)).toThrow(
+      "Invalid input, maximum_nprobes must be greater than minimum_nprobes",
+    );
+
     await tbl.dropIndex("vec_idx");
     const indices2 = await tbl.listIndices();
     expect(indices2.length).toBe(0);
@@ -1028,15 +1083,19 @@ describe("schema evolution", function () {
       { id: 1n, vector: [0.1, 0.2] },
     ]);
     // Can create a non-nullable column only through addColumns at the moment.
-    await table.addColumns([
+    const addColumnsRes = await table.addColumns([
       { name: "price", valueSql: "cast(10.0 as double)" },
     ]);
+    expect(addColumnsRes).toHaveProperty("version");
+    expect(addColumnsRes.version).toBe(2);
     expect(await table.schema()).toEqual(schema);
 
-    await table.alterColumns([
+    const alterColumnsRes = await table.alterColumns([
       { path: "id", rename: "new_id" },
       { path: "price", nullable: true },
     ]);
+    expect(alterColumnsRes).toHaveProperty("version");
+    expect(alterColumnsRes.version).toBe(3);
 
     const expectedSchema = new Schema([
       new Field("new_id", new Int64(), true),
@@ -1154,7 +1213,9 @@ describe("schema evolution", function () {
     const table = await con.createTable("vectors", [
       { id: 1n, vector: [0.1, 0.2] },
     ]);
-    await table.dropColumns(["vector"]);
+    const dropColumnsRes = await table.dropColumns(["vector"]);
+    expect(dropColumnsRes).toHaveProperty("version");
+    expect(dropColumnsRes.version).toBe(2);
 
     const expectedSchema = new Schema([new Field("id", new Int64(), true)]);
     expect(await table.schema()).toEqual(expectedSchema);
@@ -1271,6 +1332,32 @@ describe("when dealing with tags", () => {
     await table.checkoutLatest();
     expect(await table.version()).toBe(4);
   });
+
+  it("can checkout and restore tags", async () => {
+    const conn = await connect(tmpDir.name, {
+      readConsistencyInterval: 0,
+    });
+
+    const table = await conn.createTable("my_table", [
+      { id: 1n, vector: [0.1, 0.2] },
+    ]);
+    expect(await table.version()).toBe(1);
+    expect(await table.countRows()).toBe(1);
+    const tagsManager = await table.tags();
+    const tag1 = "tag1";
+    await tagsManager.create(tag1, 1);
+    await table.add([{ id: 2n, vector: [0.3, 0.4] }]);
+    const tag2 = "tag2";
+    await tagsManager.create(tag2, 2);
+    expect(await table.version()).toBe(2);
+    await table.checkout(tag1);
+    expect(await table.version()).toBe(1);
+    await table.restore();
+    expect(await table.version()).toBe(3);
+    expect(await table.countRows()).toBe(1);
+    await table.add([{ id: 3n, vector: [0.5, 0.6] }]);
+    expect(await table.countRows()).toBe(2);
+  });
 });
 
 describe("when optimizing a dataset", () => {
@@ -1450,7 +1537,9 @@ describe.each([arrow15, arrow16, arrow17, arrow18])(
       ];
       const table = await db.createTable("test", data);
       await table.createIndex("text", {
-        config: Index.fts(),
+        config: Index.fts({
+          withPosition: true,
+        }),
       });
 
       const results = await table.search("lance").toArray();
@@ -1473,6 +1562,18 @@ describe.each([arrow15, arrow16, arrow17, arrow18])(
 
       const results = await table.search("hello").toArray();
       expect(results[0].text).toBe(data[0].text);
+
+      const results2 = await table
+        .search(new MatchQuery("hello world", "text"))
+        .toArray();
+      expect(results2.length).toBe(2);
+
+      const results3 = await table
+        .search(
+          new MatchQuery("hello world", "text", { operator: Operator.And }),
+        )
+        .toArray();
+      expect(results3.length).toBe(1);
     });
 
     test("full text search without lowercase", async () => {
@@ -1503,7 +1604,9 @@ describe.each([arrow15, arrow16, arrow17, arrow18])(
       ];
       const table = await db.createTable("test", data);
       await table.createIndex("text", {
-        config: Index.fts(),
+        config: Index.fts({
+          withPosition: true,
+        }),
       });
 
       const results = await table.search("world").toArray();
@@ -1547,6 +1650,60 @@ describe.each([arrow15, arrow16, arrow17, arrow18])(
       expect(resultSet.has("fob")).toBe(true);
       expect(resultSet.has("fo")).toBe(true);
       expect(resultSet.has("food")).toBe(true);
+
+      const prefixResults = await table
+        .search(
+          new MatchQuery("foo", "text", { fuzziness: 3, prefixLength: 3 }),
+        )
+        .toArray();
+      expect(prefixResults.length).toBe(2);
+      const resultSet2 = new Set(prefixResults.map((r) => r.text));
+      expect(resultSet2.has("foo")).toBe(true);
+      expect(resultSet2.has("food")).toBe(true);
+    });
+
+    test("full text search boolean query", async () => {
+      const db = await connect(tmpDir.name);
+      const data = [
+        { text: "The cat and dog are playing" },
+        { text: "The cat is sleeping" },
+        { text: "The dog is barking" },
+        { text: "The dog chases the cat" },
+      ];
+      const table = await db.createTable("test", data);
+      await table.createIndex("text", {
+        config: Index.fts({ withPosition: false }),
+      });
+
+      const shouldResults = await table
+        .search(
+          new BooleanQuery([
+            [Occur.Should, new MatchQuery("cat", "text")],
+            [Occur.Should, new MatchQuery("dog", "text")],
+          ]),
+        )
+        .toArray();
+      expect(shouldResults.length).toBe(4);
+
+      const mustResults = await table
+        .search(
+          new BooleanQuery([
+            [Occur.Must, new MatchQuery("cat", "text")],
+            [Occur.Must, new MatchQuery("dog", "text")],
+          ]),
+        )
+        .toArray();
+      expect(mustResults.length).toBe(2);
+
+      const mustNotResults = await table
+        .search(
+          new BooleanQuery([
+            [Occur.Must, new MatchQuery("cat", "text")],
+            [Occur.MustNot, new MatchQuery("dog", "text")],
+          ]),
+        )
+        .toArray();
+      expect(mustNotResults.length).toBe(1);
     });
 
     test.each([

nodejs/lancedb/arrow.ts

@@ -417,7 +417,9 @@ function inferSchema(
         } else {
           const inferredType = inferType(value, path, opts);
           if (inferredType === undefined) {
-            throw new Error(`Failed to infer data type for field ${path.join(".")} at row ${rowI}. \
+            throw new Error(`Failed to infer data type for field ${path.join(
+              ".",
+            )} at row ${rowI}. \
                              Consider providing an explicit schema.`);
           }
           pathTree.set(path, inferredType);
@@ -799,11 +801,17 @@ async function applyEmbeddingsFromMetadata(
         `Cannot apply embedding function because the source column '${functionEntry.sourceColumn}' was not present in the data`,
       );
     }
+
+    // Check if destination column exists and handle accordingly
     if (columns[destColumn] !== undefined) {
-      throw new Error(
-        `Attempt to apply embeddings to table failed because column ${destColumn} already existed`,
-      );
+      const existingColumn = columns[destColumn];
+      // If the column exists but is all null, we can fill it with embeddings
+      if (existingColumn.nullCount !== existingColumn.length) {
+        // Column has non-null values, skip embedding application
+        continue;
+      }
     }
+
     if (table.batches.length > 1) {
       throw new Error(
         "Internal error: `makeArrowTable` unexpectedly created a table with more than one batch",
@@ -903,11 +911,23 @@ async function applyEmbeddings<T>(
       );
     }
   } else {
+    // Check if destination column exists and handle accordingly
     if (Object.prototype.hasOwnProperty.call(newColumns, destColumn)) {
-      throw new Error(
-        `Attempt to apply embeddings to table failed because column ${destColumn} already existed`,
-      );
+      const existingColumn = newColumns[destColumn];
+      // If the column exists but is all null, we can fill it with embeddings
+      if (existingColumn.nullCount !== existingColumn.length) {
+        // Column has non-null values, skip embedding application and return table as-is
+        let newTable = new ArrowTable(newColumns);
+        if (schema != null) {
+          newTable = alignTable(newTable, schema as Schema);
+        }
+        return new ArrowTable(
+          new Schema(newTable.schema.fields, schemaMetadata),
+          newTable.batches,
+        );
+      }
     }
+
     if (table.batches.length > 1) {
       throw new Error(
         "Internal error: `makeArrowTable` unexpectedly created a table with more than one batch",

nodejs/lancedb/index.ts

@@ -28,7 +28,13 @@ export {
   FragmentSummaryStats,
   Tags,
   TagContents,
-  MergeStats,
+  MergeResult,
+  AddResult,
+  AddColumnsResult,
+  AlterColumnsResult,
+  DeleteResult,
+  DropColumnsResult,
+  UpdateResult,
 } from "./native.js";
 
 export {
@@ -58,7 +64,10 @@ export {
   PhraseQuery,
   BoostQuery,
   MultiMatchQuery,
+  BooleanQuery,
   FullTextQueryType,
+  Operator,
+  Occur,
 } from "./query";
 
 export {
@@ -80,7 +89,7 @@ export {
   ColumnAlteration,
 } from "./table";
 
-export { MergeInsertBuilder } from "./merge";
+export { MergeInsertBuilder, WriteExecutionOptions } from "./merge";
 
 export * as embedding from "./embedding";
 export * as rerankers from "./rerankers";

nodejs/lancedb/merge.ts

@@ -1,7 +1,7 @@
 // SPDX-License-Identifier: Apache-2.0
 // SPDX-FileCopyrightText: Copyright The LanceDB Authors
 import { Data, Schema, fromDataToBuffer } from "./arrow";
-import { MergeStats, NativeMergeInsertBuilder } from "./native";
+import { MergeResult, NativeMergeInsertBuilder } from "./native";
 
 /** A builder used to create and run a merge insert operation */
 export class MergeInsertBuilder {
@@ -73,9 +73,12 @@ export class MergeInsertBuilder {
   /**
    * Executes the merge insert operation
    *
-   * @returns Statistics about the merge operation: counts of inserted, updated, and deleted rows
+   * @returns {Promise<MergeResult>} the merge result
    */
-  async execute(data: Data): Promise<MergeStats> {
+  async execute(
+    data: Data,
+    execOptions?: Partial<WriteExecutionOptions>,
+  ): Promise<MergeResult> {
     let schema: Schema;
     if (this.#schema instanceof Promise) {
       schema = await this.#schema;
@@ -83,7 +86,28 @@ export class MergeInsertBuilder {
     } else {
       schema = this.#schema;
     }
+
+    if (execOptions?.timeoutMs !== undefined) {
+      this.#native.setTimeout(execOptions.timeoutMs);
+    }
+
     const buffer = await fromDataToBuffer(data, undefined, schema);
     return await this.#native.execute(buffer);
   }
 }
+
+export interface WriteExecutionOptions {
+  /**
+   * Maximum time to run the operation before cancelling it.
+   *
+   * By default, there is a 30-second timeout that is only enforced after the
+   * first attempt. This is to prevent spending too long retrying to resolve
+   * conflicts. For example, if a write attempt takes 20 seconds and fails,
+   * the second attempt will be cancelled after 10 seconds, hitting the
+   * 30-second timeout. However, a write that takes one hour and succeeds on the
+   * first attempt will not be cancelled.
+   *
+   * When this is set, the timeout is enforced on all attempts, including the first.
+   */
+  timeoutMs?: number;
+}

nodejs/lancedb/query.ts

@@ -448,6 +448,10 @@ export class VectorQuery extends QueryBase<NativeVectorQuery> {
    * For best results we recommend tuning this parameter with a benchmark against
    * your actual data to find the smallest possible value that will still give
    * you the desired recall.
+   *
+   * For more fine grained control over behavior when you have a very narrow filter
+   * you can use `minimumNprobes` and `maximumNprobes`.  This method sets both
+   * the minimum and maximum to the same value.
    */
   nprobes(nprobes: number): VectorQuery {
     super.doCall((inner) => inner.nprobes(nprobes));
@@ -455,6 +459,33 @@ export class VectorQuery extends QueryBase<NativeVectorQuery> {
     return this;
   }
 
+  /**
+   * Set the minimum number of probes used.
+   *
+   * This controls the minimum number of partitions that will be searched.  This
+   * parameter will impact every query against a vector index, regardless of the
+   * filter.  See `nprobes` for more details.  Higher values will increase recall
+   * but will also increase latency.
+   */
+  minimumNprobes(minimumNprobes: number): VectorQuery {
+    super.doCall((inner) => inner.minimumNprobes(minimumNprobes));
+    return this;
+  }
+
+  /**
+   * Set the maximum number of probes used.
+   *
+   * This controls the maximum number of partitions that will be searched.  If this
+   * number is greater than minimumNprobes then the excess partitions will _only_ be
+   * searched if we have not found enough results.  This can be useful when there is
+   * a narrow filter to allow these queries to spend more time searching and avoid
+   * potential false negatives.
+   */
+  maximumNprobes(maximumNprobes: number): VectorQuery {
+    super.doCall((inner) => inner.maximumNprobes(maximumNprobes));
+    return this;
+  }
+
   /*
    * Set the distance range to use
    *
@@ -762,6 +793,31 @@ export enum FullTextQueryType {
   MatchPhrase = "match_phrase",
   Boost = "boost",
   MultiMatch = "multi_match",
+  Boolean = "boolean",
+}
+
+/**
+ * Enum representing the logical operators used in full-text queries.
+ *
+ * - `And`: All terms must match.
+ * - `Or`: At least one term must match.
+ */
+export enum Operator {
+  And = "AND",
+  Or = "OR",
+}
+
+/**
+ * Enum representing the occurrence of terms in full-text queries.
+ *
+ * - `Must`: The term must be present in the document.
+ * - `Should`: The term should contribute to the document score, but is not required.
+ * - `MustNot`: The term must not be present in the document.
+ */
+export enum Occur {
+  Should = "SHOULD",
+  Must = "MUST",
+  MustNot = "MUST_NOT",
 }
 
 /**
@@ -791,6 +847,7 @@ export function instanceOfFullTextQuery(obj: any): obj is FullTextQuery {
 export class MatchQuery implements FullTextQuery {
   /** @ignore */
   public readonly inner: JsFullTextQuery;
+
   /**
    * Creates an instance of MatchQuery.
    *
@@ -800,6 +857,8 @@ export class MatchQuery implements FullTextQuery {
    *   - `boost`: The boost factor for the query (default is 1.0).
    *   - `fuzziness`: The fuzziness level for the query (default is 0).
    *   - `maxExpansions`: The maximum number of terms to consider for fuzzy matching (default is 50).
+   *   - `operator`: The logical operator to use for combining terms in the query (default is "OR").
+   *   - `prefixLength`: The number of beginning characters being unchanged for fuzzy matching.
    */
   constructor(
     query: string,
@@ -808,6 +867,8 @@ export class MatchQuery implements FullTextQuery {
       boost?: number;
       fuzziness?: number;
       maxExpansions?: number;
+      operator?: Operator;
+      prefixLength?: number;
     },
   ) {
     let fuzziness = options?.fuzziness;
@@ -820,6 +881,8 @@ export class MatchQuery implements FullTextQuery {
       options?.boost ?? 1.0,
       fuzziness,
       options?.maxExpansions ?? 50,
+      options?.operator ?? Operator.Or,
+      options?.prefixLength ?? 0,
     );
   }
 
@@ -836,9 +899,11 @@ export class PhraseQuery implements FullTextQuery {
    *
    * @param query - The phrase to search for in the specified column.
    * @param column - The name of the column to search within.
+   * @param options - Optional parameters for the phrase query.
+   *   - `slop`: The maximum number of intervening unmatched positions allowed between words in the phrase (default is 0).
    */
-  constructor(query: string, column: string) {
-    this.inner = JsFullTextQuery.phraseQuery(query, column);
+  constructor(query: string, column: string, options?: { slop?: number }) {
+    this.inner = JsFullTextQuery.phraseQuery(query, column, options?.slop ?? 0);
   }
 
   queryType(): FullTextQueryType {
@@ -889,18 +954,21 @@ export class MultiMatchQuery implements FullTextQuery {
    * @param columns - An array of column names to search within.
    * @param options - Optional parameters for the multi-match query.
    *  - `boosts`: An array of boost factors for each column (default is 1.0 for all).
+   *  - `operator`: The logical operator to use for combining terms in the query (default is "OR").
    */
   constructor(
     query: string,
     columns: string[],
     options?: {
       boosts?: number[];
+      operator?: Operator;
     },
   ) {
     this.inner = JsFullTextQuery.multiMatchQuery(
       query,
       columns,
       options?.boosts,
+      options?.operator ?? Operator.Or,
     );
   }
 
@@ -908,3 +976,23 @@ export class MultiMatchQuery implements FullTextQuery {
     return FullTextQueryType.MultiMatch;
   }
 }
+
+export class BooleanQuery implements FullTextQuery {
+  /** @ignore */
+  public readonly inner: JsFullTextQuery;
+  /**
+   * Creates an instance of BooleanQuery.
+   *
+   * @param queries - An array of (Occur, FullTextQuery objects) to combine.
+   * Occur specifies whether the query must match, or should match.
+   */
+  constructor(queries: [Occur, FullTextQuery][]) {
+    this.inner = JsFullTextQuery.booleanQuery(
+      queries.map(([occur, query]) => [occur, query.inner]),
+    );
+  }
+
+  queryType(): FullTextQueryType {
+    return FullTextQueryType.Boolean;
+  }
+}

nodejs/lancedb/table.ts

@@ -16,12 +16,18 @@ import { EmbeddingFunctionConfig, getRegistry } from "./embedding/registry";
 import { IndexOptions } from "./indices";
 import { MergeInsertBuilder } from "./merge";
 import {
+  AddColumnsResult,
   AddColumnsSql,
+  AddResult,
+  AlterColumnsResult,
+  DeleteResult,
+  DropColumnsResult,
   IndexConfig,
   IndexStatistics,
   OptimizeStats,
   TableStatistics,
   Tags,
+  UpdateResult,
   Table as _NativeTable,
 } from "./native";
 import {
@@ -126,12 +132,19 @@ export abstract class Table {
   /**
    * Insert records into this Table.
    * @param {Data} data Records to be inserted into the Table
+   * @returns {Promise<AddResult>} A promise that resolves to an object
+   * containing the new version number of the table
    */
-  abstract add(data: Data, options?: Partial<AddDataOptions>): Promise<void>;
+  abstract add(
+    data: Data,
+    options?: Partial<AddDataOptions>,
+  ): Promise<AddResult>;
   /**
    * Update existing records in the Table
    * @param opts.values The values to update. The keys are the column names and the values
    * are the values to set.
+   * @returns {Promise<UpdateResult>} A promise that resolves to an object containing
+   * the number of rows updated and the new version number
    * @example
    * ```ts
    * table.update({where:"x = 2", values:{"vector": [10, 10]}})
@@ -141,11 +154,13 @@ export abstract class Table {
     opts: {
       values: Map<string, IntoSql> | Record<string, IntoSql>;
     } & Partial<UpdateOptions>,
-  ): Promise<void>;
+  ): Promise<UpdateResult>;
   /**
    * Update existing records in the Table
    * @param opts.valuesSql The values to update. The keys are the column names and the values
    * are the values to set. The values are SQL expressions.
+   * @returns {Promise<UpdateResult>} A promise that resolves to an object containing
+   * the number of rows updated and the new version number
    * @example
    * ```ts
    * table.update({where:"x = 2", valuesSql:{"x": "x + 1"}})
@@ -155,7 +170,7 @@ export abstract class Table {
     opts: {
       valuesSql: Map<string, string> | Record<string, string>;
     } & Partial<UpdateOptions>,
-  ): Promise<void>;
+  ): Promise<UpdateResult>;
   /**
    * Update existing records in the Table
    *
@@ -173,6 +188,8 @@ export abstract class Table {
    * repeatedly calilng this method.
    * @param {Map<string, string> | Record<string, string>} updates - the
    * columns to update
+   * @returns {Promise<UpdateResult>} A promise that resolves to an object
+   * containing the number of rows updated and the new version number
    *
    * Keys in the map should specify the name of the column to update.
    * Values in the map provide the new value of the column.  These can
@@ -184,12 +201,16 @@ export abstract class Table {
   abstract update(
     updates: Map<string, string> | Record<string, string>,
     options?: Partial<UpdateOptions>,
-  ): Promise<void>;
+  ): Promise<UpdateResult>;
 
   /** Count the total number of rows in the dataset. */
   abstract countRows(filter?: string): Promise<number>;
-  /** Delete the rows that satisfy the predicate. */
-  abstract delete(predicate: string): Promise<void>;
+  /**
+   * Delete the rows that satisfy the predicate.
+   * @returns {Promise<DeleteResult>} A promise that resolves to an object
+   * containing the new version number of the table
+   */
+  abstract delete(predicate: string): Promise<DeleteResult>;
   /**
    * Create an index to speed up queries.
    *
@@ -343,15 +364,23 @@ export abstract class Table {
    * the SQL expression to use to calculate the value of the new column. These
    * expressions will be evaluated for each row in the table, and can
    * reference existing columns in the table.
+   * @returns {Promise<AddColumnsResult>} A promise that resolves to an object
+   * containing the new version number of the table after adding the columns.
    */
-  abstract addColumns(newColumnTransforms: AddColumnsSql[]): Promise<void>;
+  abstract addColumns(
+    newColumnTransforms: AddColumnsSql[],
+  ): Promise<AddColumnsResult>;
 
   /**
    * Alter the name or nullability of columns.
    * @param {ColumnAlteration[]} columnAlterations One or more alterations to
    * apply to columns.
+   * @returns {Promise<AlterColumnsResult>} A promise that resolves to an object
+   * containing the new version number of the table after altering the columns.
    */
-  abstract alterColumns(columnAlterations: ColumnAlteration[]): Promise<void>;
+  abstract alterColumns(
+    columnAlterations: ColumnAlteration[],
+  ): Promise<AlterColumnsResult>;
   /**
    * Drop one or more columns from the dataset
    *
@@ -362,8 +391,10 @@ export abstract class Table {
    * @param {string[]} columnNames The names of the columns to drop. These can
    * be nested column references (e.g. "a.b.c") or top-level column names
    * (e.g. "a").
+   * @returns {Promise<DropColumnsResult>} A promise that resolves to an object
+   * containing the new version number of the table after dropping the columns.
    */
-  abstract dropColumns(columnNames: string[]): Promise<void>;
+  abstract dropColumns(columnNames: string[]): Promise<DropColumnsResult>;
   /** Retrieve the version of the table */
 
   abstract version(): Promise<number>;
@@ -529,12 +560,12 @@ export class LocalTable extends Table {
     return tbl.schema;
   }
 
-  async add(data: Data, options?: Partial<AddDataOptions>): Promise<void> {
+  async add(data: Data, options?: Partial<AddDataOptions>): Promise<AddResult> {
     const mode = options?.mode ?? "append";
     const schema = await this.schema();
 
     const buffer = await fromDataToBuffer(data, undefined, schema);
-    await this.inner.add(buffer, mode);
+    return await this.inner.add(buffer, mode);
   }
 
   async update(
@@ -547,7 +578,7 @@ export class LocalTable extends Table {
           valuesSql: Map<string, string> | Record<string, string>;
         } & Partial<UpdateOptions>),
     options?: Partial<UpdateOptions>,
-  ) {
+  ): Promise<UpdateResult> {
     const isValues =
       "values" in optsOrUpdates && typeof optsOrUpdates.values !== "string";
     const isValuesSql =
@@ -594,15 +625,15 @@ export class LocalTable extends Table {
         columns = Object.entries(optsOrUpdates as Record<string, string>);
         predicate = options?.where;
     }
-    await this.inner.update(predicate, columns);
+    return await this.inner.update(predicate, columns);
   }
 
   async countRows(filter?: string): Promise<number> {
     return await this.inner.countRows(filter);
   }
 
-  async delete(predicate: string): Promise<void> {
-    await this.inner.delete(predicate);
+  async delete(predicate: string): Promise<DeleteResult> {
+    return await this.inner.delete(predicate);
   }
 
   async createIndex(column: string, options?: Partial<IndexOptions>) {
@@ -690,11 +721,15 @@ export class LocalTable extends Table {
 
   // TODO: Support BatchUDF
 
-  async addColumns(newColumnTransforms: AddColumnsSql[]): Promise<void> {
-    await this.inner.addColumns(newColumnTransforms);
+  async addColumns(
+    newColumnTransforms: AddColumnsSql[],
+  ): Promise<AddColumnsResult> {
+    return await this.inner.addColumns(newColumnTransforms);
   }
 
-  async alterColumns(columnAlterations: ColumnAlteration[]): Promise<void> {
+  async alterColumns(
+    columnAlterations: ColumnAlteration[],
+  ): Promise<AlterColumnsResult> {
     const processedAlterations = columnAlterations.map((alteration) => {
       if (typeof alteration.dataType === "string") {
         return {
@@ -715,11 +750,11 @@ export class LocalTable extends Table {
       }
     });
 
-    await this.inner.alterColumns(processedAlterations);
+    return await this.inner.alterColumns(processedAlterations);
   }
 
-  async dropColumns(columnNames: string[]): Promise<void> {
-    await this.inner.dropColumns(columnNames);
+  async dropColumns(columnNames: string[]): Promise<DropColumnsResult> {
+    return await this.inner.dropColumns(columnNames);
   }
 
   async version(): Promise<number> {

nodejs/npm/darwin-arm64/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-darwin-arm64",
-	"version": "0.19.1-beta.1",
+	"version": "0.21.1-beta.1",
 	"os": ["darwin"],
 	"cpu": ["arm64"],
 	"main": "lancedb.darwin-arm64.node",

nodejs/npm/darwin-x64/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-darwin-x64",
-	"version": "0.19.1-beta.1",
+	"version": "0.21.1-beta.1",
 	"os": ["darwin"],
 	"cpu": ["x64"],
 	"main": "lancedb.darwin-x64.node",

nodejs/npm/linux-arm64-gnu/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-arm64-gnu",
-	"version": "0.19.1-beta.1",
+	"version": "0.21.1-beta.1",
 	"os": ["linux"],
 	"cpu": ["arm64"],
 	"main": "lancedb.linux-arm64-gnu.node",

nodejs/npm/linux-arm64-musl/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-arm64-musl",
-	"version": "0.19.1-beta.1",
+	"version": "0.21.1-beta.1",
 	"os": ["linux"],
 	"cpu": ["arm64"],
 	"main": "lancedb.linux-arm64-musl.node",

nodejs/npm/linux-x64-gnu/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-x64-gnu",
-	"version": "0.19.1-beta.1",
+	"version": "0.21.1-beta.1",
 	"os": ["linux"],
 	"cpu": ["x64"],
 	"main": "lancedb.linux-x64-gnu.node",

nodejs/npm/linux-x64-musl/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-x64-musl",
-	"version": "0.19.1-beta.1",
+	"version": "0.21.1-beta.1",
 	"os": ["linux"],
 	"cpu": ["x64"],
 	"main": "lancedb.linux-x64-musl.node",

nodejs/npm/win32-arm64-msvc/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@lancedb/lancedb-win32-arm64-msvc",
-  "version": "0.19.1-beta.1",
+  "version": "0.21.1-beta.1",
   "os": [
     "win32"
   ],

nodejs/npm/win32-x64-msvc/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-win32-x64-msvc",
-	"version": "0.19.1-beta.1",
+	"version": "0.21.1-beta.1",
 	"os": ["win32"],
 	"cpu": ["x64"],
 	"main": "lancedb.win32-x64-msvc.node",

nodejs/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "@lancedb/lancedb",
-  "version": "0.19.1-beta.1",
+  "version": "0.21.1-beta.1",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "@lancedb/lancedb",
-      "version": "0.19.1-beta.1",
+      "version": "0.21.1-beta.1",
       "cpu": [
         "x64",
         "arm64"

nodejs/package.json

@@ -11,7 +11,7 @@
     "ann"
   ],
   "private": false,
-  "version": "0.19.1-beta.1",
+  "version": "0.21.1-beta.1",
   "main": "dist/index.js",
   "exports": {
     ".": "./dist/index.js",

nodejs/src/index.rs

@@ -125,32 +125,30 @@ impl Index {
         ascii_folding: Option<bool>,
     ) -> Self {
         let mut opts = FtsIndexBuilder::default();
-        let mut tokenizer_configs = opts.tokenizer_configs.clone();
         if let Some(with_position) = with_position {
             opts = opts.with_position(with_position);
         }
         if let Some(base_tokenizer) = base_tokenizer {
-            tokenizer_configs = tokenizer_configs.base_tokenizer(base_tokenizer);
+            opts = opts.base_tokenizer(base_tokenizer);
         }
         if let Some(language) = language {
-            tokenizer_configs = tokenizer_configs.language(&language).unwrap();
+            opts = opts.language(&language).unwrap();
         }
         if let Some(max_token_length) = max_token_length {
-            tokenizer_configs = tokenizer_configs.max_token_length(Some(max_token_length as usize));
+            opts = opts.max_token_length(Some(max_token_length as usize));
         }
         if let Some(lower_case) = lower_case {
-            tokenizer_configs = tokenizer_configs.lower_case(lower_case);
+            opts = opts.lower_case(lower_case);
         }
         if let Some(stem) = stem {
-            tokenizer_configs = tokenizer_configs.stem(stem);
+            opts = opts.stem(stem);
         }
         if let Some(remove_stop_words) = remove_stop_words {
-            tokenizer_configs = tokenizer_configs.remove_stop_words(remove_stop_words);
+            opts = opts.remove_stop_words(remove_stop_words);
         }
         if let Some(ascii_folding) = ascii_folding {
-            tokenizer_configs = tokenizer_configs.ascii_folding(ascii_folding);
+            opts = opts.ascii_folding(ascii_folding);
         }
-        opts.tokenizer_configs = tokenizer_configs;
 
         Self {
             inner: Mutex::new(Some(LanceDbIndex::FTS(opts))),

nodejs/src/merge.rs

@@ -1,11 +1,13 @@
 // SPDX-License-Identifier: Apache-2.0
 // SPDX-FileCopyrightText: Copyright The LanceDB Authors
 
+use std::time::Duration;
+
 use lancedb::{arrow::IntoArrow, ipc::ipc_file_to_batches, table::merge::MergeInsertBuilder};
 use napi::bindgen_prelude::*;
 use napi_derive::napi;
 
-use crate::error::convert_error;
+use crate::{error::convert_error, table::MergeResult};
 
 #[napi]
 #[derive(Clone)]
@@ -36,8 +38,13 @@ impl NativeMergeInsertBuilder {
         this
     }
 
+    #[napi]
+    pub fn set_timeout(&mut self, timeout: u32) {
+        self.inner.timeout(Duration::from_millis(timeout as u64));
+    }
+
     #[napi(catch_unwind)]
-    pub async fn execute(&self, buf: Buffer) -> napi::Result<MergeStats> {
+    pub async fn execute(&self, buf: Buffer) -> napi::Result<MergeResult> {
         let data = ipc_file_to_batches(buf.to_vec())
             .and_then(IntoArrow::into_arrow)
             .map_err(|e| {
@@ -46,14 +53,13 @@ impl NativeMergeInsertBuilder {
 
         let this = self.clone();
 
-        let stats = this.inner.execute(data).await.map_err(|e| {
+        let res = this.inner.execute(data).await.map_err(|e| {
             napi::Error::from_reason(format!(
                 "Failed to execute merge insert: {}",
                 convert_error(&e)
             ))
         })?;
-
-        Ok(stats.into())
+        Ok(res.into())
     }
 }
 
@@ -62,20 +68,3 @@ impl From<MergeInsertBuilder> for NativeMergeInsertBuilder {
         Self { inner }
     }
 }
-
-#[napi(object)]
-pub struct MergeStats {
-    pub num_inserted_rows: BigInt,
-    pub num_updated_rows: BigInt,
-    pub num_deleted_rows: BigInt,
-}
-
-impl From<lancedb::table::MergeStats> for MergeStats {
-    fn from(stats: lancedb::table::MergeStats) -> Self {
-        Self {
-            num_inserted_rows: stats.num_inserted_rows.into(),
-            num_updated_rows: stats.num_updated_rows.into(),
-            num_deleted_rows: stats.num_deleted_rows.into(),
-        }
-    }
-}

nodejs/src/query.rs

@@ -4,7 +4,8 @@
 use std::sync::Arc;
 
 use lancedb::index::scalar::{
-    BoostQuery, FtsQuery, FullTextSearchQuery, MatchQuery, MultiMatchQuery, PhraseQuery,
+    BooleanQuery, BoostQuery, FtsQuery, FullTextSearchQuery, MatchQuery, MultiMatchQuery, Occur,
+    Operator, PhraseQuery,
 };
 use lancedb::query::ExecutableQuery;
 use lancedb::query::Query as LanceDbQuery;
@@ -177,6 +178,31 @@ impl VectorQuery {
         self.inner = self.inner.clone().nprobes(nprobe as usize);
     }
 
+    #[napi]
+    pub fn minimum_nprobes(&mut self, minimum_nprobe: u32) -> napi::Result<()> {
+        self.inner = self
+            .inner
+            .clone()
+            .minimum_nprobes(minimum_nprobe as usize)
+            .default_error()?;
+        Ok(())
+    }
+
+    #[napi]
+    pub fn maximum_nprobes(&mut self, maximum_nprobes: u32) -> napi::Result<()> {
+        let maximum_nprobes = if maximum_nprobes == 0 {
+            None
+        } else {
+            Some(maximum_nprobes as usize)
+        };
+        self.inner = self
+            .inner
+            .clone()
+            .maximum_nprobes(maximum_nprobes)
+            .default_error()?;
+        Ok(())
+    }
+
     #[napi]
     pub fn distance_range(&mut self, lower_bound: Option<f64>, upper_bound: Option<f64>) {
         // napi doesn't support f32, so we have to convert to f32
@@ -308,6 +334,8 @@ impl JsFullTextQuery {
         boost: f64,
         fuzziness: Option<u32>,
         max_expansions: u32,
+        operator: String,
+        prefix_length: u32,
     ) -> napi::Result<Self> {
         Ok(Self {
             inner: MatchQuery::new(query)
@@ -315,14 +343,23 @@ impl JsFullTextQuery {
                 .with_boost(boost as f32)
                 .with_fuzziness(fuzziness)
                 .with_max_expansions(max_expansions as usize)
+                .with_operator(
+                    Operator::try_from(operator.as_str()).map_err(|e| {
+                        napi::Error::from_reason(format!("Invalid operator: {}", e))
+                    })?,
+                )
+                .with_prefix_length(prefix_length)
                 .into(),
         })
     }
 
     #[napi(factory)]
-    pub fn phrase_query(query: String, column: String) -> napi::Result<Self> {
+    pub fn phrase_query(query: String, column: String, slop: u32) -> napi::Result<Self> {
         Ok(Self {
-            inner: PhraseQuery::new(query).with_column(Some(column)).into(),
+            inner: PhraseQuery::new(query)
+                .with_column(Some(column))
+                .with_slop(slop)
+                .into(),
         })
     }
 
@@ -348,6 +385,7 @@ impl JsFullTextQuery {
         query: String,
         columns: Vec<String>,
         boosts: Option<Vec<f64>>,
+        operator: String,
     ) -> napi::Result<Self> {
         let q = match boosts {
             Some(boosts) => MultiMatchQuery::try_new(query, columns)
@@ -358,7 +396,37 @@ impl JsFullTextQuery {
             napi::Error::from_reason(format!("Failed to create multi match query: {}", e))
         })?;
 
-        Ok(Self { inner: q.into() })
+        let operator = Operator::try_from(operator.as_str()).map_err(|e| {
+            napi::Error::from_reason(format!("Invalid operator for multi match query: {}", e))
+        })?;
+
+        Ok(Self {
+            inner: q.with_operator(operator).into(),
+        })
+    }
+
+    #[napi(factory)]
+    pub fn boolean_query(queries: Vec<(String, &JsFullTextQuery)>) -> napi::Result<Self> {
+        let mut sub_queries = Vec::with_capacity(queries.len());
+        for (occur, q) in queries {
+            let occur = Occur::try_from(occur.as_str())
+                .map_err(|e| napi::Error::from_reason(e.to_string()))?;
+            sub_queries.push((occur, q.inner.clone()));
+        }
+        Ok(Self {
+            inner: BooleanQuery::new(sub_queries).into(),
+        })
+    }
+
+    #[napi(getter)]
+    pub fn query_type(&self) -> String {
+        match self.inner {
+            FtsQuery::Match(_) => "match".to_string(),
+            FtsQuery::Phrase(_) => "phrase".to_string(),
+            FtsQuery::Boost(_) => "boost".to_string(),
+            FtsQuery::MultiMatch(_) => "multi_match".to_string(),
+            FtsQuery::Boolean(_) => "boolean".to_string(),
+        }
     }
 }
 

nodejs/src/table.rs

@@ -75,7 +75,7 @@ impl Table {
     }
 
     #[napi(catch_unwind)]
-    pub async fn add(&self, buf: Buffer, mode: String) -> napi::Result<()> {
+    pub async fn add(&self, buf: Buffer, mode: String) -> napi::Result<AddResult> {
         let batches = ipc_file_to_batches(buf.to_vec())
             .map_err(|e| napi::Error::from_reason(format!("Failed to read IPC file: {}", e)))?;
         let mut op = self.inner_ref()?.add(batches);
@@ -88,7 +88,8 @@ impl Table {
             return Err(napi::Error::from_reason(format!("Invalid mode: {}", mode)));
         };
 
-        op.execute().await.default_error()
+        let res = op.execute().await.default_error()?;
+        Ok(res.into())
     }
 
     #[napi(catch_unwind)]
@@ -101,8 +102,9 @@ impl Table {
     }
 
     #[napi(catch_unwind)]
-    pub async fn delete(&self, predicate: String) -> napi::Result<()> {
-        self.inner_ref()?.delete(&predicate).await.default_error()
+    pub async fn delete(&self, predicate: String) -> napi::Result<DeleteResult> {
+        let res = self.inner_ref()?.delete(&predicate).await.default_error()?;
+        Ok(res.into())
     }
 
     #[napi(catch_unwind)]
@@ -168,7 +170,7 @@ impl Table {
         &self,
         only_if: Option<String>,
         columns: Vec<(String, String)>,
-    ) -> napi::Result<u64> {
+    ) -> napi::Result<UpdateResult> {
         let mut op = self.inner_ref()?.update();
         if let Some(only_if) = only_if {
             op = op.only_if(only_if);
@@ -176,7 +178,8 @@ impl Table {
         for (column_name, value) in columns {
             op = op.column(column_name, value);
         }
-        op.execute().await.default_error()
+        let res = op.execute().await.default_error()?;
+        Ok(res.into())
     }
 
     #[napi(catch_unwind)]
@@ -190,21 +193,28 @@ impl Table {
     }
 
     #[napi(catch_unwind)]
-    pub async fn add_columns(&self, transforms: Vec<AddColumnsSql>) -> napi::Result<()> {
+    pub async fn add_columns(
+        &self,
+        transforms: Vec<AddColumnsSql>,
+    ) -> napi::Result<AddColumnsResult> {
         let transforms = transforms
             .into_iter()
             .map(|sql| (sql.name, sql.value_sql))
             .collect::<Vec<_>>();
         let transforms = NewColumnTransform::SqlExpressions(transforms);
-        self.inner_ref()?
+        let res = self
+            .inner_ref()?
             .add_columns(transforms, None)
             .await
             .default_error()?;
-        Ok(())
+        Ok(res.into())
     }
 
     #[napi(catch_unwind)]
-    pub async fn alter_columns(&self, alterations: Vec<ColumnAlteration>) -> napi::Result<()> {
+    pub async fn alter_columns(
+        &self,
+        alterations: Vec<ColumnAlteration>,
+    ) -> napi::Result<AlterColumnsResult> {
         for alteration in &alterations {
             if alteration.rename.is_none()
                 && alteration.nullable.is_none()
@@ -221,21 +231,23 @@ impl Table {
             .collect::<std::result::Result<Vec<_>, String>>()
             .map_err(napi::Error::from_reason)?;
 
-        self.inner_ref()?
+        let res = self
+            .inner_ref()?
             .alter_columns(&alterations)
             .await
             .default_error()?;
-        Ok(())
+        Ok(res.into())
     }
 
     #[napi(catch_unwind)]
-    pub async fn drop_columns(&self, columns: Vec<String>) -> napi::Result<()> {
+    pub async fn drop_columns(&self, columns: Vec<String>) -> napi::Result<DropColumnsResult> {
         let col_refs = columns.iter().map(String::as_str).collect::<Vec<_>>();
-        self.inner_ref()?
+        let res = self
+            .inner_ref()?
             .drop_columns(&col_refs)
             .await
             .default_error()?;
-        Ok(())
+        Ok(res.into())
     }
 
     #[napi(catch_unwind)]
@@ -642,6 +654,105 @@ pub struct Version {
     pub metadata: HashMap<String, String>,
 }
 
+#[napi(object)]
+pub struct UpdateResult {
+    pub rows_updated: i64,
+    pub version: i64,
+}
+
+impl From<lancedb::table::UpdateResult> for UpdateResult {
+    fn from(value: lancedb::table::UpdateResult) -> Self {
+        Self {
+            rows_updated: value.rows_updated as i64,
+            version: value.version as i64,
+        }
+    }
+}
+
+#[napi(object)]
+pub struct AddResult {
+    pub version: i64,
+}
+
+impl From<lancedb::table::AddResult> for AddResult {
+    fn from(value: lancedb::table::AddResult) -> Self {
+        Self {
+            version: value.version as i64,
+        }
+    }
+}
+
+#[napi(object)]
+pub struct DeleteResult {
+    pub version: i64,
+}
+
+impl From<lancedb::table::DeleteResult> for DeleteResult {
+    fn from(value: lancedb::table::DeleteResult) -> Self {
+        Self {
+            version: value.version as i64,
+        }
+    }
+}
+
+#[napi(object)]
+pub struct MergeResult {
+    pub version: i64,
+    pub num_inserted_rows: i64,
+    pub num_updated_rows: i64,
+    pub num_deleted_rows: i64,
+}
+
+impl From<lancedb::table::MergeResult> for MergeResult {
+    fn from(value: lancedb::table::MergeResult) -> Self {
+        Self {
+            version: value.version as i64,
+            num_inserted_rows: value.num_inserted_rows as i64,
+            num_updated_rows: value.num_updated_rows as i64,
+            num_deleted_rows: value.num_deleted_rows as i64,
+        }
+    }
+}
+
+#[napi(object)]
+pub struct AddColumnsResult {
+    pub version: i64,
+}
+
+impl From<lancedb::table::AddColumnsResult> for AddColumnsResult {
+    fn from(value: lancedb::table::AddColumnsResult) -> Self {
+        Self {
+            version: value.version as i64,
+        }
+    }
+}
+
+#[napi(object)]
+pub struct AlterColumnsResult {
+    pub version: i64,
+}
+
+impl From<lancedb::table::AlterColumnsResult> for AlterColumnsResult {
+    fn from(value: lancedb::table::AlterColumnsResult) -> Self {
+        Self {
+            version: value.version as i64,
+        }
+    }
+}
+
+#[napi(object)]
+pub struct DropColumnsResult {
+    pub version: i64,
+}
+
+impl From<lancedb::table::DropColumnsResult> for DropColumnsResult {
+    fn from(value: lancedb::table::DropColumnsResult) -> Self {
+        Self {
+            version: value.version as i64,
+        }
+    }
+}
+
 #[napi]
 pub struct TagContents {
     pub version: i64,

python/.bumpversion.toml

@@ -1,5 +1,5 @@
 [tool.bumpversion]
-current_version = "0.22.1-beta.1"
+current_version = "0.24.1-beta.1"
 parse = """(?x)
     (?P<major>0|[1-9]\\d*)\\.
     (?P<minor>0|[1-9]\\d*)\\.

python/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lancedb-python"
-version = "0.22.1-beta.1"
+version = "0.24.1-beta.1"
 edition.workspace = true
 description = "Python bindings for LanceDB"
 license.workspace = true
@@ -14,11 +14,11 @@ name = "_lancedb"
 crate-type = ["cdylib"]
 
 [dependencies]
-arrow = { version = "54.1", features = ["pyarrow"] }
+arrow = { version = "55.1", features = ["pyarrow"] }
 lancedb = { path = "../rust/lancedb", default-features = false }
 env_logger.workspace = true
-pyo3 = { version = "0.23", features = ["extension-module", "abi3-py39"] }
-pyo3-async-runtimes = { version = "0.23", features = [
+pyo3 = { version = "0.24", features = ["extension-module", "abi3-py39"] }
+pyo3-async-runtimes = { version = "0.24", features = [
     "attributes",
     "tokio-runtime",
 ] }
@@ -27,7 +27,7 @@ futures.workspace = true
 tokio = { version = "1.40", features = ["sync"] }
 
 [build-dependencies]
-pyo3-build-config = { version = "0.23", features = [
+pyo3-build-config = { version = "0.24", features = [
     "extension-module",
     "abi3-py39",
 ] }

python/pyproject.toml

@@ -60,6 +60,7 @@ tests = [
     "pyarrow-stubs",
     "pylance>=0.25",
     "requests",
+    "datafusion",
 ]
 dev = [
     "ruff",
@@ -84,7 +85,7 @@ embeddings = [
     "boto3>=1.28.57",
     "awscli>=1.29.57",
     "botocore>=1.31.57",
-    "ollama",
+    "ollama>=0.3.0",
     "ibm-watsonx-ai>=1.1.2",
 ]
 azure = ["adlfs>=2024.2.0"]

python/python/lancedb/_lancedb.pyi

@@ -36,8 +36,10 @@ class Table:
     async def schema(self) -> pa.Schema: ...
     async def add(
         self, data: pa.RecordBatchReader, mode: Literal["append", "overwrite"]
-    ) -> None: ...
-    async def update(self, updates: Dict[str, str], where: Optional[str]) -> None: ...
+    ) -> AddResult: ...
+    async def update(
+        self, updates: Dict[str, str], where: Optional[str]
+    ) -> UpdateResult: ...
     async def count_rows(self, filter: Optional[str]) -> int: ...
     async def create_index(
         self,
@@ -49,12 +51,14 @@ class Table:
     async def version(self) -> int: ...
     async def checkout(self, version: Union[int, str]): ...
     async def checkout_latest(self): ...
-    async def restore(self, version: Optional[int] = None): ...
+    async def restore(self, version: Optional[Union[int, str]] = None): ...
     async def list_indices(self) -> list[IndexConfig]: ...
-    async def delete(self, filter: str): ...
-    async def add_columns(self, columns: list[tuple[str, str]]) -> None: ...
-    async def add_columns_with_schema(self, schema: pa.Schema) -> None: ...
-    async def alter_columns(self, columns: list[dict[str, Any]]) -> None: ...
+    async def delete(self, filter: str) -> DeleteResult: ...
+    async def add_columns(self, columns: list[tuple[str, str]]) -> AddColumnsResult: ...
+    async def add_columns_with_schema(self, schema: pa.Schema) -> AddColumnsResult: ...
+    async def alter_columns(
+        self, columns: list[dict[str, Any]]
+    ) -> AlterColumnsResult: ...
     async def optimize(
         self,
         *,
@@ -139,6 +143,8 @@ class VectorQuery:
     def postfilter(self): ...
     def refine_factor(self, refine_factor: int): ...
     def nprobes(self, nprobes: int): ...
+    def minimum_nprobes(self, minimum_nprobes: int): ...
+    def maximum_nprobes(self, maximum_nprobes: int): ...
     def bypass_vector_index(self): ...
     def nearest_to_text(self, query: dict) -> HybridQuery: ...
     def to_query_request(self) -> PyQueryRequest: ...
@@ -154,6 +160,8 @@ class HybridQuery:
     def distance_type(self, distance_type: str): ...
     def refine_factor(self, refine_factor: int): ...
     def nprobes(self, nprobes: int): ...
+    def minimum_nprobes(self, minimum_nprobes: int): ...
+    def maximum_nprobes(self, maximum_nprobes: int): ...
     def bypass_vector_index(self): ...
     def to_vector_query(self) -> VectorQuery: ...
     def to_fts_query(self) -> FTSQuery: ...
@@ -161,23 +169,21 @@ class HybridQuery:
     def get_with_row_id(self) -> bool: ...
     def to_query_request(self) -> PyQueryRequest: ...
 
-class PyFullTextSearchQuery:
-    columns: Optional[List[str]]
-    query: str
-    limit: Optional[int]
-    wand_factor: Optional[float]
+class FullTextQuery:
+    pass
 
 class PyQueryRequest:
     limit: Optional[int]
     offset: Optional[int]
     filter: Optional[Union[str, bytes]]
-    full_text_search: Optional[PyFullTextSearchQuery]
+    full_text_search: Optional[FullTextQuery]
     select: Optional[Union[str, List[str]]]
     fast_search: Optional[bool]
     with_row_id: Optional[bool]
     column: Optional[str]
     query_vector: Optional[List[pa.Array]]
-    nprobes: Optional[int]
+    minimum_nprobes: Optional[int]
+    maximum_nprobes: Optional[int]
     lower_bound: Optional[float]
     upper_bound: Optional[float]
     ef: Optional[int]
@@ -208,3 +214,28 @@ class OptimizeStats:
 class Tag(TypedDict):
     version: int
     manifest_size: int
+
+class AddResult:
+    version: int
+
+class DeleteResult:
+    version: int
+
+class UpdateResult:
+    rows_updated: int
+    version: int
+
+class MergeResult:
+    version: int
+    num_updated_rows: int
+    num_inserted_rows: int
+    num_deleted_rows: int
+
+class AddColumnsResult:
+    version: int
+
+class AlterColumnsResult:
+    version: int
+
+class DropColumnsResult:
+    version: int

python/python/lancedb/embeddings/ollama.py

@@ -2,14 +2,15 @@
 # SPDX-FileCopyrightText: Copyright The LanceDB Authors
 
 from functools import cached_property
-from typing import TYPE_CHECKING, List, Optional, Union
+from typing import TYPE_CHECKING, List, Optional, Sequence, Union
+
+import numpy as np
 
 from ..util import attempt_import_or_raise
 from .base import TextEmbeddingFunction
 from .registry import register
 
 if TYPE_CHECKING:
-    import numpy as np
     import ollama
 
 
@@ -28,23 +29,21 @@ class OllamaEmbeddings(TextEmbeddingFunction):
     keep_alive: Optional[Union[float, str]] = None
     ollama_client_kwargs: Optional[dict] = {}
 
-    def ndims(self):
+    def ndims(self) -> int:
         return len(self.generate_embeddings(["foo"])[0])
 
-    def _compute_embedding(self, text) -> Union["np.array", None]:
-        return (
-            self._ollama_client.embeddings(
-                model=self.name,
-                prompt=text,
-                options=self.options,
-                keep_alive=self.keep_alive,
-            )["embedding"]
-            or None
+    def _compute_embedding(self, text: Sequence[str]) -> Sequence[Sequence[float]]:
+        response = self._ollama_client.embed(
+            model=self.name,
+            input=text,
+            options=self.options,
+            keep_alive=self.keep_alive,
         )
+        return response.embeddings
 
     def generate_embeddings(
-        self, texts: Union[List[str], "np.ndarray"]
-    ) -> list[Union["np.array", None]]:
+        self, texts: Union[List[str], np.ndarray]
+    ) -> list[Union[np.array, None]]:
         """
         Get the embeddings for the given texts
 
@@ -54,8 +53,8 @@ class OllamaEmbeddings(TextEmbeddingFunction):
             The texts to embed
         """
         # TODO retry, rate limit, token limit
-        embeddings = [self._compute_embedding(text) for text in texts]
-        return embeddings
+        embeddings = self._compute_embedding(texts)
+        return list(embeddings)
 
     @cached_property
     def _ollama_client(self) -> "ollama.Client":

python/python/lancedb/index.py

@@ -102,7 +102,7 @@ class FTS:
 
     Attributes
     ----------
-    with_position : bool, default True
+    with_position : bool, default False
         Whether to store the position of the token in the document. Setting this
         to False can reduce the size of the index and improve indexing speed,
         but it will disable support for phrase queries.
@@ -118,25 +118,25 @@ class FTS:
         ignored.
     lower_case : bool, default True
         Whether to convert the token to lower case. This makes queries case-insensitive.
-    stem : bool, default False
+    stem : bool, default True
         Whether to stem the token. Stemming reduces words to their root form.
         For example, in English "running" and "runs" would both be reduced to "run".
-    remove_stop_words : bool, default False
+    remove_stop_words : bool, default True
         Whether to remove stop words. Stop words are common words that are often
         removed from text before indexing. For example, in English "the" and "and".
-    ascii_folding : bool, default False
+    ascii_folding : bool, default True
         Whether to fold ASCII characters. This converts accented characters to
         their ASCII equivalent. For example, "café" would be converted to "cafe".
     """
 
-    with_position: bool = True
+    with_position: bool = False
     base_tokenizer: Literal["simple", "raw", "whitespace"] = "simple"
     language: str = "English"
     max_token_length: Optional[int] = 40
     lower_case: bool = True
-    stem: bool = False
-    remove_stop_words: bool = False
-    ascii_folding: bool = False
+    stem: bool = True
+    remove_stop_words: bool = True
+    ascii_folding: bool = True
 
 
 @dataclass

python/python/lancedb/merge.py

@@ -4,10 +4,14 @@
 
 from __future__ import annotations
 
+from datetime import timedelta
 from typing import TYPE_CHECKING, List, Optional
 
 if TYPE_CHECKING:
     from .common import DATA
+    from ._lancedb import (
+        MergeInsertResult,
+    )
 
 
 class LanceMergeInsertBuilder(object):
@@ -28,6 +32,7 @@ class LanceMergeInsertBuilder(object):
         self._when_not_matched_insert_all = False
         self._when_not_matched_by_source_delete = False
         self._when_not_matched_by_source_condition = None
+        self._timeout = None
 
     def when_matched_update_all(
         self, *, where: Optional[str] = None
@@ -40,6 +45,16 @@ class LanceMergeInsertBuilder(object):
         If there are multiple matches then the behavior is undefined.
         Currently this causes multiple copies of the row to be created
         but that behavior is subject to change.
+
+        Parameters
+        ----------
+        where : Optional[str], default None
+            A SQL filter expression to apply to matched rows. The filter must
+            specify whether you are referencing the source table (new data) or
+            the target table (existing data) by prefixing column names with
+            "source." or "target." respectively.
+
+            Example: "target.status = 'active'" or "source.price > target.price"
         """
         self._when_matched_update_all = True
         self._when_matched_update_all_condition = where
@@ -78,7 +93,8 @@ class LanceMergeInsertBuilder(object):
         new_data: DATA,
         on_bad_vectors: str = "error",
         fill_value: float = 0.0,
-    ):
+        timeout: Optional[timedelta] = None,
+    ) -> MergeInsertResult:
         """
         Executes the merge insert operation
 
@@ -95,5 +111,24 @@ class LanceMergeInsertBuilder(object):
             One of "error", "drop", "fill".
         fill_value: float, default 0.
             The value to use when filling vectors. Only used if on_bad_vectors="fill".
+        timeout: Optional[timedelta], default None
+            Maximum time to run the operation before cancelling it.
+
+            By default, there is a 30-second timeout that is only enforced after the
+            first attempt. This is to prevent spending too long retrying to resolve
+            conflicts. For example, if a write attempt takes 20 seconds and fails,
+            the second attempt will be cancelled after 10 seconds, hitting the
+            30-second timeout. However, a write that takes one hour and succeeds on the
+            first attempt will not be cancelled.
+
+            When this is set, the timeout is enforced on all attempts, including
+            the first.
+
+        Returns
+        -------
+        MergeInsertResult
+            version: the new version number of the table after doing merge insert.
         """
+        if timeout is not None:
+            self._timeout = timeout
         return self._table._do_merge(self, new_data, on_bad_vectors, fill_value)

python/python/lancedb/pydantic.py

@@ -415,6 +415,7 @@ class LanceModel(pydantic.BaseModel):
     >>> table.add([
     ...     TestModel(name="test", vector=[1.0, 2.0])
     ... ])
+    AddResult(version=2)
     >>> table.search([0., 0.]).limit(1).to_pydantic(TestModel)
     [TestModel(name='test', vector=FixedSizeList(dim=2))]
     """

python/python/lancedb/query.py

@@ -4,7 +4,6 @@
 from __future__ import annotations
 
 from abc import ABC, abstractmethod
-import abc
 from concurrent.futures import ThreadPoolExecutor
 from enum import Enum
 from datetime import timedelta
@@ -88,15 +87,28 @@ def ensure_vector_query(
         return val
 
 
-class FullTextQueryType(Enum):
+class FullTextQueryType(str, Enum):
     MATCH = "match"
     MATCH_PHRASE = "match_phrase"
     BOOST = "boost"
     MULTI_MATCH = "multi_match"
+    BOOLEAN = "boolean"
 
 
-class FullTextQuery(abc.ABC, pydantic.BaseModel):
-    @abc.abstractmethod
+class FullTextOperator(str, Enum):
+    AND = "AND"
+    OR = "OR"
+
+
+class Occur(str, Enum):
+    SHOULD = "SHOULD"
+    MUST = "MUST"
+    MUST_NOT = "MUST_NOT"
+
+
+@pydantic.dataclasses.dataclass
+class FullTextQuery(ABC):
+    @abstractmethod
     def query_type(self) -> FullTextQueryType:
         """
         Get the query type of the query.
@@ -106,193 +118,178 @@ class FullTextQuery(abc.ABC, pydantic.BaseModel):
         str
             The type of the query.
         """
+        pass
 
-    @abc.abstractmethod
-    def to_dict(self) -> dict:
+    def __and__(self, other: "FullTextQuery") -> "FullTextQuery":
         """
-        Convert the query to a dictionary.
-
-        Returns
-        -------
-        dict
-            The query as a dictionary.
-        """
-
-
-class MatchQuery(FullTextQuery):
-    query: str
-    column: str
-    boost: float = 1.0
-    fuzziness: int = 0
-    max_expansions: int = 50
-
-    def __init__(
-        self,
-        query: str,
-        column: str,
-        *,
-        boost: float = 1.0,
-        fuzziness: int = 0,
-        max_expansions: int = 50,
-    ):
-        """
-        Match query for full-text search.
+        Combine two queries with a logical AND operation.
 
         Parameters
         ----------
-        query : str
-            The query string to match against.
-        column : str
-            The name of the column to match against.
-        boost : float, default 1.0
-            The boost factor for the query.
-            The score of each matching document is multiplied by this value.
-        fuzziness : int, optional
-            The maximum edit distance for each term in the match query.
-            Defaults to 0 (exact match).
-            If None, fuzziness is applied automatically by the rules:
-                - 0 for terms with length <= 2
-                - 1 for terms with length <= 5
-                - 2 for terms with length > 5
-        max_expansions : int, optional
-            The maximum number of terms to consider for fuzzy matching.
-            Defaults to 50.
+        other : FullTextQuery
+            The other query to combine with.
+
+        Returns
+        -------
+        FullTextQuery
+            A new query that combines both queries with AND.
         """
-        super().__init__(
-            query=query,
-            column=column,
-            boost=boost,
-            fuzziness=fuzziness,
-            max_expansions=max_expansions,
-        )
+        return BooleanQuery([(Occur.MUST, self), (Occur.MUST, other)])
+
+    def __or__(self, other: "FullTextQuery") -> "FullTextQuery":
+        """
+        Combine two queries with a logical OR operation.
+
+        Parameters
+        ----------
+        other : FullTextQuery
+            The other query to combine with.
+
+        Returns
+        -------
+        FullTextQuery
+            A new query that combines both queries with OR.
+        """
+        return BooleanQuery([(Occur.SHOULD, self), (Occur.SHOULD, other)])
+
+
+@pydantic.dataclasses.dataclass
+class MatchQuery(FullTextQuery):
+    """
+    Match query for full-text search.
+
+    Parameters
+    ----------
+    query : str
+        The query string to match against.
+    column : str
+        The name of the column to match against.
+    boost : float, default 1.0
+        The boost factor for the query.
+        The score of each matching document is multiplied by this value.
+    fuzziness : int, optional
+        The maximum edit distance for each term in the match query.
+        Defaults to 0 (exact match).
+        If None, fuzziness is applied automatically by the rules:
+            - 0 for terms with length <= 2
+            - 1 for terms with length <= 5
+            - 2 for terms with length > 5
+    max_expansions : int, optional
+        The maximum number of terms to consider for fuzzy matching.
+        Defaults to 50.
+    operator : FullTextOperator, default OR
+        The operator to use for combining the query results.
+        Can be either `AND` or `OR`.
+        If `AND`, all terms in the query must match.
+        If `OR`, at least one term in the query must match.
+    prefix_length : int, optional
+        The number of beginning characters being unchanged for fuzzy matching.
+        This is useful to achieve prefix matching.
+    """
+
+    query: str
+    column: str
+    boost: float = pydantic.Field(1.0, kw_only=True)
+    fuzziness: int = pydantic.Field(0, kw_only=True)
+    max_expansions: int = pydantic.Field(50, kw_only=True)
+    operator: FullTextOperator = pydantic.Field(FullTextOperator.OR, kw_only=True)
+    prefix_length: int = pydantic.Field(0, kw_only=True)
 
     def query_type(self) -> FullTextQueryType:
         return FullTextQueryType.MATCH
 
-    def to_dict(self) -> dict:
-        return {
-            "match": {
-                self.column: {
-                    "query": self.query,
-                    "boost": self.boost,
-                    "fuzziness": self.fuzziness,
-                    "max_expansions": self.max_expansions,
-                }
-            }
-        }
-
 
+@pydantic.dataclasses.dataclass
 class PhraseQuery(FullTextQuery):
+    """
+    Phrase query for full-text search.
+
+    Parameters
+    ----------
+    query : str
+        The query string to match against.
+    column : str
+        The name of the column to match against.
+    """
+
     query: str
     column: str
-
-    def __init__(self, query: str, column: str):
-        """
-        Phrase query for full-text search.
-
-        Parameters
-        ----------
-        query : str
-            The query string to match against.
-        column : str
-            The name of the column to match against.
-        """
-        super().__init__(query=query, column=column)
+    slop: int = pydantic.Field(0, kw_only=True)
 
     def query_type(self) -> FullTextQueryType:
         return FullTextQueryType.MATCH_PHRASE
 
-    def to_dict(self) -> dict:
-        return {
-            "match_phrase": {
-                self.column: self.query,
-            }
-        }
-
 
+@pydantic.dataclasses.dataclass
 class BoostQuery(FullTextQuery):
+    """
+    Boost query for full-text search.
+
+    Parameters
+    ----------
+    positive : dict
+        The positive query object.
+    negative : dict
+        The negative query object.
+    negative_boost : float, default 0.5
+        The boost factor for the negative query.
+    """
+
     positive: FullTextQuery
     negative: FullTextQuery
-    negative_boost: float = 0.5
-
-    def __init__(
-        self,
-        positive: FullTextQuery,
-        negative: FullTextQuery,
-        *,
-        negative_boost: float = 0.5,
-    ):
-        """
-        Boost query for full-text search.
-
-        Parameters
-        ----------
-        positive : dict
-            The positive query object.
-        negative : dict
-            The negative query object.
-        negative_boost : float
-            The boost factor for the negative query.
-        """
-        super().__init__(
-            positive=positive, negative=negative, negative_boost=negative_boost
-        )
+    negative_boost: float = pydantic.Field(0.5, kw_only=True)
 
     def query_type(self) -> FullTextQueryType:
         return FullTextQueryType.BOOST
 
-    def to_dict(self) -> dict:
-        return {
-            "boost": {
-                "positive": self.positive.to_dict(),
-                "negative": self.negative.to_dict(),
-                "negative_boost": self.negative_boost,
-            }
-        }
-
 
+@pydantic.dataclasses.dataclass
 class MultiMatchQuery(FullTextQuery):
+    """
+    Multi-match query for full-text search.
+
+    Parameters
+    ----------
+    query : str | list[Query]
+        If a string, the query string to match against.
+    columns : list[str]
+        The list of columns to match against.
+    boosts : list[float], optional
+        The list of boost factors for each column. If not provided,
+        all columns will have the same boost factor.
+    operator : FullTextOperator, default OR
+        The operator to use for combining the query results.
+        Can be either `AND` or `OR`.
+        It would be applied to all columns individually.
+        For example, if the operator is `AND`,
+        then the query "hello world" is equal to
+        `match("hello AND world", column1) OR match("hello AND world", column2)`.
+    """
+
     query: str
     columns: list[str]
-    boosts: list[float]
-
-    def __init__(
-        self,
-        query: str,
-        columns: list[str],
-        *,
-        boosts: Optional[list[float]] = None,
-    ):
-        """
-        Multi-match query for full-text search.
-
-        Parameters
-        ----------
-        query : str
-            The query string to match against.
-
-        columns : list[str]
-            The list of columns to match against.
-
-        boosts : list[float], optional
-            The list of boost factors for each column. If not provided,
-            all columns will have the same boost factor.
-        """
-        if boosts is None:
-            boosts = [1.0] * len(columns)
-        super().__init__(query=query, columns=columns, boosts=boosts)
+    boosts: Optional[list[float]] = pydantic.Field(None, kw_only=True)
+    operator: FullTextOperator = pydantic.Field(FullTextOperator.OR, kw_only=True)
 
     def query_type(self) -> FullTextQueryType:
         return FullTextQueryType.MULTI_MATCH
 
-    def to_dict(self) -> dict:
-        return {
-            "multi_match": {
-                "query": self.query,
-                "columns": self.columns,
-                "boost": self.boosts,
-            }
-        }
+
+@pydantic.dataclasses.dataclass
+class BooleanQuery(FullTextQuery):
+    """
+    Boolean query for full-text search.
+
+    Parameters
+    ----------
+    queries : list[tuple(Occur, FullTextQuery)]
+        The list of queries with their occurrence requirements.
+    """
+
+    queries: list[tuple[Occur, FullTextQuery]]
+
+    def query_type(self) -> FullTextQueryType:
+        return FullTextQueryType.BOOLEAN
 
 
 class FullTextSearchQuery(pydantic.BaseModel):
@@ -445,8 +442,18 @@ class Query(pydantic.BaseModel):
     # which columns to return in the results
     columns: Optional[Union[List[str], Dict[str, str]]] = None
 
-    # number of IVF partitions to search
-    nprobes: Optional[int] = None
+    # minimum number of IVF partitions to search
+    #
+    # If None then a default value (20) will be used.
+    minimum_nprobes: Optional[int] = None
+
+    # maximum number of IVF partitions to search
+    #
+    # If None then a default value (20) will be used.
+    #
+    # If 0 then no limit will be applied and all partitions could be searched
+    # if needed to satisfy the limit.
+    maximum_nprobes: Optional[int] = None
 
     # lower bound for distance search
     lower_bound: Optional[float] = None
@@ -484,7 +491,8 @@ class Query(pydantic.BaseModel):
         query.vector_column = req.column
         query.vector = req.query_vector
         query.distance_type = req.distance_type
-        query.nprobes = req.nprobes
+        query.minimum_nprobes = req.minimum_nprobes
+        query.maximum_nprobes = req.maximum_nprobes
         query.lower_bound = req.lower_bound
         query.upper_bound = req.upper_bound
         query.ef = req.ef
@@ -493,10 +501,8 @@ class Query(pydantic.BaseModel):
         query.postfilter = req.postfilter
         if req.full_text_search is not None:
             query.full_text_query = FullTextSearchQuery(
-                columns=req.full_text_search.columns,
-                query=req.full_text_search.query,
-                limit=req.full_text_search.limit,
-                wand_factor=req.full_text_search.wand_factor,
+                columns=None,
+                query=req.full_text_search,
             )
         return query
 
@@ -1047,7 +1053,8 @@ class LanceVectorQueryBuilder(LanceQueryBuilder):
         super().__init__(table)
         self._query = query
         self._distance_type = None
-        self._nprobes = None
+        self._minimum_nprobes = None
+        self._maximum_nprobes = None
         self._lower_bound = None
         self._upper_bound = None
         self._refine_factor = None
@@ -1110,6 +1117,10 @@ class LanceVectorQueryBuilder(LanceQueryBuilder):
         See discussion in [Querying an ANN Index][querying-an-ann-index] for
         tuning advice.
 
+        This method sets both the minimum and maximum number of probes to the same
+        value. See `minimum_nprobes` and `maximum_nprobes` for more fine-grained
+        control.
+
         Parameters
         ----------
         nprobes: int
@@ -1120,7 +1131,36 @@ class LanceVectorQueryBuilder(LanceQueryBuilder):
         LanceVectorQueryBuilder
             The LanceQueryBuilder object.
         """
-        self._nprobes = nprobes
+        self._minimum_nprobes = nprobes
+        self._maximum_nprobes = nprobes
+        return self
+
+    def minimum_nprobes(self, minimum_nprobes: int) -> LanceVectorQueryBuilder:
+        """Set the minimum number of probes to use.
+
+        See `nprobes` for more details.
+
+        These partitions will be searched on every vector query and will increase recall
+        at the expense of latency.
+        """
+        self._minimum_nprobes = minimum_nprobes
+        return self
+
+    def maximum_nprobes(self, maximum_nprobes: int) -> LanceVectorQueryBuilder:
+        """Set the maximum number of probes to use.
+
+        See `nprobes` for more details.
+
+        If this value is greater than `minimum_nprobes` then the excess partitions
+        will be searched only if we have not found enough results.
+
+        This can be useful when there is a narrow filter to allow these queries to
+        spend more time searching and avoid potential false negatives.
+
+        If this value is 0 then no limit will be applied and all partitions could be
+        searched if needed to satisfy the limit.
+        """
+        self._maximum_nprobes = maximum_nprobes
         return self
 
     def distance_range(
@@ -1224,7 +1264,8 @@ class LanceVectorQueryBuilder(LanceQueryBuilder):
             limit=self._limit,
             distance_type=self._distance_type,
             columns=self._columns,
-            nprobes=self._nprobes,
+            minimum_nprobes=self._minimum_nprobes,
+            maximum_nprobes=self._maximum_nprobes,
             lower_bound=self._lower_bound,
             upper_bound=self._upper_bound,
             refine_factor=self._refine_factor,
@@ -1410,10 +1451,13 @@ class LanceFtsQueryBuilder(LanceQueryBuilder):
 
         query = self._query
         if self._phrase_query:
-            raise NotImplementedError(
-                "Phrase query is not yet supported in Lance FTS. "
-                "Use tantivy-based index instead for now."
-            )
+            if isinstance(query, str):
+                if not query.startswith('"') or not query.endswith('"'):
+                    query = f'"{query}"'
+            elif isinstance(query, FullTextQuery) and not isinstance(
+                query, PhraseQuery
+            ):
+                raise TypeError("Please use PhraseQuery for phrase queries.")
         query = self.to_query_object()
         results = self._table._execute_query(query, timeout=timeout)
         results = results.read_all()
@@ -1588,7 +1632,8 @@ class LanceHybridQueryBuilder(LanceQueryBuilder):
         self._fts_columns = fts_columns
         self._norm = None
         self._reranker = None
-        self._nprobes = None
+        self._minimum_nprobes = None
+        self._maximum_nprobes = None
         self._refine_factor = None
         self._distance_type = None
         self._phrase_query = None
@@ -1820,7 +1865,24 @@ class LanceHybridQueryBuilder(LanceQueryBuilder):
         LanceHybridQueryBuilder
             The LanceHybridQueryBuilder object.
         """
-        self._nprobes = nprobes
+        self._minimum_nprobes = nprobes
+        self._maximum_nprobes = nprobes
+        return self
+
+    def minimum_nprobes(self, minimum_nprobes: int) -> LanceHybridQueryBuilder:
+        """Set the minimum number of probes to use.
+
+        See `nprobes` for more details.
+        """
+        self._minimum_nprobes = minimum_nprobes
+        return self
+
+    def maximum_nprobes(self, maximum_nprobes: int) -> LanceHybridQueryBuilder:
+        """Set the maximum number of probes to use.
+
+        See `nprobes` for more details.
+        """
+        self._maximum_nprobes = maximum_nprobes
         return self
 
     def distance_range(
@@ -2049,8 +2111,10 @@ class LanceHybridQueryBuilder(LanceQueryBuilder):
             self._fts_query.phrase_query(True)
         if self._distance_type:
             self._vector_query.metric(self._distance_type)
-        if self._nprobes:
-            self._vector_query.nprobes(self._nprobes)
+        if self._minimum_nprobes:
+            self._vector_query.minimum_nprobes(self._minimum_nprobes)
+        if self._maximum_nprobes is not None:
+            self._vector_query.maximum_nprobes(self._maximum_nprobes)
         if self._refine_factor:
             self._vector_query.refine_factor(self._refine_factor)
         if self._ef:
@@ -2513,7 +2577,7 @@ class AsyncQuery(AsyncQueryBase):
                 self._inner.nearest_to_text({"query": query, "columns": columns})
             )
         # FullTextQuery object
-        return AsyncFTSQuery(self._inner.nearest_to_text({"query": query.to_dict()}))
+        return AsyncFTSQuery(self._inner.nearest_to_text({"query": query}))
 
 
 class AsyncFTSQuery(AsyncQueryBase):
@@ -2661,6 +2725,34 @@ class AsyncVectorQueryBase:
         self._inner.nprobes(nprobes)
         return self
 
+    def minimum_nprobes(self, minimum_nprobes: int) -> Self:
+        """Set the minimum number of probes to use.
+
+        See `nprobes` for more details.
+
+        These partitions will be searched on every indexed vector query and will
+        increase recall at the expense of latency.
+        """
+        self._inner.minimum_nprobes(minimum_nprobes)
+        return self
+
+    def maximum_nprobes(self, maximum_nprobes: int) -> Self:
+        """Set the maximum number of probes to use.
+
+        See `nprobes` for more details.
+
+        If this value is greater than `minimum_nprobes` then the excess partitions
+        will be searched only if we have not found enough results.
+
+        This can be useful when there is a narrow filter to allow these queries to
+        spend more time searching and avoid potential false negatives.
+
+        If this value is 0 then no limit will be applied and all partitions could be
+        searched if needed to satisfy the limit.
+        """
+        self._inner.maximum_nprobes(maximum_nprobes)
+        return self
+
     def distance_range(
         self, lower_bound: Optional[float] = None, upper_bound: Optional[float] = None
     ) -> Self:
@@ -2835,7 +2927,7 @@ class AsyncVectorQuery(AsyncQueryBase, AsyncVectorQueryBase):
                 self._inner.nearest_to_text({"query": query, "columns": columns})
             )
         # FullTextQuery object
-        return AsyncHybridQuery(self._inner.nearest_to_text({"query": query.to_dict()}))
+        return AsyncHybridQuery(self._inner.nearest_to_text({"query": query}))
 
     async def to_batches(
         self,
@@ -2950,15 +3042,21 @@ class AsyncHybridQuery(AsyncQueryBase, AsyncVectorQueryBase):
         >>> asyncio.run(doctest_example()) # doctest: +ELLIPSIS, +NORMALIZE_WHITESPACE
         Vector Search Plan:
         ProjectionExec: expr=[vector@0 as vector, text@3 as text, _distance@2 as _distance]
-            Take: columns="vector, _rowid, _distance, (text)"
-                CoalesceBatchesExec: target_batch_size=1024
-                GlobalLimitExec: skip=0, fetch=10
-                    FilterExec: _distance@2 IS NOT NULL
-                    SortExec: TopK(fetch=10), expr=[_distance@2 ASC NULLS LAST], preserve_partitioning=[false]
-                        KNNVectorDistance: metric=l2
-                        LanceScan: uri=..., projection=[vector], row_id=true, row_addr=false, ordered=false
+          Take: columns="vector, _rowid, _distance, (text)"
+            CoalesceBatchesExec: target_batch_size=1024
+              GlobalLimitExec: skip=0, fetch=10
+                FilterExec: _distance@2 IS NOT NULL
+                  SortExec: TopK(fetch=10), expr=[_distance@2 ASC NULLS LAST], preserve_partitioning=[false]
+                    KNNVectorDistance: metric=l2
+                      LanceScan: uri=..., projection=[vector], row_id=true, row_addr=false, ordered=false
+        <BLANKLINE>
         FTS Search Plan:
-        LanceScan: uri=..., projection=[vector, text], row_id=false, row_addr=false, ordered=true
+        ProjectionExec: expr=[vector@2 as vector, text@3 as text, _score@1 as _score]
+          Take: columns="_rowid, _score, (vector), (text)"
+            CoalesceBatchesExec: target_batch_size=1024
+              GlobalLimitExec: skip=0, fetch=10
+                MatchQuery: query=hello
+        <BLANKLINE>
 
         Parameters
         ----------

python/python/lancedb/remote/table.py

@@ -7,9 +7,18 @@ from functools import cached_property
 from typing import Dict, Iterable, List, Optional, Union, Literal
 import warnings
 
-from lancedb._lancedb import IndexConfig
+from lancedb._lancedb import (
+    AddColumnsResult,
+    AddResult,
+    AlterColumnsResult,
+    DeleteResult,
+    DropColumnsResult,
+    IndexConfig,
+    MergeResult,
+    UpdateResult,
+)
 from lancedb.embeddings.base import EmbeddingFunctionConfig
-from lancedb.index import FTS, BTree, Bitmap, HnswPq, HnswSq, IvfFlat, IvfPq, LabelList
+from lancedb.index import FTS, BTree, Bitmap, HnswSq, IvfFlat, IvfPq, LabelList
 from lancedb.remote.db import LOOP
 import pyarrow as pa
 
@@ -38,9 +47,6 @@ class RemoteTable(Table):
     def __repr__(self) -> str:
         return f"RemoteTable({self.db_name}.{self.name})"
 
-    def __len__(self) -> int:
-        self.count_rows(None)
-
     @property
     def schema(self) -> pa.Schema:
         """The [Arrow Schema](https://arrow.apache.org/docs/python/api/datatypes.html#)
@@ -83,7 +89,7 @@ class RemoteTable(Table):
 
     def to_pandas(self):
         """to_pandas() is not yet supported on LanceDB cloud."""
-        return NotImplementedError("to_pandas() is not yet supported on LanceDB cloud.")
+        raise NotImplementedError("to_pandas() is not yet supported on LanceDB cloud.")
 
     def checkout(self, version: Union[int, str]):
         return LOOP.run(self._table.checkout(version))
@@ -91,7 +97,7 @@ class RemoteTable(Table):
     def checkout_latest(self):
         return LOOP.run(self._table.checkout_latest())
 
-    def restore(self, version: Optional[int] = None):
+    def restore(self, version: Optional[Union[int, str]] = None):
         return LOOP.run(self._table.restore(version))
 
     def list_indices(self) -> Iterable[IndexConfig]:
@@ -143,15 +149,15 @@ class RemoteTable(Table):
         *,
         replace: bool = False,
         wait_timeout: timedelta = None,
-        with_position: bool = True,
+        with_position: bool = False,
         # tokenizer configs:
         base_tokenizer: str = "simple",
         language: str = "English",
         max_token_length: Optional[int] = 40,
         lower_case: bool = True,
-        stem: bool = False,
-        remove_stop_words: bool = False,
-        ascii_folding: bool = False,
+        stem: bool = True,
+        remove_stop_words: bool = True,
+        ascii_folding: bool = True,
     ):
         config = FTS(
             with_position=with_position,
@@ -180,6 +186,8 @@ class RemoteTable(Table):
         accelerator: Optional[str] = None,
         index_type="vector",
         wait_timeout: Optional[timedelta] = None,
+        *,
+        num_bits: int = 8,
     ):
         """Create an index on the table.
         Currently, the only parameters that matter are
@@ -214,11 +222,6 @@ class RemoteTable(Table):
         >>> table.create_index("l2", "vector") # doctest: +SKIP
         """
 
-        if num_partitions is not None:
-            logging.warning(
-                "num_partitions is not supported on LanceDB cloud."
-                "This parameter will be tuned automatically."
-            )
         if num_sub_vectors is not None:
             logging.warning(
                 "num_sub_vectors is not supported on LanceDB cloud."
@@ -238,13 +241,21 @@ class RemoteTable(Table):
 
         index_type = index_type.upper()
         if index_type == "VECTOR" or index_type == "IVF_PQ":
-            config = IvfPq(distance_type=metric)
+            config = IvfPq(
+                distance_type=metric,
+                num_partitions=num_partitions,
+                num_sub_vectors=num_sub_vectors,
+                num_bits=num_bits,
+            )
         elif index_type == "IVF_HNSW_PQ":
-            config = HnswPq(distance_type=metric)
+            raise ValueError(
+                "IVF_HNSW_PQ is not supported on LanceDB cloud."
+                "Please use IVF_HNSW_SQ instead."
+            )
         elif index_type == "IVF_HNSW_SQ":
-            config = HnswSq(distance_type=metric)
+            config = HnswSq(distance_type=metric, num_partitions=num_partitions)
         elif index_type == "IVF_FLAT":
-            config = IvfFlat(distance_type=metric)
+            config = IvfFlat(distance_type=metric, num_partitions=num_partitions)
         else:
             raise ValueError(
                 f"Unknown vector index type: {index_type}. Valid options are"
@@ -263,7 +274,7 @@ class RemoteTable(Table):
         mode: str = "append",
         on_bad_vectors: str = "error",
         fill_value: float = 0.0,
-    ) -> int:
+    ) -> AddResult:
         """Add more data to the [Table](Table). It has the same API signature as
         the OSS version.
 
@@ -286,8 +297,12 @@ class RemoteTable(Table):
         fill_value: float, default 0.
             The value to use when filling vectors. Only used if on_bad_vectors="fill".
 
+        Returns
+        -------
+        AddResult
+            An object containing the new version number of the table after adding data.
         """
-        LOOP.run(
+        return LOOP.run(
             self._table.add(
                 data, mode=mode, on_bad_vectors=on_bad_vectors, fill_value=fill_value
             )
@@ -413,10 +428,12 @@ class RemoteTable(Table):
         new_data: DATA,
         on_bad_vectors: str,
         fill_value: float,
-    ):
-        LOOP.run(self._table._do_merge(merge, new_data, on_bad_vectors, fill_value))
+    ) -> MergeResult:
+        return LOOP.run(
+            self._table._do_merge(merge, new_data, on_bad_vectors, fill_value)
+        )
 
-    def delete(self, predicate: str):
+    def delete(self, predicate: str) -> DeleteResult:
         """Delete rows from the table.
 
         This can be used to delete a single row, many rows, all rows, or
@@ -431,6 +448,11 @@ class RemoteTable(Table):
 
             The filter must not be empty, or it will error.
 
+        Returns
+        -------
+        DeleteResult
+            An object containing the new version number of the table after deletion.
+
         Examples
         --------
         >>> import lancedb
@@ -463,7 +485,7 @@ class RemoteTable(Table):
            x      vector  _distance # doctest: +SKIP
         0  2  [3.0, 4.0]       85.0 # doctest: +SKIP
         """
-        LOOP.run(self._table.delete(predicate))
+        return LOOP.run(self._table.delete(predicate))
 
     def update(
         self,
@@ -471,7 +493,7 @@ class RemoteTable(Table):
         values: Optional[dict] = None,
         *,
         values_sql: Optional[Dict[str, str]] = None,
-    ):
+    ) -> UpdateResult:
         """
         This can be used to update zero to all rows depending on how many
         rows match the where clause.
@@ -489,6 +511,12 @@ class RemoteTable(Table):
             reference existing columns. For example, {"x": "x + 1"} will increment
             the x column by 1.
 
+        Returns
+        -------
+        UpdateResult
+            - rows_updated: The number of rows that were updated
+            - version: The new version number of the table after the update
+
         Examples
         --------
         >>> import lancedb
@@ -513,7 +541,7 @@ class RemoteTable(Table):
         2  2  [10.0, 10.0] # doctest: +SKIP
 
         """
-        LOOP.run(
+        return LOOP.run(
             self._table.update(where=where, updates=values, updates_sql=values_sql)
         )
 
@@ -561,13 +589,15 @@ class RemoteTable(Table):
     def count_rows(self, filter: Optional[str] = None) -> int:
         return LOOP.run(self._table.count_rows(filter))
 
-    def add_columns(self, transforms: Dict[str, str]):
+    def add_columns(self, transforms: Dict[str, str]) -> AddColumnsResult:
         return LOOP.run(self._table.add_columns(transforms))
 
-    def alter_columns(self, *alterations: Iterable[Dict[str, str]]):
+    def alter_columns(
+        self, *alterations: Iterable[Dict[str, str]]
+    ) -> AlterColumnsResult:
         return LOOP.run(self._table.alter_columns(*alterations))
 
-    def drop_columns(self, columns: Iterable[str]):
+    def drop_columns(self, columns: Iterable[str]) -> DropColumnsResult:
         return LOOP.run(self._table.drop_columns(columns))
 
     def drop_index(self, index_name: str):

python/python/lancedb/table.py

@@ -78,6 +78,13 @@ if TYPE_CHECKING:
         CleanupStats,
         CompactionStats,
         Tag,
+        AddColumnsResult,
+        AddResult,
+        AlterColumnsResult,
+        DeleteResult,
+        DropColumnsResult,
+        MergeResult,
+        UpdateResult,
     )
     from .db import LanceDBConnection
     from .index import IndexConfig
@@ -550,6 +557,7 @@ class Table(ABC):
     Can append new data with [Table.add()][lancedb.table.Table.add].
 
     >>> table.add([{"vector": [0.5, 1.3], "b": 4}])
+    AddResult(version=2)
 
     Can query the table with [Table.search][lancedb.table.Table.search].
 
@@ -612,6 +620,10 @@ class Table(ABC):
         """
         raise NotImplementedError
 
+    def __len__(self) -> int:
+        """The number of rows in this Table"""
+        return self.count_rows(None)
+
     @property
     @abstractmethod
     def embedding_functions(self) -> Dict[str, EmbeddingFunctionConfig]:
@@ -815,17 +827,17 @@ class Table(ABC):
         ordering_field_names: Optional[Union[str, List[str]]] = None,
         replace: bool = False,
         writer_heap_size: Optional[int] = 1024 * 1024 * 1024,
-        use_tantivy: bool = True,
+        use_tantivy: bool = False,
         tokenizer_name: Optional[str] = None,
-        with_position: bool = True,
+        with_position: bool = False,
         # tokenizer configs:
         base_tokenizer: BaseTokenizerType = "simple",
         language: str = "English",
         max_token_length: Optional[int] = 40,
         lower_case: bool = True,
-        stem: bool = False,
-        remove_stop_words: bool = False,
-        ascii_folding: bool = False,
+        stem: bool = True,
+        remove_stop_words: bool = True,
+        ascii_folding: bool = True,
         wait_timeout: Optional[timedelta] = None,
     ):
         """Create a full-text search index on the table.
@@ -852,10 +864,10 @@ class Table(ABC):
             The tokenizer to use for the index. Can be "raw", "default" or the 2 letter
             language code followed by "_stem". So for english it would be "en_stem".
             For available languages see: https://docs.rs/tantivy/latest/tantivy/tokenizer/enum.Language.html
-        use_tantivy: bool, default True
+        use_tantivy: bool, default False
             If True, use the legacy full-text search implementation based on tantivy.
             If False, use the new full-text search implementation based on lance-index.
-        with_position: bool, default True
+        with_position: bool, default False
             Only available with use_tantivy=False
             If False, do not store the positions of the terms in the text.
             This can reduce the size of the index and improve indexing speed.
@@ -873,13 +885,13 @@ class Table(ABC):
         lower_case : bool, default True
             Whether to convert the token to lower case. This makes queries
             case-insensitive.
-        stem : bool, default False
+        stem : bool, default True
             Whether to stem the token. Stemming reduces words to their root form.
             For example, in English "running" and "runs" would both be reduced to "run".
-        remove_stop_words : bool, default False
+        remove_stop_words : bool, default True
             Whether to remove stop words. Stop words are common words that are often
             removed from text before indexing. For example, in English "the" and "and".
-        ascii_folding : bool, default False
+        ascii_folding : bool, default True
             Whether to fold ASCII characters. This converts accented characters to
             their ASCII equivalent. For example, "café" would be converted to "cafe".
         wait_timeout: timedelta, optional
@@ -894,7 +906,7 @@ class Table(ABC):
         mode: AddMode = "append",
         on_bad_vectors: OnBadVectorsType = "error",
         fill_value: float = 0.0,
-    ):
+    ) -> AddResult:
         """Add more data to the [Table](Table).
 
         Parameters
@@ -916,6 +928,10 @@ class Table(ABC):
         fill_value: float, default 0.
             The value to use when filling vectors. Only used if on_bad_vectors="fill".
 
+        Returns
+        -------
+        AddResult
+            An object containing the new version number of the table after adding data.
         """
         raise NotImplementedError
 
@@ -962,12 +978,12 @@ class Table(ABC):
         >>> table = db.create_table("my_table", data)
         >>> new_data = pa.table({"a": [2, 3, 4], "b": ["x", "y", "z"]})
         >>> # Perform a "upsert" operation
-        >>> stats = table.merge_insert("a")     \\
+        >>> res = table.merge_insert("a")     \\
         ...      .when_matched_update_all()     \\
         ...      .when_not_matched_insert_all() \\
         ...      .execute(new_data)
-        >>> stats
-        {'num_inserted_rows': 1, 'num_updated_rows': 2, 'num_deleted_rows': 0}
+        >>> res
+        MergeResult(version=2, num_updated_rows=2, num_inserted_rows=1, num_deleted_rows=0)
         >>> # The order of new rows is non-deterministic since we use
         >>> # a hash-join as part of this operation and so we sort here
         >>> table.to_arrow().sort_by("a").to_pandas()
@@ -976,7 +992,7 @@ class Table(ABC):
         1  2  x
         2  3  y
         3  4  z
-        """
+        """  # noqa: E501
         on = [on] if isinstance(on, str) else list(iter(on))
 
         return LanceMergeInsertBuilder(self, on)
@@ -1091,10 +1107,10 @@ class Table(ABC):
         new_data: DATA,
         on_bad_vectors: OnBadVectorsType,
         fill_value: float,
-    ): ...
+    ) -> MergeResult: ...
 
     @abstractmethod
-    def delete(self, where: str):
+    def delete(self, where: str) -> DeleteResult:
         """Delete rows from the table.
 
         This can be used to delete a single row, many rows, all rows, or
@@ -1109,6 +1125,11 @@ class Table(ABC):
 
             The filter must not be empty, or it will error.
 
+        Returns
+        -------
+        DeleteResult
+            An object containing the new version number of the table after deletion.
+
         Examples
         --------
         >>> import lancedb
@@ -1125,6 +1146,7 @@ class Table(ABC):
         1  2  [3.0, 4.0]
         2  3  [5.0, 6.0]
         >>> table.delete("x = 2")
+        DeleteResult(version=2)
         >>> table.to_pandas()
            x      vector
         0  1  [1.0, 2.0]
@@ -1138,6 +1160,7 @@ class Table(ABC):
         >>> to_remove
         '1, 5'
         >>> table.delete(f"x IN ({to_remove})")
+        DeleteResult(version=3)
         >>> table.to_pandas()
            x      vector
         0  3  [5.0, 6.0]
@@ -1151,7 +1174,7 @@ class Table(ABC):
         values: Optional[dict] = None,
         *,
         values_sql: Optional[Dict[str, str]] = None,
-    ):
+    ) -> UpdateResult:
         """
         This can be used to update zero to all rows depending on how many
         rows match the where clause. If no where clause is provided, then
@@ -1173,6 +1196,12 @@ class Table(ABC):
             reference existing columns. For example, {"x": "x + 1"} will increment
             the x column by 1.
 
+        Returns
+        -------
+        UpdateResult
+            - rows_updated: The number of rows that were updated
+            - version: The new version number of the table after the update
+
         Examples
         --------
         >>> import lancedb
@@ -1186,12 +1215,14 @@ class Table(ABC):
         1  2  [3.0, 4.0]
         2  3  [5.0, 6.0]
         >>> table.update(where="x = 2", values={"vector": [10.0, 10]})
+        UpdateResult(rows_updated=1, version=2)
         >>> table.to_pandas()
            x        vector
         0  1    [1.0, 2.0]
         1  3    [5.0, 6.0]
         2  2  [10.0, 10.0]
         >>> table.update(values_sql={"x": "x + 1"})
+        UpdateResult(rows_updated=3, version=3)
         >>> table.to_pandas()
            x        vector
         0  2    [1.0, 2.0]
@@ -1354,6 +1385,11 @@ class Table(ABC):
             Alternatively, a pyarrow Field or Schema can be provided to add
             new columns with the specified data types. The new columns will
             be initialized with null values.
+
+        Returns
+        -------
+        AddColumnsResult
+            version: the new version number of the table after adding columns.
         """
 
     @abstractmethod
@@ -1379,10 +1415,15 @@ class Table(ABC):
                 nullability is not changed. Only non-nullable columns can be changed
                 to nullable. Currently, you cannot change a nullable column to
                 non-nullable.
+
+        Returns
+        -------
+        AlterColumnsResult
+            version: the new version number of the table after the alteration.
         """
 
     @abstractmethod
-    def drop_columns(self, columns: Iterable[str]):
+    def drop_columns(self, columns: Iterable[str]) -> DropColumnsResult:
         """
         Drop columns from the table.
 
@@ -1390,6 +1431,11 @@ class Table(ABC):
         ----------
         columns : Iterable[str]
             The names of the columns to drop.
+
+        Returns
+        -------
+        DropColumnsResult
+            version: the new version number of the table dropping the columns.
         """
 
     @abstractmethod
@@ -1428,7 +1474,7 @@ class Table(ABC):
         """
 
     @abstractmethod
-    def restore(self, version: Optional[int] = None):
+    def restore(self, version: Optional[Union[int, str]] = None):
         """Restore a version of the table. This is an in-place operation.
 
         This creates a new version where the data is equivalent to the
@@ -1436,9 +1482,10 @@ class Table(ABC):
 
         Parameters
         ----------
-        version : int, default None
-            The version to restore. If unspecified then restores the currently
-            checked out version. If the currently checked out version is the
+        version : int or str, default None
+            The version number or version tag to restore.
+            If unspecified then restores the currently checked out version.
+            If the currently checked out version is the
             latest version then this is a no-op.
         """
 
@@ -1611,6 +1658,7 @@ class LanceTable(Table):
         ...    [{"vector": [1.1, 0.9], "type": "vector"}])
         >>> table.tags.create("v1", table.version)
         >>> table.add([{"vector": [0.5, 0.2], "type": "vector"}])
+        AddResult(version=2)
         >>> tags = table.tags.list()
         >>> print(tags["v1"]["version"])
         1
@@ -1649,6 +1697,7 @@ class LanceTable(Table):
                vector    type
         0  [1.1, 0.9]  vector
         >>> table.add([{"vector": [0.5, 0.2], "type": "vector"}])
+        AddResult(version=2)
         >>> table.version
         2
         >>> table.checkout(1)
@@ -1666,7 +1715,7 @@ class LanceTable(Table):
         """
         LOOP.run(self._table.checkout_latest())
 
-    def restore(self, version: Optional[int] = None):
+    def restore(self, version: Optional[Union[int, str]] = None):
         """Restore a version of the table. This is an in-place operation.
 
         This creates a new version where the data is equivalent to the
@@ -1674,9 +1723,10 @@ class LanceTable(Table):
 
         Parameters
         ----------
-        version : int, default None
-            The version to restore. If unspecified then restores the currently
-            checked out version. If the currently checked out version is the
+        version : int or str, default None
+            The version number or version tag to restore.
+            If unspecified then restores the currently checked out version.
+            If the currently checked out version is the
             latest version then this is a no-op.
 
         Examples
@@ -1691,14 +1741,23 @@ class LanceTable(Table):
                vector    type
         0  [1.1, 0.9]  vector
         >>> table.add([{"vector": [0.5, 0.2], "type": "vector"}])
+        AddResult(version=2)
         >>> table.version
         2
+        >>> table.tags.create("v2", 2)
         >>> table.restore(1)
         >>> table.to_pandas()
                vector    type
         0  [1.1, 0.9]  vector
         >>> len(table.list_versions())
         3
+        >>> table.restore("v2")
+        >>> table.to_pandas()
+               vector    type
+        0  [1.1, 0.9]  vector
+        1  [0.5, 0.2]  vector
+        >>> len(table.list_versions())
+        4
         """
         if version is not None:
             LOOP.run(self._table.checkout(version))
@@ -1707,9 +1766,6 @@ class LanceTable(Table):
     def count_rows(self, filter: Optional[str] = None) -> int:
         return LOOP.run(self._table.count_rows(filter))
 
-    def __len__(self) -> int:
-        return self.count_rows()
-
     def __repr__(self) -> str:
         val = f"{self.__class__.__name__}(name={self.name!r}, version={self.version}"
         if self._conn.read_consistency_interval is not None:
@@ -1914,17 +1970,17 @@ class LanceTable(Table):
         ordering_field_names: Optional[Union[str, List[str]]] = None,
         replace: bool = False,
         writer_heap_size: Optional[int] = 1024 * 1024 * 1024,
-        use_tantivy: bool = True,
+        use_tantivy: bool = False,
         tokenizer_name: Optional[str] = None,
-        with_position: bool = True,
+        with_position: bool = False,
         # tokenizer configs:
         base_tokenizer: BaseTokenizerType = "simple",
         language: str = "English",
         max_token_length: Optional[int] = 40,
         lower_case: bool = True,
-        stem: bool = False,
-        remove_stop_words: bool = False,
-        ascii_folding: bool = False,
+        stem: bool = True,
+        remove_stop_words: bool = True,
+        ascii_folding: bool = True,
     ):
         if not use_tantivy:
             if not isinstance(field_names, str):
@@ -1934,6 +1990,7 @@ class LanceTable(Table):
                 tokenizer_configs = {
                     "base_tokenizer": base_tokenizer,
                     "language": language,
+                    "with_position": with_position,
                     "max_token_length": max_token_length,
                     "lower_case": lower_case,
                     "stem": stem,
@@ -1944,7 +2001,6 @@ class LanceTable(Table):
                 tokenizer_configs = self.infer_tokenizer_configs(tokenizer_name)
 
             config = FTS(
-                with_position=with_position,
                 **tokenizer_configs,
             )
 
@@ -2055,7 +2111,7 @@ class LanceTable(Table):
         mode: AddMode = "append",
         on_bad_vectors: OnBadVectorsType = "error",
         fill_value: float = 0.0,
-    ):
+    ) -> AddResult:
         """Add data to the table.
         If vector columns are missing and the table
         has embedding functions, then the vector columns
@@ -2079,7 +2135,7 @@ class LanceTable(Table):
         int
             The number of vectors in the table.
         """
-        LOOP.run(
+        return LOOP.run(
             self._table.add(
                 data, mode=mode, on_bad_vectors=on_bad_vectors, fill_value=fill_value
             )
@@ -2409,8 +2465,8 @@ class LanceTable(Table):
         )
         return self
 
-    def delete(self, where: str):
-        LOOP.run(self._table.delete(where))
+    def delete(self, where: str) -> DeleteResult:
+        return LOOP.run(self._table.delete(where))
 
     def update(
         self,
@@ -2418,7 +2474,7 @@ class LanceTable(Table):
         values: Optional[dict] = None,
         *,
         values_sql: Optional[Dict[str, str]] = None,
-    ):
+    ) -> UpdateResult:
         """
         This can be used to update zero to all rows depending on how many
         rows match the where clause.
@@ -2436,6 +2492,12 @@ class LanceTable(Table):
             reference existing columns. For example, {"x": "x + 1"} will increment
             the x column by 1.
 
+        Returns
+        -------
+        UpdateResult
+            - rows_updated: The number of rows that were updated
+            - version: The new version number of the table after the update
+
         Examples
         --------
         >>> import lancedb
@@ -2449,6 +2511,7 @@ class LanceTable(Table):
         1  2  [3.0, 4.0]
         2  3  [5.0, 6.0]
         >>> table.update(where="x = 2", values={"vector": [10.0, 10]})
+        UpdateResult(rows_updated=1, version=2)
         >>> table.to_pandas()
            x        vector
         0  1    [1.0, 2.0]
@@ -2456,7 +2519,7 @@ class LanceTable(Table):
         2  2  [10.0, 10.0]
 
         """
-        LOOP.run(self._table.update(values, where=where, updates_sql=values_sql))
+        return LOOP.run(self._table.update(values, where=where, updates_sql=values_sql))
 
     def _execute_query(
         self,
@@ -2490,7 +2553,7 @@ class LanceTable(Table):
         new_data: DATA,
         on_bad_vectors: OnBadVectorsType,
         fill_value: float,
-    ):
+    ) -> MergeResult:
         return LOOP.run(
             self._table._do_merge(merge, new_data, on_bad_vectors, fill_value)
         )
@@ -2635,14 +2698,16 @@ class LanceTable(Table):
 
     def add_columns(
         self, transforms: Dict[str, str] | pa.field | List[pa.field] | pa.Schema
-    ):
-        LOOP.run(self._table.add_columns(transforms))
+    ) -> AddColumnsResult:
+        return LOOP.run(self._table.add_columns(transforms))
 
-    def alter_columns(self, *alterations: Iterable[Dict[str, str]]):
-        LOOP.run(self._table.alter_columns(*alterations))
+    def alter_columns(
+        self, *alterations: Iterable[Dict[str, str]]
+    ) -> AlterColumnsResult:
+        return LOOP.run(self._table.alter_columns(*alterations))
 
-    def drop_columns(self, columns: Iterable[str]):
-        LOOP.run(self._table.drop_columns(columns))
+    def drop_columns(self, columns: Iterable[str]) -> DropColumnsResult:
+        return LOOP.run(self._table.drop_columns(columns))
 
     def uses_v2_manifest_paths(self) -> bool:
         """
@@ -3197,7 +3262,7 @@ class AsyncTable:
         mode: Optional[Literal["append", "overwrite"]] = "append",
         on_bad_vectors: Optional[OnBadVectorsType] = None,
         fill_value: Optional[float] = None,
-    ):
+    ) -> AddResult:
         """Add more data to the [Table](Table).
 
         Parameters
@@ -3236,7 +3301,7 @@ class AsyncTable:
         if isinstance(data, pa.Table):
             data = data.to_reader()
 
-        await self._inner.add(data, mode or "append")
+        return await self._inner.add(data, mode or "append")
 
     def merge_insert(self, on: Union[str, Iterable[str]]) -> LanceMergeInsertBuilder:
         """
@@ -3281,12 +3346,12 @@ class AsyncTable:
         >>> table = db.create_table("my_table", data)
         >>> new_data = pa.table({"a": [2, 3, 4], "b": ["x", "y", "z"]})
         >>> # Perform a "upsert" operation
-        >>> stats = table.merge_insert("a")     \\
+        >>> res = table.merge_insert("a")     \\
         ...      .when_matched_update_all()     \\
         ...      .when_not_matched_insert_all() \\
         ...      .execute(new_data)
-        >>> stats
-        {'num_inserted_rows': 1, 'num_updated_rows': 2, 'num_deleted_rows': 0}
+        >>> res
+        MergeResult(version=2, num_updated_rows=2, num_inserted_rows=1, num_deleted_rows=0)
         >>> # The order of new rows is non-deterministic since we use
         >>> # a hash-join as part of this operation and so we sort here
         >>> table.to_arrow().sort_by("a").to_pandas()
@@ -3295,7 +3360,7 @@ class AsyncTable:
         1  2  x
         2  3  y
         3  4  z
-        """
+        """  # noqa: E501
         on = [on] if isinstance(on, str) else list(iter(on))
 
         return LanceMergeInsertBuilder(self, on)
@@ -3572,8 +3637,10 @@ class AsyncTable:
             )
             if query.distance_type is not None:
                 async_query = async_query.distance_type(query.distance_type)
-            if query.nprobes is not None:
-                async_query = async_query.nprobes(query.nprobes)
+            if query.minimum_nprobes is not None:
+                async_query = async_query.minimum_nprobes(query.minimum_nprobes)
+            if query.maximum_nprobes is not None:
+                async_query = async_query.maximum_nprobes(query.maximum_nprobes)
             if query.refine_factor is not None:
                 async_query = async_query.refine_factor(query.refine_factor)
             if query.vector_column:
@@ -3626,7 +3693,7 @@ class AsyncTable:
         new_data: DATA,
         on_bad_vectors: OnBadVectorsType,
         fill_value: float,
-    ):
+    ) -> MergeResult:
         schema = await self.schema()
         if on_bad_vectors is None:
             on_bad_vectors = "error"
@@ -3651,10 +3718,11 @@ class AsyncTable:
                 when_not_matched_insert_all=merge._when_not_matched_insert_all,
                 when_not_matched_by_source_delete=merge._when_not_matched_by_source_delete,
                 when_not_matched_by_source_condition=merge._when_not_matched_by_source_condition,
+                timeout=merge._timeout,
             ),
         )
 
-    async def delete(self, where: str):
+    async def delete(self, where: str) -> DeleteResult:
         """Delete rows from the table.
 
         This can be used to delete a single row, many rows, all rows, or
@@ -3685,6 +3753,7 @@ class AsyncTable:
         1  2  [3.0, 4.0]
         2  3  [5.0, 6.0]
         >>> table.delete("x = 2")
+        DeleteResult(version=2)
         >>> table.to_pandas()
            x      vector
         0  1  [1.0, 2.0]
@@ -3698,6 +3767,7 @@ class AsyncTable:
         >>> to_remove
         '1, 5'
         >>> table.delete(f"x IN ({to_remove})")
+        DeleteResult(version=3)
         >>> table.to_pandas()
            x      vector
         0  3  [5.0, 6.0]
@@ -3710,7 +3780,7 @@ class AsyncTable:
         *,
         where: Optional[str] = None,
         updates_sql: Optional[Dict[str, str]] = None,
-    ):
+    ) -> UpdateResult:
         """
         This can be used to update zero to all rows in the table.
 
@@ -3732,6 +3802,13 @@ class AsyncTable:
             literals (e.g. "7" or "'foo'") or they can be expressions based on the
             previous value of the row (e.g. "x + 1" to increment the x column by 1)
 
+        Returns
+        -------
+        UpdateResult
+            An object containing:
+            - rows_updated: The number of rows that were updated
+            - version: The new version number of the table after the update
+
         Examples
         --------
         >>> import asyncio
@@ -3760,7 +3837,7 @@ class AsyncTable:
 
     async def add_columns(
         self, transforms: dict[str, str] | pa.field | List[pa.field] | pa.Schema
-    ):
+    ) -> AddColumnsResult:
         """
         Add new columns with defined values.
 
@@ -3772,6 +3849,12 @@ class AsyncTable:
             each row in the table, and can reference existing columns.
             Alternatively, you can pass a pyarrow field or schema to add
             new columns with NULLs.
+
+        Returns
+        -------
+        AddColumnsResult
+            version: the new version number of the table after adding columns.
+
         """
         if isinstance(transforms, pa.Field):
             transforms = [transforms]
@@ -3780,11 +3863,13 @@ class AsyncTable:
         ):
             transforms = pa.schema(transforms)
         if isinstance(transforms, pa.Schema):
-            await self._inner.add_columns_with_schema(transforms)
+            return await self._inner.add_columns_with_schema(transforms)
         else:
-            await self._inner.add_columns(list(transforms.items()))
+            return await self._inner.add_columns(list(transforms.items()))
 
-    async def alter_columns(self, *alterations: Iterable[dict[str, Any]]):
+    async def alter_columns(
+        self, *alterations: Iterable[dict[str, Any]]
+    ) -> AlterColumnsResult:
         """
         Alter column names and nullability.
 
@@ -3804,8 +3889,13 @@ class AsyncTable:
                 nullability is not changed. Only non-nullable columns can be changed
                 to nullable. Currently, you cannot change a nullable column to
                 non-nullable.
+
+        Returns
+        -------
+        AlterColumnsResult
+            version: the new version number of the table after the alteration.
         """
-        await self._inner.alter_columns(alterations)
+        return await self._inner.alter_columns(alterations)
 
     async def drop_columns(self, columns: Iterable[str]):
         """
@@ -3816,7 +3906,7 @@ class AsyncTable:
         columns : Iterable[str]
             The names of the columns to drop.
         """
-        await self._inner.drop_columns(columns)
+        return await self._inner.drop_columns(columns)
 
     async def version(self) -> int:
         """
@@ -3886,7 +3976,7 @@ class AsyncTable:
         """
         await self._inner.checkout_latest()
 
-    async def restore(self, version: Optional[int] = None):
+    async def restore(self, version: Optional[int | str] = None):
         """
         Restore the table to the currently checked out version
 

python/python/tests/docs/test_guide_tables.py

@@ -25,6 +25,10 @@ import numpy as np
 from lancedb.pydantic import Vector, LanceModel
 
 # --8<-- [end:import-lancedb-pydantic]
+# --8<-- [start:import-session-context]
+from datafusion import SessionContext
+
+# --8<-- [end:import-session-context]
 # --8<-- [start:import-datetime]
 from datetime import timedelta
 
@@ -33,6 +37,10 @@ from datetime import timedelta
 from lancedb.embeddings import get_registry
 
 # --8<-- [end:import-embeddings]
+# --8<-- [start:import-ffi-dataset]
+from lance import FFILanceTableProvider
+
+# --8<-- [end:import-ffi-dataset]
 # --8<-- [start:import-pydantic-basemodel]
 from pydantic import BaseModel
 
@@ -341,6 +349,27 @@ def test_table_with_embedding():
     # --8<-- [end:create_table_with_embedding]
 
 
+def test_sql_query():
+    db = lancedb.connect("data/sample-lancedb")
+    data = [
+        {"vector": [1.1, 1.2], "lat": 45.5, "long": -122.7},
+        {"vector": [0.2, 1.8], "lat": 40.1, "long": -74.1},
+    ]
+    table = db.create_table("lance_table", data)
+
+    # --8<-- [start:lance_sql_basic]
+    ctx = SessionContext()
+    ffi_lance_table = FFILanceTableProvider(
+        table.to_lance(), with_row_id=False, with_row_addr=False
+    )
+
+    ctx.register_table_provider("ffi_lance_table", ffi_lance_table)
+    ctx.table("ffi_lance_table")
+
+    ctx.sql("SELECT vector FROM ffi_lance_table")
+    # --8<-- [end:lance_sql_basic]
+
+
 @pytest.mark.skip
 async def test_table_with_embedding_async():
     async_db = await lancedb.connect_async("data/sample-lancedb")

python/python/tests/docs/test_merge_insert.py

@@ -18,19 +18,19 @@ def test_upsert(mem_db):
         {"id": 1, "name": "Bobby"},
         {"id": 2, "name": "Charlie"},
     ]
-    stats = (
+    res = (
         table.merge_insert("id")
         .when_matched_update_all()
         .when_not_matched_insert_all()
         .execute(new_users)
     )
     table.count_rows()  # 3
-    stats  # {'num_inserted_rows': 1, 'num_updated_rows': 1, 'num_deleted_rows': 0}
+    res  # {'num_inserted_rows': 1, 'num_updated_rows': 1, 'num_deleted_rows': 0}
     # --8<-- [end:upsert_basic]
     assert table.count_rows() == 3
-    assert stats["num_inserted_rows"] == 1
-    assert stats["num_updated_rows"] == 1
-    assert stats["num_deleted_rows"] == 0
+    assert res.num_inserted_rows == 1
+    assert res.num_deleted_rows == 0
+    assert res.num_updated_rows == 1
 
 
 @pytest.mark.asyncio
@@ -48,19 +48,22 @@ async def test_upsert_async(mem_db_async):
         {"id": 1, "name": "Bobby"},
         {"id": 2, "name": "Charlie"},
     ]
-    stats = await (
+    res = await (
         table.merge_insert("id")
         .when_matched_update_all()
         .when_not_matched_insert_all()
         .execute(new_users)
     )
     await table.count_rows()  # 3
-    stats  # {'num_inserted_rows': 1, 'num_updated_rows': 1, 'num_deleted_rows': 0}
+    res
+    # MergeResult(version=2, num_updated_rows=1,
+    # num_inserted_rows=1, num_deleted_rows=0)
     # --8<-- [end:upsert_basic_async]
     assert await table.count_rows() == 3
-    assert stats["num_inserted_rows"] == 1
-    assert stats["num_updated_rows"] == 1
-    assert stats["num_deleted_rows"] == 0
+    assert res.version == 2
+    assert res.num_inserted_rows == 1
+    assert res.num_deleted_rows == 0
+    assert res.num_updated_rows == 1
 
 
 def test_insert_if_not_exists(mem_db):
@@ -77,16 +80,19 @@ def test_insert_if_not_exists(mem_db):
         {"domain": "google.com", "name": "Google"},
         {"domain": "facebook.com", "name": "Facebook"},
     ]
-    stats = (
+    res = (
         table.merge_insert("domain").when_not_matched_insert_all().execute(new_domains)
     )
     table.count_rows()  # 3
-    stats  # {'num_inserted_rows': 1, 'num_updated_rows': 0, 'num_deleted_rows': 0}
+    res
+    # MergeResult(version=2, num_updated_rows=0,
+    # num_inserted_rows=1, num_deleted_rows=0)
     # --8<-- [end:insert_if_not_exists]
     assert table.count_rows() == 3
-    assert stats["num_inserted_rows"] == 1
-    assert stats["num_updated_rows"] == 0
-    assert stats["num_deleted_rows"] == 0
+    assert res.version == 2
+    assert res.num_inserted_rows == 1
+    assert res.num_deleted_rows == 0
+    assert res.num_updated_rows == 0
 
 
 @pytest.mark.asyncio
@@ -104,16 +110,19 @@ async def test_insert_if_not_exists_async(mem_db_async):
         {"domain": "google.com", "name": "Google"},
         {"domain": "facebook.com", "name": "Facebook"},
     ]
-    stats = await (
+    res = await (
         table.merge_insert("domain").when_not_matched_insert_all().execute(new_domains)
     )
     await table.count_rows()  # 3
-    stats  # {'num_inserted_rows': 1, 'num_updated_rows': 0, 'num_deleted_rows': 0}
-    # --8<-- [end:insert_if_not_exists_async]
+    res
+    # MergeResult(version=2, num_updated_rows=0,
+    # num_inserted_rows=1, num_deleted_rows=0)
+    # --8<-- [end:insert_if_not_exists]
     assert await table.count_rows() == 3
-    assert stats["num_inserted_rows"] == 1
-    assert stats["num_updated_rows"] == 0
-    assert stats["num_deleted_rows"] == 0
+    assert res.version == 2
+    assert res.num_inserted_rows == 1
+    assert res.num_deleted_rows == 0
+    assert res.num_updated_rows == 0
 
 
 def test_replace_range(mem_db):
@@ -131,7 +140,7 @@ def test_replace_range(mem_db):
     new_chunks = [
         {"doc_id": 1, "chunk_id": 0, "text": "Baz"},
     ]
-    stats = (
+    res = (
         table.merge_insert(["doc_id", "chunk_id"])
         .when_matched_update_all()
         .when_not_matched_insert_all()
@@ -139,12 +148,15 @@ def test_replace_range(mem_db):
         .execute(new_chunks)
     )
     table.count_rows("doc_id = 1")  # 1
-    stats  # {'num_inserted_rows': 0, 'num_updated_rows': 1, 'num_deleted_rows': 1}
-    # --8<-- [end:replace_range]
+    res
+    # MergeResult(version=2, num_updated_rows=1,
+    # num_inserted_rows=0, num_deleted_rows=1)
+    # --8<-- [end:insert_if_not_exists]
     assert table.count_rows("doc_id = 1") == 1
-    assert stats["num_inserted_rows"] == 0
-    assert stats["num_updated_rows"] == 1
-    assert stats["num_deleted_rows"] == 1
+    assert res.version == 2
+    assert res.num_inserted_rows == 0
+    assert res.num_deleted_rows == 1
+    assert res.num_updated_rows == 1
 
 
 @pytest.mark.asyncio
@@ -163,7 +175,7 @@ async def test_replace_range_async(mem_db_async):
     new_chunks = [
         {"doc_id": 1, "chunk_id": 0, "text": "Baz"},
     ]
-    stats = await (
+    res = await (
         table.merge_insert(["doc_id", "chunk_id"])
         .when_matched_update_all()
         .when_not_matched_insert_all()
@@ -171,9 +183,12 @@ async def test_replace_range_async(mem_db_async):
         .execute(new_chunks)
     )
     await table.count_rows("doc_id = 1")  # 1
-    stats  # {'num_inserted_rows': 0, 'num_updated_rows': 1, 'num_deleted_rows': 1}
-    # --8<-- [end:replace_range_async]
+    res
+    # MergeResult(version=2, num_updated_rows=1,
+    # num_inserted_rows=0, num_deleted_rows=1)
+    # --8<-- [end:insert_if_not_exists]
     assert await table.count_rows("doc_id = 1") == 1
-    assert stats["num_inserted_rows"] == 0
-    assert stats["num_updated_rows"] == 1
-    assert stats["num_deleted_rows"] == 1
+    assert res.version == 2
+    assert res.num_inserted_rows == 0
+    assert res.num_deleted_rows == 1
+    assert res.num_updated_rows == 1

python/python/tests/docs/test_search.py

@@ -6,7 +6,7 @@ import lancedb
 
 # --8<-- [end:import-lancedb]
 # --8<-- [start:import-numpy]
-from lancedb.query import BoostQuery, MatchQuery
+from lancedb.query import BooleanQuery, BoostQuery, MatchQuery, Occur
 import numpy as np
 import pyarrow as pa
 
@@ -156,6 +156,9 @@ async def test_vector_search_async():
     # --8<-- [end:search_result_async_as_list]
 
 
+@pytest.mark.skipif(
+    os.name == "nt", reason="Need to fix https://github.com/lancedb/lance/issues/3905"
+)
 def test_fts_fuzzy_query():
     uri = "data/fuzzy-example"
     db = lancedb.connect(uri)
@@ -188,7 +191,19 @@ def test_fts_fuzzy_query():
         "food",  # 1 insertion
     }
 
+    results = table.search(
+        MatchQuery("foo", "text", fuzziness=1, prefix_length=3)
+    ).to_pandas()
+    assert len(results) == 2
+    assert set(results["text"].to_list()) == {
+        "foo",
+        "food",
+    }
 
+
+@pytest.mark.skipif(
+    os.name == "nt", reason="Need to fix https://github.com/lancedb/lance/issues/3905"
+)
 def test_fts_boost_query():
     uri = "data/boost-example"
     db = lancedb.connect(uri)
@@ -234,6 +249,63 @@ def test_fts_boost_query():
     )
 
 
+@pytest.mark.skipif(
+    os.name == "nt", reason="Need to fix https://github.com/lancedb/lance/issues/3905"
+)
+def test_fts_boolean_query(tmp_path):
+    uri = tmp_path / "boolean-example"
+    db = lancedb.connect(uri)
+    table = db.create_table(
+        "my_table_fts_boolean",
+        data=[
+            {"text": "The cat and dog are playing"},
+            {"text": "The cat is sleeping"},
+            {"text": "The dog is barking"},
+            {"text": "The dog chases the cat"},
+        ],
+        mode="overwrite",
+    )
+    table.create_fts_index("text", use_tantivy=False, replace=True)
+
+    # SHOULD
+    results = table.search(
+        MatchQuery("cat", "text") | MatchQuery("dog", "text")
+    ).to_pandas()
+    assert len(results) == 4
+    assert set(results["text"].to_list()) == {
+        "The cat and dog are playing",
+        "The cat is sleeping",
+        "The dog is barking",
+        "The dog chases the cat",
+    }
+    # MUST
+    results = table.search(
+        MatchQuery("cat", "text") & MatchQuery("dog", "text")
+    ).to_pandas()
+    assert len(results) == 2
+    assert set(results["text"].to_list()) == {
+        "The cat and dog are playing",
+        "The dog chases the cat",
+    }
+
+    # MUST NOT
+    results = table.search(
+        BooleanQuery(
+            [
+                (Occur.MUST, MatchQuery("cat", "text")),
+                (Occur.MUST_NOT, MatchQuery("dog", "text")),
+            ]
+        )
+    ).to_pandas()
+    assert len(results) == 1
+    assert set(results["text"].to_list()) == {
+        "The cat is sleeping",
+    }
+
+
+@pytest.mark.skipif(
+    os.name == "nt", reason="Need to fix https://github.com/lancedb/lance/issues/3905"
+)
 def test_fts_native():
     # --8<-- [start:basic_fts]
     uri = "data/sample-lancedb"
@@ -282,6 +354,9 @@ def test_fts_native():
     # --8<-- [end:fts_incremental_index]
 
 
+@pytest.mark.skipif(
+    os.name == "nt", reason="Need to fix https://github.com/lancedb/lance/issues/3905"
+)
 @pytest.mark.asyncio
 async def test_fts_native_async():
     # --8<-- [start:basic_fts_async]

python/python/tests/test_fts.py

@@ -215,6 +215,19 @@ def test_search_fts(table, use_tantivy):
         assert len(results) == 5
         assert len(results[0]) == 3  # id, text, _score
 
+        # Test boolean query
+        results = (
+            table.search(MatchQuery("puppy", "text") & MatchQuery("runs", "text"))
+            .select(["id", "text"])
+            .limit(5)
+            .to_list()
+        )
+        assert len(results) == 5
+        assert len(results[0]) == 3  # id, text, _score
+        for r in results:
+            assert "puppy" in r["text"]
+            assert "runs" in r["text"]
+
 
 @pytest.mark.asyncio
 async def test_fts_select_async(async_table):
@@ -287,7 +300,7 @@ def test_search_fts_phrase_query(table):
         assert False
     except Exception:
         pass
-    table.create_fts_index("text", use_tantivy=False, replace=True)
+    table.create_fts_index("text", use_tantivy=False, with_position=True, replace=True)
     results = table.search("puppy").limit(100).to_list()
     phrase_results = table.search('"puppy runs"').limit(100).to_list()
     assert len(results) > len(phrase_results)
@@ -312,7 +325,7 @@ async def test_search_fts_phrase_query_async(async_table):
         assert False
     except Exception:
         pass
-    await async_table.create_index("text", config=FTS())
+    await async_table.create_index("text", config=FTS(with_position=True))
     results = await async_table.query().nearest_to_text("puppy").limit(100).to_list()
     phrase_results = (
         await async_table.query().nearest_to_text('"puppy runs"').limit(100).to_list()
@@ -649,7 +662,7 @@ def test_fts_on_list(mem_db: DBConnection):
         }
     )
     table = mem_db.create_table("test", data=data)
-    table.create_fts_index("text", use_tantivy=False)
+    table.create_fts_index("text", use_tantivy=False, with_position=True)
 
     res = table.search("lance").limit(5).to_list()
     assert len(res) == 3

python/python/tests/test_query.py

@@ -25,6 +25,8 @@ from lancedb.query import (
     AsyncQueryBase,
     AsyncVectorQuery,
     LanceVectorQueryBuilder,
+    MatchQuery,
+    PhraseQuery,
     Query,
     FullTextSearchQuery,
 )
@@ -437,6 +439,33 @@ def test_query_builder_with_filter(table):
     assert all(np.array(rs[0]["vector"]) == [3, 4])
 
 
+def test_invalid_nprobes_sync(table):
+    with pytest.raises(ValueError, match="minimum_nprobes must be greater than 0"):
+        LanceVectorQueryBuilder(table, [0, 0], "vector").minimum_nprobes(0).to_list()
+    with pytest.raises(
+        ValueError, match="maximum_nprobes must be greater than minimum_nprobes"
+    ):
+        LanceVectorQueryBuilder(table, [0, 0], "vector").maximum_nprobes(5).to_list()
+    with pytest.raises(
+        ValueError, match="minimum_nprobes must be less or equal to maximum_nprobes"
+    ):
+        LanceVectorQueryBuilder(table, [0, 0], "vector").minimum_nprobes(100).to_list()
+
+
+@pytest.mark.asyncio
+async def test_invalid_nprobes_async(table_async: AsyncTable):
+    with pytest.raises(ValueError, match="minimum_nprobes must be greater than 0"):
+        await table_async.vector_search([0, 0]).minimum_nprobes(0).to_list()
+    with pytest.raises(
+        ValueError, match="maximum_nprobes must be greater than minimum_nprobes"
+    ):
+        await table_async.vector_search([0, 0]).maximum_nprobes(5).to_list()
+    with pytest.raises(
+        ValueError, match="minimum_nprobes must be less or equal to maximum_nprobes"
+    ):
+        await table_async.vector_search([0, 0]).minimum_nprobes(100).to_list()
+
+
 def test_query_builder_with_prefilter(table):
     df = (
         LanceVectorQueryBuilder(table, [0, 0], "vector")
@@ -583,6 +612,21 @@ async def test_query_async(table_async: AsyncTable):
         table_async.query().nearest_to(pa.array([1, 2])).nprobes(10),
         expected_num_rows=2,
     )
+    await check_query(
+        table_async.query().nearest_to(pa.array([1, 2])).minimum_nprobes(10),
+        expected_num_rows=2,
+    )
+    await check_query(
+        table_async.query().nearest_to(pa.array([1, 2])).maximum_nprobes(30),
+        expected_num_rows=2,
+    )
+    await check_query(
+        table_async.query()
+        .nearest_to(pa.array([1, 2]))
+        .minimum_nprobes(10)
+        .maximum_nprobes(20),
+        expected_num_rows=2,
+    )
     await check_query(
         table_async.query().nearest_to(pa.array([1, 2])).bypass_vector_index(),
         expected_num_rows=2,
@@ -731,6 +775,82 @@ async def test_explain_plan_async(table_async: AsyncTable):
     assert "KNN" in plan
 
 
+@pytest.mark.asyncio
+async def test_explain_plan_fts(table_async: AsyncTable):
+    """Test explain plan for FTS queries"""
+    # Create FTS index
+    from lancedb.index import FTS
+
+    await table_async.create_index("text", config=FTS())
+
+    # Test pure FTS query
+    query = await table_async.search("dog", query_type="fts", fts_columns="text")
+    plan = await query.explain_plan()
+    # Should show FTS details (issue #2465 is now fixed)
+    assert "MatchQuery: query=dog" in plan
+    assert "GlobalLimitExec" in plan  # Default limit
+
+    # Test FTS query with limit
+    query_with_limit = await table_async.search(
+        "dog", query_type="fts", fts_columns="text"
+    )
+    plan_with_limit = await query_with_limit.limit(1).explain_plan()
+    assert "MatchQuery: query=dog" in plan_with_limit
+    assert "GlobalLimitExec: skip=0, fetch=1" in plan_with_limit
+
+    # Test FTS query with offset and limit
+    query_with_offset = await table_async.search(
+        "dog", query_type="fts", fts_columns="text"
+    )
+    plan_with_offset = await query_with_offset.offset(1).limit(1).explain_plan()
+    assert "MatchQuery: query=dog" in plan_with_offset
+    assert "GlobalLimitExec: skip=1, fetch=1" in plan_with_offset
+
+
+@pytest.mark.asyncio
+async def test_explain_plan_vector_with_limit_offset(table_async: AsyncTable):
+    """Test explain plan for vector queries with limit and offset"""
+    # Test vector query with limit
+    plan_with_limit = await (
+        table_async.query().nearest_to(pa.array([1, 2])).limit(1).explain_plan()
+    )
+    assert "KNN" in plan_with_limit
+    assert "GlobalLimitExec: skip=0, fetch=1" in plan_with_limit
+
+    # Test vector query with offset and limit
+    plan_with_offset = await (
+        table_async.query()
+        .nearest_to(pa.array([1, 2]))
+        .offset(1)
+        .limit(1)
+        .explain_plan()
+    )
+    assert "KNN" in plan_with_offset
+    assert "GlobalLimitExec: skip=1, fetch=1" in plan_with_offset
+
+
+@pytest.mark.asyncio
+async def test_explain_plan_with_filters(table_async: AsyncTable):
+    """Test explain plan for queries with filters"""
+    # Test vector query with filter
+    plan_with_filter = await (
+        table_async.query().nearest_to(pa.array([1, 2])).where("id = 1").explain_plan()
+    )
+    assert "KNN" in plan_with_filter
+    assert "FilterExec" in plan_with_filter
+
+    # Test FTS query with filter
+    from lancedb.index import FTS
+
+    await table_async.create_index("text", config=FTS())
+    query_fts_filter = await table_async.search(
+        "dog", query_type="fts", fts_columns="text"
+    )
+    plan_fts_filter = await query_fts_filter.where("id = 1").explain_plan()
+    assert "MatchQuery: query=dog" in plan_fts_filter
+    assert "FilterExec: id@" in plan_fts_filter  # Should show filter details
+
+
 @pytest.mark.asyncio
 async def test_query_camelcase_async(tmp_path):
     db = await lancedb.connect_async(tmp_path)
@@ -909,7 +1029,39 @@ def test_query_serialization_sync(table: lancedb.table.Table):
 
     q = table.search([5.0, 6.0]).nprobes(10).refine_factor(5).to_query_object()
     check_set_props(
-        q, vector_column="vector", vector=[5.0, 6.0], nprobes=10, refine_factor=5
+        q,
+        vector_column="vector",
+        vector=[5.0, 6.0],
+        minimum_nprobes=10,
+        maximum_nprobes=10,
+        refine_factor=5,
+    )
+
+    q = table.search([5.0, 6.0]).minimum_nprobes(10).to_query_object()
+    check_set_props(
+        q,
+        vector_column="vector",
+        vector=[5.0, 6.0],
+        minimum_nprobes=10,
+        maximum_nprobes=None,
+    )
+
+    q = table.search([5.0, 6.0]).nprobes(50).to_query_object()
+    check_set_props(
+        q,
+        vector_column="vector",
+        vector=[5.0, 6.0],
+        minimum_nprobes=50,
+        maximum_nprobes=50,
+    )
+
+    q = table.search([5.0, 6.0]).maximum_nprobes(10).to_query_object()
+    check_set_props(
+        q,
+        vector_column="vector",
+        vector=[5.0, 6.0],
+        maximum_nprobes=10,
+        minimum_nprobes=None,
     )
 
     q = table.search([5.0, 6.0]).distance_range(0.0, 1.0).to_query_object()
@@ -961,7 +1113,8 @@ async def test_query_serialization_async(table_async: AsyncTable):
         limit=10,
         vector=sample_vector,
         postfilter=False,
-        nprobes=20,
+        minimum_nprobes=20,
+        maximum_nprobes=20,
         with_row_id=False,
         bypass_vector_index=False,
     )
@@ -971,7 +1124,20 @@ async def test_query_serialization_async(table_async: AsyncTable):
         q,
         vector=sample_vector,
         postfilter=False,
-        nprobes=20,
+        minimum_nprobes=20,
+        maximum_nprobes=20,
+        with_row_id=False,
+        bypass_vector_index=False,
+        limit=10,
+    )
+
+    q = (await table_async.search([5.0, 6.0])).nprobes(50).to_query_object()
+    check_set_props(
+        q,
+        vector=sample_vector,
+        postfilter=False,
+        minimum_nprobes=50,
+        maximum_nprobes=50,
         with_row_id=False,
         bypass_vector_index=False,
         limit=10,
@@ -990,7 +1156,8 @@ async def test_query_serialization_async(table_async: AsyncTable):
         filter="id = 1",
         postfilter=True,
         vector=sample_vector,
-        nprobes=20,
+        minimum_nprobes=20,
+        maximum_nprobes=20,
         with_row_id=False,
         bypass_vector_index=False,
     )
@@ -1004,7 +1171,8 @@ async def test_query_serialization_async(table_async: AsyncTable):
     check_set_props(
         q,
         vector=sample_vector,
-        nprobes=10,
+        minimum_nprobes=10,
+        maximum_nprobes=10,
         refine_factor=5,
         postfilter=False,
         with_row_id=False,
@@ -1012,6 +1180,18 @@ async def test_query_serialization_async(table_async: AsyncTable):
         limit=10,
     )
 
+    q = (await table_async.search([5.0, 6.0])).minimum_nprobes(5).to_query_object()
+    check_set_props(
+        q,
+        vector=sample_vector,
+        minimum_nprobes=5,
+        maximum_nprobes=20,
+        postfilter=False,
+        with_row_id=False,
+        bypass_vector_index=False,
+        limit=10,
+    )
+
     q = (
         (await table_async.search([5.0, 6.0]))
         .distance_range(0.0, 1.0)
@@ -1023,7 +1203,8 @@ async def test_query_serialization_async(table_async: AsyncTable):
         lower_bound=0.0,
         upper_bound=1.0,
         postfilter=False,
-        nprobes=20,
+        minimum_nprobes=20,
+        maximum_nprobes=20,
         with_row_id=False,
         bypass_vector_index=False,
         limit=10,
@@ -1035,7 +1216,8 @@ async def test_query_serialization_async(table_async: AsyncTable):
         distance_type="cosine",
         vector=sample_vector,
         postfilter=False,
-        nprobes=20,
+        minimum_nprobes=20,
+        maximum_nprobes=20,
         with_row_id=False,
         bypass_vector_index=False,
         limit=10,
@@ -1047,7 +1229,8 @@ async def test_query_serialization_async(table_async: AsyncTable):
         ef=7,
         vector=sample_vector,
         postfilter=False,
-        nprobes=20,
+        minimum_nprobes=20,
+        maximum_nprobes=20,
         with_row_id=False,
         bypass_vector_index=False,
         limit=10,
@@ -1059,24 +1242,34 @@ async def test_query_serialization_async(table_async: AsyncTable):
         bypass_vector_index=True,
         vector=sample_vector,
         postfilter=False,
-        nprobes=20,
+        minimum_nprobes=20,
+        maximum_nprobes=20,
         with_row_id=False,
         limit=10,
     )
 
     # FTS queries
-    q = (await table_async.search("foo")).limit(10).to_query_object()
+    match_query = MatchQuery("foo", "text")
+    q = (await table_async.search(match_query)).limit(10).to_query_object()
     check_set_props(
         q,
         limit=10,
-        full_text_query=FullTextSearchQuery(columns=[], query="foo"),
+        full_text_query=FullTextSearchQuery(columns=None, query=match_query),
         with_row_id=False,
     )
 
-    q = (await table_async.search("foo", query_type="fts")).to_query_object()
+    q = (await table_async.search(match_query)).to_query_object()
     check_set_props(
         q,
-        full_text_query=FullTextSearchQuery(columns=[], query="foo"),
+        full_text_query=FullTextSearchQuery(columns=None, query=match_query),
+        with_row_id=False,
+    )
+
+    phrase_query = PhraseQuery("foo", "text", slop=1)
+    q = (await table_async.search(phrase_query)).to_query_object()
+    check_set_props(
+        q,
+        full_text_query=FullTextSearchQuery(columns=None, query=phrase_query),
         with_row_id=False,
     )
 

python/python/tests/test_remote_db.py

@@ -149,6 +149,24 @@ async def test_async_checkout():
         assert await table.count_rows() == 300
 
 
+def test_table_len_sync():
+    def handler(request):
+        if request.path == "/v1/table/test/create/?mode=create":
+            request.send_response(200)
+            request.send_header("Content-Type", "application/json")
+            request.end_headers()
+            request.wfile.write(b"{}")
+
+        request.send_response(200)
+        request.send_header("Content-Type", "application/json")
+        request.end_headers()
+        request.wfile.write(json.dumps(1).encode())
+
+    with mock_lancedb_connection(handler) as db:
+        table = db.create_table("test", [{"id": 1}])
+        assert len(table) == 1
+
+
 @pytest.mark.asyncio
 async def test_http_error():
     request_id_holder = {"request_id": None}
@@ -192,6 +210,25 @@ async def test_retry_error():
         assert cause.status_code == 429
 
 
+def test_table_unimplemented_functions():
+    def handler(request):
+        if request.path == "/v1/table/test/create/?mode=create":
+            request.send_response(200)
+            request.send_header("Content-Type", "application/json")
+            request.end_headers()
+            request.wfile.write(b"{}")
+        else:
+            request.send_response(404)
+            request.end_headers()
+
+    with mock_lancedb_connection(handler) as db:
+        table = db.create_table("test", [{"id": 1}])
+        with pytest.raises(NotImplementedError):
+            table.to_arrow()
+        with pytest.raises(NotImplementedError):
+            table.to_pandas()
+
+
 def test_table_add_in_threadpool():
     def handler(request):
         if request.path == "/v1/table/test/insert/":
@@ -478,6 +515,8 @@ def test_query_sync_minimal():
             "ef": None,
             "vector": [1.0, 2.0, 3.0],
             "nprobes": 20,
+            "minimum_nprobes": 20,
+            "maximum_nprobes": 20,
             "version": None,
         }
 
@@ -518,6 +557,8 @@ def test_query_sync_maximal():
             "refine_factor": 10,
             "vector": [1.0, 2.0, 3.0],
             "nprobes": 5,
+            "minimum_nprobes": 5,
+            "maximum_nprobes": 5,
             "lower_bound": None,
             "upper_bound": None,
             "ef": None,
@@ -546,6 +587,66 @@ def test_query_sync_maximal():
         )
 
 
+def test_query_sync_nprobes():
+    def handler(body):
+        assert body == {
+            "distance_type": "l2",
+            "k": 10,
+            "prefilter": True,
+            "fast_search": True,
+            "vector_column": "vector2",
+            "refine_factor": None,
+            "lower_bound": None,
+            "upper_bound": None,
+            "ef": None,
+            "vector": [1.0, 2.0, 3.0],
+            "nprobes": 5,
+            "minimum_nprobes": 5,
+            "maximum_nprobes": 15,
+            "version": None,
+        }
+
+        return pa.table({"id": [1, 2, 3], "name": ["a", "b", "c"]})
+
+    with query_test_table(handler) as table:
+        (
+            table.search([1, 2, 3], vector_column_name="vector2", fast_search=True)
+            .minimum_nprobes(5)
+            .maximum_nprobes(15)
+            .to_list()
+        )
+
+
+def test_query_sync_no_max_nprobes():
+    def handler(body):
+        assert body == {
+            "distance_type": "l2",
+            "k": 10,
+            "prefilter": True,
+            "fast_search": True,
+            "vector_column": "vector2",
+            "refine_factor": None,
+            "lower_bound": None,
+            "upper_bound": None,
+            "ef": None,
+            "vector": [1.0, 2.0, 3.0],
+            "nprobes": 5,
+            "minimum_nprobes": 5,
+            "maximum_nprobes": 0,
+            "version": None,
+        }
+
+        return pa.table({"id": [1, 2, 3], "name": ["a", "b", "c"]})
+
+    with query_test_table(handler) as table:
+        (
+            table.search([1, 2, 3], vector_column_name="vector2", fast_search=True)
+            .minimum_nprobes(5)
+            .maximum_nprobes(0)
+            .to_list()
+        )
+
+
 @pytest.mark.parametrize("server_version", [Version("0.1.0"), Version("0.2.0")])
 def test_query_sync_batch_queries(server_version):
     def handler(body):
@@ -648,6 +749,8 @@ def test_query_sync_hybrid():
                 "refine_factor": None,
                 "vector": [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
                 "nprobes": 20,
+                "minimum_nprobes": 20,
+                "maximum_nprobes": 20,
                 "lower_bound": None,
                 "upper_bound": None,
                 "ef": None,

python/python/tests/test_s3.py

@@ -245,7 +245,7 @@ def test_s3_dynamodb_sync(s3_bucket: str, commit_table: str, monkeypatch):
         NotImplementedError,
         match="Full-text search is only supported on the local filesystem",
     ):
-        table.create_fts_index("x")
+        table.create_fts_index("x", use_tantivy=True)
 
     # make sure list tables still works
     assert db.table_names() == ["test_ddb_sync"]

python/python/tests/test_table.py

@@ -106,15 +106,22 @@ async def test_update_async(mem_db_async: AsyncConnection):
     table = await mem_db_async.create_table("some_table", data=[{"id": 0}])
     assert await table.count_rows("id == 0") == 1
     assert await table.count_rows("id == 7") == 0
-    await table.update({"id": 7})
+    update_res = await table.update({"id": 7})
+    assert update_res.rows_updated == 1
+    assert update_res.version == 2
     assert await table.count_rows("id == 7") == 1
     assert await table.count_rows("id == 0") == 0
-    await table.add([{"id": 2}])
-    await table.update(where="id % 2 == 0", updates_sql={"id": "5"})
+    add_res = await table.add([{"id": 2}])
+    assert add_res.version == 3
+    update_res = await table.update(where="id % 2 == 0", updates_sql={"id": "5"})
+    assert update_res.rows_updated == 1
+    assert update_res.version == 4
     assert await table.count_rows("id == 7") == 1
     assert await table.count_rows("id == 2") == 0
     assert await table.count_rows("id == 5") == 1
-    await table.update({"id": 10}, where="id == 5")
+    update_res = await table.update({"id": 10}, where="id == 5")
+    assert update_res.rows_updated == 1
+    assert update_res.version == 5
     assert await table.count_rows("id == 10") == 1
 
 
@@ -437,7 +444,8 @@ def test_add_pydantic_model(mem_db: DBConnection):
             content="foo", meta=Metadata(source="bar", timestamp=datetime.now())
         ),
     )
-    tbl.add([expected])
+    add_res = tbl.add([expected])
+    assert add_res.version == 2
 
     result = tbl.search([0.0, 0.0]).limit(1).to_pydantic(LanceSchema)[0]
     assert result == expected
@@ -459,11 +467,12 @@ async def test_add_async(mem_db_async: AsyncConnection):
         ],
     )
     assert await table.count_rows() == 2
-    await table.add(
+    add_res = await table.add(
         data=[
             {"vector": [10.0, 11.0], "item": "baz", "price": 30.0},
         ],
     )
+    assert add_res.version == 2
     assert await table.count_rows() == 3
 
 
@@ -760,6 +769,29 @@ def test_restore(mem_db: DBConnection):
         table.restore(0)
 
 
+def test_restore_with_tags(mem_db: DBConnection):
+    table = mem_db.create_table(
+        "my_table",
+        data=[{"vector": [1.1, 0.9], "type": "vector"}],
+    )
+    tag = "tag1"
+    table.tags.create(tag, 1)
+    table.add([{"vector": [0.5, 0.2], "type": "vector"}])
+    table.restore(tag)
+    assert len(table.list_versions()) == 3
+    assert len(table) == 1
+    expected = table.to_arrow()
+
+    table.add([{"vector": [0.3, 0.3], "type": "vector"}])
+    table.checkout("tag1")
+    table.restore()
+    assert len(table.list_versions()) == 5
+    assert table.to_arrow() == expected
+
+    with pytest.raises(ValueError):
+        table.restore("tag_unknown")
+
+
 def test_merge(tmp_db: DBConnection, tmp_path):
     pytest.importorskip("lance")
     import lance
@@ -795,7 +827,8 @@ def test_delete(mem_db: DBConnection):
     )
     assert len(table) == 2
     assert len(table.list_versions()) == 1
-    table.delete("id=0")
+    delete_res = table.delete("id=0")
+    assert delete_res.version == 2
     assert len(table.list_versions()) == 2
     assert table.version == 2
     assert len(table) == 1
@@ -809,7 +842,9 @@ def test_update(mem_db: DBConnection):
     )
     assert len(table) == 2
     assert len(table.list_versions()) == 1
-    table.update(where="id=0", values={"vector": [1.1, 1.1]})
+    update_res = table.update(where="id=0", values={"vector": [1.1, 1.1]})
+    assert update_res.version == 2
+    assert update_res.rows_updated == 1
     assert len(table.list_versions()) == 2
     assert table.version == 2
     assert len(table) == 2
@@ -898,9 +933,16 @@ def test_merge_insert(mem_db: DBConnection):
     new_data = pa.table({"a": [2, 3, 4], "b": ["x", "y", "z"]})
 
     # upsert
-    table.merge_insert(
-        "a"
-    ).when_matched_update_all().when_not_matched_insert_all().execute(new_data)
+    merge_insert_res = (
+        table.merge_insert("a")
+        .when_matched_update_all()
+        .when_not_matched_insert_all()
+        .execute(new_data, timeout=timedelta(seconds=10))
+    )
+    assert merge_insert_res.version == 2
+    assert merge_insert_res.num_inserted_rows == 1
+    assert merge_insert_res.num_updated_rows == 2
+    assert merge_insert_res.num_deleted_rows == 0
 
     expected = pa.table({"a": [1, 2, 3, 4], "b": ["a", "x", "y", "z"]})
     assert table.to_arrow().sort_by("a") == expected
@@ -908,17 +950,28 @@ def test_merge_insert(mem_db: DBConnection):
     table.restore(version)
 
     # conditional update
-    table.merge_insert("a").when_matched_update_all(where="target.b = 'b'").execute(
-        new_data
+    merge_insert_res = (
+        table.merge_insert("a")
+        .when_matched_update_all(where="target.b = 'b'")
+        .execute(new_data)
     )
+    assert merge_insert_res.version == 4
+    assert merge_insert_res.num_inserted_rows == 0
+    assert merge_insert_res.num_updated_rows == 1
+    assert merge_insert_res.num_deleted_rows == 0
     expected = pa.table({"a": [1, 2, 3], "b": ["a", "x", "c"]})
     assert table.to_arrow().sort_by("a") == expected
 
     table.restore(version)
 
     # insert-if-not-exists
-    table.merge_insert("a").when_not_matched_insert_all().execute(new_data)
-
+    merge_insert_res = (
+        table.merge_insert("a").when_not_matched_insert_all().execute(new_data)
+    )
+    assert merge_insert_res.version == 6
+    assert merge_insert_res.num_inserted_rows == 1
+    assert merge_insert_res.num_updated_rows == 0
+    assert merge_insert_res.num_deleted_rows == 0
     expected = pa.table({"a": [1, 2, 3, 4], "b": ["a", "b", "c", "z"]})
     assert table.to_arrow().sort_by("a") == expected
 
@@ -927,13 +980,17 @@ def test_merge_insert(mem_db: DBConnection):
     new_data = pa.table({"a": [2, 4], "b": ["x", "z"]})
 
     # replace-range
-    (
+    merge_insert_res = (
         table.merge_insert("a")
         .when_matched_update_all()
         .when_not_matched_insert_all()
         .when_not_matched_by_source_delete("a > 2")
         .execute(new_data)
     )
+    assert merge_insert_res.version == 8
+    assert merge_insert_res.num_inserted_rows == 1
+    assert merge_insert_res.num_updated_rows == 1
+    assert merge_insert_res.num_deleted_rows == 1
 
     expected = pa.table({"a": [1, 2, 4], "b": ["a", "x", "z"]})
     assert table.to_arrow().sort_by("a") == expected
@@ -941,15 +998,27 @@ def test_merge_insert(mem_db: DBConnection):
     table.restore(version)
 
     # replace-range no condition
-    table.merge_insert(
-        "a"
-    ).when_matched_update_all().when_not_matched_insert_all().when_not_matched_by_source_delete().execute(
-        new_data
+    merge_insert_res = (
+        table.merge_insert("a")
+        .when_matched_update_all()
+        .when_not_matched_insert_all()
+        .when_not_matched_by_source_delete()
+        .execute(new_data)
     )
+    assert merge_insert_res.version == 10
+    assert merge_insert_res.num_inserted_rows == 1
+    assert merge_insert_res.num_updated_rows == 1
+    assert merge_insert_res.num_deleted_rows == 2
 
     expected = pa.table({"a": [2, 4], "b": ["x", "z"]})
     assert table.to_arrow().sort_by("a") == expected
 
+    # timeout
+    with pytest.raises(Exception, match="merge insert timed out"):
+        table.merge_insert("a").when_matched_update_all().execute(
+            new_data, timeout=timedelta(0)
+        )
+
 
 # We vary the data format because there are slight differences in how
 # subschemas are handled in different formats
@@ -1478,11 +1547,13 @@ def test_restore_consistency(tmp_path):
 def test_add_columns(mem_db: DBConnection):
     data = pa.table({"id": [0, 1]})
     table = LanceTable.create(mem_db, "my_table", data=data)
-    table.add_columns({"new_col": "id + 2"})
+    add_columns_res = table.add_columns({"new_col": "id + 2"})
+    assert add_columns_res.version == 2
     assert table.to_arrow().column_names == ["id", "new_col"]
     assert table.to_arrow()["new_col"].to_pylist() == [2, 3]
 
-    table.add_columns({"null_int": "cast(null as bigint)"})
+    add_columns_res = table.add_columns({"null_int": "cast(null as bigint)"})
+    assert add_columns_res.version == 3
     assert table.schema.field("null_int").type == pa.int64()
 
 
@@ -1490,7 +1561,8 @@ def test_add_columns(mem_db: DBConnection):
 async def test_add_columns_async(mem_db_async: AsyncConnection):
     data = pa.table({"id": [0, 1]})
     table = await mem_db_async.create_table("my_table", data=data)
-    await table.add_columns({"new_col": "id + 2"})
+    add_columns_res = await table.add_columns({"new_col": "id + 2"})
+    assert add_columns_res.version == 2
     data = await table.to_arrow()
     assert data.column_names == ["id", "new_col"]
     assert data["new_col"].to_pylist() == [2, 3]
@@ -1500,9 +1572,10 @@ async def test_add_columns_async(mem_db_async: AsyncConnection):
 async def test_add_columns_with_schema(mem_db_async: AsyncConnection):
     data = pa.table({"id": [0, 1]})
     table = await mem_db_async.create_table("my_table", data=data)
-    await table.add_columns(
+    add_columns_res = await table.add_columns(
         [pa.field("x", pa.int64()), pa.field("vector", pa.list_(pa.float32(), 8))]
     )
+    assert add_columns_res.version == 2
 
     assert await table.schema() == pa.schema(
         [
@@ -1513,11 +1586,12 @@ async def test_add_columns_with_schema(mem_db_async: AsyncConnection):
     )
 
     table = await mem_db_async.create_table("table2", data=data)
-    await table.add_columns(
+    add_columns_res = await table.add_columns(
         pa.schema(
             [pa.field("y", pa.int64()), pa.field("emb", pa.list_(pa.float32(), 8))]
         )
     )
+    assert add_columns_res.version == 2
     assert await table.schema() == pa.schema(
         [
             pa.field("id", pa.int64()),
@@ -1530,7 +1604,8 @@ async def test_add_columns_with_schema(mem_db_async: AsyncConnection):
 def test_alter_columns(mem_db: DBConnection):
     data = pa.table({"id": [0, 1]})
     table = mem_db.create_table("my_table", data=data)
-    table.alter_columns({"path": "id", "rename": "new_id"})
+    alter_columns_res = table.alter_columns({"path": "id", "rename": "new_id"})
+    assert alter_columns_res.version == 2
     assert table.to_arrow().column_names == ["new_id"]
 
 
@@ -1538,9 +1613,13 @@ def test_alter_columns(mem_db: DBConnection):
 async def test_alter_columns_async(mem_db_async: AsyncConnection):
     data = pa.table({"id": [0, 1]})
     table = await mem_db_async.create_table("my_table", data=data)
-    await table.alter_columns({"path": "id", "rename": "new_id"})
+    alter_columns_res = await table.alter_columns({"path": "id", "rename": "new_id"})
+    assert alter_columns_res.version == 2
     assert (await table.to_arrow()).column_names == ["new_id"]
-    await table.alter_columns(dict(path="new_id", data_type=pa.int16(), nullable=True))
+    alter_columns_res = await table.alter_columns(
+        dict(path="new_id", data_type=pa.int16(), nullable=True)
+    )
+    assert alter_columns_res.version == 3
     data = await table.to_arrow()
     assert data.column(0).type == pa.int16()
     assert data.schema.field(0).nullable
@@ -1549,7 +1628,8 @@ async def test_alter_columns_async(mem_db_async: AsyncConnection):
 def test_drop_columns(mem_db: DBConnection):
     data = pa.table({"id": [0, 1], "category": ["a", "b"]})
     table = mem_db.create_table("my_table", data=data)
-    table.drop_columns(["category"])
+    drop_columns_res = table.drop_columns(["category"])
+    assert drop_columns_res.version == 2
     assert table.to_arrow().column_names == ["id"]
 
 
@@ -1557,7 +1637,8 @@ def test_drop_columns(mem_db: DBConnection):
 async def test_drop_columns_async(mem_db_async: AsyncConnection):
     data = pa.table({"id": [0, 1], "category": ["a", "b"]})
     table = await mem_db_async.create_table("my_table", data=data)
-    await table.drop_columns(["category"])
+    drop_columns_res = await table.drop_columns(["category"])
+    assert drop_columns_res.version == 2
     assert (await table.to_arrow()).column_names == ["id"]
 
 

python/src/index.rs

@@ -3,7 +3,7 @@
 
 use lancedb::index::vector::IvfFlatIndexBuilder;
 use lancedb::index::{
-    scalar::{BTreeIndexBuilder, FtsIndexBuilder, TokenizerConfig},
+    scalar::{BTreeIndexBuilder, FtsIndexBuilder},
     vector::{IvfHnswPqIndexBuilder, IvfHnswSqIndexBuilder, IvfPqIndexBuilder},
     Index as LanceDbIndex,
 };
@@ -38,19 +38,17 @@ pub fn extract_index_params(source: &Option<Bound<'_, PyAny>>) -> PyResult<Lance
             "LabelList" => Ok(LanceDbIndex::LabelList(Default::default())),
             "FTS" => {
                 let params = source.extract::<FtsParams>()?;
-                let inner_opts = TokenizerConfig::default()
+                let inner_opts = FtsIndexBuilder::default()
                     .base_tokenizer(params.base_tokenizer)
                     .language(&params.language)
                     .map_err(|_| PyValueError::new_err(format!("LanceDB does not support the requested language: '{}'", params.language)))?
+                    .with_position(params.with_position)
                     .lower_case(params.lower_case)
                     .max_token_length(params.max_token_length)
                     .remove_stop_words(params.remove_stop_words)
                     .stem(params.stem)
                     .ascii_folding(params.ascii_folding);
-                let mut opts = FtsIndexBuilder::default()
-                    .with_position(params.with_position);
-                opts.tokenizer_configs = inner_opts;
-                Ok(LanceDbIndex::FTS(opts))
+                Ok(LanceDbIndex::FTS(inner_opts))
             },
             "IvfFlat" => {
                 let params = source.extract::<IvfFlatParams>()?;

python/src/lib.rs

@@ -11,7 +11,10 @@ use pyo3::{
     wrap_pyfunction, Bound, PyResult, Python,
 };
 use query::{FTSQuery, HybridQuery, Query, VectorQuery};
-use table::Table;
+use table::{
+    AddColumnsResult, AddResult, AlterColumnsResult, DeleteResult, DropColumnsResult, MergeResult,
+    Table, UpdateResult,
+};
 
 pub mod arrow;
 pub mod connection;
@@ -35,6 +38,13 @@ pub fn _lancedb(_py: Python, m: &Bound<'_, PyModule>) -> PyResult<()> {
     m.add_class::<HybridQuery>()?;
     m.add_class::<VectorQuery>()?;
     m.add_class::<RecordBatchStream>()?;
+    m.add_class::<AddColumnsResult>()?;
+    m.add_class::<AlterColumnsResult>()?;
+    m.add_class::<AddResult>()?;
+    m.add_class::<MergeResult>()?;
+    m.add_class::<DeleteResult>()?;
+    m.add_class::<DropColumnsResult>()?;
+    m.add_class::<UpdateResult>()?;
     m.add_function(wrap_pyfunction!(connect, m)?)?;
     m.add_function(wrap_pyfunction!(util::validate_table_name, m)?)?;
     m.add("__version__", env!("CARGO_PKG_VERSION"))?;

python/src/query.rs

@@ -9,15 +9,16 @@ use arrow::array::Array;
 use arrow::array::ArrayData;
 use arrow::pyarrow::FromPyArrow;
 use arrow::pyarrow::IntoPyArrow;
-use lancedb::index::scalar::{FtsQuery, FullTextSearchQuery, MatchQuery, PhraseQuery};
+use lancedb::index::scalar::{
+    BooleanQuery, BoostQuery, FtsQuery, FullTextSearchQuery, MatchQuery, MultiMatchQuery, Occur,
+    Operator, PhraseQuery,
+};
 use lancedb::query::QueryExecutionOptions;
 use lancedb::query::QueryFilter;
 use lancedb::query::{
     ExecutableQuery, Query as LanceDbQuery, QueryBase, Select, VectorQuery as LanceDbVectorQuery,
 };
 use lancedb::table::AnyQuery;
-use pyo3::exceptions::PyRuntimeError;
-use pyo3::exceptions::{PyNotImplementedError, PyValueError};
 use pyo3::prelude::{PyAnyMethods, PyDictMethods};
 use pyo3::pymethods;
 use pyo3::types::PyList;
@@ -27,30 +28,172 @@ use pyo3::IntoPyObject;
 use pyo3::PyAny;
 use pyo3::PyRef;
 use pyo3::PyResult;
+use pyo3::{exceptions::PyRuntimeError, FromPyObject};
+use pyo3::{
+    exceptions::{PyNotImplementedError, PyValueError},
+    intern,
+};
 use pyo3::{pyclass, PyErr};
 use pyo3_async_runtimes::tokio::future_into_py;
 
-use crate::arrow::RecordBatchStream;
-use crate::error::PythonErrorExt;
-use crate::util::{parse_distance_type, parse_fts_query};
+use crate::util::parse_distance_type;
+use crate::{arrow::RecordBatchStream, util::PyLanceDB};
+use crate::{error::PythonErrorExt, index::class_name};
 
-// Python representation of full text search parameters
-#[derive(Clone)]
-#[pyclass(get_all)]
-pub struct PyFullTextSearchQuery {
-    pub columns: Vec<String>,
-    pub query: String,
-    pub limit: Option<i64>,
-    pub wand_factor: Option<f32>,
+impl FromPyObject<'_> for PyLanceDB<FtsQuery> {
+    fn extract_bound(ob: &Bound<'_, PyAny>) -> PyResult<Self> {
+        match class_name(ob)?.as_str() {
+            "MatchQuery" => {
+                let query = ob.getattr("query")?.extract()?;
+                let column = ob.getattr("column")?.extract()?;
+                let boost = ob.getattr("boost")?.extract()?;
+                let fuzziness = ob.getattr("fuzziness")?.extract()?;
+                let max_expansions = ob.getattr("max_expansions")?.extract()?;
+                let operator = ob.getattr("operator")?.extract::<String>()?;
+                let prefix_length = ob.getattr("prefix_length")?.extract()?;
+
+                Ok(Self(
+                    MatchQuery::new(query)
+                        .with_column(Some(column))
+                        .with_boost(boost)
+                        .with_fuzziness(fuzziness)
+                        .with_max_expansions(max_expansions)
+                        .with_operator(Operator::try_from(operator.as_str()).map_err(|e| {
+                            PyValueError::new_err(format!("Invalid operator: {}", e))
+                        })?)
+                        .with_prefix_length(prefix_length)
+                        .into(),
+                ))
+            }
+            "PhraseQuery" => {
+                let query = ob.getattr("query")?.extract()?;
+                let column = ob.getattr("column")?.extract()?;
+                let slop = ob.getattr("slop")?.extract()?;
+
+                Ok(Self(
+                    PhraseQuery::new(query)
+                        .with_column(Some(column))
+                        .with_slop(slop)
+                        .into(),
+                ))
+            }
+            "BoostQuery" => {
+                let positive: Self = ob.getattr("positive")?.extract()?;
+                let negative: Self = ob.getattr("negative")?.extract()?;
+                let negative_boost = ob.getattr("negative_boost")?.extract()?;
+                Ok(Self(
+                    BoostQuery::new(positive.0, negative.0, negative_boost).into(),
+                ))
+            }
+            "MultiMatchQuery" => {
+                let query = ob.getattr("query")?.extract()?;
+                let columns = ob.getattr("columns")?.extract()?;
+                let boosts: Option<Vec<f32>> = ob.getattr("boosts")?.extract()?;
+                let operator: String = ob.getattr("operator")?.extract()?;
+
+                let q = MultiMatchQuery::try_new(query, columns)
+                    .map_err(|e| PyValueError::new_err(format!("Invalid query: {}", e)))?;
+                let q = if let Some(boosts) = boosts {
+                    q.try_with_boosts(boosts)
+                        .map_err(|e| PyValueError::new_err(format!("Invalid boosts: {}", e)))?
+                } else {
+                    q
+                };
+
+                let op = Operator::try_from(operator.as_str())
+                    .map_err(|e| PyValueError::new_err(format!("Invalid operator: {}", e)))?;
+
+                Ok(Self(q.with_operator(op).into()))
+            }
+            "BooleanQuery" => {
+                let queries: Vec<(String, Self)> = ob.getattr("queries")?.extract()?;
+                let mut sub_queries = Vec::with_capacity(queries.len());
+                for (occur, q) in queries {
+                    let occur = Occur::try_from(occur.as_str())
+                        .map_err(|e| PyValueError::new_err(e.to_string()))?;
+                    sub_queries.push((occur, q.0));
+                }
+                Ok(Self(BooleanQuery::new(sub_queries).into()))
+            }
+            name => Err(PyValueError::new_err(format!(
+                "Unsupported FTS query type: {}",
+                name
+            ))),
+        }
+    }
 }
 
-impl From<FullTextSearchQuery> for PyFullTextSearchQuery {
-    fn from(query: FullTextSearchQuery) -> Self {
-        Self {
-            columns: query.columns().into_iter().collect(),
-            query: query.query.query().to_owned(),
-            limit: query.limit,
-            wand_factor: query.wand_factor,
+impl<'py> IntoPyObject<'py> for PyLanceDB<FtsQuery> {
+    type Target = PyAny;
+    type Output = Bound<'py, Self::Target>;
+    type Error = PyErr;
+
+    fn into_pyobject(self, py: pyo3::Python<'py>) -> PyResult<Self::Output> {
+        let namespace = py
+            .import(intern!(py, "lancedb"))
+            .and_then(|m| m.getattr(intern!(py, "query")))
+            .expect("Failed to import namespace");
+
+        match self.0 {
+            FtsQuery::Match(query) => {
+                let kwargs = PyDict::new(py);
+                kwargs.set_item("boost", query.boost)?;
+                kwargs.set_item("fuzziness", query.fuzziness)?;
+                kwargs.set_item("max_expansions", query.max_expansions)?;
+                kwargs.set_item::<_, &str>("operator", query.operator.into())?;
+                kwargs.set_item("prefix_length", query.prefix_length)?;
+                namespace
+                    .getattr(intern!(py, "MatchQuery"))?
+                    .call((query.terms, query.column.unwrap()), Some(&kwargs))
+            }
+            FtsQuery::Phrase(query) => {
+                let kwargs = PyDict::new(py);
+                kwargs.set_item("slop", query.slop)?;
+                namespace
+                    .getattr(intern!(py, "PhraseQuery"))?
+                    .call((query.terms, query.column.unwrap()), Some(&kwargs))
+            }
+            FtsQuery::Boost(query) => {
+                let positive = Self(query.positive.as_ref().clone()).into_pyobject(py)?;
+                let negative = Self(query.negative.as_ref().clone()).into_pyobject(py)?;
+                let kwargs = PyDict::new(py);
+                kwargs.set_item("negative_boost", query.negative_boost)?;
+                namespace
+                    .getattr(intern!(py, "BoostQuery"))?
+                    .call((positive, negative), Some(&kwargs))
+            }
+            FtsQuery::MultiMatch(query) => {
+                let first = &query.match_queries[0];
+                let (columns, boosts): (Vec<_>, Vec<_>) = query
+                    .match_queries
+                    .iter()
+                    .map(|q| (q.column.as_ref().unwrap().clone(), q.boost))
+                    .unzip();
+                let kwargs = PyDict::new(py);
+                kwargs.set_item("boosts", boosts)?;
+                kwargs.set_item::<_, &str>("operator", first.operator.into())?;
+                namespace
+                    .getattr(intern!(py, "MultiMatchQuery"))?
+                    .call((first.terms.clone(), columns), Some(&kwargs))
+            }
+            FtsQuery::Boolean(query) => {
+                let mut queries: Vec<(&str, Bound<'py, PyAny>)> = Vec::with_capacity(
+                    query.should.len() + query.must.len() + query.must_not.len(),
+                );
+                for q in query.should {
+                    queries.push((Occur::Should.into(), Self(q).into_pyobject(py)?));
+                }
+                for q in query.must {
+                    queries.push((Occur::Must.into(), Self(q).into_pyobject(py)?));
+                }
+                for q in query.must_not {
+                    queries.push((Occur::MustNot.into(), Self(q).into_pyobject(py)?));
+                }
+
+                namespace
+                    .getattr(intern!(py, "BooleanQuery"))?
+                    .call1((queries,))
+            }
         }
     }
 }
@@ -80,13 +223,16 @@ pub struct PyQueryRequest {
     pub limit: Option<usize>,
     pub offset: Option<usize>,
     pub filter: Option<PyQueryFilter>,
-    pub full_text_search: Option<PyFullTextSearchQuery>,
+    pub full_text_search: Option<PyLanceDB<FtsQuery>>,
     pub select: PySelect,
     pub fast_search: Option<bool>,
     pub with_row_id: Option<bool>,
     pub column: Option<String>,
     pub query_vector: Option<PyQueryVectors>,
-    pub nprobes: Option<usize>,
+    pub minimum_nprobes: Option<usize>,
+    // None means user did not set it and default shoud be used (currenty 20)
+    // Some(0) means user set it to None and there is no limit
+    pub maximum_nprobes: Option<usize>,
     pub lower_bound: Option<f32>,
     pub upper_bound: Option<f32>,
     pub ef: Option<usize>,
@@ -106,13 +252,14 @@ impl From<AnyQuery> for PyQueryRequest {
                 filter: query_request.filter.map(PyQueryFilter),
                 full_text_search: query_request
                     .full_text_search
-                    .map(PyFullTextSearchQuery::from),
+                    .map(|fts| PyLanceDB(fts.query)),
                 select: PySelect(query_request.select),
                 fast_search: Some(query_request.fast_search),
                 with_row_id: Some(query_request.with_row_id),
                 column: None,
                 query_vector: None,
-                nprobes: None,
+                minimum_nprobes: None,
+                maximum_nprobes: None,
                 lower_bound: None,
                 upper_bound: None,
                 ef: None,
@@ -132,7 +279,11 @@ impl From<AnyQuery> for PyQueryRequest {
                 with_row_id: Some(vector_query.base.with_row_id),
                 column: vector_query.column,
                 query_vector: Some(PyQueryVectors(vector_query.query_vector)),
-                nprobes: Some(vector_query.nprobes),
+                minimum_nprobes: Some(vector_query.minimum_nprobes),
+                maximum_nprobes: match vector_query.maximum_nprobes {
+                    None => Some(0),
+                    Some(value) => Some(value),
+                },
                 lower_bound: vector_query.lower_bound,
                 upper_bound: vector_query.upper_bound,
                 ef: vector_query.ef,
@@ -269,8 +420,8 @@ impl Query {
                 }
             };
             let mut query = FullTextSearchQuery::new_query(query);
-            if let Some(cols) = columns {
-                if !cols.is_empty() {
+            match columns {
+                Some(cols) if !cols.is_empty() => {
                     query = query.with_columns(&cols).map_err(|e| {
                         PyValueError::new_err(format!(
                             "Failed to set full text search columns: {}",
@@ -278,15 +429,12 @@ impl Query {
                         ))
                     })?;
                 }
+                _ => {}
             }
             query
-        } else if let Ok(query) = fts_query.downcast::<PyDict>() {
-            let query = parse_fts_query(query)?;
-            FullTextSearchQuery::new_query(query)
         } else {
-            return Err(PyValueError::new_err(
-                "query must be a string or a Query object",
-            ));
+            let query = fts_query.extract::<PyLanceDB<FtsQuery>>()?;
+            FullTextSearchQuery::new_query(query.0)
         };
 
         Ok(FTSQuery {
@@ -414,7 +562,10 @@ impl FTSQuery {
     }
 
     pub fn explain_plan(self_: PyRef<'_, Self>, verbose: bool) -> PyResult<Bound<'_, PyAny>> {
-        let inner = self_.inner.clone();
+        let inner = self_
+            .inner
+            .clone()
+            .full_text_search(self_.fts_query.clone());
         future_into_py(self_.py(), async move {
             inner
                 .explain_plan(verbose)
@@ -424,7 +575,10 @@ impl FTSQuery {
     }
 
     pub fn analyze_plan(self_: PyRef<'_, Self>) -> PyResult<Bound<'_, PyAny>> {
-        let inner = self_.inner.clone();
+        let inner = self_
+            .inner
+            .clone()
+            .full_text_search(self_.fts_query.clone());
         future_into_py(self_.py(), async move {
             inner
                 .analyze_plan()
@@ -509,6 +663,29 @@ impl VectorQuery {
         self.inner = self.inner.clone().nprobes(nprobe as usize);
     }
 
+    pub fn minimum_nprobes(&mut self, minimum_nprobes: u32) -> PyResult<()> {
+        self.inner = self
+            .inner
+            .clone()
+            .minimum_nprobes(minimum_nprobes as usize)
+            .infer_error()?;
+        Ok(())
+    }
+
+    pub fn maximum_nprobes(&mut self, maximum_nprobes: u32) -> PyResult<()> {
+        let maximum_nprobes = if maximum_nprobes == 0 {
+            None
+        } else {
+            Some(maximum_nprobes as usize)
+        };
+        self.inner = self
+            .inner
+            .clone()
+            .maximum_nprobes(maximum_nprobes)
+            .infer_error()?;
+        Ok(())
+    }
+
     #[pyo3(signature = (lower_bound=None, upper_bound=None))]
     pub fn distance_range(&mut self, lower_bound: Option<f32>, upper_bound: Option<f32>) {
         self.inner = self.inner.clone().distance_range(lower_bound, upper_bound);

python/src/table.rs

@@ -17,10 +17,10 @@ use lancedb::table::{
     Table as LanceDbTable,
 };
 use pyo3::{
-    exceptions::{PyIOError, PyKeyError, PyRuntimeError, PyValueError},
+    exceptions::{PyKeyError, PyRuntimeError, PyValueError},
     pyclass, pymethods,
-    types::{IntoPyDict, PyAnyMethods, PyDict, PyDictMethods, PyInt, PyString},
-    Bound, FromPyObject, PyAny, PyObject, PyRef, PyResult, Python,
+    types::{IntoPyDict, PyAnyMethods, PyDict, PyDictMethods},
+    Bound, FromPyObject, PyAny, PyRef, PyResult, Python,
 };
 use pyo3_async_runtimes::tokio::future_into_py;
 
@@ -58,6 +58,170 @@ pub struct OptimizeStats {
     pub prune: RemovalStats,
 }
 
+#[pyclass(get_all)]
+#[derive(Clone, Debug)]
+pub struct UpdateResult {
+    pub rows_updated: u64,
+    pub version: u64,
+}
+
+#[pymethods]
+impl UpdateResult {
+    pub fn __repr__(&self) -> String {
+        format!(
+            "UpdateResult(rows_updated={}, version={})",
+            self.rows_updated, self.version
+        )
+    }
+}
+
+impl From<lancedb::table::UpdateResult> for UpdateResult {
+    fn from(result: lancedb::table::UpdateResult) -> Self {
+        Self {
+            rows_updated: result.rows_updated,
+            version: result.version,
+        }
+    }
+}
+
+#[pyclass(get_all)]
+#[derive(Clone, Debug)]
+pub struct AddResult {
+    pub version: u64,
+}
+
+#[pymethods]
+impl AddResult {
+    pub fn __repr__(&self) -> String {
+        format!("AddResult(version={})", self.version)
+    }
+}
+
+impl From<lancedb::table::AddResult> for AddResult {
+    fn from(result: lancedb::table::AddResult) -> Self {
+        Self {
+            version: result.version,
+        }
+    }
+}
+
+#[pyclass(get_all)]
+#[derive(Clone, Debug)]
+pub struct DeleteResult {
+    pub version: u64,
+}
+
+#[pymethods]
+impl DeleteResult {
+    pub fn __repr__(&self) -> String {
+        format!("DeleteResult(version={})", self.version)
+    }
+}
+
+impl From<lancedb::table::DeleteResult> for DeleteResult {
+    fn from(result: lancedb::table::DeleteResult) -> Self {
+        Self {
+            version: result.version,
+        }
+    }
+}
+
+#[pyclass(get_all)]
+#[derive(Clone, Debug)]
+pub struct MergeResult {
+    pub version: u64,
+    pub num_updated_rows: u64,
+    pub num_inserted_rows: u64,
+    pub num_deleted_rows: u64,
+}
+
+#[pymethods]
+impl MergeResult {
+    pub fn __repr__(&self) -> String {
+        format!(
+            "MergeResult(version={}, num_updated_rows={}, num_inserted_rows={}, num_deleted_rows={})",
+            self.version,
+            self.num_updated_rows,
+            self.num_inserted_rows,
+            self.num_deleted_rows
+        )
+    }
+}
+
+impl From<lancedb::table::MergeResult> for MergeResult {
+    fn from(result: lancedb::table::MergeResult) -> Self {
+        Self {
+            version: result.version,
+            num_updated_rows: result.num_updated_rows,
+            num_inserted_rows: result.num_inserted_rows,
+            num_deleted_rows: result.num_deleted_rows,
+        }
+    }
+}
+
+#[pyclass(get_all)]
+#[derive(Clone, Debug)]
+pub struct AddColumnsResult {
+    pub version: u64,
+}
+
+#[pymethods]
+impl AddColumnsResult {
+    pub fn __repr__(&self) -> String {
+        format!("AddColumnsResult(version={})", self.version)
+    }
+}
+
+impl From<lancedb::table::AddColumnsResult> for AddColumnsResult {
+    fn from(result: lancedb::table::AddColumnsResult) -> Self {
+        Self {
+            version: result.version,
+        }
+    }
+}
+
+#[pyclass(get_all)]
+#[derive(Clone, Debug)]
+pub struct AlterColumnsResult {
+    pub version: u64,
+}
+
+#[pymethods]
+impl AlterColumnsResult {
+    pub fn __repr__(&self) -> String {
+        format!("AlterColumnsResult(version={})", self.version)
+    }
+}
+
+impl From<lancedb::table::AlterColumnsResult> for AlterColumnsResult {
+    fn from(result: lancedb::table::AlterColumnsResult) -> Self {
+        Self {
+            version: result.version,
+        }
+    }
+}
+
+#[pyclass(get_all)]
+#[derive(Clone, Debug)]
+pub struct DropColumnsResult {
+    pub version: u64,
+}
+
+#[pymethods]
+impl DropColumnsResult {
+    pub fn __repr__(&self) -> String {
+        format!("DropColumnsResult(version={})", self.version)
+    }
+}
+
+impl From<lancedb::table::DropColumnsResult> for DropColumnsResult {
+    fn from(result: lancedb::table::DropColumnsResult) -> Self {
+        Self {
+            version: result.version,
+        }
+    }
+}
+
 #[pyclass]
 pub struct Table {
     // We keep a copy of the name to use if the inner table is dropped
@@ -132,15 +296,16 @@ impl Table {
         }
 
         future_into_py(self_.py(), async move {
-            op.execute().await.infer_error()?;
-            Ok(())
+            let result = op.execute().await.infer_error()?;
+            Ok(AddResult::from(result))
         })
     }
 
     pub fn delete(self_: PyRef<'_, Self>, condition: String) -> PyResult<Bound<'_, PyAny>> {
         let inner = self_.inner_ref()?.clone();
         future_into_py(self_.py(), async move {
-            inner.delete(&condition).await.infer_error()
+            let result = inner.delete(&condition).await.infer_error()?;
+            Ok(DeleteResult::from(result))
         })
     }
 
@@ -160,8 +325,8 @@ impl Table {
             op = op.column(column_name, value);
         }
         future_into_py(self_.py(), async move {
-            op.execute().await.infer_error()?;
-            Ok(())
+            let result = op.execute().await.infer_error()?;
+            Ok(UpdateResult::from(result))
         })
     }
 
@@ -355,25 +520,15 @@ impl Table {
         })
     }
 
-    pub fn checkout(self_: PyRef<'_, Self>, version: PyObject) -> PyResult<Bound<'_, PyAny>> {
+    pub fn checkout(self_: PyRef<'_, Self>, version: LanceVersion) -> PyResult<Bound<'_, PyAny>> {
         let inner = self_.inner_ref()?.clone();
         let py = self_.py();
-        let (is_int, int_value, string_value) = if let Ok(i) = version.downcast_bound::<PyInt>(py) {
-            let num: u64 = i.extract()?;
-            (true, num, String::new())
-        } else if let Ok(s) = version.downcast_bound::<PyString>(py) {
-            let str_value = s.to_string();
-            (false, 0, str_value)
-        } else {
-            return Err(PyIOError::new_err(
-                "version must be an integer or a string.",
-            ));
-        };
         future_into_py(py, async move {
-            if is_int {
-                inner.checkout(int_value).await.infer_error()
-            } else {
-                inner.checkout_tag(&string_value).await.infer_error()
+            match version {
+                LanceVersion::Version(version_num) => {
+                    inner.checkout(version_num).await.infer_error()
+                }
+                LanceVersion::Tag(tag) => inner.checkout_tag(&tag).await.infer_error(),
             }
         })
     }
@@ -386,12 +541,19 @@ impl Table {
     }
 
     #[pyo3(signature = (version=None))]
-    pub fn restore(self_: PyRef<'_, Self>, version: Option<u64>) -> PyResult<Bound<'_, PyAny>> {
+    pub fn restore(
+        self_: PyRef<'_, Self>,
+        version: Option<LanceVersion>,
+    ) -> PyResult<Bound<'_, PyAny>> {
         let inner = self_.inner_ref()?.clone();
+        let py = self_.py();
 
-        future_into_py(self_.py(), async move {
+        future_into_py(py, async move {
             if let Some(version) = version {
-                inner.checkout(version).await.infer_error()?;
+                match version {
+                    LanceVersion::Version(num) => inner.checkout(num).await.infer_error()?,
+                    LanceVersion::Tag(tag) => inner.checkout_tag(&tag).await.infer_error()?,
+                }
             }
             inner.restore().await.infer_error()
         })
@@ -487,16 +649,13 @@ impl Table {
             builder
                 .when_not_matched_by_source_delete(parameters.when_not_matched_by_source_condition);
         }
+        if let Some(timeout) = parameters.timeout {
+            builder.timeout(timeout);
+        }
 
         future_into_py(self_.py(), async move {
-            let stats = builder.execute(Box::new(batches)).await.infer_error()?;
-            Python::with_gil(|py| {
-                let dict = PyDict::new(py);
-                dict.set_item("num_inserted_rows", stats.num_inserted_rows)?;
-                dict.set_item("num_updated_rows", stats.num_updated_rows)?;
-                dict.set_item("num_deleted_rows", stats.num_deleted_rows)?;
-                Ok(dict.unbind())
-            })
+            let res = builder.execute(Box::new(batches)).await.infer_error()?;
+            Ok(MergeResult::from(res))
         })
     }
 
@@ -532,8 +691,8 @@ impl Table {
 
         let inner = self_.inner_ref()?.clone();
         future_into_py(self_.py(), async move {
-            inner.add_columns(definitions, None).await.infer_error()?;
-            Ok(())
+            let result = inner.add_columns(definitions, None).await.infer_error()?;
+            Ok(AddColumnsResult::from(result))
         })
     }
 
@@ -546,8 +705,8 @@ impl Table {
 
         let inner = self_.inner_ref()?.clone();
         future_into_py(self_.py(), async move {
-            inner.add_columns(transform, None).await.infer_error()?;
-            Ok(())
+            let result = inner.add_columns(transform, None).await.infer_error()?;
+            Ok(AddColumnsResult::from(result))
         })
     }
 
@@ -590,8 +749,8 @@ impl Table {
 
         let inner = self_.inner_ref()?.clone();
         future_into_py(self_.py(), async move {
-            inner.alter_columns(&alterations).await.infer_error()?;
-            Ok(())
+            let result = inner.alter_columns(&alterations).await.infer_error()?;
+            Ok(AlterColumnsResult::from(result))
         })
     }
 
@@ -599,8 +758,8 @@ impl Table {
         let inner = self_.inner_ref()?.clone();
         future_into_py(self_.py(), async move {
             let column_refs = columns.iter().map(String::as_str).collect::<Vec<&str>>();
-            inner.drop_columns(&column_refs).await.infer_error()?;
-            Ok(())
+            let result = inner.drop_columns(&column_refs).await.infer_error()?;
+            Ok(DropColumnsResult::from(result))
         })
     }
 
@@ -636,6 +795,12 @@ impl Table {
     }
 }
 
+#[derive(FromPyObject)]
+pub enum LanceVersion {
+    Version(u64),
+    Tag(String),
+}
+
 #[derive(FromPyObject)]
 #[pyo3(from_item_all)]
 pub struct MergeInsertParams {
@@ -645,6 +810,7 @@ pub struct MergeInsertParams {
     when_not_matched_insert_all: bool,
     when_not_matched_by_source_delete: bool,
     when_not_matched_by_source_condition: Option<String>,
+    timeout: Option<std::time::Duration>,
 }
 
 #[pyclass]

python/src/util.rs

@@ -3,15 +3,11 @@
 
 use std::sync::Mutex;
 
-use lancedb::index::scalar::{BoostQuery, FtsQuery, MatchQuery, MultiMatchQuery, PhraseQuery};
 use lancedb::DistanceType;
-use pyo3::prelude::{PyAnyMethods, PyDictMethods, PyListMethods};
-use pyo3::types::PyDict;
 use pyo3::{
     exceptions::{PyRuntimeError, PyValueError},
     pyfunction, PyResult,
 };
-use pyo3::{Bound, PyAny};
 
 /// A wrapper around a rust builder
 ///
@@ -64,116 +60,6 @@ pub fn validate_table_name(table_name: &str) -> PyResult<()> {
         .map_err(|e| PyValueError::new_err(e.to_string()))
 }
 
-pub fn parse_fts_query(query: &Bound<'_, PyDict>) -> PyResult<FtsQuery> {
-    let query_type = query.keys().get_item(0)?.extract::<String>()?;
-    let query_value = query
-        .get_item(&query_type)?
-        .ok_or(PyValueError::new_err(format!(
-            "Query type {} not found",
-            query_type
-        )))?;
-    let query_value = query_value.downcast::<PyDict>()?;
-
-    match query_type.as_str() {
-        "match" => {
-            let column = query_value.keys().get_item(0)?.extract::<String>()?;
-            let params = query_value
-                .get_item(&column)?
-                .ok_or(PyValueError::new_err(format!(
-                    "column {} not found",
-                    column
-                )))?;
-            let params = params.downcast::<PyDict>()?;
-
-            let query = params
-                .get_item("query")?
-                .ok_or(PyValueError::new_err("query not found"))?
-                .extract::<String>()?;
-            let boost = params
-                .get_item("boost")?
-                .ok_or(PyValueError::new_err("boost not found"))?
-                .extract::<f32>()?;
-            let fuzziness = params
-                .get_item("fuzziness")?
-                .ok_or(PyValueError::new_err("fuzziness not found"))?
-                .extract::<Option<u32>>()?;
-            let max_expansions = params
-                .get_item("max_expansions")?
-                .ok_or(PyValueError::new_err("max_expansions not found"))?
-                .extract::<usize>()?;
-
-            let query = MatchQuery::new(query)
-                .with_column(Some(column))
-                .with_boost(boost)
-                .with_fuzziness(fuzziness)
-                .with_max_expansions(max_expansions);
-            Ok(query.into())
-        }
-
-        "match_phrase" => {
-            let column = query_value.keys().get_item(0)?.extract::<String>()?;
-            let query = query_value
-                .get_item(&column)?
-                .ok_or(PyValueError::new_err(format!(
-                    "column {} not found",
-                    column
-                )))?
-                .extract::<String>()?;
-
-            let query = PhraseQuery::new(query).with_column(Some(column));
-            Ok(query.into())
-        }
-
-        "boost" => {
-            let positive: Bound<'_, PyAny> = query_value
-                .get_item("positive")?
-                .ok_or(PyValueError::new_err("positive not found"))?;
-            let positive = positive.downcast::<PyDict>()?;
-
-            let negative = query_value
-                .get_item("negative")?
-                .ok_or(PyValueError::new_err("negative not found"))?;
-            let negative = negative.downcast::<PyDict>()?;
-
-            let negative_boost = query_value
-                .get_item("negative_boost")?
-                .ok_or(PyValueError::new_err("negative_boost not found"))?
-                .extract::<f32>()?;
-
-            let positive_query = parse_fts_query(positive)?;
-            let negative_query = parse_fts_query(negative)?;
-            let query = BoostQuery::new(positive_query, negative_query, Some(negative_boost));
-
-            Ok(query.into())
-        }
-
-        "multi_match" => {
-            let query = query_value
-                .get_item("query")?
-                .ok_or(PyValueError::new_err("query not found"))?
-                .extract::<String>()?;
-
-            let columns = query_value
-                .get_item("columns")?
-                .ok_or(PyValueError::new_err("columns not found"))?
-                .extract::<Vec<String>>()?;
-
-            let boost = query_value
-                .get_item("boost")?
-                .ok_or(PyValueError::new_err("boost not found"))?
-                .extract::<Vec<f32>>()?;
-
-            let query = MultiMatchQuery::try_new(query, columns)
-                .and_then(|q| q.try_with_boosts(boost))
-                .map_err(|e| {
-                    PyValueError::new_err(format!("Error creating MultiMatchQuery: {}", e))
-                })?;
-            Ok(query.into())
-        }
-
-        _ => Err(PyValueError::new_err(format!(
-            "Unsupported query type: {}",
-            query_type
-        ))),
-    }
-}
+/// A wrapper around a LanceDB type to allow it to be used in Python
+#[derive(Debug, Clone)]
+pub struct PyLanceDB<T>(pub T);

rust-toolchain.toml

@@ -1,2 +1,2 @@
 [toolchain]
-channel = "1.83.0"
+channel = "1.86.0"

rust/ffi/node/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lancedb-node"
-version = "0.19.1-beta.1"
+version = "0.21.1-beta.1"
 description = "Serverless, low-latency vector database for AI applications"
 license.workspace = true
 edition.workspace = true

rust/lancedb/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lancedb"
-version = "0.19.1-beta.1"
+version = "0.21.1-beta.1"
 edition.workspace = true
 description = "LanceDB: A serverless, low-latency vector database for AI applications"
 license.workspace = true
@@ -60,15 +60,15 @@ reqwest = { version = "0.12.0", default-features = false, features = [
     "macos-system-configuration",
     "stream",
 ], optional = true }
-rand = { version = "0.8.3", features = ["small_rng"], optional = true }
+rand = { version = "0.9", features = ["small_rng"], optional = true }
 http = { version = "1", optional = true } # Matching what is in reqwest
 uuid = { version = "1.7.0", features = ["v4"], optional = true }
 polars-arrow = { version = ">=0.37,<0.40.0", optional = true }
 polars = { version = ">=0.37,<0.40.0", optional = true }
 hf-hub = { version = "0.4.1", optional = true, default-features = false, features = ["rustls-tls", "tokio", "ureq"]}
-candle-core = { version = "0.6.0", optional = true }
-candle-transformers = { version = "0.6.0", optional = true }
-candle-nn = { version = "0.6.0", optional = true }
+candle-core = { version = "0.9.1", optional = true }
+candle-transformers = { version = "0.9.1", optional = true }
+candle-nn = { version = "0.9.1", optional = true }
 tokenizers = { version = "0.19.1", optional = true }
 semver = { workspace = true }
 
@@ -78,7 +78,7 @@ bytemuck_derive.workspace = true
 
 [dev-dependencies]
 tempfile = "3.5.0"
-rand = { version = "0.8.3", features = ["small_rng"] }
+rand = { version = "0.9", features = ["small_rng"] }
 random_word = { version = "0.4.3", features = ["en"] }
 uuid = { version = "1.7.0", features = ["v4"] }
 walkdir = "2"