[python] Bump version: 0.5.7 → 0.6.0

Signed-off-by: BubbleCal <bubble-cal@outlook.com>

.github/workflows/build_linux_wheel/action.yml

@@ -0,0 +1,58 @@
+# We create a composite action to be re-used both for testing and for releasing
+name: build-linux-wheel
+description: "Build a manylinux wheel for lance"
+inputs:
+  python-minor-version:
+    description: "8, 9, 10, 11, 12"
+    required: true
+  args:
+    description: "--release"
+    required: false
+    default: ""
+  arm-build:
+    description: "Build for arm64 instead of x86_64"
+    # Note: this does *not* mean the host is arm64, since we might be cross-compiling.
+    required: false
+    default: "false"
+runs:
+  using: "composite"
+  steps:
+    - name: CONFIRM ARM BUILD
+      shell: bash
+      run: |
+        echo "ARM BUILD: ${{ inputs.arm-build }}"
+    - name: Build x86_64 Manylinux wheel
+      if: ${{ inputs.arm-build == 'false' }}
+      uses: PyO3/maturin-action@v1
+      with:
+        command: build
+        working-directory: python
+        target: x86_64-unknown-linux-gnu
+        manylinux: "2_17"
+        args: ${{ inputs.args }}
+        before-script-linux: |
+          set -e
+          yum install -y openssl-devel \
+            && curl -L https://github.com/protocolbuffers/protobuf/releases/download/v24.4/protoc-24.4-linux-$(uname -m).zip > /tmp/protoc.zip \
+            && unzip /tmp/protoc.zip -d /usr/local \
+            && rm /tmp/protoc.zip
+    - name: Build Arm Manylinux Wheel
+      if: ${{ inputs.arm-build == 'true' }}
+      uses: PyO3/maturin-action@v1
+      with:
+        command: build
+        working-directory: python
+        target: aarch64-unknown-linux-gnu
+        manylinux: "2_24"
+        args: ${{ inputs.args }}
+        before-script-linux: |
+          set -e
+          apt install -y unzip
+          if [ $(uname -m) = "x86_64" ]; then
+            PROTOC_ARCH="x86_64"
+          else
+            PROTOC_ARCH="aarch_64"
+          fi
+          curl -L https://github.com/protocolbuffers/protobuf/releases/download/v24.4/protoc-24.4-linux-$PROTOC_ARCH.zip > /tmp/protoc.zip \
+            && unzip /tmp/protoc.zip -d /usr/local \
+            && rm /tmp/protoc.zip

.github/workflows/build_mac_wheel/action.yml

@@ -0,0 +1,25 @@
+# We create a composite action to be re-used both for testing and for releasing
+name: build_wheel
+description: "Build a lance wheel"
+inputs:
+  python-minor-version:
+    description: "8, 9, 10, 11"
+    required: true
+  args:
+    description: "--release"
+    required: false
+    default: ""
+runs:
+  using: "composite"
+  steps:
+    - name: Install macos dependency
+      shell: bash
+      run: |
+        brew install protobuf
+    - name: Build wheel
+      uses: PyO3/maturin-action@v1
+      with:
+        command: build
+        args: ${{ inputs.args }}
+        working-directory: python
+        interpreter: 3.${{ inputs.python-minor-version }}

.github/workflows/build_windows_wheel/action.yml

@@ -0,0 +1,33 @@
+# We create a composite action to be re-used both for testing and for releasing
+name: build_wheel
+description: "Build a lance wheel"
+inputs:
+  python-minor-version:
+    description: "8, 9, 10, 11"
+    required: true
+  args:
+    description: "--release"
+    required: false
+    default: ""
+runs:
+  using: "composite"
+  steps:
+    - name: Install Protoc v21.12
+      working-directory: C:\
+      run: |
+        New-Item -Path 'C:\protoc' -ItemType Directory
+        Set-Location C:\protoc
+        Invoke-WebRequest https://github.com/protocolbuffers/protobuf/releases/download/v21.12/protoc-21.12-win64.zip -OutFile C:\protoc\protoc.zip
+        7z x protoc.zip
+        Add-Content $env:GITHUB_PATH "C:\protoc\bin"
+      shell: powershell
+    - name: Build wheel
+      uses: PyO3/maturin-action@v1
+      with:
+        command: build
+        args: ${{ inputs.args }}
+        working-directory: python
+    - uses: actions/upload-artifact@v3
+      with:
+        name: windows-wheels
+        path: python\target\wheels

.github/workflows/pypi-publish.yml

@@ -2,30 +2,91 @@ name: PyPI Publish
 
 on:
   release:
-    types: [ published ]
+    types: [published]
 
 jobs:
-  publish:
+  linux:
-    runs-on: ubuntu-latest
+    timeout-minutes: 60
-    # Only runs on tags that matches the python-make-release action
+    strategy:
-    if: startsWith(github.ref, 'refs/tags/python-v')
+      matrix:
-    defaults:
+        python-minor-version: ["8"]
-      run:
+        platform:
-        shell: bash
+          - x86_64
-        working-directory: python
+          - aarch64
+    runs-on: "ubuntu-22.04"
     steps:
       - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
       - name: Set up Python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v4
         with:
-          python-version: "3.8"
+          python-version: 3.${{ matrix.python-minor-version }}
-      - name: Build distribution
+      - uses: ./.github/workflows/build_linux_wheel
-        run: |
-          ls -la
-          pip install wheel setuptools --upgrade
-          python setup.py sdist bdist_wheel
-      - name: Publish
-        uses: pypa/gh-action-pypi-publish@v1.8.5
         with:
-          password: ${{ secrets.LANCEDB_PYPI_API_TOKEN }}
+          python-minor-version: ${{ matrix.python-minor-version }}
-          packages-dir: python/dist
+          args: "--release --strip"
+          arm-build: ${{ matrix.platform == 'aarch64' }}
+      - uses: ./.github/workflows/upload_wheel
+        with:
+          token: ${{ secrets.LANCEDB_PYPI_API_TOKEN }}
+          repo: "pypi"
+  mac:
+    timeout-minutes: 60
+    runs-on: ${{ matrix.config.runner }}
+    strategy:
+      matrix:
+        python-minor-version: ["8"]
+        config:
+          - target: x86_64-apple-darwin
+            runner: macos-13
+          - target: aarch64-apple-darwin
+            runner: macos-14
+    env:
+      MACOSX_DEPLOYMENT_TARGET: 10.15
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          ref: ${{ inputs.ref }}
+          fetch-depth: 0
+          lfs: true
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: 3.12
+      - uses: ./.github/workflows/build_mac_wheel
+        with:
+          python-minor-version: ${{ matrix.python-minor-version }}
+          args: "--release --strip --target ${{ matrix.config.target }}"
+      - uses: ./.github/workflows/upload_wheel
+        with:
+          python-minor-version: ${{ matrix.python-minor-version }}
+          token: ${{ secrets.LANCEDB_PYPI_API_TOKEN }}
+          repo: "pypi"
+  windows:
+    timeout-minutes: 60
+    runs-on: windows-latest
+    strategy:
+      matrix:
+        python-minor-version: ["8"]
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          ref: ${{ inputs.ref }}
+          fetch-depth: 0
+          lfs: true
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: 3.${{ matrix.python-minor-version }}
+      - uses: ./.github/workflows/build_windows_wheel
+        with:
+          python-minor-version: ${{ matrix.python-minor-version }}
+          args: "--release --strip"
+          vcpkg_token: ${{ secrets.VCPKG_GITHUB_PACKAGES }}
+      - uses: ./.github/workflows/upload_wheel
+        with:
+          python-minor-version: ${{ matrix.python-minor-version }}
+          token: ${{ secrets.LANCEDB_PYPI_API_TOKEN }}
+          repo: "pypi"

.github/workflows/python.yml

@@ -14,49 +14,133 @@ concurrency:
   cancel-in-progress: true
 
 jobs:
-  linux:
+  lint:
+    name: "Lint"
     timeout-minutes: 30
-    strategy:
-      matrix:
-        python-minor-version: [ "8", "11" ]
     runs-on: "ubuntu-22.04"
     defaults:
       run:
         shell: bash
         working-directory: python
     steps:
-    - uses: actions/checkout@v4
+      - uses: actions/checkout@v4
-      with:
+        with:
-        fetch-depth: 0
+          fetch-depth: 0
-        lfs: true
+          lfs: true
-    - name: Set up Python
+      - name: Set up Python
-      uses: actions/setup-python@v5
+        uses: actions/setup-python@v5
-      with:
+        with:
-        python-version: 3.${{ matrix.python-minor-version }}
+          python-version: "3.11"
-    - name: Install lancedb
+      - name: Install ruff
-      run: |
+        run: |
-        pip install -e .[tests]
+          pip install ruff
-        pip install tantivy@git+https://github.com/quickwit-oss/tantivy-py#164adc87e1a033117001cf70e38c82a53014d985
+      - name: Format check
-        pip install pytest pytest-mock ruff
+        run: ruff format --check .
-    - name: Format check
+      - name: Lint
-      run: ruff format --check .
+        run: ruff .
-    - name: Lint
+  doctest:
-      run: ruff .
+    name: "Doctest"
-    - name: Run tests
+    timeout-minutes: 30
-      run: pytest -m "not slow" -x -v --durations=30 tests
+    runs-on: "ubuntu-22.04"
-    - name: doctest
+    defaults:
-      run: pytest --doctest-modules lancedb
+      run:
+        shell: bash
+        working-directory: python
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+          cache: "pip"
+      - name: Install protobuf
+        run: |
+          sudo apt update
+          sudo apt install -y protobuf-compiler
+      - uses: Swatinem/rust-cache@v2
+        with:
+          workspaces: python
+      - name: Install
+        run: |
+          pip install -e .[tests,dev,embeddings]
+          pip install tantivy@git+https://github.com/quickwit-oss/tantivy-py#164adc87e1a033117001cf70e38c82a53014d985
+          pip install mlx
+      - name: Doctest
+        run: pytest --doctest-modules python/lancedb
+  linux:
+    name: "Linux: python-3.${{ matrix.python-minor-version }}"
+    timeout-minutes: 30
+    strategy:
+      matrix:
+        python-minor-version: ["8", "11"]
+    runs-on: "ubuntu-22.04"
+    defaults:
+      run:
+        shell: bash
+        working-directory: python
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Install protobuf
+        run: |
+          sudo apt update
+          sudo apt install -y protobuf-compiler
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: 3.${{ matrix.python-minor-version }}
+      - uses: Swatinem/rust-cache@v2
+        with:
+          workspaces: python
+      - uses: ./.github/workflows/build_linux_wheel
+      - uses: ./.github/workflows/run_tests
+      # Make sure wheels are not included in the Rust cache
+      - name: Delete wheels
+        run: rm -rf target/wheels
   platform:
-    name: "Platform: ${{ matrix.config.name }}"
+    name: "Mac: ${{ matrix.config.name }}"
     timeout-minutes: 30
     strategy:
       matrix:
         config:
-          - name: x86 Mac
+          - name: x86
             runner: macos-13
-          - name: Arm Mac
+          - name: Arm
             runner: macos-14
-          - name: x86 Windows
+    runs-on: "${{ matrix.config.runner }}"
+    defaults:
+      run:
+        shell: bash
+        working-directory: python
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+      - uses: Swatinem/rust-cache@v2
+        with:
+          workspaces: python
+      - uses: ./.github/workflows/build_mac_wheel
+      - uses: ./.github/workflows/run_tests
+      # Make sure wheels are not included in the Rust cache
+      - name: Delete wheels
+        run: rm -rf target/wheels
+  windows:
+    name: "Windows: ${{ matrix.config.name }}"
+    timeout-minutes: 30
+    strategy:
+      matrix:
+        config:
+          - name: x86
             runner: windows-latest
     runs-on: "${{ matrix.config.runner }}"
     defaults:
@@ -64,21 +148,22 @@ jobs:
         shell: bash
         working-directory: python
     steps:
-    - uses: actions/checkout@v4
+      - uses: actions/checkout@v4
-      with:
+        with:
-        fetch-depth: 0
+          fetch-depth: 0
-        lfs: true
+          lfs: true
-    - name: Set up Python
+      - name: Set up Python
-      uses: actions/setup-python@v5
+        uses: actions/setup-python@v5
-      with:
+        with:
-        python-version: "3.11"
+          python-version: "3.11"
-    - name: Install lancedb
+      - uses: Swatinem/rust-cache@v2
-      run: |
+        with:
-        pip install -e .[tests]
+          workspaces: python
-        pip install tantivy@git+https://github.com/quickwit-oss/tantivy-py#164adc87e1a033117001cf70e38c82a53014d985
+      - uses: ./.github/workflows/build_windows_wheel
-        pip install pytest pytest-mock
+      - uses: ./.github/workflows/run_tests
-    - name: Run tests
+      # Make sure wheels are not included in the Rust cache
-      run: pytest -m "not slow" -x -v --durations=30 tests
+      - name: Delete wheels
+        run: rm -rf target/wheels
   pydantic1x:
     timeout-minutes: 30
     runs-on: "ubuntu-22.04"
@@ -87,21 +172,22 @@ jobs:
         shell: bash
         working-directory: python
     steps:
-    - uses: actions/checkout@v4
+      - uses: actions/checkout@v4
-      with:
+        with:
-        fetch-depth: 0
+          fetch-depth: 0
-        lfs: true
+          lfs: true
-    - name: Set up Python
+      - name: Install dependencies
-      uses: actions/setup-python@v5
+        run: |
-      with:
+          sudo apt update
-        python-version: 3.9
+          sudo apt install -y protobuf-compiler
-    - name: Install lancedb
+      - name: Set up Python
-      run: |
+        uses: actions/setup-python@v5
-        pip install "pydantic<2"
+        with:
-        pip install -e .[tests]
+          python-version: 3.9
-        pip install tantivy@git+https://github.com/quickwit-oss/tantivy-py#164adc87e1a033117001cf70e38c82a53014d985
+      - name: Install lancedb
-        pip install pytest pytest-mock
+        run: |
-    - name: Run tests
+          pip install "pydantic<2"
-      run: pytest -m "not slow" -x -v --durations=30 tests
+          pip install -e .[tests]
-    - name: doctest
+          pip install tantivy@git+https://github.com/quickwit-oss/tantivy-py#164adc87e1a033117001cf70e38c82a53014d985
-      run: pytest --doctest-modules lancedb
+      - name: Run tests
+        run: pytest -m "not slow" -x -v --durations=30 python/tests

.github/workflows/run_tests/action.yml

@@ -0,0 +1,17 @@
+name: run-tests
+
+description: "Install lance wheel and run unit tests"
+inputs:
+  python-minor-version:
+    required: true
+    description: "8 9 10 11 12"
+runs:
+  using: "composite"
+  steps:
+    - name: Install lancedb
+      shell: bash
+      run: |
+        pip3 install $(ls target/wheels/lancedb-*.whl)[tests,dev,embeddings]
+    - name: pytest
+      shell: bash
+      run: pytest -m "not slow" -x -v --durations=30 python/python/tests

.github/workflows/upload_wheel/action.yml

@@ -0,0 +1,29 @@
+name: upload-wheel
+
+description: "Upload wheels to Pypi"
+inputs:
+  os:
+    required: true
+    description: "ubuntu-22.04 or macos-13"
+  repo:
+    required: false
+    description: "pypi or testpypi"
+    default: "pypi"
+  token:
+    required: true
+    description: "release token for the repo"
+
+runs:
+  using: "composite"
+  steps:
+    - name: Install dependencies
+      shell: bash
+      run: |
+        python -m pip install --upgrade pip
+        pip install twine
+    - name: Publish wheel
+      env:
+        TWINE_USERNAME: __token__
+        TWINE_PASSWORD: ${{ inputs.token }}
+      shell: bash
+      run: twine upload --repository ${{ inputs.repo }} target/wheels/lancedb-*.whl

.gitignore

@@ -22,6 +22,11 @@ python/dist
 
 **/.hypothesis
 
+# Compiled Dynamic libraries
+*.so
+*.dylib
+*.dll
+
 ## Javascript
 *.node
 **/node_modules

Cargo.toml

@@ -1,5 +1,5 @@
 [workspace]
-members = ["rust/ffi/node", "rust/lancedb", "nodejs"]
+members = ["rust/ffi/node", "rust/lancedb", "nodejs", "python"]
 # Python package needs to be built by maturin.
 exclude = ["python"]
 resolver = "2"
@@ -14,10 +14,10 @@ keywords = ["lancedb", "lance", "database", "vector", "search"]
 categories = ["database-implementations"]
 
 [workspace.dependencies]
-lance = { "version" = "=0.9.18", "features" = ["dynamodb"] }
+lance = { "version" = "=0.10.1", "features" = ["dynamodb"] }
-lance-index = { "version" = "=0.9.18" }
+lance-index = { "version" = "=0.10.1" }
-lance-linalg = { "version" = "=0.9.18" }
+lance-linalg = { "version" = "=0.10.1" }
-lance-testing = { "version" = "=0.9.18" }
+lance-testing = { "version" = "=0.10.1" }
 # Note that this one does not include pyarrow
 arrow = { version = "50.0", optional = false }
 arrow-array = "50.0"

dockerfiles/Dockerfile

@@ -0,0 +1,27 @@
+#Simple base dockerfile that supports basic dependencies required to run lance with FTS and Hybrid Search
+#Usage docker build -t lancedb:latest -f Dockerfile .
+FROM python:3.10-slim-buster
+
+# Install Rust
+RUN apt-get update && apt-get install -y curl build-essential && \
+  curl https://sh.rustup.rs -sSf | sh -s -- -y
+
+# Set the environment variable for Rust
+ENV PATH="/root/.cargo/bin:${PATH}"
+
+# Install protobuf compiler
+RUN apt-get install -y protobuf-compiler && \
+  apt-get clean && \
+  rm -rf /var/lib/apt/lists/*
+
+RUN apt-get -y update &&\
+  apt-get -y upgrade && \
+  apt-get -y install git
+
+
+# Verify installations
+RUN python --version && \
+  rustc --version && \
+  protoc --version
+
+RUN pip install tantivy lancedb

node/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "vectordb",
-  "version": "0.4.10",
+  "version": "0.4.11",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "vectordb",
-      "version": "0.4.10",
+      "version": "0.4.11",
       "cpu": [
         "x64",
         "arm64"
@@ -53,11 +53,11 @@
         "uuid": "^9.0.0"
       },
       "optionalDependencies": {
-        "@lancedb/vectordb-darwin-arm64": "0.4.10",
+        "@lancedb/vectordb-darwin-arm64": "0.4.11",
-        "@lancedb/vectordb-darwin-x64": "0.4.10",
+        "@lancedb/vectordb-darwin-x64": "0.4.11",
-        "@lancedb/vectordb-linux-arm64-gnu": "0.4.10",
+        "@lancedb/vectordb-linux-arm64-gnu": "0.4.11",
-        "@lancedb/vectordb-linux-x64-gnu": "0.4.10",
+        "@lancedb/vectordb-linux-x64-gnu": "0.4.11",
-        "@lancedb/vectordb-win32-x64-msvc": "0.4.10"
+        "@lancedb/vectordb-win32-x64-msvc": "0.4.11"
       }
     },
     "node_modules/@75lb/deep-merge": {
@@ -329,9 +329,9 @@
       }
     },
     "node_modules/@lancedb/vectordb-darwin-arm64": {
-      "version": "0.4.10",
+      "version": "0.4.11",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-arm64/-/vectordb-darwin-arm64-0.4.10.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-arm64/-/vectordb-darwin-arm64-0.4.11.tgz",
-      "integrity": "sha512-y/uHOGb0g15pvqv5tdTyZ6oN+0QVpBmZDzKFWW6pPbuSZjB2uPqcs+ti0RB+AUdmS21kavVQqaNsw/HLKEGrHA==",
+      "integrity": "sha512-JDOKmFnuJPFkA7ZmrzBJolROwSjWr7yMvAbi40uLBc25YbbVezodd30u2EFtIwWwtk1GqNYRZ49FZOElKYeC/Q==",
       "cpu": [
         "arm64"
       ],
@@ -341,9 +341,9 @@
       ]
     },
     "node_modules/@lancedb/vectordb-darwin-x64": {
-      "version": "0.4.10",
+      "version": "0.4.11",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-x64/-/vectordb-darwin-x64-0.4.10.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-x64/-/vectordb-darwin-x64-0.4.11.tgz",
-      "integrity": "sha512-XbfR58OkQpAe0xMSTrwJh9ZjGSzG9EZ7zwO6HfYem8PxcLYAcC6eWRWoSG/T0uObyrPTcYYyvHsp0eNQWYBFAQ==",
+      "integrity": "sha512-iy6r+8tp2v1EFgJV52jusXtxgO6NY6SkpOdX41xPqN2mQWMkfUAR9Xtks1mgknjPOIKH4MRc8ZS0jcW/UWmilQ==",
       "cpu": [
         "x64"
       ],
@@ -353,9 +353,9 @@
       ]
     },
     "node_modules/@lancedb/vectordb-linux-arm64-gnu": {
-      "version": "0.4.10",
+      "version": "0.4.11",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-arm64-gnu/-/vectordb-linux-arm64-gnu-0.4.10.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-arm64-gnu/-/vectordb-linux-arm64-gnu-0.4.11.tgz",
-      "integrity": "sha512-x40WKH9b+KxorRmKr9G7fv8p5mMj8QJQvRMA0v6v+nbZHr2FLlAZV+9mvhHOnm4AGIkPP5335cUgv6Qz6hgwkQ==",
+      "integrity": "sha512-5K6IVcTMuH0SZBjlqB5Gg39WC889FpTwIWKufxzQMMXrzxo5J3lKUHVoR28RRlNhDF2d9kZXBEyCpIfDFsV9iQ==",
       "cpu": [
         "arm64"
       ],
@@ -365,9 +365,9 @@
       ]
     },
     "node_modules/@lancedb/vectordb-linux-x64-gnu": {
-      "version": "0.4.10",
+      "version": "0.4.11",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-x64-gnu/-/vectordb-linux-x64-gnu-0.4.10.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-x64-gnu/-/vectordb-linux-x64-gnu-0.4.11.tgz",
-      "integrity": "sha512-CTGPpuzlqq2nVjUxI9gAJOT1oBANIovtIaFsOmBSnEAHgX7oeAxKy2b6L/kJzsgqSzvR5vfLwYcWFrr6ZmBxSA==",
+      "integrity": "sha512-hF9ZChsdqKqqnivOzd9mE7lC3PmhZadXtwThi2RrsPiOLoEaGDfmr6Ni3amVQnB3bR8YEJtTxdQxe0NC4uW/8g==",
       "cpu": [
         "x64"
       ],
@@ -377,9 +377,9 @@
       ]
     },
     "node_modules/@lancedb/vectordb-win32-x64-msvc": {
-      "version": "0.4.10",
+      "version": "0.4.11",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-win32-x64-msvc/-/vectordb-win32-x64-msvc-0.4.10.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-win32-x64-msvc/-/vectordb-win32-x64-msvc-0.4.11.tgz",
-      "integrity": "sha512-Fd7r74coZyrKzkfXg4WthqOL+uKyJyPTia6imcrMNqKOlTGdKmHf02Qi2QxWZrFaabkRYo4Tpn5FeRJ3yYX8CA==",
+      "integrity": "sha512-0+9ut1ccKoqIyGxsVixwx3771Z+DXpl5WfSmOeA8kf3v3jlOg2H+0YUahiXLDid2ju+yeLPrAUYm7A1gKHVhew==",
       "cpu": [
         "x64"
       ],

node/package.json

@@ -61,11 +61,13 @@
     "uuid": "^9.0.0"
   },
   "dependencies": {
-    "@apache-arrow/ts": "^14.0.2",
     "@neon-rs/load": "^0.0.74",
-    "apache-arrow": "^14.0.2",
     "axios": "^1.4.0"
   },
+  "peerDependencies": {
+    "@apache-arrow/ts": "^14.0.2",
+    "apache-arrow": "^14.0.2"
+  },
   "os": [
     "darwin",
     "linux",

node/src/index.ts

@@ -42,7 +42,10 @@ const {
   tableCompactFiles,
   tableListIndices,
   tableIndexStats,
-  tableSchema
+  tableSchema,
+  tableAddColumns,
+  tableAlterColumns,
+  tableDropColumns
   // eslint-disable-next-line @typescript-eslint/no-var-requires
 } = require('../native.js')
 
@@ -500,6 +503,59 @@ export interface Table<T = number[]> {
   filter(value: string): Query<T>
 
   schema: Promise<Schema>
+
+  // TODO: Support BatchUDF
+  /**
+   * Add new columns with defined values.
+   *
+   * @param newColumnTransforms pairs of column names and 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.
+   */
+  addColumns(newColumnTransforms: Array<{ name: string, valueSql: string }>): Promise<void>
+
+  /**
+   * Alter the name or nullability of columns.
+   *
+   * @param columnAlterations One or more alterations to apply to columns.
+   */
+  alterColumns(columnAlterations: ColumnAlteration[]): Promise<void>
+
+  /**
+   * Drop one or more columns from the dataset
+   *
+   * This is a metadata-only operation and does not remove the data from the
+   * underlying storage. In order to remove the data, you must subsequently
+   * call ``compact_files`` to rewrite the data without the removed columns and
+   * then call ``cleanup_files`` to remove the old files.
+   *
+   * @param 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").
+   */
+  dropColumns(columnNames: string[]): Promise<void>
+}
+
+/**
+ * A definition of a column alteration. The alteration changes the column at
+ * `path` to have the new name `name`, to be nullable if `nullable` is true,
+ * and to have the data type `data_type`. At least one of `rename` or `nullable`
+ * must be provided.
+ */
+export interface ColumnAlteration {
+  /**
+   * The path to the column to alter. This is a dot-separated path to the column.
+   * If it is a top-level column then it is just the name of the column. If it is
+   * a nested column then it is the path to the column, e.g. "a.b.c" for a column
+   * `c` nested inside a column `b` nested inside a column `a`.
+   */
+  path: string
+  rename?: string
+  /**
+   * Set the new nullability. Note that a nullable column cannot be made non-nullable.
+   */
+  nullable?: boolean
 }
 
 export interface UpdateArgs {
@@ -1028,6 +1084,18 @@ export class LocalTable<T = number[]> implements Table<T> {
       return false
     }
   }
+
+  async addColumns (newColumnTransforms: Array<{ name: string, valueSql: string }>): Promise<void> {
+    return tableAddColumns.call(this._tbl, newColumnTransforms)
+  }
+
+  async alterColumns (columnAlterations: ColumnAlteration[]): Promise<void> {
+    return tableAlterColumns.call(this._tbl, columnAlterations)
+  }
+
+  async dropColumns (columnNames: string[]): Promise<void> {
+    return tableDropColumns.call(this._tbl, columnNames)
+  }
 }
 
 export interface CleanupStats {

node/src/remote/index.ts

@@ -25,7 +25,8 @@ import {
   type UpdateArgs,
   type UpdateSqlArgs,
   makeArrowTable,
-  type MergeInsertArgs
+  type MergeInsertArgs,
+  type ColumnAlteration
 } from '../index'
 import { Query } from '../query'
 
@@ -474,4 +475,16 @@ export class RemoteTable<T = number[]> implements Table<T> {
       numUnindexedRows: results.data.num_unindexed_rows
     }
   }
+
+  async addColumns (newColumnTransforms: Array<{ name: string, valueSql: string }>): Promise<void> {
+    throw new Error('Add columns is not yet supported in LanceDB Cloud.')
+  }
+
+  async alterColumns (columnAlterations: ColumnAlteration[]): Promise<void> {
+    throw new Error('Alter columns is not yet supported in LanceDB Cloud.')
+  }
+
+  async dropColumns (columnNames: string[]): Promise<void> {
+    throw new Error('Drop columns is not yet supported in LanceDB Cloud.')
+  }
 }

node/src/test/test.ts

@@ -37,8 +37,10 @@ import {
   Utf8,
   Table as ArrowTable,
   vectorFromArray,
+  Float64,
   Float32,
-  Float16
+  Float16,
+  Int64
 } from 'apache-arrow'
 
 const expect = chai.expect
@@ -196,7 +198,7 @@ describe('LanceDB client', function () {
       const table = await con.openTable('vectors')
       const results = await table
         .search([0.1, 0.1])
-        .select(['is_active'])
+        .select(['is_active', 'vector'])
         .execute()
       assert.equal(results.length, 2)
       // vector and _distance are always returned
@@ -1057,3 +1059,63 @@ describe('Compact and cleanup', function () {
     assert.equal(await table.countRows(), 3)
   })
 })
+
+describe('schema evolution', function () {
+  // Create a new sample table
+  it('can add a new column to the schema', async function () {
+    const dir = await track().mkdir('lancejs')
+    const con = await lancedb.connect(dir)
+    const table = await con.createTable('vectors', [
+      { id: 1n, vector: [0.1, 0.2] }
+    ])
+
+    await table.addColumns([{ name: 'price', valueSql: 'cast(10.0 as float)' }])
+
+    const expectedSchema = new Schema([
+      new Field('id', new Int64()),
+      new Field('vector', new FixedSizeList(2, new Field('item', new Float32(), true))),
+      new Field('price', new Float32())
+    ])
+    expect(await table.schema).to.deep.equal(expectedSchema)
+  })
+
+  it('can alter the columns in the schema', async function () {
+    const dir = await track().mkdir('lancejs')
+    const con = await lancedb.connect(dir)
+    const schema = new Schema([
+      new Field('id', new Int64(), false),
+      new Field('vector', new FixedSizeList(2, new Field('item', new Float32(), true))),
+      new Field('price', new Float64(), false)
+    ])
+    const table = await con.createTable('vectors', [
+      { id: 1n, vector: [0.1, 0.2], price: 10.0 }
+    ])
+    expect(await table.schema).to.deep.equal(schema)
+
+    await table.alterColumns([
+      { path: 'id', rename: 'new_id' },
+      { path: 'price', nullable: true }
+    ])
+
+    const expectedSchema = new Schema([
+      new Field('new_id', new Int64(), false),
+      new Field('vector', new FixedSizeList(2, new Field('item', new Float32(), true))),
+      new Field('price', new Float64(), true)
+    ])
+    expect(await table.schema).to.deep.equal(expectedSchema)
+  })
+
+  it('can drop a column from the schema', async function () {
+    const dir = await track().mkdir('lancejs')
+    const con = await lancedb.connect(dir)
+    const table = await con.createTable('vectors', [
+      { id: 1n, vector: [0.1, 0.2] }
+    ])
+    await table.dropColumns(['vector'])
+
+    const expectedSchema = new Schema([
+      new Field('id', new Int64(), false)
+    ])
+    expect(await table.schema).to.deep.equal(expectedSchema)
+  })
+})

nodejs/__test__/table.test.ts

@@ -17,7 +17,7 @@ import * as path from "path";
 import * as fs from "fs";
 
 import { connect } from "../dist";
-import { Schema, Field, Float32, Int32, FixedSizeList } from "apache-arrow";
+import { Schema, Field, Float32, Int32, FixedSizeList, Int64, Float64 } from "apache-arrow";
 import { makeArrowTable } from "../dist/arrow";
 
 describe("Test creating index", () => {
@@ -214,4 +214,69 @@ describe("Read consistency interval", () => {
       expect(await table2.countRows()).toEqual(2n);
     }
   });
+});
+
+
+describe('schema evolution', function () {
+  let tmpDir: string;
+  beforeEach(() => {
+    tmpDir = fs.mkdtempSync(path.join(os.tmpdir(), "schema-evolution-"));
+  });
+
+  // Create a new sample table
+  it('can add a new column to the schema', async function () {
+    const con = await connect(tmpDir)
+    const table = await con.createTable('vectors', [
+      { id: 1n, vector: [0.1, 0.2] }
+    ])
+
+    await table.addColumns([{ name: 'price', valueSql: 'cast(10.0 as float)' }])
+
+    const expectedSchema = new Schema([
+      new Field('id', new Int64(), true),
+      new Field('vector', new FixedSizeList(2, new Field('item', new Float32(), true)), true),
+      new Field('price', new Float32(), false)
+    ])
+    expect(await table.schema()).toEqual(expectedSchema)
+  });
+
+  it('can alter the columns in the schema', async function () {
+    const con = await connect(tmpDir)
+    const schema = new Schema([
+      new Field('id', new Int64(), true),
+      new Field('vector', new FixedSizeList(2, new Field('item', new Float32(), true)), true),
+      new Field('price', new Float64(), false)
+    ])
+    const table = await con.createTable('vectors', [
+      { id: 1n, vector: [0.1, 0.2] }
+    ])
+    // Can create a non-nullable column only through addColumns at the moment.
+    await table.addColumns([{ name: 'price', valueSql: 'cast(10.0 as double)' }])
+    expect(await table.schema()).toEqual(schema)
+
+    await table.alterColumns([
+      { path: 'id', rename: 'new_id' },
+      { path: 'price', nullable: true }
+    ])
+
+    const expectedSchema = new Schema([
+      new Field('new_id', new Int64(), true),
+      new Field('vector', new FixedSizeList(2, new Field('item', new Float32(), true)), true),
+      new Field('price', new Float64(), true)
+    ])
+    expect(await table.schema()).toEqual(expectedSchema)
+  });
+
+  it('can drop a column from the schema', async function () {
+    const con = await connect(tmpDir)
+    const table = await con.createTable('vectors', [
+      { id: 1n, vector: [0.1, 0.2] }
+    ])
+    await table.dropColumns(['vector'])
+
+    const expectedSchema = new Schema([
+      new Field('id', new Int64(), true)
+    ])
+    expect(await table.schema()).toEqual(expectedSchema)
+  });
 });

nodejs/lancedb/native.d.ts

@@ -12,6 +12,38 @@ export const enum MetricType {
   Cosine = 1,
   Dot = 2
 }
+/**
+ *  A definition of a column alteration. The alteration changes the column at
+ * `path` to have the new name `name`, to be nullable if `nullable` is true,
+ * and to have the data type `data_type`. At least one of `rename` or `nullable`
+ * must be provided.
+ */
+export interface ColumnAlteration {
+  /**
+   * The path to the column to alter. This is a dot-separated path to the column.
+   * If it is a top-level column then it is just the name of the column. If it is
+   * a nested column then it is the path to the column, e.g. "a.b.c" for a column
+   * `c` nested inside a column `b` nested inside a column `a`.
+   */
+  path: string
+  /**
+   * The new name of the column. If not provided then the name will not be changed.
+   * This must be distinct from the names of all other columns in the table.
+   */
+  rename?: string
+  /** Set the new nullability. Note that a nullable column cannot be made non-nullable. */
+  nullable?: boolean
+}
+/** A definition of a new column to add to a table. */
+export interface AddColumnsSql {
+  /** The name of the new column. */
+  name: string
+  /**
+   * The values to populate the new column with, as a SQL expression.
+   * The expression can reference other columns in the table.
+   */
+  valueSql: string
+}
 export interface ConnectionOptions {
   uri: string
   apiKey?: string
@@ -89,4 +121,7 @@ export class Table {
   delete(predicate: string): Promise<void>
   createIndex(): IndexBuilder
   query(): Query
+  addColumns(transforms: Array<AddColumnsSql>): Promise<void>
+  alterColumns(alterations: Array<ColumnAlteration>): Promise<void>
+  dropColumns(columns: Array<string>): Promise<void>
 }

nodejs/lancedb/table.ts

@@ -13,7 +13,7 @@
 // limitations under the License.
 
 import { Schema, tableFromIPC } from "apache-arrow";
-import { Table as _NativeTable } from "./native";
+import { AddColumnsSql, ColumnAlteration, Table as _NativeTable } from "./native";
 import { toBuffer, Data } from "./arrow";
 import { Query } from "./query";
 import { IndexBuilder } from "./indexer";
@@ -150,4 +150,42 @@ export class Table {
     }
     return q;
   }
+
+  // TODO: Support BatchUDF
+  /**
+   * Add new columns with defined values.
+   *
+   * @param newColumnTransforms pairs of column names and 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.
+   */
+  async addColumns(newColumnTransforms: AddColumnsSql[]): Promise<void> {
+    await this.inner.addColumns(newColumnTransforms);
+  }
+
+  /**
+   * Alter the name or nullability of columns.
+   *
+   * @param columnAlterations One or more alterations to apply to columns.
+   */
+  async alterColumns(columnAlterations: ColumnAlteration[]): Promise<void> {
+    await this.inner.alterColumns(columnAlterations);
+  }
+
+  /**
+   * Drop one or more columns from the dataset
+   *
+   * This is a metadata-only operation and does not remove the data from the
+   * underlying storage. In order to remove the data, you must subsequently
+   * call ``compact_files`` to rewrite the data without the removed columns and
+   * then call ``cleanup_files`` to remove the old files.
+   *
+   * @param 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").
+   */
+  async dropColumns(columnNames: string[]): Promise<void> {
+    await this.inner.dropColumns(columnNames);
+  }
 }

nodejs/package.json

@@ -62,7 +62,7 @@
     "lancedb-linux-arm64-gnu": "0.4.3",
     "lancedb-linux-x64-gnu": "0.4.3"
   },
-  "dependencies": {
+  "peerDependencies": {
     "apache-arrow": "^15.0.0"
   }
 }

nodejs/src/table.rs

@@ -13,8 +13,11 @@
 // limitations under the License.
 
 use arrow_ipc::writer::FileWriter;
-use lancedb::table::AddDataOptions;
+use lance::dataset::ColumnAlteration as LanceColumnAlteration;
-use lancedb::{ipc::ipc_file_to_batches, table::TableRef};
+use lancedb::{
+    ipc::ipc_file_to_batches,
+    table::{AddDataOptions, TableRef},
+};
 use napi::bindgen_prelude::*;
 use napi_derive::napi;
 
@@ -93,4 +96,106 @@ impl Table {
     pub fn query(&self) -> Query {
         Query::new(self)
     }
+
+    #[napi]
+    pub async fn add_columns(&self, transforms: Vec<AddColumnsSql>) -> napi::Result<()> {
+        let transforms = transforms
+            .into_iter()
+            .map(|sql| (sql.name, sql.value_sql))
+            .collect::<Vec<_>>();
+        let transforms = lance::dataset::NewColumnTransform::SqlExpressions(transforms);
+        self.table
+            .add_columns(transforms, None)
+            .await
+            .map_err(|err| {
+                napi::Error::from_reason(format!(
+                    "Failed to add columns to table {}: {}",
+                    self.table, err
+                ))
+            })?;
+        Ok(())
+    }
+
+    #[napi]
+    pub async fn alter_columns(&self, alterations: Vec<ColumnAlteration>) -> napi::Result<()> {
+        for alteration in &alterations {
+            if alteration.rename.is_none() && alteration.nullable.is_none() {
+                return Err(napi::Error::from_reason(
+                    "Alteration must have a 'rename' or 'nullable' field.",
+                ));
+            }
+        }
+        let alterations = alterations
+            .into_iter()
+            .map(LanceColumnAlteration::from)
+            .collect::<Vec<_>>();
+
+        self.table
+            .alter_columns(&alterations)
+            .await
+            .map_err(|err| {
+                napi::Error::from_reason(format!(
+                    "Failed to alter columns in table {}: {}",
+                    self.table, err
+                ))
+            })?;
+        Ok(())
+    }
+
+    #[napi]
+    pub async fn drop_columns(&self, columns: Vec<String>) -> napi::Result<()> {
+        let col_refs = columns.iter().map(String::as_str).collect::<Vec<_>>();
+        self.table.drop_columns(&col_refs).await.map_err(|err| {
+            napi::Error::from_reason(format!(
+                "Failed to drop columns from table {}: {}",
+                self.table, err
+            ))
+        })?;
+        Ok(())
+    }
+}
+
+///  A definition of a column alteration. The alteration changes the column at
+/// `path` to have the new name `name`, to be nullable if `nullable` is true,
+/// and to have the data type `data_type`. At least one of `rename` or `nullable`
+/// must be provided.
+#[napi(object)]
+pub struct ColumnAlteration {
+    /// The path to the column to alter. This is a dot-separated path to the column.
+    /// If it is a top-level column then it is just the name of the column. If it is
+    /// a nested column then it is the path to the column, e.g. "a.b.c" for a column
+    /// `c` nested inside a column `b` nested inside a column `a`.
+    pub path: String,
+    /// The new name of the column. If not provided then the name will not be changed.
+    /// This must be distinct from the names of all other columns in the table.
+    pub rename: Option<String>,
+    /// Set the new nullability. Note that a nullable column cannot be made non-nullable.
+    pub nullable: Option<bool>,
+}
+
+impl From<ColumnAlteration> for LanceColumnAlteration {
+    fn from(js: ColumnAlteration) -> Self {
+        let ColumnAlteration {
+            path,
+            rename,
+            nullable,
+        } = js;
+        Self {
+            path,
+            rename,
+            nullable,
+            // TODO: wire up this field
+            data_type: None,
+        }
+    }
+}
+
+/// A definition of a new column to add to a table.
+#[napi(object)]
+pub struct AddColumnsSql {
+    /// The name of the new column.
+    pub name: String,
+    /// The values to populate the new column with, as a SQL expression.
+    /// The expression can reference other columns in the table.
+    pub value_sql: String,
 }

python/.bumpversion.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.5.7
+current_version = 0.6.0
 commit = True
 message = [python] Bump version: {current_version} → {new_version}
 tag = True

python/Cargo.toml

@@ -0,0 +1,26 @@
+[package]
+name = "lancedb-python"
+version = "0.4.10"
+edition.workspace = true
+description = "Python bindings for LanceDB"
+license.workspace = true
+repository.workspace = true
+keywords.workspace = true
+categories.workspace = true
+
+
+[lib]
+name = "_lancedb"
+crate-type = ["cdylib"]
+
+[dependencies]
+lancedb = { path = "../rust/lancedb" }
+env_logger = "0.10"
+pyo3 = { version = "0.20", features = ["extension-module", "abi3-py38"] }
+pyo3-asyncio = { version = "0.20", features = ["attributes", "tokio-runtime"] }
+
+# Prevent dynamic linking of lzma, which comes from datafusion
+lzma-sys = { version = "*", features = ["static"] }
+
+[build-dependencies]
+pyo3-build-config = { version = "0.20.3", features = ["extension-module", "abi3-py38"] }

python/README.md

@@ -20,10 +20,10 @@ results = table.search([0.1, 0.3]).limit(20).to_list()
 print(results)
 ```
 
-
 ## Development
 
-Create a virtual environment and activate it:
+LanceDb is based on the rust crate `lancedb` and is built with maturin.  In order to build with maturin
+you will either need a conda environment or a virtual environment (venv).
 
 ```bash
 python -m venv venv
@@ -33,7 +33,15 @@ python -m venv venv
 Install the necessary packages:
 
 ```bash
-python -m pip install .
+python -m pip install .[tests,dev]
+```
+
+To build the python package you can use maturin:
+
+```bash
+# This will build the rust bindings and place them in the appropriate place
+# in your venv or conda environment
+matruin develop
 ```
 
 To run the unit tests:
@@ -45,7 +53,7 @@ pytest
 To run the doc tests:
 
 ```bash
-pytest --doctest-modules lancedb
+pytest --doctest-modules python/lancedb
 ```
 
 To run linter and automatically fix all errors:
@@ -61,31 +69,27 @@ If any packages are missing, install them with:
 pip install <PACKAGE_NAME>
 ```
 
-
 ___
 For **Windows** users, there may be errors when installing packages, so these commands may be helpful:
 
 Activate the virtual environment:
+
 ```bash
 . .\venv\Scripts\activate
 ```
 
 You may need to run the installs separately:
+
 ```bash
 pip install -e .[tests]
 pip install -e .[dev]
 ```
 
-
 `tantivy` requires `rust` to be installed, so install it with `conda`, as it doesn't support windows installation:
+
 ```bash
 pip install wheel
 pip install cargo
 conda install rust
 pip install tantivy
 ```
-
-To run the unit tests:
-```bash
-pytest
-```

python/build.rs

@@ -0,0 +1,3 @@
+fn main() {
+    pyo3_build_config::add_extension_module_link_args();
+}

python/pyproject.toml

@@ -1,9 +1,9 @@
 [project]
 name = "lancedb"
-version = "0.5.7"
+version = "0.6.0"
 dependencies = [
     "deprecation",
-    "pylance==0.9.18",
+    "pylance==0.10.1",
     "ratelimiter~=1.0",
     "retry>=0.9.2",
     "tqdm>=4.27.0",
@@ -14,7 +14,7 @@ dependencies = [
     "pyyaml>=6.0",
     "click>=8.1.7",
     "requests>=2.31.0",
-    "overrides>=0.7"
+    "overrides>=0.7",
 ]
 description = "lancedb"
 authors = [{ name = "LanceDB Devs", email = "dev@lancedb.com" }]
@@ -26,7 +26,7 @@ keywords = [
     "data-science",
     "machine-learning",
     "arrow",
-    "data-analytics"
+    "data-analytics",
 ]
 classifiers = [
     "Development Status :: 3 - Alpha",
@@ -48,21 +48,53 @@ classifiers = [
 repository = "https://github.com/lancedb/lancedb"
 
 [project.optional-dependencies]
-tests = ["aiohttp", "pandas>=1.4", "pytest", "pytest-mock", "pytest-asyncio", "duckdb", "pytz", "polars>=0.19"]
+tests = [
+    "aiohttp",
+    "pandas>=1.4",
+    "pytest",
+    "pytest-mock",
+    "pytest-asyncio",
+    "duckdb",
+    "pytz",
+    "polars>=0.19",
+]
 dev = ["ruff", "pre-commit"]
-docs = ["mkdocs", "mkdocs-jupyter", "mkdocs-material", "mkdocstrings[python]", "mkdocs-ultralytics-plugin==0.0.44"]
+docs = [
+    "mkdocs",
+    "mkdocs-jupyter",
+    "mkdocs-material",
+    "mkdocstrings[python]",
+    "mkdocs-ultralytics-plugin==0.0.44",
+]
 clip = ["torch", "pillow", "open-clip"]
-embeddings = ["openai>=1.6.1", "sentence-transformers", "torch", "pillow", "open-clip-torch", "cohere", "huggingface_hub",
+embeddings = [
-                "InstructorEmbedding", "google.generativeai", "boto3>=1.28.57", "awscli>=1.29.57", "botocore>=1.31.57"]
+    "openai>=1.6.1",
+    "sentence-transformers",
+    "torch",
+    "pillow",
+    "open-clip-torch",
+    "cohere",
+    "huggingface_hub",
+    "InstructorEmbedding",
+    "google.generativeai",
+    "boto3>=1.28.57",
+    "awscli>=1.29.57",
+    "botocore>=1.31.57",
+]
+
+[tool.maturin]
+python-source = "python"
+module-name = "lancedb._lancedb"
 
 [project.scripts]
 lancedb = "lancedb.cli.cli:cli"
 
 [build-system]
-requires = ["setuptools", "wheel"]
+requires = ["maturin>=1.4"]
-build-backend = "setuptools.build_meta"
+build-backend = "maturin"
 
-[tool.ruff]
+
+[tool.ruff.lint]
 select = ["F", "E", "W", "I", "G", "TCH", "PERF"]
 
 [tool.pytest.ini_options]
@@ -70,5 +102,5 @@ addopts = "--strict-markers --ignore-glob=lancedb/embeddings/*.py"
 
 markers = [
     "slow: marks tests as slow (deselect with '-m \"not slow\"')",
-    "asyncio"
+    "asyncio",
 ]

python/python/lancedb/__init__.py

@@ -19,8 +19,9 @@ from typing import Optional, Union
 
 __version__ = importlib.metadata.version("lancedb")
 
-from .common import URI
+from ._lancedb import connect as lancedb_connect
-from .db import DBConnection, LanceDBConnection
+from .common import URI, sanitize_uri
+from .db import AsyncConnection, AsyncLanceDBConnection, DBConnection, LanceDBConnection
 from .remote.db import RemoteDBConnection
 from .schema import vector  # noqa: F401
 from .utils import sentry_log  # noqa: F401
@@ -101,3 +102,74 @@ def connect(
             uri, api_key, region, host_override, request_thread_pool=request_thread_pool
         )
     return LanceDBConnection(uri, read_consistency_interval=read_consistency_interval)
+
+
+async def connect_async(
+    uri: URI,
+    *,
+    api_key: Optional[str] = None,
+    region: str = "us-east-1",
+    host_override: Optional[str] = None,
+    read_consistency_interval: Optional[timedelta] = None,
+    request_thread_pool: Optional[Union[int, ThreadPoolExecutor]] = None,
+) -> AsyncConnection:
+    """Connect to a LanceDB database.
+
+    Parameters
+    ----------
+    uri: str or Path
+        The uri of the database.
+    api_key: str, optional
+        If present, connect to LanceDB cloud.
+        Otherwise, connect to a database on file system or cloud storage.
+        Can be set via environment variable `LANCEDB_API_KEY`.
+    region: str, default "us-east-1"
+        The region to use for LanceDB Cloud.
+    host_override: str, optional
+        The override url for LanceDB Cloud.
+    read_consistency_interval: timedelta, default None
+        (For LanceDB OSS only)
+        The interval at which to check for updates to the table from other
+        processes. If None, then consistency is not checked. For performance
+        reasons, this is the default. For strong consistency, set this to
+        zero seconds. Then every read will check for updates from other
+        processes. As a compromise, you can set this to a non-zero timedelta
+        for eventual consistency. If more than that interval has passed since
+        the last check, then the table will be checked for updates. Note: this
+        consistency only applies to read operations. Write operations are
+        always consistent.
+    request_thread_pool: int or ThreadPoolExecutor, optional
+        The thread pool to use for making batch requests to the LanceDB Cloud API.
+        If an integer, then a ThreadPoolExecutor will be created with that
+        number of threads. If None, then a ThreadPoolExecutor will be created
+        with the default number of threads. If a ThreadPoolExecutor, then that
+        executor will be used for making requests. This is for LanceDB Cloud
+        only and is only used when making batch requests (i.e., passing in
+        multiple queries to the search method at once).
+
+    Examples
+    --------
+
+    For a local directory, provide a path for the database:
+
+    >>> import lancedb
+    >>> db = lancedb.connect("~/.lancedb")
+
+    For object storage, use a URI prefix:
+
+    >>> db = lancedb.connect("s3://my-bucket/lancedb")
+
+    Connect to LancdDB cloud:
+
+    >>> db = lancedb.connect("db://my_database", api_key="ldb_...")
+
+    Returns
+    -------
+    conn : DBConnection
+        A connection to a LanceDB database.
+    """
+    return AsyncLanceDBConnection(
+        await lancedb_connect(
+            sanitize_uri(uri), api_key, region, host_override, read_consistency_interval
+        )
+    )

python/python/lancedb/_lancedb.pyi

@@ -0,0 +1,12 @@
+from typing import Optional
+
+class Connection(object):
+    async def table_names(self) -> list[str]: ...
+
+async def connect(
+    uri: str,
+    api_key: Optional[str],
+    region: Optional[str],
+    host_override: Optional[str],
+    read_consistency_interval: Optional[float],
+) -> Connection: ...

python/python/lancedb/common.py

@@ -34,3 +34,7 @@ class Credential(str):
 
     def __str__(self) -> str:
         return "********"
+
+
+def sanitize_uri(uri: URI) -> str:
+    return str(uri)

python/python/lancedb/db.py

@@ -28,6 +28,7 @@ from .util import fs_from_uri, get_uri_location, get_uri_scheme, join_uri
 if TYPE_CHECKING:
     from datetime import timedelta
 
+    from ._lancedb import Connection as LanceDbConnection
     from .common import DATA, URI
     from .embeddings import EmbeddingFunctionConfig
     from .pydantic import LanceModel
@@ -40,14 +41,21 @@ class DBConnection(EnforceOverrides):
     def table_names(
         self, page_token: Optional[str] = None, limit: int = 10
     ) -> Iterable[str]:
-        """List all table in this database
+        """List all tables in this database, in sorted order
 
         Parameters
         ----------
         page_token: str, optional
             The token to use for pagination. If not present, start from the beginning.
+            Typically, this token is last table name from the previous page.
+            Only supported by LanceDb Cloud.
         limit: int, default 10
             The size of the page to return.
+            Only supported by LanceDb Cloud.
+
+        Returns
+        -------
+        Iterable of str
         """
         pass
 
@@ -412,3 +420,254 @@ class LanceDBConnection(DBConnection):
     def drop_database(self):
         filesystem, path = fs_from_uri(self.uri)
         filesystem.delete_dir(path)
+
+
+class AsyncConnection(EnforceOverrides):
+    """An active LanceDB connection interface."""
+
+    @abstractmethod
+    async def table_names(
+        self, *, page_token: Optional[str] = None, limit: int = 10
+    ) -> Iterable[str]:
+        """List all tables in this database, in sorted order
+
+        Parameters
+        ----------
+        page_token: str, optional
+            The token to use for pagination. If not present, start from the beginning.
+            Typically, this token is last table name from the previous page.
+            Only supported by LanceDb Cloud.
+        limit: int, default 10
+            The size of the page to return.
+            Only supported by LanceDb Cloud.
+
+        Returns
+        -------
+        Iterable of str
+        """
+        pass
+
+    @abstractmethod
+    async def create_table(
+        self,
+        name: str,
+        data: Optional[DATA] = None,
+        schema: Optional[Union[pa.Schema, LanceModel]] = None,
+        mode: str = "create",
+        exist_ok: bool = False,
+        on_bad_vectors: str = "error",
+        fill_value: float = 0.0,
+        embedding_functions: Optional[List[EmbeddingFunctionConfig]] = None,
+    ) -> Table:
+        """Create a [Table][lancedb.table.Table] in the database.
+
+        Parameters
+        ----------
+        name: str
+            The name of the table.
+        data: The data to initialize the table, *optional*
+            User must provide at least one of `data` or `schema`.
+            Acceptable types are:
+
+            - dict or list-of-dict
+
+            - pandas.DataFrame
+
+            - pyarrow.Table or pyarrow.RecordBatch
+        schema: The schema of the table, *optional*
+            Acceptable types are:
+
+            - pyarrow.Schema
+
+            - [LanceModel][lancedb.pydantic.LanceModel]
+        mode: str; default "create"
+            The mode to use when creating the table.
+            Can be either "create" or "overwrite".
+            By default, if the table already exists, an exception is raised.
+            If you want to overwrite the table, use mode="overwrite".
+        exist_ok: bool, default False
+            If a table by the same name already exists, then raise an exception
+            if exist_ok=False. If exist_ok=True, then open the existing table;
+            it will not add the provided data but will validate against any
+            schema that's specified.
+        on_bad_vectors: str, default "error"
+            What to do if any of the vectors are not the same size or contains NaNs.
+            One of "error", "drop", "fill".
+        fill_value: float
+            The value to use when filling vectors. Only used if on_bad_vectors="fill".
+
+        Returns
+        -------
+        LanceTable
+            A reference to the newly created table.
+
+        !!! note
+
+            The vector index won't be created by default.
+            To create the index, call the `create_index` method on the table.
+
+        Examples
+        --------
+
+        Can create with list of tuples or dictionaries:
+
+        >>> import lancedb
+        >>> db = lancedb.connect("./.lancedb")
+        >>> data = [{"vector": [1.1, 1.2], "lat": 45.5, "long": -122.7},
+        ...         {"vector": [0.2, 1.8], "lat": 40.1, "long":  -74.1}]
+        >>> db.create_table("my_table", data)
+        LanceTable(connection=..., name="my_table")
+        >>> db["my_table"].head()
+        pyarrow.Table
+        vector: fixed_size_list<item: float>[2]
+          child 0, item: float
+        lat: double
+        long: double
+        ----
+        vector: [[[1.1,1.2],[0.2,1.8]]]
+        lat: [[45.5,40.1]]
+        long: [[-122.7,-74.1]]
+
+        You can also pass a pandas DataFrame:
+
+        >>> import pandas as pd
+        >>> data = pd.DataFrame({
+        ...    "vector": [[1.1, 1.2], [0.2, 1.8]],
+        ...    "lat": [45.5, 40.1],
+        ...    "long": [-122.7, -74.1]
+        ... })
+        >>> db.create_table("table2", data)
+        LanceTable(connection=..., name="table2")
+        >>> db["table2"].head()
+        pyarrow.Table
+        vector: fixed_size_list<item: float>[2]
+          child 0, item: float
+        lat: double
+        long: double
+        ----
+        vector: [[[1.1,1.2],[0.2,1.8]]]
+        lat: [[45.5,40.1]]
+        long: [[-122.7,-74.1]]
+
+        Data is converted to Arrow before being written to disk. For maximum
+        control over how data is saved, either provide the PyArrow schema to
+        convert to or else provide a [PyArrow Table](pyarrow.Table) directly.
+
+        >>> custom_schema = pa.schema([
+        ...   pa.field("vector", pa.list_(pa.float32(), 2)),
+        ...   pa.field("lat", pa.float32()),
+        ...   pa.field("long", pa.float32())
+        ... ])
+        >>> db.create_table("table3", data, schema = custom_schema)
+        LanceTable(connection=..., name="table3")
+        >>> db["table3"].head()
+        pyarrow.Table
+        vector: fixed_size_list<item: float>[2]
+          child 0, item: float
+        lat: float
+        long: float
+        ----
+        vector: [[[1.1,1.2],[0.2,1.8]]]
+        lat: [[45.5,40.1]]
+        long: [[-122.7,-74.1]]
+
+
+        It is also possible to create an table from `[Iterable[pa.RecordBatch]]`:
+
+
+        >>> import pyarrow as pa
+        >>> def make_batches():
+        ...     for i in range(5):
+        ...         yield pa.RecordBatch.from_arrays(
+        ...             [
+        ...                 pa.array([[3.1, 4.1], [5.9, 26.5]],
+        ...                     pa.list_(pa.float32(), 2)),
+        ...                 pa.array(["foo", "bar"]),
+        ...                 pa.array([10.0, 20.0]),
+        ...             ],
+        ...             ["vector", "item", "price"],
+        ...         )
+        >>> schema=pa.schema([
+        ...     pa.field("vector", pa.list_(pa.float32(), 2)),
+        ...     pa.field("item", pa.utf8()),
+        ...     pa.field("price", pa.float32()),
+        ... ])
+        >>> db.create_table("table4", make_batches(), schema=schema)
+        LanceTable(connection=..., name="table4")
+
+        """
+        raise NotImplementedError
+
+    async def open_table(self, name: str) -> Table:
+        """Open a Lance Table in the database.
+
+        Parameters
+        ----------
+        name: str
+            The name of the table.
+
+        Returns
+        -------
+        A LanceTable object representing the table.
+        """
+        raise NotImplementedError
+
+    async def drop_table(self, name: str):
+        """Drop a table from the database.
+
+        Parameters
+        ----------
+        name: str
+            The name of the table.
+        """
+        raise NotImplementedError
+
+    async def drop_database(self):
+        """
+        Drop database
+        This is the same thing as dropping all the tables
+        """
+        raise NotImplementedError
+
+
+class AsyncLanceDBConnection(AsyncConnection):
+    def __init__(self, connection: LanceDbConnection):
+        self._inner = connection
+
+    async def __repr__(self) -> str:
+        pass
+
+    @override
+    async def table_names(
+        self,
+        *,
+        page_token=None,
+        limit=None,
+    ) -> Iterable[str]:
+        return await self._inner.table_names()
+
+    @override
+    async def create_table(
+        self,
+        name: str,
+        data: Optional[DATA] = None,
+        schema: Optional[Union[pa.Schema, LanceModel]] = None,
+        mode: str = "create",
+        exist_ok: bool = False,
+        on_bad_vectors: str = "error",
+        fill_value: float = 0.0,
+        embedding_functions: Optional[List[EmbeddingFunctionConfig]] = None,
+    ) -> LanceTable:
+        raise NotImplementedError
+
+    @override
+    async def open_table(self, name: str) -> LanceTable:
+        raise NotImplementedError
+
+    @override
+    async def drop_table(self, name: str, ignore_missing: bool = False):
+        raise NotImplementedError
+
+    @override
+    async def drop_database(self):
+        raise NotImplementedError

python/python/lancedb/query.py

@@ -16,7 +16,7 @@ from __future__ import annotations
 from abc import ABC, abstractmethod
 from concurrent.futures import ThreadPoolExecutor
 from pathlib import Path
-from typing import TYPE_CHECKING, List, Literal, Optional, Tuple, Type, Union
+from typing import TYPE_CHECKING, Dict, List, Literal, Optional, Tuple, Type, Union
 
 import deprecation
 import numpy as np
@@ -93,7 +93,7 @@ class Query(pydantic.BaseModel):
     metric: str = "L2"
 
     # which columns to return in the results
-    columns: Optional[List[str]] = None
+    columns: Optional[Union[List[str], Dict[str, str]]] = None
 
     # optional query parameters for tuning the results,
     # e.g. `{"nprobes": "10", "refine_factor": "10"}`
@@ -321,20 +321,27 @@ class LanceQueryBuilder(ABC):
             self._limit = limit
         return self
 
-    def select(self, columns: list) -> LanceQueryBuilder:
+    def select(self, columns: Union[list[str], dict[str, str]]) -> LanceQueryBuilder:
         """Set the columns to return.
 
         Parameters
         ----------
-        columns: list
+        columns: list of str, or dict of str to str default None
-            The columns to return.
+            List of column names to be fetched.
+            Or a dictionary of column names to SQL expressions.
+            All columns are fetched if None or unspecified.
 
         Returns
         -------
         LanceQueryBuilder
             The LanceQueryBuilder object.
         """
-        self._columns = columns
+        if isinstance(columns, list):
+            self._columns = columns
+        elif isinstance(columns, dict):
+            self._columns = list(columns.items())
+        else:
+            raise ValueError("columns must be a list or a dictionary")
         return self
 
     def where(self, where: str, prefilter: bool = False) -> LanceQueryBuilder:
@@ -392,7 +399,7 @@ class LanceVectorQueryBuilder(LanceQueryBuilder):
     >>> (table.search([0.4, 0.4])
     ...       .metric("cosine")
     ...       .where("b < 10")
-    ...       .select(["b"])
+    ...       .select(["b", "vector"])
     ...       .limit(2)
     ...       .to_pandas())
        b      vector  _distance

python/python/lancedb/remote/table.py

@@ -15,7 +15,7 @@ import logging
 import uuid
 from concurrent.futures import Future
 from functools import cached_property
-from typing import Dict, Optional, Union
+from typing import Dict, Iterable, Optional, Union
 
 import pyarrow as pa
 from lance import json_to_schema
@@ -473,6 +473,21 @@ class RemoteTable(Table):
             "count_rows() is not yet supported on the LanceDB cloud"
         )
 
+    def add_columns(self, transforms: Dict[str, str]):
+        raise NotImplementedError(
+            "add_columns() is not yet supported on the LanceDB cloud"
+        )
+
+    def alter_columns(self, alterations: Iterable[Dict[str, str]]):
+        raise NotImplementedError(
+            "alter_columns() is not yet supported on the LanceDB cloud"
+        )
+
+    def drop_columns(self, columns: Iterable[str]):
+        raise NotImplementedError(
+            "drop_columns() is not yet supported on the LanceDB cloud"
+        )
+
 
 def add_index(tbl: pa.Table, i: int) -> pa.Table:
     return tbl.add_column(

python/python/lancedb/table.py

@@ -160,7 +160,7 @@ class Table(ABC):
 
     Can query the table with [Table.search][lancedb.table.Table.search].
 
-    >>> table.search([0.4, 0.4]).select(["b"]).to_pandas()
+    >>> table.search([0.4, 0.4]).select(["b", "vector"]).to_pandas()
        b      vector  _distance
     0  4  [0.5, 1.3]       0.82
     1  2  [1.1, 1.2]       1.13
@@ -436,7 +436,7 @@ class Table(ABC):
         >>> query = [0.4, 1.4, 2.4]
         >>> (table.search(query)
         ...     .where("original_width > 1000", prefilter=True)
-        ...     .select(["caption", "original_width"])
+        ...     .select(["caption", "original_width", "vector"])
         ...     .limit(2)
         ...     .to_pandas())
           caption  original_width           vector  _distance
@@ -660,6 +660,56 @@ class Table(ABC):
         For most cases, the default should be fine.
         """
 
+    @abstractmethod
+    def add_columns(self, transforms: Dict[str, str]):
+        """
+        Add new columns with defined values.
+
+        This is not yet available in LanceDB Cloud.
+
+        Parameters
+        ----------
+        transforms: Dict[str, str]
+            A map of column name to a 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.
+        """
+
+    @abstractmethod
+    def alter_columns(self, alterations: Iterable[Dict[str, str]]):
+        """
+        Alter column names and nullability.
+
+        This is not yet available in LanceDB Cloud.
+
+        alterations : Iterable[Dict[str, Any]]
+            A sequence of dictionaries, each with the following keys:
+            - "path": str
+                The column path to alter. For a top-level column, this is the name.
+                For a nested column, this is the dot-separated path, e.g. "a.b.c".
+            - "name": str, optional
+                The new name of the column. If not specified, the column name is
+                not changed.
+            - "nullable": bool, optional
+                Whether the column should be nullable. If not specified, the column
+                nullability is not changed. Only non-nullable columns can be changed
+                to nullable. Currently, you cannot change a nullable column to
+                non-nullable.
+        """
+
+    @abstractmethod
+    def drop_columns(self, columns: Iterable[str]):
+        """
+        Drop columns from the table.
+
+        This is not yet available in LanceDB Cloud.
+
+        Parameters
+        ----------
+        columns : Iterable[str]
+            The names of the columns to drop.
+        """
+
 
 class _LanceDatasetRef(ABC):
     @property
@@ -1219,7 +1269,7 @@ class LanceTable(Table):
         >>> query = [0.4, 1.4, 2.4]
         >>> (table.search(query)
         ...     .where("original_width > 1000", prefilter=True)
-        ...     .select(["caption", "original_width"])
+        ...     .select(["caption", "original_width", "vector"])
         ...     .limit(2)
         ...     .to_pandas())
           caption  original_width           vector  _distance
@@ -1536,6 +1586,22 @@ class LanceTable(Table):
         """
         return self.to_lance().optimize.compact_files(*args, **kwargs)
 
+    def add_columns(self, transforms: Dict[str, str]):
+        self._dataset_mut.add_columns(transforms)
+
+    def alter_columns(self, *alterations: Iterable[Dict[str, str]]):
+        modified = []
+        # I called this name in pylance, but I think I regret that now. So we
+        # allow both name and rename.
+        for alter in alterations:
+            if "rename" in alter:
+                alter["name"] = alter.pop("rename")
+            modified.append(alter)
+        self._dataset_mut.alter_columns(*modified)
+
+    def drop_columns(self, columns: Iterable[str]):
+        self._dataset_mut.drop_columns(columns)
+
 
 def _sanitize_schema(
     data: pa.Table,

python/python/tests/test_cli.py

@@ -1,5 +1,4 @@
 from click.testing import CliRunner
-
 from lancedb.cli.cli import cli
 from lancedb.utils import CONFIG
 

python/python/tests/test_context.py

@@ -13,7 +13,6 @@
 
 import pandas as pd
 import pytest
-
 from lancedb.context import contextualize
 
 

python/python/tests/test_db.py

@@ -11,12 +11,11 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 
+import lancedb
 import numpy as np
 import pandas as pd
 import pyarrow as pa
 import pytest
-
-import lancedb
 from lancedb.pydantic import LanceModel, Vector
 
 
@@ -166,6 +165,24 @@ def test_table_names(tmp_path):
     assert db.table_names() == ["test1", "test2", "test3"]
 
 
+@pytest.mark.asyncio
+async def test_table_names_async(tmp_path):
+    db = lancedb.connect(tmp_path)
+    data = pd.DataFrame(
+        {
+            "vector": [[3.1, 4.1], [5.9, 26.5]],
+            "item": ["foo", "bar"],
+            "price": [10.0, 20.0],
+        }
+    )
+    db.create_table("test2", data=data)
+    db.create_table("test1", data=data)
+    db.create_table("test3", data=data)
+
+    db = await lancedb.connect_async(tmp_path)
+    assert await db.table_names() == ["test1", "test2", "test3"]
+
+
 def test_create_mode(tmp_path):
     db = lancedb.connect(tmp_path)
     data = pd.DataFrame(

python/python/tests/test_e2e_remote_db.py

@@ -13,7 +13,6 @@
 
 import numpy as np
 import pytest
-
 from lancedb import LanceDBConnection
 
 # TODO: setup integ test mark and script

python/python/tests/test_embeddings.py

@@ -13,11 +13,10 @@
 import sys
 
 import lance
+import lancedb
 import numpy as np
 import pyarrow as pa
 import pytest
-
-import lancedb
 from lancedb.conftest import MockTextEmbeddingFunction
 from lancedb.embeddings import (
     EmbeddingFunctionConfig,

python/python/tests/test_embeddings_slow.py

@@ -14,12 +14,11 @@ import importlib
 import io
 import os
 
+import lancedb
 import numpy as np
 import pandas as pd
 import pytest
 import requests
-
-import lancedb
 from lancedb.embeddings import get_registry
 from lancedb.pydantic import LanceModel, Vector
 
@@ -185,10 +184,9 @@ def test_imagebind(tmp_path):
     import shutil
     import tempfile
 
+    import lancedb.embeddings.imagebind
     import pandas as pd
     import requests
-
-    import lancedb.embeddings.imagebind
     from lancedb.embeddings import get_registry
     from lancedb.pydantic import LanceModel, Vector
 

python/python/tests/test_fts.py

@@ -14,13 +14,13 @@ import os
 import random
 from unittest import mock
 
+import lancedb as ldb
 import numpy as np
 import pandas as pd
 import pytest
-import tantivy
 
-import lancedb as ldb
+pytest.importorskip("lancedb.fts")
-import lancedb.fts
+tantivy = pytest.importorskip("tantivy")
 
 
 @pytest.fixture

python/python/tests/test_io.py

@@ -13,9 +13,8 @@
 
 import os
 
-import pytest
-
 import lancedb
+import pytest
 
 # You need to setup AWS credentials an a base path to run this test. Example
 #    AWS_PROFILE=default TEST_S3_BASE_URL=s3://my_bucket/dataset pytest tests/test_io.py

python/python/tests/test_pydantic.py

@@ -20,9 +20,8 @@ from typing import List, Optional, Tuple
 import pyarrow as pa
 import pydantic
 import pytest
-from pydantic import Field
-
 from lancedb.pydantic import PYDANTIC_VERSION, LanceModel, Vector, pydantic_to_schema
+from pydantic import Field
 
 
 @pytest.mark.skipif(

python/python/tests/test_query.py

@@ -18,7 +18,6 @@ import numpy as np
 import pandas.testing as tm
 import pyarrow as pa
 import pytest
-
 from lancedb.db import LanceDBConnection
 from lancedb.pydantic import LanceModel, Vector
 from lancedb.query import LanceVectorQueryBuilder, Query
@@ -88,7 +87,7 @@ def test_query_builder(table):
     rs = (
         LanceVectorQueryBuilder(table, [0, 0], "vector")
         .limit(1)
-        .select(["id"])
+        .select(["id", "vector"])
         .to_list()
     )
     assert rs[0]["id"] == 1

python/python/tests/test_remote_client.py

@@ -17,7 +17,6 @@ import pandas as pd
 import pyarrow as pa
 import pytest
 from aiohttp import web
-
 from lancedb.remote.client import RestfulLanceDBClient, VectorQuery
 
 

python/python/tests/test_remote_db.py

@@ -11,9 +11,8 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 
-import pyarrow as pa
-
 import lancedb
+import pyarrow as pa
 from lancedb.remote.client import VectorQuery, VectorQueryResult
 
 

python/python/tests/test_rerankers.py

@@ -1,9 +1,8 @@
 import os
 
+import lancedb
 import numpy as np
 import pytest
-
-import lancedb
 from lancedb.conftest import MockTextEmbeddingFunction  # noqa
 from lancedb.embeddings import EmbeddingFunctionRegistry
 from lancedb.pydantic import LanceModel, Vector
@@ -15,6 +14,9 @@ from lancedb.rerankers import (
 )
 from lancedb.table import LanceTable
 
+# Tests rely on FTS index
+pytest.importorskip("lancedb.fts")
+
 
 def get_test_table(tmp_path):
     db = lancedb.connect(tmp_path)

python/python/tests/test_table.py

@@ -20,19 +20,18 @@ from typing import List
 from unittest.mock import PropertyMock, patch
 
 import lance
+import lancedb
 import numpy as np
 import pandas as pd
 import polars as pl
 import pyarrow as pa
 import pytest
-from pydantic import BaseModel
-
-import lancedb
 from lancedb.conftest import MockTextEmbeddingFunction
 from lancedb.db import LanceDBConnection
 from lancedb.embeddings import EmbeddingFunctionConfig, EmbeddingFunctionRegistry
 from lancedb.pydantic import LanceModel, Vector
 from lancedb.table import LanceTable
+from pydantic import BaseModel
 
 
 class MockDB:
@@ -804,6 +803,9 @@ def test_count_rows(db):
 
 
 def test_hybrid_search(db, tmp_path):
+    # This test uses an FTS index
+    pytest.importorskip("lancedb.fts")
+
     db = MockDB(str(tmp_path))
     # Create a LanceDB table schema with a vector and a text column
     emb = EmbeddingFunctionRegistry.get_instance().get("test")()
@@ -898,3 +900,29 @@ def test_restore_consistency(tmp_path):
     table.add([{"id": 2}])
     assert table_fixed.version == table.version - 1
     assert table_ref_latest.version == table.version
+
+
+# Schema evolution
+def test_add_columns(tmp_path):
+    db = lancedb.connect(tmp_path)
+    data = pa.table({"id": [0, 1]})
+    table = LanceTable.create(db, "my_table", data=data)
+    table.add_columns({"new_col": "id + 2"})
+    assert table.to_arrow().column_names == ["id", "new_col"]
+    assert table.to_arrow()["new_col"].to_pylist() == [2, 3]
+
+
+def test_alter_columns(tmp_path):
+    db = lancedb.connect(tmp_path)
+    data = pa.table({"id": [0, 1]})
+    table = LanceTable.create(db, "my_table", data=data)
+    table.alter_columns({"path": "id", "rename": "new_id"})
+    assert table.to_arrow().column_names == ["new_id"]
+
+
+def test_drop_columns(tmp_path):
+    db = lancedb.connect(tmp_path)
+    data = pa.table({"id": [0, 1], "category": ["a", "b"]})
+    table = LanceTable.create(db, "my_table", data=data)
+    table.drop_columns(["category"])
+    assert table.to_arrow().column_names == ["id"]

python/python/tests/test_telemetry.py

@@ -1,8 +1,7 @@
 import json
 
-import pytest
-
 import lancedb
+import pytest
 from lancedb.utils.events import _Events
 
 

python/python/tests/test_util.py

@@ -15,7 +15,6 @@ import os
 import pathlib
 
 import pytest
-
 from lancedb.util import get_uri_scheme, join_uri
 
 

python/setup.py

@@ -1,17 +0,0 @@
-#  Copyright 2023 LanceDB Developers
-#
-#  Licensed under the Apache License, Version 2.0 (the "License");
-#  you may not use this file except in compliance with the License.
-#  You may obtain a copy of the License at
-#      http://www.apache.org/licenses/LICENSE-2.0
-#
-#  Unless required by applicable law or agreed to in writing, software
-#  distributed under the License is distributed on an "AS IS" BASIS,
-#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-#  See the License for the specific language governing permissions and
-#  limitations under the License.
-
-import setuptools
-
-if __name__ == "__main__":
-    setuptools.setup()

python/src/connection.rs

@@ -0,0 +1,66 @@
+// Copyright 2024 Lance Developers.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+use std::time::Duration;
+
+use lancedb::connection::Connection as LanceConnection;
+use pyo3::{pyclass, pyfunction, pymethods, PyAny, PyRef, PyResult, Python};
+use pyo3_asyncio::tokio::future_into_py;
+
+use crate::error::PythonErrorExt;
+
+#[pyclass]
+pub struct Connection {
+    inner: LanceConnection,
+}
+
+#[pymethods]
+impl Connection {
+    pub fn table_names(self_: PyRef<'_, Self>) -> PyResult<&PyAny> {
+        let inner = self_.inner.clone();
+        future_into_py(self_.py(), async move {
+            inner.table_names().await.infer_error()
+        })
+    }
+}
+
+#[pyfunction]
+pub fn connect(
+    py: Python,
+    uri: String,
+    api_key: Option<String>,
+    region: Option<String>,
+    host_override: Option<String>,
+    read_consistency_interval: Option<f64>,
+) -> PyResult<&PyAny> {
+    future_into_py(py, async move {
+        let mut builder = lancedb::connect(&uri);
+        if let Some(api_key) = api_key {
+            builder = builder.api_key(&api_key);
+        }
+        if let Some(region) = region {
+            builder = builder.region(&region);
+        }
+        if let Some(host_override) = host_override {
+            builder = builder.host_override(&host_override);
+        }
+        if let Some(read_consistency_interval) = read_consistency_interval {
+            let read_consistency_interval = Duration::from_secs_f64(read_consistency_interval);
+            builder = builder.read_consistency_interval(read_consistency_interval);
+        }
+        Ok(Connection {
+            inner: builder.execute().await.infer_error()?,
+        })
+    })
+}

python/src/error.rs

@@ -0,0 +1,61 @@
+// Copyright 2024 Lance Developers.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+use pyo3::{
+    exceptions::{PyOSError, PyRuntimeError, PyValueError},
+    PyResult,
+};
+
+use lancedb::error::Error as LanceError;
+
+pub trait PythonErrorExt<T> {
+    /// Convert to a python error based on the Lance error type
+    fn infer_error(self) -> PyResult<T>;
+    /// Convert to OSError
+    fn os_error(self) -> PyResult<T>;
+    /// Convert to RuntimeError
+    fn runtime_error(self) -> PyResult<T>;
+    /// Convert to ValueError
+    fn value_error(self) -> PyResult<T>;
+}
+
+impl<T> PythonErrorExt<T> for std::result::Result<T, LanceError> {
+    fn infer_error(self) -> PyResult<T> {
+        match &self {
+            Ok(_) => Ok(self.unwrap()),
+            Err(err) => match err {
+                LanceError::InvalidTableName { .. } => self.value_error(),
+                LanceError::TableNotFound { .. } => self.value_error(),
+                LanceError::TableAlreadyExists { .. } => self.runtime_error(),
+                LanceError::CreateDir { .. } => self.os_error(),
+                LanceError::Store { .. } => self.runtime_error(),
+                LanceError::Lance { .. } => self.runtime_error(),
+                LanceError::Schema { .. } => self.value_error(),
+                LanceError::Runtime { .. } => self.runtime_error(),
+            },
+        }
+    }
+
+    fn os_error(self) -> PyResult<T> {
+        self.map_err(|err| PyOSError::new_err(err.to_string()))
+    }
+
+    fn runtime_error(self) -> PyResult<T> {
+        self.map_err(|err| PyRuntimeError::new_err(err.to_string()))
+    }
+
+    fn value_error(self) -> PyResult<T> {
+        self.map_err(|err| PyValueError::new_err(err.to_string()))
+    }
+}

python/src/lib.rs

@@ -0,0 +1,32 @@
+// Copyright 2024 Lance Developers.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+use connection::{connect, Connection};
+use env_logger::Env;
+use pyo3::{pymodule, types::PyModule, wrap_pyfunction, PyResult, Python};
+
+pub mod connection;
+pub(crate) mod error;
+
+#[pymodule]
+pub fn _lancedb(_py: Python, m: &PyModule) -> PyResult<()> {
+    let env = Env::new()
+        .filter_or("LANCEDB_LOG", "warn")
+        .write_style("LANCEDB_LOG_STYLE");
+    env_logger::init_from_env(env);
+    m.add_class::<Connection>()?;
+    m.add_function(wrap_pyfunction!(connect, m)?)?;
+    m.add("__version__", env!("CARGO_PKG_VERSION"))?;
+    Ok(())
+}

rust/ffi/node/src/lib.rs

@@ -286,5 +286,8 @@ fn main(mut cx: ModuleContext) -> NeonResult<()> {
         index::vector::table_create_vector_index,
     )?;
     cx.export_function("tableSchema", JsTable::js_schema)?;
+    cx.export_function("tableAddColumns", JsTable::js_add_columns)?;
+    cx.export_function("tableAlterColumns", JsTable::js_alter_columns)?;
+    cx.export_function("tableDropColumns", JsTable::js_drop_columns)?;
     Ok(())
 }

rust/ffi/node/src/table.rs

@@ -16,7 +16,7 @@ use std::ops::Deref;
 
 use arrow_array::{RecordBatch, RecordBatchIterator};
 use lance::dataset::optimize::CompactionOptions;
-use lance::dataset::{WriteMode, WriteParams};
+use lance::dataset::{ColumnAlteration, NewColumnTransform, WriteMode, WriteParams};
 use lance::io::ObjectStoreParams;
 use lancedb::table::{AddDataOptions, OptimizeAction, WriteOptions};
 
@@ -544,4 +544,118 @@ impl JsTable {
         });
         Ok(promise)
     }
+
+    pub(crate) fn js_add_columns(mut cx: FunctionContext) -> JsResult<JsPromise> {
+        let expressions = cx
+            .argument::<JsArray>(0)?
+            .to_vec(&mut cx)?
+            .into_iter()
+            .map(|val| {
+                let obj = val.downcast_or_throw::<JsObject, _>(&mut cx)?;
+                let name = obj.get::<JsString, _, _>(&mut cx, "name")?.value(&mut cx);
+                let sql = obj
+                    .get::<JsString, _, _>(&mut cx, "valueSql")?
+                    .value(&mut cx);
+                Ok((name, sql))
+            })
+            .collect::<NeonResult<Vec<(String, String)>>>()?;
+
+        let transforms = NewColumnTransform::SqlExpressions(expressions);
+
+        let js_table = cx.this().downcast_or_throw::<JsBox<Self>, _>(&mut cx)?;
+        let rt = runtime(&mut cx)?;
+
+        let (deferred, promise) = cx.promise();
+        let channel = cx.channel();
+        let table = js_table.table.clone();
+
+        rt.spawn(async move {
+            let result = table.add_columns(transforms, None).await;
+            deferred.settle_with(&channel, move |mut cx| {
+                result.or_throw(&mut cx)?;
+                Ok(cx.undefined())
+            })
+        });
+
+        Ok(promise)
+    }
+
+    pub(crate) fn js_alter_columns(mut cx: FunctionContext) -> JsResult<JsPromise> {
+        let alterations = cx
+            .argument::<JsArray>(0)?
+            .to_vec(&mut cx)?
+            .into_iter()
+            .map(|val| {
+                let obj = val.downcast_or_throw::<JsObject, _>(&mut cx)?;
+                let path = obj.get::<JsString, _, _>(&mut cx, "path")?.value(&mut cx);
+                let rename = obj
+                    .get_opt::<JsString, _, _>(&mut cx, "rename")?
+                    .map(|val| val.value(&mut cx));
+                let nullable = obj
+                    .get_opt::<JsBoolean, _, _>(&mut cx, "nullable")?
+                    .map(|val| val.value(&mut cx));
+                // TODO: support data type here. Will need to do some serialization/deserialization
+
+                if rename.is_none() && nullable.is_none() {
+                    return cx.throw_error("At least one of 'name' or 'nullable' must be provided");
+                }
+
+                Ok(ColumnAlteration {
+                    path,
+                    rename,
+                    nullable,
+                    // TODO: wire up this field
+                    data_type: None,
+                })
+            })
+            .collect::<NeonResult<Vec<ColumnAlteration>>>()?;
+
+        let js_table = cx.this().downcast_or_throw::<JsBox<Self>, _>(&mut cx)?;
+        let rt = runtime(&mut cx)?;
+
+        let (deferred, promise) = cx.promise();
+        let channel = cx.channel();
+        let table = js_table.table.clone();
+
+        rt.spawn(async move {
+            let result = table.alter_columns(&alterations).await;
+            deferred.settle_with(&channel, move |mut cx| {
+                result.or_throw(&mut cx)?;
+                Ok(cx.undefined())
+            })
+        });
+
+        Ok(promise)
+    }
+
+    pub(crate) fn js_drop_columns(mut cx: FunctionContext) -> JsResult<JsPromise> {
+        let columns = cx
+            .argument::<JsArray>(0)?
+            .to_vec(&mut cx)?
+            .into_iter()
+            .map(|val| {
+                Ok(val
+                    .downcast_or_throw::<JsString, _>(&mut cx)?
+                    .value(&mut cx))
+            })
+            .collect::<NeonResult<Vec<String>>>()?;
+
+        let js_table = cx.this().downcast_or_throw::<JsBox<Self>, _>(&mut cx)?;
+        let rt = runtime(&mut cx)?;
+
+        let (deferred, promise) = cx.promise();
+        let channel = cx.channel();
+        let table = js_table.table.clone();
+
+        rt.spawn(async move {
+            let col_refs = columns.iter().map(|s| s.as_str()).collect::<Vec<_>>();
+            let result = table.drop_columns(&col_refs).await;
+            deferred.settle_with(&channel, move |mut cx| {
+                result.or_throw(&mut cx)?;
+                Ok(cx.undefined())
+            })
+        });
+
+        Ok(promise)
+    }
 }

rust/lancedb/src/connection.rs

@@ -623,7 +623,17 @@ impl ConnectionInternal for Database {
     async fn drop_table(&self, name: &str) -> Result<()> {
         let dir_name = format!("{}.{}", name, LANCE_EXTENSION);
         let full_path = self.base_path.child(dir_name.clone());
-        self.object_store.remove_dir_all(full_path).await?;
+        self.object_store
+            .remove_dir_all(full_path)
+            .await
+            .map_err(|err| match err {
+                // this error is not lance::Error::DatasetNotFound,
+                // as the method `remove_dir_all` may be used to remove something not be a dataset
+                lance::Error::NotFound { .. } => Error::TableNotFound {
+                    name: name.to_owned(),
+                },
+                _ => Error::from(err),
+            })?;
         Ok(())
     }
 
@@ -634,8 +644,6 @@ impl ConnectionInternal for Database {
 
 #[cfg(test)]
 mod tests {
-    use std::fs::create_dir_all;
-
     use arrow_schema::{DataType, Field, Schema};
     use tempfile::tempdir;
 
@@ -691,13 +699,46 @@ mod tests {
         // let db = Database::connect("s3://bucket/path/to/database").await.unwrap();
     }
 
+    #[tokio::test]
+    #[ignore = "this can't pass due to https://github.com/lancedb/lancedb/issues/1019, enable it after the bug fixed"]
+    async fn test_open_table() {
+        let tmp_dir = tempdir().unwrap();
+        let uri = tmp_dir.path().to_str().unwrap();
+        let db = connect(uri).execute().await.unwrap();
+
+        assert_eq!(db.table_names().await.unwrap().len(), 0);
+        // open non-exist table
+        assert!(matches!(
+            db.open_table("invalid_table").execute().await,
+            Err(crate::Error::TableNotFound { .. })
+        ));
+
+        assert_eq!(db.table_names().await.unwrap().len(), 0);
+
+        let schema = Arc::new(Schema::new(vec![Field::new("x", DataType::Int32, false)]));
+        db.create_empty_table("table1", schema)
+            .execute()
+            .await
+            .unwrap();
+        db.open_table("table1").execute().await.unwrap();
+        let tables = db.table_names().await.unwrap();
+        assert_eq!(tables, vec!["table1".to_owned()]);
+    }
+
     #[tokio::test]
     async fn drop_table() {
         let tmp_dir = tempdir().unwrap();
-        create_dir_all(tmp_dir.path().join("table1.lance")).unwrap();
 
         let uri = tmp_dir.path().to_str().unwrap();
         let db = connect(uri).execute().await.unwrap();
+
+        // drop non-exist table
+        assert!(matches!(
+            db.drop_table("invalid_table").await,
+            Err(crate::Error::TableNotFound { .. }),
+        ));
+
+        create_dir_all(tmp_dir.path().join("table1.lance")).unwrap();
         db.drop_table("table1").await.unwrap();
 
         let tables = db.table_names().await.unwrap();

rust/lancedb/src/query.rs

@@ -24,6 +24,13 @@ use crate::Error;
 
 const DEFAULT_TOP_K: usize = 10;
 
+#[derive(Debug, Clone)]
+pub enum Select {
+    All,
+    Simple(Vec<String>),
+    Projection(Vec<(String, String)>),
+}
+
 /// A builder for nearest neighbor queries for LanceDB.
 #[derive(Clone)]
 pub struct Query {
@@ -44,7 +51,7 @@ pub struct Query {
     /// Apply filter to the returned rows.
     filter: Option<String>,
     /// Select column projection.
-    select: Option<Vec<String>>,
+    select: Select,
 
     /// Default is true. Set to false to enforce a brute force search.
     use_index: bool,
@@ -70,7 +77,7 @@ impl Query {
             metric_type: None,
             use_index: true,
             filter: None,
-            select: None,
+            select: Select::All,
             prefilter: false,
         }
     }
@@ -115,7 +122,16 @@ impl Query {
         scanner.use_index(self.use_index);
         scanner.prefilter(self.prefilter);
 
-        self.select.as_ref().map(|p| scanner.project(p.as_slice()));
+        match &self.select {
+            Select::Simple(select) => {
+                scanner.project(select.as_slice())?;
+            }
+            Select::Projection(select_with_transform) => {
+                scanner.project_with_transform(select_with_transform.as_slice())?;
+            }
+            Select::All => { /* Do nothing */ }
+        }
+
         self.filter.as_ref().map(|f| scanner.filter(f));
         self.refine_factor.map(|rf| scanner.refine(rf));
         self.metric_type.map(|mt| scanner.distance_metric(mt));
@@ -206,7 +222,23 @@ impl Query {
     ///
     /// Only select the specified columns. If not specified, all columns will be returned.
     pub fn select(mut self, columns: &[impl AsRef<str>]) -> Self {
-        self.select = Some(columns.iter().map(|c| c.as_ref().to_string()).collect());
+        self.select = Select::Simple(columns.iter().map(|c| c.as_ref().to_string()).collect());
+        self
+    }
+
+    /// Return only the specified columns.
+    ///
+    /// Only select the specified columns. If not specified, all columns will be returned.
+    pub fn select_with_projection(
+        mut self,
+        columns: &[(impl AsRef<str>, impl AsRef<str>)],
+    ) -> Self {
+        self.select = Select::Projection(
+            columns
+                .iter()
+                .map(|(c, t)| (c.as_ref().to_string(), t.as_ref().to_string()))
+                .collect(),
+        );
         self
     }
 
@@ -226,7 +258,7 @@ mod tests {
         RecordBatchReader,
     };
     use arrow_schema::{DataType, Field as ArrowField, Schema as ArrowSchema};
-    use futures::StreamExt;
+    use futures::{StreamExt, TryStreamExt};
     use lance::dataset::Dataset;
     use lance_testing::datagen::{BatchGenerator, IncrementingInt32, RandomVector};
     use tempfile::tempdir;
@@ -294,6 +326,38 @@ mod tests {
         }
     }
 
+    #[tokio::test]
+    async fn test_select_with_transform() {
+        let batches = make_non_empty_batches();
+        let ds = Dataset::write(batches, "memory://foo", None).await.unwrap();
+
+        let ds = DatasetConsistencyWrapper::new_latest(ds, None);
+
+        let query = Query::new(ds)
+            .limit(10)
+            .select_with_projection(&[("id2", "id * 2"), ("id", "id")]);
+        let result = query.execute_stream().await;
+        let mut batches = result
+            .expect("should have result")
+            .try_collect::<Vec<_>>()
+            .await
+            .unwrap();
+        assert_eq!(batches.len(), 1);
+        let batch = batches.pop().unwrap();
+
+        // id, and id2
+        assert_eq!(batch.num_columns(), 2);
+
+        let id: &Int32Array = batch.column_by_name("id").unwrap().as_primitive();
+        let id2: &Int32Array = batch.column_by_name("id2").unwrap().as_primitive();
+
+        id.iter().zip(id2.iter()).for_each(|(id, id2)| {
+            let id = id.unwrap();
+            let id2 = id2.unwrap();
+            assert_eq!(id * 2, id2);
+        });
+    }
+
     #[tokio::test]
     async fn test_execute_no_vector() {
         // test that it's ok to not specify a query vector (just filter / limit)

rust/lancedb/src/table.rs

@@ -27,7 +27,10 @@ use lance::dataset::optimize::{
     compact_files, CompactionMetrics, CompactionOptions, IndexRemapperOptions,
 };
 pub use lance::dataset::ReadParams;
-use lance::dataset::{Dataset, UpdateBuilder, WhenMatched, WriteMode, WriteParams};
+use lance::dataset::{
+    ColumnAlteration, Dataset, NewColumnTransform, UpdateBuilder, WhenMatched, WriteMode,
+    WriteParams,
+};
 use lance::dataset::{MergeInsertBuilder as LanceMergeInsertBuilder, WhenNotMatchedBySource};
 use lance::io::WrappingObjectStore;
 use lance_index::{optimize::OptimizeOptions, DatasetIndexExt};
@@ -376,6 +379,19 @@ pub trait Table: std::fmt::Display + Send + Sync {
     /// Modeled after ``VACUUM`` in PostgreSQL.
     /// Not all implementations support explicit optimization.
     async fn optimize(&self, action: OptimizeAction) -> Result<OptimizeStats>;
+
+    /// Add new columns to the table, providing values to fill in.
+    async fn add_columns(
+        &self,
+        transforms: NewColumnTransform,
+        read_columns: Option<Vec<String>>,
+    ) -> Result<()>;
+
+    /// Change a column's name or nullability.
+    async fn alter_columns(&self, alterations: &[ColumnAlteration]) -> Result<()>;
+
+    /// Remove columns from the table.
+    async fn drop_columns(&self, columns: &[&str]) -> Result<()>;
 }
 
 /// Reference to a Table pointer.
@@ -902,6 +918,33 @@ impl Table for NativeTable {
         }
         Ok(stats)
     }
+
+    async fn add_columns(
+        &self,
+        transforms: NewColumnTransform,
+        read_columns: Option<Vec<String>>,
+    ) -> Result<()> {
+        self.dataset
+            .get_mut()
+            .await?
+            .add_columns(transforms, read_columns)
+            .await?;
+        Ok(())
+    }
+
+    async fn alter_columns(&self, alterations: &[ColumnAlteration]) -> Result<()> {
+        self.dataset
+            .get_mut()
+            .await?
+            .alter_columns(alterations)
+            .await?;
+        Ok(())
+    }
+
+    async fn drop_columns(&self, columns: &[&str]) -> Result<()> {
+        self.dataset.get_mut().await?.drop_columns(columns).await?;
+        Ok(())
+    }
 }
 
 #[cfg(test)]